FINDINGS AND RECOMMENDATION REGARDING STATUTE OF LIMITATIONS ISSUES

MICHAEL J. SENG, Magistrate Judge.

I. INTRODUCTION

Petitioner is a state prisoner proceeding with a petition for writ of habeas corpus pursuant to 28 U.S.C. § 2254. Petitioner is represented by Jimmy S. McBirney, Megan G. Crane, and Shannon C. Leong of Orrick, Herrington & Suitcliffe LLP, and Linda Starr of the Northern California Innocence Project. Respondent, Anthony Hedgpeth, as warden of Salinas Valley State Prison, is hereby substituted as the proper named respondent pursuant to Rule 25(d) of the Federal Rules of Civil Procedure. Respondent is represented by Kathleen A. McKenna of the Office of the Attorney General of California.

This matter arises from the following facts:

On January 15, 1997, a fire occurred at rental property owned by Petitioner. Three tenants died in the fire. Petitioner was prosecuted for starting the fire. At Petitioner's trial, fire investigators testified with conviction that the fire was intentionally set. A state criminalist testified that flammable compounds used as accelerants and found at the scene were also found on Petitioner's shoes. An eyewitness testified that she saw Petitioner at the scene just before the fire. A chain of circumstantial evidence was woven around these facts. A jury found Petitioner guilty of arson-murder. He was sentenced to life in prison without the possibility of parole.

Petitioner, through counsel, pursued post-conviction relief in both state and federal courts. However, his counsel mistakenly calculated the one year statutory deadline for filing a federal petition for habeas corpus. The petition was filed five days late. Petitioner is not entitled to equitable tolling. Instead he presents new evidence of actual innocence in an attempt to qualify for an equitable exception to the statute of limitations.

The new evidence upon which Petitioner relies establishes unequivocally that the chemical compound found on Petitioner's shoes was not the same as that found at the fire scene. Respondent now agrees there is no chemical evidence linking Petitioner to the fire scene. Respondent also now acknowledges that previously-introduced scientific evidence of arson was in fact unreliable; both parties agree it cannot be determined if the fire was intentionally or accidentally initiated.

Petitioner claims that the new scientific evidence and the unreliability of the eyewitness constitute a sufficient showing of actual innocence to equitably excuse the untimeliness of his federal habeas petition. His collateral appeals have culminated in the present evidentiary hearing in which this Court has been directed to examine all relevant evidence and make a determination as to whether any reasonable juror would find Petitioner guilty beyond a reasonable doubt.

II. FACTUAL SUMMARY

A. Petitioner's Conviction

As described by the California Court of Appeal, Fifth Appellate District, Petitioner was convicted of arson and three counts of murder:1

People v. Souliotes, 2002 Cal. App. Unpub. LEXIS 7379, 1-3 (Cal. App. 5th Dist., Aug. 5, 2002).

B. Evidence and Theories Presented at Trial

1. Overview

At trial, the prosecution's arson case against Petitioner rested primarily on three sources: (1) fire cause and origin analysis that proved the fire was arson, (2) scientific evidence that Petitioner's shoes were present at the fire scene because they tested positive for medium petroleum distillates ("MPDs") found at the scene, and (3) eyewitness Monica Sandoval's testimony that she saw Petitioner in his motor home at the scene of the fire just before it started. The prosecution also presented evidence Petitioner had motive to destroy the home.

2. Arson

At trial, fire investigators from the Modesto County Fire Department testified with certainty that the fire was caused by arson. They relied on several factors to support their conclusion: The fire was described as being unusually hot; there were "pour patterns" on the floor where flammable liquids obviously had been poured and ignited; there was "deep charring" on the walls; there was insufficient combustible material ("fuel load") in the house to sustain such an intense fire unless an ignitable liquid, i.e., an accelerant, had been added; a hand-held hydrocarbon detector indicated the presence of ignitable liquids at the scene; and the eyewitness testified that a suspicious person had surreptitiously visited the house just before the fire started. Based on this evidence, the primary fire investigator testified at trial that he had "no doubt . . . that this was an arson fire" involving an ignitable liquid, and "[t]he ignition device was a human hand." (RT 6722, 6957.)2

3. MPD Evidence

At trial, a state criminialist testified that MPDs were found on samples from the fire scene and also on Petitioner's shoes. Many MPDs are ignitable liquids; common household items such as charcoal lighter fluid, some camp fuels, and some solvents are MPDs. (RT 8885.)

This evidence created what the district attorney claimed was a very strong physical link between Petitioner and the fire scene. In his closing argument, he argued that finding MPDs at the fire scene was very unusual and that the presence of MPDs on Petitioner's shoes was the "most conclusive scientific evidence." (RT 9033, 9049.) He finished by explaining how this evidence was enough to find Petitioner guilty of arson:

(RT 9050.)

C. Petitioner's Newly Discovered Evidence

1. New MPD Testing Results

In September 2005, Petitioner's sister contacted John Lentini, a criminalist, chemist, and arson investigator, to see if there was any new technique or information regarding the testing of MPD samples from Petitioner's case. Lentini had originally tested the samples for Petitioner at trial and, like the state criminalist, concluded that both Petitioner's shoes and samples from the fire exhibited the presence of an ignitable liquid belonging to the class of compounds known as medium petroleum distillates and, further, that the MPDs in these samples could not be distinguished. (See Pet., Ex. B, ECF No. 1.) Thus, at the time of Petitioner's second trial, the MPDs found on Petitioner's shoes could not be excluded as having come from the same source of MPDs found in the samples from the fire scene. (Id.)

However, upon further research in 2005, Lentini devised a method, previously unknown to him and not described in scientific literature, to distinguish chemical differences between the MPDs found on the samples from the fire scene and Petitioner's shoes. Using that method, he found that there was no chemical match between the residue on Petitioner's shoes and the MPDs at the fire scene. Lentini alerted Petitioner and his counsel to his findings. Petitioner thereupon filed state and federal petitions seeking relief based on the new findings.

Prior to the evidentiary hearing on actual innocence, Respondent stipulated: "The MPD on Petitioner's shoes is chemically distinguishable from the MPD found on the carpet samples taken from the fire scene, and the MPDs did not originate from a common source," and "Detectable MPDs are commonly found on many household products and consumer goods, including the solvents in glues and adhesives used in floor coverings and footwear, residues of dry cleaning solvents, insecticides and cleaning agents." (See Joint Pretrial Statement, Undisputed Facts 13-14, ECF No. 108.)

Thus, this Court is presented with new evidence which annuls a key evidentiary link leading to Petitioner's conviction. Contrary to what the jury had been told, Petitioner's shoes do not link Petitioner to the scene of the fire. Respondent, despite challenging Petitioner's diligence in raising this argument, does not dispute that the chemicals are distinguishable.

2. Significant Advances in Fire Science

Petitioner also challenges the conclusions of the arson investigators and asserts that their trial findings can no longer be considered reliable. Here too, Respondent now agrees that, prior expert testimony to the contrary notwithstanding, it is not possible to determine the cause of the fire.3 The parties have so stipulated.

Specifically, while arson cannot be ruled out, the parties agree that experts cannot determine whether accident or arson was the more likely cause of the fire. "The parties' experts all agree that they cannot determine the cause and origin of the fire based on the available evidence and record as it exists today, including whether the fire was accidental or the result of arson." (See Undisputed Fact 7.) They also have stipulated that many of the factors relied upon by the experts to prove arson — burn patterns, extreme heat of the fire, lack of other fuel — are now known not to be indicators of arson. (See Undisputed Fact 3.) Indeed, the parties further stipulated that the floor damage and burn patterns in this fire appear to have resulted from flashover burning4 and the collapse of the roof rather than an ignitable liquid; that fires involving liquid accelerants do not burn any hotter than other fires; that there was a sufficient fuel load present in the house without the presence of a flammable liquid; and that hydrocarbon testers commonly give false positives, are only to be used for presumptive testing, and, standing alone, are never reliable evidence of the presence of a liquid accelerant. (See Undisputed Fact 4-10.)

III. PROCEDURAL HISTORY

A. State Court Proceedings

On October 20, 2000, Petitioner was convicted in Stanislaus Superior Court of three counts of murder relating to the arson of an inhabited structure and was sentenced to serve a term of life imprisonment without the possibility of parole. (See Abstract of Judgment, Lodged Doc. 1.5)

Petitioner filed a direct appeal with the California Court of Appeal, Fifth Appellate District. The Court of Appeal affirmed the conviction on August 5, 2002. (Lodged Doc. 2.) Petitioner filed a petition for review with the California Supreme Court. It was denied October 16, 2002.

Starting on December 10, 2003, Petitioner filed three petitions for writ of habeas corpus in state court. On January 6, 2004, the Stanislaus County Superior Court denied the petition. (Lodged Doc. 6.) The court's opinion, in its entirety holds:

The petition fails to show that appeal is inadequate.

(Id.)

Petitioner then filed a habeas petition with the Court of Appeal on March 1, 2004. It was denied summarily on August 26, 2004. Petitioner filed a petition with California Supreme Court on October 12, 2004. The California Supreme Court summarily denied the petition on April 19, 2006.

B. Federal Habeas Proceedings

1. District Court Decision and Relevant Findings

The present petition was filed on May 20, 2006. (Pet., ECF No. 1.) The petition presents five claims for relief: (1) ineffective assistance of trial counsel for failing to present an arson expert in Petitioner's defense, (2) ineffective assistance of trial counsel for failing to present additional witnesses in Petitioner's defense, (3) failure to notify Petitioner of his right to consular assistance, (4) jury misconduct, and (5) actual innocence.6

On May 29, 2007, the Court ordered Respondent to file a response to the Petition. (Order, ECF No. 8.) On July 27, 2007, Respondent filed a motion to dismiss the petition as untimely and for failure to exhaust Petitioner's actual innocence claim in state court. (Mot. To Dismiss, ECF No. 17.)

On March 20, 2008, the District Court found the petition to be untimely under the one year statute of limitations of the Antiterrorism and Effective Death Penalty Act of 1996 ("AEDPA") (28 U.S.C. § 2244(d)), did not decide the exhaustion issue7, and dismissed the petition. (Order, ECF No. 36.) Specifically, the Court found that, even with the benefit of statutory tolling under 28 U.S.C. § 2244(d)(2) during the time Petitioner filed his petitions for habeas corpus in state court, the present petition was five days late. (Order at 6.)

The Court further found that Petitioner was not entitled to statutory tolling under 28 U.S.C. § 2244(d)(1)(D) based on the newly discovered evidence distinguishing MPDs on Petitioner's shoes from those found at the fire scene; the Court concluded Petitioner had not been sufficiently diligent in discovering that evidence. (Id. at 9.) The Court based its conclusion on the fact that Petitioner had not earlier asked the expert to retest the MPD evidence. Specifically, the Court concluded Petitioner had failed to establish that the new evidence was undiscoverable prior to retesting at Petitioner's request in 2005. (Id.) The Court also rejected Petitioner's claim of equitable tolling based on his counsel's reliance on a misleading docket entry by the California Supreme Court. (Id. at 11. "[T]he court finds this to be the type of attorney error which has [been] repeatedly held to be insufficient to justify equitable tolling.") Finally, the Court denied Petitioner's actual innocence gateway claim established under Schlup v. Delo, 513 U.S. 298 (1995), as the Ninth Circuit had not yet determined if the gateway applied to a statue of limitations bar. (Order at 13.) Based on the above reasoning, the Court dismissed the petition as untimely.

2. Ninth Circuit Decision and Relevant Findings

Petitioner appealed the order to the Ninth Circuit Court of Appeals. It affirmed in part, reversed in part, and remanded the matter for further adjudication. See Souliotes v. Evans, 622 F.3d 1173 (9th Cir. 2010). The Ninth Circuit agreed that Petitioner filed his federal habeas petition five days late and was not entitled to equitable tolling. The Ninth Circuit, relying on its then-recent opinion in Lee v. Lampert, 610 F.3d 1125, 1128-31 (9th Cir. 2010), also held that the Schlup `actual innocence gateway' did not apply to challenges of the AEDPA statute of limitations and affirmed the lower court decision on such grounds. However, the Ninth Circuit held that the District Court applied an inappropriately stringent standard regarding Petitioner's diligence in presenting newly discovered evidence under 28 U.S.C. § 2244(d)(1)(D).

Specifically, the Ninth Circuit gave the following instruction regarding Petitioner's required showing of diligence:

Souliotes v. Evans, 622 F.3d at 1178-1179.

While the Ninth Circuit remanded the case to resolve the issue of diligence in presenting newly discovered evidence, it held that the statutory tolling only applied to Petitioner's substantive actual innocence claim. Id. at 1180. Additionally, the Ninth Circuit ordered this Court to hold an "expedited hearing so that Souliotes, who is now almost seventy years old, and who has been incarcerated since 1997, may have an opportunity for meaningful review of his innocence claim." Id. at 1178 n.3.

Judge Zilly, sitting by designation, dissented. He found that the majority failed to provide meaningful guidance to the District Court. Id. at 1183. Further, he disagreed with relying on Lee to foreclose Petitioner's actual innocence gateway claim under Schlup. Specifically, he found that Petitioner presented a strong case for relief:

Souliotes v. Evans, 622 F.3d at 1187 (Zilly, J., dissenting).

On May 25, 2011, the Ninth Circuit issued a second order regarding Petitioner's claims. At that time, the Ninth Circuit had voted to rehear Lee v. Lampert en banc, a case with the potential to affect Petitioner's ability to present foreclosed claims in his petition. Regardless, the Ninth Circuit again ordered this Court to conduct an expedited evidentiary hearing regarding the newly discovered evidence. The Court stated:

Souliotes v. Evans, 434 Fed. Appx. 660 (9th Cir. 2011).

In light of the explicit instruction of the Ninth Circuit, the Court ordered an evidentiary hearing to take place on August 23, 2011 in order to determine if Petitioner was diligent in presenting the newly discovered evidence. (Min. Order, ECF No. 63.)

3. Impact of Intervening Ninth Circuit Authority — Lee v. Lampert (en banc)

On August 2, 2011, the Ninth Circuit revisited Lee v. Lampert, and in an en banc decision overturned the decision of the three judge panel. See Lee v. Lampert, 653 F.3d 929 (9th Cir. 2011) (en banc). The Ninth Circuit held "that a credible claim of actual innocence constitutes an equitable exception to AEDPA's limitations period, and a petitioner who makes such a showing may pass through the Schlup gateway and have his otherwise time-barred claims heard on the merits." Id. at 931-32. The Ninth Circuit joined the Sixth, Tenth, and Eleventh Circuit in recognizing an equitable exception to the statute of limitations. Id. at 932-33.

4. Instructions Upon Remand

Two weeks later, on August 17, 2011, the Ninth Circuit vacated its decision in Souliotes and issued the following order:

Souliotes v. Evans, 654 F.3d 902 (9th Cir. 2011).

In light of the August 17, 2011 order, the Court vacated the limited evidentiary hearing regarding diligence in discovering the new MPD evidence set for August 23, 2011. Instead, the Court scheduled an evidentiary hearing to hear Petitioner's actual innocence gateway claim under Schulp and his claims of statutory tolling based on diligence. The evidentiary hearing was set so as to promptly hear Petitioner's claims while providing the parties adequate time to conduct discovery regarding the issues to be addressed at the hearing.

A three day evidentiary hearing commenced on January 24, 2012. Both parties presented witnesses in support of their respective positions. After the hearing the Court provided the parties an opportunity to file post-hearing briefing. Petitioner and Respondent filed post-hearing briefs and Petitioner filed a reply to Respondent's brief. (See ECF Nos. 138-40.) The matter then stood ready for adjudication.

IV. ISSUES

As described, Petitioner's federal petition was filed five days late. Unless Petitioner can show that he is entitled to statutory tolling or that he is otherwise equitably excepted from the statute of limitations set forth by AEDPA, his claims must be denied as time barred. Accordingly, the Court held an evidentiary hearing to determine the following two issues:

V. STATEMENT OF RELEVANT LAW

A. Actual Innocence

The United States Supreme Court has determined that a petitioner whose petition for habeas corpus is procedurally barred may still obtain consideration of the merits of his petition if he can establish his "actual innocence." Specifically, the Supreme Court, in Schlup v. Delo, established an equitable exception to allow review of procedurally defaulted "constitutional claims only if [petitioner] falls within the narrow class of cases . . . implicating a fundamental miscarriage of justice." 513 U.S. 298, 314-315 (citations omitted.) A showing of actual innocence under Schlup is "not itself a constitutional claim, but instead a gateway through which a habeas petitioner must pass to have his otherwise barred constitutional claim considered on the merits." 513 U.S. at 315 (citing Herrera, 506 U.S. at 404).

This exception, equitable in nature, was created in light of the "concern about the injustice that results from the conviction of an innocent person." Schlup, 513 U.S. at 325. "That concern is reflected, for example, in the `fundamental value determination of our society that it is far worse to convict an innocent man than to let a guilty man go free.'" Id. (citing In re Winship, 397 U.S. 358, 372 (1970) (Harlan, J., concurring)).

"To ensure that the fundamental miscarriage of justice exception would remain `rare' and would only be applied in the `extraordinary case,' while at the same time ensuring that the exception would extend relief to those who were truly deserving," the Supreme Court explicitly limited the equitable exception to cases where a petitioner has made a showing of innocence. Schlup, 513 U.S. at 321. "The Supreme Court did not hold that a petitioner may invoke Schlup whenever he wants a trial do-over." Lee v. Lampert, 653 F.3d at 946 (Kozinski, J., concurring).

The Supreme Court has yet to determine whether the Schulp actual innocence gateway applies to other procedural bars. However the Ninth Circuit in Lee v. Lampert, joined three other circuit courts in holding that the equitable exception also applies to claims barred by AEDPA's statue of limitations. 653 F.3d at 932 ("We hold that a credible claim of actual innocence constitutes an equitable exception to AEDPA's limitations period, and a petitioner who makes such a showing may pass through the Schlup gateway and have his otherwise time-barred claims heard on the merits."). The Ninth Circuit found "[a]bsent evidence of Congress's contrary intent, there is no articulable reason for treating habeas claims barred by the federal statute of limitations differently" than claims involving procedural default. Id. (citing Souter v. Jones, 395 F.3d 577, 599 (6th Cir. 2005)).

It is presumed "that Congress knew of the exception when it drafted AEDPA, and absent the clearest command, we will not construe the statute to displace that equitable authority." Lee, 653 F.3d at 934 (citing Holland v. Florida, 130 S.Ct. 2549, 2560-61 (2010)). Strong equitable concerns outweigh other interests in allowing claims based on showings of innocence to proceed:

As the [Supreme] Court warned in Holland:

Lee, 653 F.3d at 934-935.

Accordingly, if Petitioner can make a showing of actual innocence, that showing may implicate a fundamental miscarriage of justice entitling Petitioner to pass through the Schulp gateway and serve as an equitable exception to the AEDPA statute of limitations. In allowing the equitable exception, the courts have provided guidance regarding the relevant criteria for making a sufficient showing of "actual innocence."

1. Credible New Evidence

The Supreme Court has required actual innocence gateway claims to be supported by credible "new reliable evidence . . . not presented at trial." Schlup, 513 U.S. at 324. The Supreme Court enumerated three specific categories of new reliable evidence: "exculpatory scientific evidence, trustworthy eyewitness accounts, or critical physical evidence." Id.

Assessing the reliability of new evidence is a typical fact finding role, requiring credibility determinations and a weighing of the probative force of the new evidence in light of the evidence of guilt adduced at trial. House, 547 U.S. at 560 (Roberts, C.J., concurring in part and dissenting in part). The district court must act as a fact finder and determine if the new evidence presented is reliable. Id. at 557 ("As House presented his new evidence, and as the State rebutted it, the District Court observed the witnesses' demeanor, examined physical evidence, and made findings about whether House's new evidence was in fact reliable. This fact finding role is familiar to a district court. `The trial judge's major role is the determination of fact, and with experience in fulfilling that role comes expertise.'"); Anderson v. Bessemer City, 470 U.S. 564, 574 (1985)).

Here, Petitioner has presented exculpatory scientific evidence that the MPDs on his shoes and from the fire scene are distinguishable. Although Respondent asserts that Petitioner was not diligent in presenting the scientific evidence, he concedes that the new scientific evidence regarding MPDs is correct. (See Undisputed Fact 13.) Having presented the Court with reliable exculpatory scientific evidence, Petitioner was granted a right to hold an evidentiary hearing and present his claim of actual innocence to the Court.

2. Scope of Evidence to be Considered

While Petitioner must present credible new evidence to pass through the Schulp actual innocence gateway, he is not limited to presenting only the new evidence in making a showing of innocence. With respect to an actual innocence gateway inquiry "Schlup makes plain that the habeas court must consider all the evidence, old and new, incriminating and exculpatory, without regard to whether it would necessarily be admitted under rules of admissibility that would govern at trial." House v. Bell, 547 U.S. at 537-538 (citing Schlup, 513 U.S. at 327-328).

3. Petitioner's Burden of Proof

As the actual innocence gateway under Schlup is limited to extraordinary cases, a petitioner must overcome a stringent standard of proof to qualify for the equitable exception. "When a petitioner has been `tried before a jury of his peers, with the full panoply of protections that our Constitution affords criminal defendants,' it is appropriate to apply an `extraordinarily high' standard of review." Schlup, 513 U.S. at 315-316 (citing Herrera, 506 U.S. at 419, 426 (O'Connor, J., concurring). However, when a petitioner accompanies his claim of innocence with an assertion of constitutional error at trial, his conviction may not be entitled to the same degree of respect as one that is the product of an error-free trial. Id.

If a Petitioner "presents evidence of innocence so strong that a court cannot have confidence in the outcome of the trial unless the court is also satisfied that the trial was free of nonharmless constitutional error, the petitioner should be allowed to pass through the gateway and argue the merits of his underlying claims." Schlup, 513 U.S. at 316. "[T]he petitioner must show that it is more likely than not that no reasonable juror would have convicted him in the light of the new evidence." Id. at 327. In other words, a petitioner must show "that more likely than not any reasonable juror would have reasonable doubt." House, 547 U.S. at 538.

The Supreme Court noted that "[t]he word `reasonable' in that formulation is not without meaning." Schlup, 513 U.S. at 329. "It must be presumed that a reasonable juror would consider fairly all of the evidence presented. It must also be presumed that such a juror would conscientiously obey the instructions of the trial court requiring proof beyond a reasonable doubt." Id.

"Because a Schlup claim involves evidence the trial jury did not have before it, the inquiry requires the federal court to assess how reasonable jurors would react to the overall, newly supplemented record."8 House, 547 U.S. at 538 (citing Schlup, 513 U.S. at 330). "If new evidence so requires, this may include consideration of the credibility of the witnesses presented at trial." Id.

House, 547 U.S. at 538 (internal quotation marks omitted) (citing Schlup, 513 U.S. at 329); Lee, 653 F.3d at 938.

"[T]he Schlup inquiry . . . requires a holistic judgment about `all the evidence,' and its likely effect on reasonable jurors applying the reasonable-doubt standard." House, 547 U.S. at 539-40 (citing Schlup, 513 U.S. at 328-29). "As a general rule, the inquiry does not turn on discrete findings regarding disputed points of fact, and `it is not the district court's independent judgment as to whether reasonable doubt exists that the standard addresses.'" Id. "Predicting what reasonable jurors would do is not easy. It's much harder than combing the record and picking at the state's case. But it's what Schlup demands." Carriger v. Stewart, 132 F.3d 463, 486 (9th Cir. 1997) (Kozinski, J., dissenting).

4. Diligence

Neither the Supreme Court nor the Ninth Circuit have held that a showing of diligence is needed to pass through the actual innocence gateway. See Lee, 653 F.3d at 935 n.9 ("Because this case does not present the question, we need not—and do not—decide what diligence, if any, a petitioner must demonstrate in order to qualify for the actual innocence exception recognized in this opinion.").

Respondent asserts that the majority of circuits have concluded that an actual innocence gateway claim requires a showing of diligence. (Resp. Post Hr'g Brief at 19.) This assertion overstates the holdings of the federal circuit courts.

First, it is necessary that a court must determine that the actual innocence gateway applies to the statute of limitations created by AEDPA. Only then does the issue of diligence arise.

At least three federal circuits do not recognize an actual innocence exception under any circumstances. Lee, 653 F.3d at 933 n.6 (noting that in addition to the Ninth Circuit, three other circuits have found such an exception, and three other circuits have not recognized an actual innocence exception under any circumstances.)

The Seventh and Eighth Circuits have held that a petitioner must be diligent in presenting an actual innocence claim. See Escamilla v. Jungwirth, 426 F.3d 868, 872 (7th Cir. 2005) ("Section 2244(d) has a rule for when new factual discoveries provide a fresh period for litigation; unless that standard is met, a contention that the new discoveries add up to actual innocence is unavailing. Prisoners claiming to be innocent, like those contending that other events spoil the conviction, must meet the statutory requirement of timely action."); Flanders v. Graves, 299 F.3d 974, 978 (8th Cir. 2002). In Flanders, the Eighth circuit held that the limitations period was statutorily created and "[i]t is not our place to engraft an additional judge-made exception onto congressional language that is clear on its face." 299 F.3d at 977. However, the court reasoned that equitable tolling has been allowed in cases where a petitioner, though no fault of his own, is delayed in filing. To better conform the actual innocence gateway within the equitable tolling doctrine, the court imposed a diligence requirement. Id. at 978. ("We do not hold that actual innocence can never be relevant to a claim that the habeas statute of limitations should be equitably tolled. For such a claim to be viable, though, a petitioner would have to show some action or inaction on the part of the respondent that prevented him from discovering the relevant facts in a timely fashion, or, at the very least, that a reasonably diligent petitioner could not have discovered these facts in time to file a petition within the period of limitations.")

The Second Circuit has expressly not decided whether diligence is required. Whitley v. Senkowski, 317 F.3d 223, 225-226 (2d Cir. 2003). Instead the court remanded the issue to the district court with instructions to determine if diligence is required, and if so, whether the petitioner was diligent. Id. (remanding to district court to determine "[i]f [petitioner] did not pursue the claim with reasonable diligence, must an actual innocence claim be pursued with reasonable diligence in order to raise the issue of whether the United States Constitution requires an `actual innocence' exception to the AEDPA statute of limitations?").

Both the Sixth and Tenth Circuit have found that no diligence is required to present an actual innocence gateway claim under Schlup. See Souter v. Jones, 395 F.3d 577, 600-601 (6th Cir. 2005); Perkins v. McQuiggin, 2012 U.S. App. LEXIS 4160 (6th Cir. Mar. 1, 2012) (affirming the holding of Souter); Lopez v. Trani, 628 F.3d 1228, 1230-1231 (10th Cir. 2010).

In Souter, the Sixth Circuit explained that the presumption is that Congress enacts statutes against the background of existing jurisprudence unless it specifically negates that jurisprudence. Souter, 395 F.3d at 598. When Congress enacted AEDPA it was aware of judicially created equitable exceptions to the statute of limitations which survived its enactment. Id. Congress was also aware it was "working against the jurisprudential background of Murray v. Carrier and Schlup v. Delo in which the Supreme Court held that a showing of actual innocence, defined as more likely than not that no reasonable juror would vote to convict, was sufficient to overcome a procedurally-defaulted habeas petition." Id.

Based on the Supreme Court law existing at the time of AEDPA, the Sixth Circuit held that the miscarriage of justice exception should apply to the newly created federal time limitation:

Souter, 395 F.3d at 599. Additionally, the Sixth Circuit declined to apply the diligence requirement imposed by the Eighth Circuit in Flanders. Souter, 395 F.3d at 601 n.16. In light of "the grave constitutional concerns which are raised by the incarceration of one who is actually innocent, [the Sixth Circuit] decline[d] to impose additional requirements upon a petitioner beyond those which the Supreme Court has set forth in its habeas corpus jurisprudence." Id.

In Lopez, the Tenth Circuit differentiated between the actual innocence gateway claim which falls under the fundamental miscarriage of justice exception, and other reasons for equitable tolling that provide an excusable cause for the failure to timely file. 628 F.3d at 1230. Based on the different rationale for equitable tolling and the creation of an equitable exception for actual innocence gateway claims, the court held that diligence is not required for the latter. Id. at 1230-31. ("We thus reject the reading of our precedent that would require a habeas petitioner seeking equitable tolling on actual innocence grounds to demonstrate that he diligently pursued his actual innocence claim. . . . Where . . . a petitioner argues that he is entitled to equitable tolling because he is actually innocent, this argument is premised on the same fundamental miscarriage exception that was discussed by the Supreme Court in Schlup and Coleman, and as such the petitioner need make no showing of cause for the delay.")

Finally, the Eleventh Circuit, without reaching whether there is an actual innocence exception to the AEDPA statute of limitations, has distinguished the actual innocence exception based on the Suspension Clause from claims of equitable tolling and its diligence requirement. See Rozzelle v. Fla. Dep't of Corr., 2012 U.S. App. LEXIS 4114, n.14 (11th Cir. Feb. 29, 2012) ("This Court has generally framed the AEDPA-actual-innocence question not as an equitable tolling issue, which has a due diligence component, but as whether the Suspension Clause necessitates a constitutional exception to AEDPA's one-year time-bar where a habeas petitioner has shown actual innocence.").

While the Ninth Circuit in Lee did not address whether a diligence requirement was imposed on an actual innocence gateway claim, its underlying reasoning for finding that the equitable exception applies to the statute of limitations of AEDPA provides guidance in making the present determination. First, in Lee, the Ninth Circuit was careful to differentiate actual innocence claims from equitable tolling claims. Lee, 653 F.3d at 933 n.5 ("We note that, in many cases, the phrase `equitable tolling' is used in describing the use of equitable power to allow the untimely filing of a habeas petition in an actual innocence case. The more accurate characterization is `equitable exception,' because equitable tolling involves different theoretical underpinnings.") Relying on the reasoning set forth by the Sixth Circuit in Souter, the Ninth Circuit found that Congress enacted AEDPA in light of the existing actual innocence exception, and that the doctrine survives the enactment:

Id. at 934.

The Ninth Circuit relied heavily on the Supreme Court's precedent to explain the importance of maintaining equitable claims, especially ones based on actual innocence to allow habeas claims to proceed. See Holland v. Florida, 130 S.Ct. 2549, 2562 (2010) ("The importance of the Great Writ, the only writ explicitly protected by the Constitution, . . . along with congressional efforts to harmonize the new statute with prior law, counsels hesitancy before interpreting AEDPA's statutory silence as indicating a congressional intent to close courthouse doors that a strong equitable claim would ordinarily keep open."); Lee, 653 F.3d at 935-936 ("An actual innocence exception to the limitations provisions does not foster abuse or delay, but instead recognizes that in extraordinary cases, the societal interests of finality, comity, and conserving judicial resources `must yield to the imperative of correcting a fundamentally unjust incarceration.'") (citing Murray v. Carrier, 477 U.S. 478, 495 (1986)); Schlup, 513 U.S. at 324 ("[T]he fundamental miscarriage of justice exception seeks to balance the societal interests in finality, comity, and conservation of scarce judicial resources with the individual interest in justice that arises in the extraordinary case.").

Based on the Ninth Circuit's clear distinction between equitable tolling and the actual innocence gateway and its strong respect for the fundamental miscarriage of justice exception, this Court adopts the reasoning of Souter and will not require a statutorily created showing of diligence such as in Flanders. See Lee, 653 F.3d at 933 ("[E]quitable principles have traditionally governed the substantive law of habeas corpus, and federal courts will not construe a statute to displace courts' traditional equitable authority absent the clearest command.") (citation omitted.) In light of the extraordinary showing required for the actual innocence gateway, there is little concern that the exception will be abused or greatly affects the functioning of the statutory scheme enacted under AEDPA. Souter, 395 F.3d at 600-601. ("Because one must meet a significantly greater burden to pass through the gateway, no petitioner would forego filing within the one-year period under § 2244(d)(1)(D) if possible. The actual innocence exception would be limited to the rare and extraordinary case where a petitioner can demonstrate a credible claim of actual innocence and the one-year limitations window has closed. Therefore, an actual innocence exception would not be inconsistent with the AEDPA limitations provision itself.")

Even though Petitioner need not make a showing of diligence with regard to his actual innocence gateway claim, he also presents a substantive actual innocence claim. If there is statutory tolling of that latter claim, it may be considered timely filed. Mardesich v. Cate, 668 F.3d 1164 (9th Cir. 2012); Souliotes, 622 F.3d at 1181. The analysis involves the same facts and showing of diligence required under 28 U.S.C. 2244(d)(1)(D) and Flanders. Accordingly, even though this Court finds Petitioner need not show he was diligent in presenting his actual innocence gateway claim, he may in fact do so when presenting evidence of statutory tolling of his substantive actual innocence claim based on the discovery of new MPD evidence.

B. Statutory Tolling

1. Newly Discovered Evidence

Petitioner's federal petition was filed five days late even after he was given the benefit of statutory tolling under 28 U.S.C. § 2244(d)(2) for the period in which he was properly seeking state post-conviction relief. However, with regard to Petitioner's substantive actual innocence claim under Herrera v. Collins, 506 U.S. at 417, Petitioner asserts tolling under § 2244(d)(1)(D) based on his discovery of new MPD evidence.

Section 2244(d)(1) provides:

2. Reasonable Diligence in Discovering Evidence

"Section 2244(d)(1)(D) does not demand the maximum diligence possible, but only `due' or `reasonable' diligence." Souliotes, 622 F.3d at 1178 (citing Starns v. Andrews, 524 F.3d 612, 618-19 (5th Cir. 2008); Wilson v. Beard, 426 F.3d 653, 660-62 (3d Cir. 2005); Moore v. Knight, 368 F.3d 936, 939-40 (7th Cir. 2004); Wims v. United States, 225 F.3d 186, 190 n.4 (2d Cir. 2000)); see also Holland v. Florida, 130 S.Ct. 2549, 2565 (2010); Mathis v. Thaler, 616 F.3d 461, 474 (5th Cir. 2010); DiCenzi v. Rose, 452 F.3d 465, 470 (6th Cir. 2006).

Specifically, the Ninth Circuit stated with respect to Petitioner's case:

Souliotes, 622 F.3d at 1178 (emphasis in original).

The Ninth Circuit also explained that "§ 2244(d)(1)(D)'s due diligence requirement is an objective standard that considers the petitioner's specific situation." Id. "[A] due diligence inquiry should take into account that prisoners are limited by their physical confinement." See Moore, 368 F.3d at 940. "The proper focus of the inquiry is therefore when Souliotes himself would have learned of the new evidence had he exercised reasonable care." Souliotes, 622 F.3d at 1178.

Should Petitioner make the required showing of diligence, he would only be entitled to present his claim of substantive actual innocence. "[Section] 2244(d)(1) requires consideration of the appropriate triggering date for each claim presented in the application." Id. at 1180; Mardesich, 668 F.3d 1164. "Accordingly, here § 2244(d)(1)(D) provides the triggering date for Souliotes's substantive actual innocence claim, which is based on newly discovered evidence, but § 2244(d)(1)(A) cabins the limitations period as to Souliotes's other habeas claims." Souliotes, 622 F.3d at 1180.

VI. ANALYSIS OF ACTUAL INNOCENCE CLAIM

A. Criteria

As the Supreme Court has explained, Petitioner "must show that it is more likely than not that no reasonable juror would have convicted him in the light of the new evidence." Schlup, 513 U.S. at 327. The new evidence here is exculpatory scientific evidence that the MPDs found on Petitioner's shoes and at the fire scene are distinguishable and provide no evidence that Petitioner was physically present at the fire scene. Petitioner also challenges other evidence used to convict him, most notably, the findings of the fire investigators and the reliability and credibility of the eyewitness Monica Sandoval and of the other circumstantial evidence.

This Court must review all the evidence implicating Petitioner's guilt, not just the new evidence, to determine if "no reasonable juror would have convicted him." Id. "Schlup makes plain that the habeas court must consider all the evidence, old and new, incriminating and exculpatory, without regard to whether it would necessarily be admitted under rules of admissibility that would govern at trial." House, 547 U.S. at 537-538 (citing Schlup, 513 U.S. at 327-328). Accordingly, Petitioner's challenge to the conclusions of the original fire investigators and his renewed questioning of the reliability of the eyewitness and circumstantial evidence is relevant to the present inquiry. "Based on this total record, the court must make a `probabilistic determination about what reasonable, properly instructed jurors would do.'" House, 547 U.S. at 538 (citing Schlup, 513 U.S. at 329).

Respondent asserts, correctly, that Petitioner made similar challenges to much of the evidence at trial. (Resp. Post-Hr'g Brief at 17.) However, Petitioner's challenges to that evidence at trial do not preclude him from asking the Court to review it again in light of the new exculpatory evidence. Petitioner's conviction informs this Court only that the evidence presented at trial satisfied the jury Petitioner was guilty beyond a reasonable doubt; it does not reflect what weight the jury gave to the various pieces of evidence the prosecution presented or suggest that the jury found all of that evidence reliable and probative of Petitioner's guilt. Now with the probative value of key pieces of that evidence having been stipulated away, the Court must reexamine and evaluate through the eyes of a reasonable juror that which remains. Though Petitioner has presented new evidence only regarding MPDs and fire science techniques, this Court must examine all relevant evidence and make a probabilistic determination regarding its combined effect on a reasonable jury. Accordingly, the Court shall discuss the evidence by category, in detail, below.

B. Evidence of Arson

1. Trial Testimony Regarding Cause and Origin of Fire

At trial, fire investigators testified with certainty that the fire was intentionally started with ignitable liquids (also referred to as accelerants). Facts giving rise to that conclusion are outlined below:

In the very early morning of January 15, 1997, and within minutes of being called, the Modesto Fire Department arrived at the scene of the fire at 1319 Ronald Avenue. According to the first firemen and civilian witnesses on scene, flames were already emanating at floor level from the area of the automobile door of the garage. (Ex. J-XII, Resucher Rpt. at 1-2.) Upon entry into the house through the front doors, the firemen immediately observed very hot conditions, intense smoke, and extensive fire activity in the family room, kitchen, and garage. (Id.)

The damage to the structure was severe. Examination of the house by an investigator within a half hour or so after the fire department arrived found that fire had completely gutted the garage, melted the garage door, and caused the garage roof to collapse. (Reuscher Rpt. at 7.) Charring on a tree approximately fifteen feet from the garage indicated intense radiant heat had emanated from the garage door. (Id.) A redwood deck showed surface burning twelve feet from a melted sliding glass door on the rear of the house. (Id.) The roof over the family room had burned through and collapsed. (Id.)

At trial the prosecution presented testimony of two fire investigators, Thomas Reuscher and Robert Evers. Reuscher was a retired captain of the Modesto Fire Department with 24 years of service who had investigated hundreds of fires, perhaps over a thousand, to determine cause and origin. (RT 6538-39, 6558.) Evers was a retired Modesto Fire Department fire captain who had 33 years of service and had been the primary investigator for over 500 fires. (RT 8461-64; 8475.) During trial, these two arson investigators testified repeatedly and essentially without reservation that the nature of the damage to the floor of the house constituted clear evidence that an ignitable liquid had been poured on the floor and intentionally lit. (See, e.g., RT 6576:7-6577:3, 6597:13-6598:15, 6609:16-6610:8, 6613:4-6614:24, 6615:27-6616:14, 6633:1-6634:25, 6683:10-6689:9, 6693:16-6694:3, 6704:18-6705:7, 6710:3-7, 6721:4-6722:22, 6818:23-6819:2, 6823:7-18, 6825:15-20, 6941:10-6942:14, 8474:10-26, 8499:23-8502:28, 9247:6-25, 9050:7-21.) They identified marks on the floor as clear "pour patterns" showing precisely where an accelerant had been deposited. (Id.) They asserted that other damage, such as holes in the floorboards and curled floor tiles, also reliably indicated an ignitable liquid had been used. (Id.)

The jury was told that deep charring inside the residence and garage and witness accounts of "an extremely hot fire" reflected intense heat that could only occur in a fire started with an ignitable liquid. (See, e.g., RT 6574:14-27, 6576:7-6577:3, 6579:24-6580:24, 6586:1-5, 6709:21-6710:7, 6947:23-6948:4, 8501:24-8502:28.) The jury was told also that there were insufficient other sources of fuel in the house to sustain a fire of that intensity unless an ignitable liquid had been added. (See, e.g., RT 6576:7-6577:3, 6718:19-6719:23, 6764:12-27, 8483:17-8484:2.) The investigators also described using a hand-held "hydrocarbon detector" to search the fire scene for traces of ignitable liquid, and reported that the detector gave positive reactions which constituted further evidence that a liquid accelerant had been used. (See, e.g., RT 6574:19-6576:6, 6627:1-6, 6763:20-6766:28, 8485:25-8488:11.) Moreover, two of the many materials taken from the fire scene for laboratory testing tested positive for ignitable liquids — a piece of carpet foam and floorboard wood from the living room. (Undisputed Fact 11.) These opinions were subject to intense cross-examination, but at Petitioner's second trial no defense experts were proffered to counter or respond to the prosecution's experts.

As will be discussed herein, virtually all of the investigators' opinions have been shown to be unfounded or erroneous.

2. Advances in Fire Science

Traditional fire scene investigation has been described as a mixture of art and science. Carman, The State of Forensic Science: Science Trumps Art in Fire Investigation, 74 Tex. B. J. 587 (2011). The identification and interpretation of burn patterns was considered an art with little foundation in scientific principle. (Id.) Unlike other forensic science professionals who normally have graduate and post-graduate educations, fire investigators historically were fire fighters and law enforcement officers who had risen through the ranks, learned from predecessors who had done likewise, and had no formal scientific training.9 (Id. at 588.) Few investigators had any serious training in scientific methodology. Neither Reuscher nor Evans had a post-secondary degree in fire science or had studied thermodynamics, chemistry, or physics. (RT 6756-57, 8841.)

Recognizing the need for scientific validation of fire investigatory principles, the National Fire Protection Association's Technical Committee on Fire Investigations released a manual in 1992, NFPA 921, which provided guidance to investigators based on scientific principles. Nat'l Fire Protection Ass'n, NFPA 921: Guide for Fire & Explosion Investigations (1992). NFPA 921 explained that widespread beliefs regarding the infallibility of burn indicators were simply incorrect. Nevertheless, the fire investigation community as a whole has been slow to accept and adopt the new guidelines. Wolf, Habeas Relief From Bad Science: Does Federal Habeas Corpus Provide Relief for Prisoners Possibly Convicted on Misunderstood Fire Science?, 10(1) Minn. J. L. Sci. & Tech. 213, 218-19 (2009).

Thus, courts and others have been taking a more critical look at fire investigation techniques, and many courts have excluded expert witnesses who failed to follow guidelines set forth in NFPA 921. See United States v. Hebshie, 754 F.Supp.2d 89, 114-15 (D. Mass. 2010) (listing cases and explaining that, "While most of the above cases are civil, it cannot be that science is different in criminal cases than in civil ones. Bad science is bad science; unreliable methodologies are unreliable methodologies, no matter the side of the docket."). Other courts have permitted experts who did not follow the methodology set forth in NFPA 921 to testify, but allowed the failure to follow the guidelines to be considered in evaluating the weight of the evidence. Schlesinger v. United States, 2012 U.S. Dist. LEXIS 15754 (E.D.N.Y. Feb. 6, 2012).

a. Stipulations Regarding Fire Science, MPDs and Hydrocarbon Detectors

To his credit, Respondent has acknowledged and accepted these scientific advancements in arson investigation and agreed with Petitioner that the fire investigators' trial conclusions can no longer be considered correct. Specifically, the parties stipulated as follows in connection with the evidentiary hearing:

b. Two Points of Origin

(1) Testimony of Carman

At the evidentiary hearing, Petitioner presented Steven Carman, a former special agent and certified fire investigator for the United States Bureau of Alcohol, Tobacco, Firearms & Explosives with over twenty years of experience investigating approximately 550 fires.10 (HT at 48-49.) Carman, having reviewed the reports of the original fire investigators, opined that the investigators had relied on indicators of arson that have proven to be unreliable.11 (HT at 52.) When asked at what point it became widely known that such indicators were not accurate, Carman explained that fire scientists and engineers came to that conclusion before "on-the-ground fire investigators" (i.e., investigators who rose through the ranks to become investigators without formal scientific training.):

(HT at 52-53.)

Because the fire investigators in this case concluded very early that the evidence strongly supported a finding of arson, Carman believed that accidental causes were only "fleetingly considered" and "not investigated completely." (HT at 54.) For example, Carman explained, it was common practice at the time to have the investigator document why he had eliminated appliances as possible accidental causes of a fire. (HT at 75.) He found no documentation in this case explaining why the washer or dryer were eliminated as potential causes. (Id.) He also explained that mere visual inspection of electrical outlets for obvious damage (as was done here) was an inadequate basis for eliminating an electrical cause of fire. (HT at 76-77.) Finally, Carman testified that while he could not definitively rule out multiple points of origin of this fire, he thought it unlikely. (HT at 69.) He believed it more likely, based on the evidence, that there was one origin and the fire burned for "a sufficiently long period of time and aggressively enough that it ended up in this situation." (Id.)

(2) Testimony of DeHaan

Respondent presented John DeHaan, a criminalist with over 40 years of experience with the Alameda County Sheriff's Department Criminalistics Lab, the California Department of Justice, and Bureau of Alcohol, Tobacco, and Firearms. (HT at 425.) DeHaan agreed that, given the severe damage to the structure and limited investigation documentation, the cause and origin of this fire could not now be determined.12 (HT at 453.) Nevertheless, believing it would have been difficult for the fire to spread across the firewall separating the living room from the garage, DeHaan hypothesized that the fire may have started at multiple points (while conceding that with enough time, the same type of damage could have resulted from a single point of origin). (HT at 455-61.)

Based the original inspectors' investigation, DeHaan did not think that the stove was a possible accidental cause of the fire. He also thought that the insurance company investigator's evaluation of the water heater and electrical outlets satisfactorily eliminated them as sources of ignition. (HT at 467-68, 470.)

On cross-examination, DeHaan acknowledged having found arson based on multiple points of origin in other cases only to later revise his opinion or see arson charges dropped because of a lack of evidence. (HT 495, 500-501.) He also acknowledged that the ethics committee of the California Association of Criminalists had concluded that he had acted unethically in connection with his failure voluntarily and promptly to disclose that his multiple points of origin theory had proven unsupportable in one case.13 (HT 503-04.)

(3) Testimony of Schilling

At the evidentiary hearing, Respondent also presented testimony from David Shilling who investigated the scene of the fire on behalf of Petitioner's insurance carrier. Shilling had been a Fremont city firefighter for 25 years, the last five years of which he had worked as a fire investigator. (HT 397-98.) After retiring from the fire department and continuing until about 2001, he was employed by a private fire investigation company. (Id.) He attended one seminar on NFPA 921, but its guidelines were just being adopted at about the time of his retirement, and so he never incorporated them into his training. (HT at 402.)

Shilling described his investigation of the fire scene. Upon arriving, he met a police detective who informed him that the owner of the house had been arrested for murder and arson. (HT 420-21.) He then entered the house through the garage where he examined distinct patterns on the floor and checked electrical outlets. (HT 408-09.) Shilling concluded that the pour patterns in the garage were indications of the presence of a flammable liquid. (Schilling Rpt., Ex. R-H. at 5.) ("This fire appears to be incendiary, based on the burn patterns which were located on the garage floor and the fact that the burns were fresh and left tail sooting marks around them, which is indicative of a flammable liquid, freshly poured.") He also observed distinct patterns leading into the kitchen and in the living room and a large burn hole he found suspicious. (HT 414-15.) As he entered each room, he visually inspected electrical outlets and other sources and did not see any accidental causes of fire. (HT 410-411.)

Shilling testified that he examined the refrigerator, washer, dryer, water heater and forced air heater from the garage as they lay in a large pile of debris the fire department had removed from the house. (HT at 421-22; Shilling Rpt. at 4, 14, Photo 6.) While Shilling specifically remembered examining the water heater, he could not similarly recall if he had examined the washer and dryer. (HT 413.) He explained that it was his usual practice to include in his report photographs of each appliance he examined. (HT 417.) Other than the photograph of the debris pile (in which appliances, if identifiable at all, are either partially or completely obscured by other rubble) and a photograph of a wall heater in the house, there are no pictures of appliances in his report.

C. Shoes and MPDs

1. Facts Regarding MPD Evidence

At trial Sarah Yoshida, a Department of Justice criminalist, testified as to her examination of fire debris samples collected from the scene and Petitioner's shoes. (RT 8884-85.) She found that two samples from the fire scene (one from burned wood and another from burned foam and carpeting) tested positive for MPDs as did Petitioner's shoes. (RT 8860, 8864-65, 8874-78, 8883-84.)14

Yoshida acknowledged that scientific literature reflected that some shoes contained materials producing patterns identified as petroleum distillates. (RT 8886-87.) She also explained that the presence of an MPD did not necessarily indicate its use to start the fire. (RT 8888.) While admitting that she could not identify the particular substance detected in each sample and that the MPDs could be from any number of substances, Yoshida concluded that there was a correlation between the shoe and fire scene samples: all contained MPDs. (RT 8923-24.) She also pointed out that in her eight years of conducting arson analyses she had only very rarely found MPDs. (RT 8895.)

The government emphasized what it obviously believed was the highly probative value of the MPD evidence. In discussing other evidence, the government returned to the shoes as proof that its theory was correct. Thus, commenting on inconsistencies in eye witness testimony, the prosecutor argued:

(RT 9049, 9250.) In closing, the prosecution reminded the jury that MPDs had been found both on Petitioner's shoes and at the fire scene and argued:

(RT 9050.)

2. New Stipulated MPD Evidence

Respondent now concedes: "The MPD on Petitioner's shoes is chemically distinguishable from the MPD found on the carpet samples taken from the fire scene and the MPDs did not originate from a common source." (Undisputed Fact 13).

The government's contrary trial assertions notwithstanding, Petitioner's shoes tell no tale; they do not point "the finger of guilt" at Petitioner. There is no link between Petitioner's shoes and the fire. There is no MPD evidence placing Petitioner at the scene of the fire.

D. Monica Sandoval's Eyewitness Testimony

1. Introduction

The other direct evidence of Petitioner's participation in the act of arson was the testimony from Monica Sandoval. Sandoval described seeing a recreational vehicle or motorhome ("RV15") drive back and forth past the Ronald Avenue house ten to twenty times as she stood on her nearby apartment balcony on the early morning of the fire.16 On one occasion the driver stopped across from Petitioner's rental at 1319 Ronald, got out and walked to the rental carrying a white sack and then soon returned to the RV without the sack. Sandoval later identified Petitioner as the driver and Petitioner's RV as the one she had seen that morning.

2. Factual Summary

The following circumstances surrounding Monica Sandoval's identifications are effectively undisputed:

Sandoval, a then-nineteen year old woman, was standing on her second story apartment balcony at 1839 Tully Avenue17 almost continuously for several hours on a cold misty night18 angrily awaiting her boyfriend's overdue return home. (See Req. For Judicial Notice, Ex. G.) Sandoval testified that she observed the suspect vehicle drive by repeatedly during the period from 1:00 a.m. to about 3:30 a.m. During this time, she left the balcony only once, at 2:30 a.m., to put on warmer clothes and turn off her lights. (CT 90.) While on the balcony, she noticed the vehicle drive back and forth, east and west, along Ronald approximately ten to twenty times. At one point, the driver stopped across from 1319 Ronald, crossed the street carrying a white sack, and then returned to the vehicle a few minutes later without the sack. The vehicle then drove east past Sandoval's balcony and turned left, heading north at the intersection with Tully. A moment later it reappeared traveling south on Tully. As it came into the intersection with Ronald, the vehicle slowed and the driver turned his head, leaned forward and looked west down Ronald. At this time, Sandoval saw the bottom of the driver's face, from his eyeglasses down, through the windshield. Neither view of the driver was sufficient, she told investigators later that morning, to identify the driver.

A few minutes after the vehicle left the area, Sandoval noticed the fire at 1319 Ronald. At about 3:30 a.m., after fire and police personnel had arrived, she approached Officer Joseph Pimentel who was directing traffic near the scene and told him she had observed a "white or off-white RV" driving up and down Ronald Avenue. (RT 7167-69.)

She was interviewed by other officers later that morning. By then officials had learned Petitioner was the owner of the rental property, was trying to sell it, and that he owned an RV. The police drove Sandoval by Petitioner's residence to see if she could identify his RV as the one she had seen. It was parked on the side of his house and only the back of it was visible from the street. She could not identify it. The police then took Sandoval directly to an RV sales lot (one with several Winnebago RVs) to see if she could find any that looked like the one she had seen. None of the vehicles on the lot reminded her of the one which had driven up and down Ronald. While there, the officer escorting Sandoval was told to bring her back to Petitioner's residence to look again at Petitioner's RV, now parked on the street where Petitioner had moved it at the request of the police. There were several police cars in the area. This time, upon seeing the side and front of the vehicle, Sandoval identified it as the one that had driven up and down Ronald before the fire.

Sandoval became emotional after making the identification, and the police took her home. Later that day, the police presented Sandoval with a photographic lineup of Petitioner and five other individuals. Sandoval could not find the driver of the vehicle in the lineup.

In the months following the fire, the case generated a significant amount of press.19 The news coverage included many articles, including a full page story on April 27, 1997, in the local newspaper, the Modesto Bee, regarding murders in the area and including a picture of Petitioner. (RT 210-11.) Sandoval saw articles regarding the fire in the paper. (RT 5956.)

Six months after the fire, Sandoval testified at Petitioner's preliminary hearing. The district attorney questioned Sandoval, but initially did not ask her to identify the RV driver. After he concluded his direct examination of her, a recess was taken. After the recess, during which the district attorney possibly had a conversation with Sandoval, the district attorney asked the Court permission to question Sandoval further.20 Sandoval retook the stand and, for the first time in the six months since the fire, identified Petitioner, the man sitting in the red jail jumpsuit at the table with defense counsel, as the driver of the suspicious RV. (RT 5960-62.)

Given the importance of the identification testimony now that the fire science and MPD evidence has been removed from the equation, the Court shall undertake to outline in some detail the various issues which have been raised regarding Sandoval's identification of Petitioner and his RV. The Court shall summarize expert testimony presented at the evidentiary hearing regarding witnesses' ability to observe, retain, and recall evidence accurately in general and in circumstances such as Sandoval experienced. The Court also shall review expert testimony regarding investigatory practices in this case which, Petitioner claims, may have influenced the identification.

3. Facts Potentially Influencing Sandoval's Identification of Petitioner

a. Introduction

The above scenario raises factors that, according to Petitioner, call into question Sandoval's ability to observe Petitioner. She testified she saw only part of his face, for a brief period of time, through a windshield, from a significant distance, in the dark. She was particularly upset and distracted at the time. On the day of the fire, Sandoval said she would not be able to identify the driver. She, in fact, was unable to identify Petitioner in a photographic lineup. Six months later she had developed the ability to do so. She could not account for the change.

b. Distracted Mental State

At trial, Sandoval explained that as she waited on her balcony that morning for her boyfriend to come home, she was quite upset that he was late. She later described herself variously as: "angry," "agitated," "mad," "pretty mad," and "seething." (RT 5857, 5892.) She said her boyfriend was going to be in "hot water" and she planned on yelling at him when he came home. (RT 1517, 5857, 5897.) Sandoval had a "bad attitude" and temper which apparently accounted for her having stabbed her boyfriend with a switchblade during an earlier similar occasion on which he returned home in the middle of the night. (RT 1518-19, 5892-93.)21 Sandoval acknowledged that she was preoccupied with waiting for her boyfriend and was not paying much attention to the RV for much of the time it was driving around. (RT 5861.)

c. Limited View of the Driver

Significant trial testimony focused on Sandoval's opportunity to observe the driver's face.

Sandoval acknowledged she could not see the face of the RV driver when he walked from the vehicle to 1319 Ronald or on any of the multiple occasions on which he drove past the front of her balcony. (RT 5907.) Her only asserted opportunity to see his face was when the vehicle traveled south on Tully into its intersection with Ronald and the driver slowed and leaned to look down Ronald. (RT 1497-98, 5868.) At that time, Sandoval partially observed the driver's face from her vantage point over 100 feet away. The driver's action "sort of gave him away . . . some of his facial features." (RT 5936.) While so testifying, Sandoval put her hand across her nose to illustrate that she was only able to see the bottom half of the driver's face:

(RT 5936-37.) The bottom of his face was visible only for a brief period of time.22 (RT 1540.) Though it was dark, a street light on the southwest side of Ronald helped illuminate the driver. (RT 5868.)

Sandoval told the police about the details of the driver's face at the time of the fire. They took notes and possibly recorded the interview. (CT 115.) No such notes or recording have been produced. Police reports prepared days after the fire reported only Sandoval stating that the driver wore glasses. (Exs. J-X, J-XIII.)

Sandoval described additional details, specifically the driver as having a narrow face and long pointy nose, for the first time at the preliminary hearing, six months later. (CT 66-67.) Despite her obscured view, Sandoval felt it had been sufficient to enable her to make an identification at the preliminary hearing.23 (RT 5867.)

d. Estimation of the Driver's Age

According to police reports, Sandoval described the driver as being in his 30's or 40's. (Petitioner was 56 at the time of the fire.)

At the preliminary hearing with Petitioner sitting in front of her, Sandoval revised her age estimate upward and said the driver was in his 50's. (CT 67, 122.) She did not remember having stated initially that the driver was in his 30's, but did acknowledge, somewhat reluctantly, that possibility.24 (RT 5914-15.) Sandoval also admitted having difficulty determining age. (RT 5962-63.) Her responses to trial cross-examination were not reassuring:

(RT 5962-63.)

Investigators who had interviewed Sandoval after the incident were questioned at trial regarding her description of the driver's age. According to Captain Reuscher's contemporaneous handwritten notes, Sandoval had said the driver was in his thirties. (Joint Ex. X.) At trial, Reuscher remembered Sandoval stating that the driver was in his thirties "to forties." (RT 2197.) Officer Pimentel, who was present when Reuscher interviewed Sandoval, noted in his report that the driver was between the ages of 35-45. (Joint Ex. XI.) At trial, Pimentel stated that he did not remember Sandoval describing the driver as someone in his thirties. (RT 7182.) He testified that Sandoval described the individual only as an "older male." (RT 7182.)

e. Evolved Ability to Identify Petitioner

On the day of the fire, Sandoval stated that she had not had a sufficient view of the driver of the vehicle to identify him.25 (RT 5953.) She did not pick anybody out of the lineup because she could not tell if any of the men in it had been the driver of the vehicle. (RT 5954.) She found the individuals in the lineup photos to have similar features and looks. (Id.)

Although she could not identify Petitioner on the day of the fire, she did identify him at the preliminary hearing six months later. She was cross-examined at length about her improved ability to identify Petitioner:

(RT 5955-5962.)

f. Description of Driver Exiting the Vehicle

At the first trial, Sandoval testified that she had a clear view of the RV when it stopped across from 1319 Ronald Avenue. (RT 1493.) Nevertheless, she testified at the preliminary hearing, the distance was such that she could not tell if the driver was a man or woman. (CT 58.) Then at trial, she stated she felt that the person who exited the vehicle was most likely a man:

(RT 1493-94.)

Despite uncertainty as to the driver's sex, Sandoval described the person as wearing blue denim jeans, a blue and white checkered Pendleton-type shirt, and dark shoes. (RT 1494; 5915.) She could not describe the size of the checked pattern on the shirt: "I wouldn't know. As far as where, I was standing, there's no way of knowing." (RT 5916.) When pressed, she responded:

(RT 5916-17.)26

g. Video Reenactment

At the evidentiary hearing, Petitioner presented a witness, Carly Balletto, who had created a video designed to depict Sandoval's opportunity to identify a driver of Petitioner's RV27 as it traveled south on Tully. Balletto, while standing on what had been Sandoval's apartment balcony at a time when natural and artificial lighting was similar to that at the time of the fire,28 filmed the RV driving through the intersection of Tully and Ronald. In the video the RV slowed near the area where Sandoval said she had observed Petitioner's face through the windshield. (HT 239.) The driver of the RV then leaned forward and turned his head in the same manner as had been described by Sandoval. (HT 336-38.) Balletto testified that the video accurately depicted what she was able to see with regard to the face of the driver. (Id. at 259.) Despite her best efforts, Balletto was unable to observe the driver's facial features. (Id. at 249-50.)

On cross-examination, Balletto acknowledged that a nighttime photograph of the intersection portrayed a much brighter scene than that in the videotape reenactment. (HT 255; Pet'r's Exs. 1, 5.) However, as noted, she maintains that the video accurately reflected her view. (HT 259.)

h. Testimony of Dysart

Jennifer Dysart is an Associate Professor of Psychology at John Jay College of Criminal Justice of the City University of New York. She specializes in eyewitness identification research.29 Dysart testified regarding potential impediments to the accuracy of Sandoval's identification of Petitioner.30

Dysart acknowledged that, not having been present at the scene, she could not say with certainty whether Sandoval's identification was accurate or not. Regardless, she concluded, based on the consideration of factors research has shown to influence the reliability of memory, that Sandoval's identification of Petitioner presented "perhaps one of the most compelling cases I've ever worked on" and did raise "issues to be concerned with respect to reliability." (HT 131.) Based on the totality of circumstances surrounding Sandoval's identification of Pettioner, as described below, Dysart concluded the identification was "wholly unreliable." (HT 157-58.)

As will be seen, much of Dysart's testimony was consistent not only with what one would expect from common experience, but with California jury instructions which were given in Petitioner's case. Nevertheless, her testimony is enlightening in confirming, and to some extent quantifying, common sense principles with objective empirical study.

(1) Distance, Light, and Length of Exposure

Dysart testified that scientific studies have confirmed common perceptions about how distance, amount of light, and length of time to view a suspect strongly affect the reliability of an identification.

With regard to distance, studies have demonstrated what is intuitive, i.e., that faces become "coarse or fuzzy" as they move farther away from the witness. (HT 139.) Some researchers consider unreliable any identification made from beyond 15 meters or roughly 50 feet. (Id. at 140.) Dysart does not subscribe to such a rigid rule, but agrees that identification at longer distances negatively affect the ability to identify a face. (Id.)

Here, Sandoval observed the driver from over 300 feet when he exited the vehicle and from about 120 feet when she saw a portion of his face through his windshield as he drove by. Certainly these distances had a limiting impact on Sandoval's ability to observe the driver.

While not surprising that scientific studies have confirmed people have reduced visual acuity and color perception at night, they indicate too that darkness reduces the amount of information acquired and processed into memory. (Dysart Rpt. 9.) Sandoval observed the driver of the vehicle in the middle of the night, albeit with the benefit of some level of light from a nearby street light.31

According to Dysart, the length of time the suspect is exposed to the witness is also a significant factor in the reliability of identifications, perhaps the "most critical thing in an identification scenario." Dysart described a study in which some participants saw a perpetrator's face for 12 seconds and others for 45 seconds; a significant increase in identification errors were shown in the group with the lesser exposure time. (HT 145-46.) Here, Sandoval observed the driver's face for no more than a few seconds. According to Dysart this very brief exposure had a significant impact on reliability of the identification. (HT 170.)

(2) Effects of and Stress and Fear

Dysart also noted that stress, anxiety and fear negatively impact one's ability to process information and decrease the accuracy of information observed. (HT 139.) In her report, Dysart described a study of military personnel placed in a mock prisoner of war camp for twelve hours. One group of participants experienced a forty minute high-stress interrogation; the other experienced a low-stress interrogation. The high-stress interrogation group's accuracy at later identifying the interrogator was much reduced compared to the low-stress group. (Dysert Rpt. at 13-14.) Here, as noted, Sandoval was particularly upset and angry with her boyfriend at the time she saw the suspect and was "shooken up," "in shock" and "trembling" after witnessing the fire. (Dysert Rpt. at 13.)

(3) Failure to Identify Petitioner in Lineup

Empirical studies confirm that the failure to identify a person in a lineup is a very reliable indicator that the person observed is not actually present. (Dysart Rpt. at 15.) Studies show eyewitnesses give non-identification responses far more often when the suspect was absent from the lineup than when the target was present in the lineup. (Id.) Dysart opined to a significantly greater likelihood that Sandoval was telling the truth when she said she was unable to identify the driver on the day of the fire.

(4) Effect of Post-Event Information

Dysart testified that post-event information can strongly affect a witness' ability to accurately retain information. (HT 151-52.) Everyone, including eye witnesses, forget large amounts of information quickly after an event. Post-event information may be subconsciously transferred to fill in a witness' memory gaps. (Id.) People are very poor at remembering the context in which they observed something. (Id.) A witness who remembers seeing a picture of a suspect later, such as in a newspaper, may subconsciously begin to associate and blend the new information with the original viewing of the suspect. (HT 155.) Empirical evidence confirms witnesses are more likely to pick an individual they have previously viewed from a lineup. (Dysert Rpt. at 21.)

Here, Sandoval was presented with several opportunities to see Petitioner's face after the event. She saw his picture as one of six in the photographic lineup and there had a better and longer opportunity to observe Petitioner's face than at the scene. (HT 157-58.) The photograph provided Petitioner's complete face as opposed to her partial view of the driver's face at a distance in the middle of the night. Sandoval also saw Petitioner's photograph in the newspaper prior to her identification at the preliminary hearing. And, as discussed below, he was sitting in a red jail jumpsuit with defense counsel when she identified him at the preliminary hearing.

(5) Suggestive In-Court Identification

Dysart described the in-court identification as a highly suggestive show up procedure in which the witness is presented with only one person to identify and asked if that is the person he saw. (HT 157-58.) According to Dysart, such a procedure is much more likely to produce an identification even where the suspects are innocent. She referenced a 2006 study in which nearly one-third of positive identifications in show up procedures were false identifications of an innocent suspect. (Dysart Rpt. at 20.) Dysart also finds in-court show up procedures to be far more suggestive than out-of-court show ups. In the latter case, the witness is presented a potential suspect at a time, pre-investigation, when the witness understands the person may well not be the suspect. However, by the time of the preliminary hearing, the police and the government have begun prosecuting the suspect for the crime. That knowledge would have a suggestive impact on a witness' identification. (HT 158-59.)

Other information also raises questions about Sandoval's in-court identification of Pettioner: Sandoval changed her original statements regarding the age of the driver to better match Petitioner's age; she did not provide details regarding the features of Petitioner's face when originally questioned, but did so six months later; Sandoval admitted her memory was, as it should have been, best on the night of the fire and that she had difficulty remembering the facts of the incident six months before.32 (CT 155; RT 1523.)

(6) Repeated Identification After the Preliminary Hearing

Researchers have found that once a witness commits to a choice, he tends to continue to pick the same person in future identifications regardless of whether the identification was correct in the first place. (HT 168-69.) Such a "commitment effect is pervasive and strong." (Id. at 170.) In one study, 72% of eyewitnesses identified the same suspect in two identification procedures regardless as to whether the suspect was the actual perpetrator. (Dysert Rpt. at 22.) Based on the commitment effect, the continued identification of Petitioner by Sandoval may not be an indication of the reliability of the identification.

i. Fallibility of Eyewitness Identifications and Judicial Safeguards

Concerns with eye witness testimony as the basis for convicting innocent people has been acknowledged by the Supreme Court for some time. Perry v. New Hampshire, 132 S.Ct. 716, 728 (2012) ("We do not doubt either the importance or the fallibility of eyewitness identifications. . . . [W]e observed that the annals of criminal law are rife with instances of mistaken identification.") (internal quotations and citation omitted).

The concepts described by Dysart are not new; many confirm common sense principles. In California, criminal jury instructions instruct jurors of potential signs of unreliable testimony. See CALCRIM 315. The factors in the jury instruction are quite comparable to those raised by Dysart in her report and testimony. In determining the weight reasonable jurors would give Sandoval's identification testimony, the Court shall consider the effect of jury instructions as well as expert testimony on issues of unreliability.

The Supreme Court has recently reiterated the important role eyewitness identification plays in criminal cases despite its fallibility. Perry, 132 S. Ct. at 728. Moreover, scientific evidence suggests that eyewitness testimony, while potentially unreliable, can be very influential to the determinations of juries. "[D]espite its inherent unreliability, much eyewitness identification evidence has a powerful impact on juries. Juries seem most receptive to, and not inclined to discredit, testimony of a witness who states that he saw the defendant commit the crime." Watkins v. Sowders, 449 U.S. 341, 352 (1981) (Brennan, J., dissenting).

Id. (citations omitted).

The Supreme Court explains that reliability of witness identification is a significant factor in determining admissibility of the identification testimony. Manson v. Brathwaite, 432 U.S. 98, 114 (1977) ("[R]eliability is the linchpin in determining the admissibility of identification testimony for both pre- and post-Stovall confrontations. The factors to be considered are set out in Biggers. 409 U.S., at 199-200. These include the opportunity of the witness to view the criminal at the time of the crime, the witness' degree of attention, the accuracy of his prior description of the criminal, the level of certainty demonstrated at the confrontation, and the time between the crime and the confrontation. Against these factors is to be weighed the corrupting effect of the suggestive identification itself."); see also Perry, 132 S. Ct. at 733-34, n.3. (Sotomayor, J., dissenting). The Supreme Court was addressing reliability concerns in the case of suggestive law enforcement techniques as possible constitutional violations. Perry, 132 S. Ct. at 721, n.1. Regardless, reliability factors remain important considerations in determining the probative value of identification testimony. It can be argued that given current scientific knowledge, witness reliability should face greater scrutiny. As Justice Sotomayor recently explained:

Perry, 132 S. Ct. at 738-39 (Sotomayor, J., dissenting).

In order to protect against juries relying on eyewitness testimony of questionable reliability, criminal defendants are provided protections in the form of the right to confront the witness, assistance of effective counsel, and eyewitness-specific jury instructions that "warn the jury to take care in appraising identification evidence." Perry, 132 S. Ct. at 728-729.

Here, the Court is charged with determining in light of the present state of evidence if any reasonable juror would find Petitioner guilty beyond a reasonable doubt. Under Schlup, this Court must make a determination regarding what a reasonable jury would find in light of all of the evidence including expert witness testimony regarding the reliability of witness identifications and relevant instructions regarding reliability of eyewitness identifications. See People v. Wright, 45 Cal.3d 1126 (1988). Such instruction "should be given when requested in a case in which identification is a crucial issue and there is no substantial corroborative evidence." Id. at 1144. Presently, Judicial Council of California Criminal Jury Instruction No. 315 provides a jury a description of the factors that may influence the reliability of an eyewitness. It states, in relevant part:

In evaluating identification testimony, consider the following questions:

1-300 CALCRIM 315.

4. Identification of Winnebago

In addition to identifying Petitioner, Sandoval also identified Petitioner's motor home as the vehicle she had seen driving up and down and stopping in front of 1319 Ronald just before the fire. Sandoval's identification rested primarily on two features: the presence of a printed "W" on the vehicle and its slanted front end or windshield. Her descriptions of these features were relatively consistent. Her descriptions of many other features of the vehicle were not and in fact were erroneous.

The errors would not be so significant if Sandoval had consistently described an RV such as the one Petitioner owned. She apparently did not. According to the handwritten notes of the investigator who questioned Sandoval on the morning of the fire, she described the vehicle as a "Dodge RV" or "Dodge Caravan." Moreover, although she described many of the features of the back of the vehicle she had seen, she had been unable to identify the rear of Petitioner's RV on the morning of the fire. Many of the features she described did not exist on Petitioner's RV.

Sandoval ultimately testified that Petitioner's RV was the one she had seen on the day of the fire. However, Petitioner suggests that the credibility of her identification is undermined by demonstrated confusion on her part, inconsistencies within her statements and testimony, conflicting investigator notes about her comments, and her inability to accurately describe the vehicle. Those factors are discussed below.

b. Sandoval's Identification of the Winnebago

Sandoval had several opportunities to note the features of the recreational-type vehicle which repeatedly drove up and down in front of her balcony in the middle of the night. However, she reportedly did not pay much attention to it at first. It was only after she saw the vehicle stop across from 1319 Ronald and the driver exit the vehicle, go toward the house with a sack and then return without it, that she began to pay more attention.33 After seeing the vehicle drive around and then observing the driver looking down Ronald from Tully, she saw flames erupt from 1319 Ronald.

Some time later, around 3:00 a.m., Sandoval advised Modesto police officer Pimentel, who was directing traffic at the fire scene, that she had potentially relevant information. Once Captain Reuscher, the lead Modesto Fire Department investigator arrived about thirty minutes later, he and Pimentel interviewed Sandoval. Reuscher's handwritten notes from this interview were preserved. No other contemporaneous investigator notes have been produced.

Later, likely around 6:00 a.m., Modesto Police Officers Lee, Ridenour and Pimentel re-interviewed Sandoval. Sandoval and Lee remember that the interview was recorded electronically. No such recording has been produced.

During the interview, the officers attempted to sketch the shape of the vehicle Sandoval described. No such sketches have been produced.

Having learned that Petitioner owned the burned house, that the house had been for sale, and also that Petitioner owned an RV, the officers drove Sandoval by Petitioner's residence to see if his RV was the one she had seen. The RV was parked so that only the rear of the vehicle was visible from the street. Sandoval could not identify the vehicle.

The police then took her to look for similar RVs at a sales lot that had several Winnebegos with "W"s on them. None of those RVs reminded her of the one she had seen.

The police next took Sandoval back to Petitioner's house where his RV was parked on the street so she could see the front and side. She then identified it as the one she had seen before the fire.

c. The Record of Sandoval's Descriptions

Sandoval's descriptions of the vehicle were recorded in notes of the investigators, investigative reports prepared shortly after Petitioner was arrested and in testimony of Sandoval and fire and police investigators at the preliminary hearing and first and second trials. These sources reflect a consistent lack of consistency. They are here reviewed in chronological order of preparation:

(1) Reuscher's Notes

As noted, Sandoval was first interviewed on the morning of the fire by Captain Reuscher and Officer Pimentel. According to Reuscher's handwritten notes from the interview, Sandoval stated she had seen a "Beige RV Dodge Van type"; in the margin he wrote "Dodge Caravan."34 (Hr'g. Ex J-X.) There was no mention of the letter "W" or a "Winnebago." (Id.)

(2) Investigatory Reports

Later prepared and still existing police reports state that Sandoval described a "white or off-white RV" with a "W" on it. (See Exs. J-XI, XII, XIII.) Officer Buehler escorted Sandoval to the RV lot and was with her when she later identified Petitioner's RV. (Ex. J-XV.) His report states that Sandoval described the vehicle as having a slanted front end, but there is no mention of a letter "W." (Id.) He also reported that Sandoval identified Petitioner's RV based on its front end, not the "W."35 (Id.) No other features of the vehicle were prominently discussed in the police and fire reports.

(3) Sandoval's Preliminary Hearing Testimony

During the preliminary hearing, Sandoval refers to the vehicle as a Winnebago. (CT 60.) On direct examination, she stated that she identified the Winnebago, but had not provided a physical description of it. On cross-examination, she explained that she identified the vehicle based on the letter "W" and the front windshield. (CT 144.) Later, she added the back of the vehicle as reflecting distinctive features she used to make the identification. (CT 145.)

Sandoval testified that the vehicle had an eighteen inch blue "W" on the passenger side.36 (CT 103.) She did not remember a "W" on the driver side.37 (Id.) Sandoval knew the vehicle was a Winnebago from the "W" and believed she told the police it was a Winnebago. (CT 108.) She called the vehicle a motorhome or Winnebago; she would not have used the term "recreational vehicle" or "RV." (CT 109.)

Sandoval also remembered that the vehicle was beige or cream in color and had a stripe. (CT 104.) She somewhat hesitatingly, but correctly, described its two doors on the passenger side, but was not sure whether there were two doors on the driver side. (CT 101, 116.)

Sandoval could not describe, or inaccurately described, several features of the RV. She said the vehicle she had seen had a ladder, white license plates, no spare tire, nothing on top, no running lights, and a white curtain or drape in the back window. (CT 101; 104-07.) Petitioner's RV had none of these features. (See Ex. J-XX.)

When asked how she knew the vehicle parked in the street was the same vehicle that earlier had been parked in the driveway, she said she remembered how the front of the vehicle, which she had not seen originally, looked. (CT 145.) When defense counsel clarified the question and asked again, Sandoval was unable to formulate an answer. (CT 146.)

Sandoval stated she had not seen any RVs with a "W" at the sales lot; if she did, the "W" was smaller than the one on Petitioner's RV. (CT 140.)

(4) Sandoval's Testimony at the First Trial

Most of Sandoval's testimony at the first trial repeated the descriptions she had given at the preliminary hearing. She testified that she was not able to recognize the suspicious RV from the back, and so did not identify Petitioner's RV when only the back was visible. (RT 1506-1507.) Her identification of Petitioner's RV was based on the solid color "W" and the way its front windshield sloped down. (RT 1509, 1555-56.) Despite focusing on the front windshield as an identifying feature of the vehicle, she did not see the word Winnebago printed just below that windshield. (RT 1561.)

Sandoval was asked why she may have identified the vehicle as a Dodge Caravan, if, as she admitted, she knew the difference between a Winnebago and a Dodge Caravan. (RT 1530-34.) She responded that while she might have described the vehicle as a Dodge Caravan, she knew it was a Winnebago because of the "W."38 (Id.)

At the first trial, Sandoval again incorrectly stated that the RV had a ladder and nothing on top.39 (RT 1523, 1525.)

When shown a picture of the back of Petitioner's RV, Sandoval identified it as the vehicle she had seen that night based on the size of the rear window and the presence of a drape or curtain in the window. (RT 1526.) Later she agreed that there was no drape or curtain in the back window of Petitioner's vehicle. (Id.) She also acknowledged that the RV in the picture had a blue license plate and no ladder as compared to the white license plate and ladder on the RV she had seen. (RT 1529-30.) Based on the differences, Sandoval concluded that the photograph of Petitioner's RV depicted a different vehicle than the one she had observed. (RT 1530.)

When questioned about differences between the features on the rear of the vehicle in the photo versus the vehicle she had seen on the night of the fire, Sandoval again alluded to the solid "W" on the side of the vehicle.40 (RT 1527.) Defense counsel directed Sandoval's attention to the features on the rear of the vehicle in the picture; she explained that she was not good at explaining features of an RV. She agreed that the photo may have depicted an RV from the sales lot.41 (RT 1527-28.)

(5) Sandoval's Testimony at the Second Trial

At the second trial, Sandoval explained again that she had seen a "W" on the RV and communicated that to the investigating officers. (RT 5865, 77.) However, she expressed significant confusion when trying to explain how she knew it was a Winnebago.42 (RT 5877-78.) She also did not remember which investigators she told abut the "W." (RT 5917.)

Sandoval again explained that she had identified the RV based on seeing the "W." (RT 5883.) She did not dispute that she had used the term "Dodge Caravan," but, for the first time, explained that she had used that term only to describe the front of the vehicle. (RT 5913-14.) Upon further questioning she protestingly admitted that she may have called it a Dodge Caravan.43 (Id.)

Sandoval's inability to correctly identify many features of the RV was probed deeply on cross-examination during the second trial.

(RT 5948-49.)

d. Admitted Difficulty in Describing Winnebago

At trial, Sandoval repeatedly admitted having difficulty describing the features of the vehicle she had seen on the night of the fire:

(RT 5879.)

When asked to describe the difference between a Winnebago and a Dodge Caravan, Sandoval explained:

(RT 5969.) She did not mention any of the distinct physical features of the vehicles.

e. Winnebago or Dodge Caravan?

Sandoval explained that she observed the "W" and knew that it stood for Winnebago. (RT 5877.) However, she could not explain how she knew the "W" stood for a Winnebago. (RT 5877-78.) She offered only an agreement with the District Attorney that she recognized the vehicle in the same way that a young child might recognize his mother. (RT 5878-79.)

Sandoval was uncertain if and when she told investigators about seeing the "W." She did not know who first described the vehicle as having a "W."44 (RT 1517.) She could not remember if she told the first officers she talked to about the "W." (RT 1505.) She could not remember if she said that the "W" was on the back of the vehicle; she also acknowledged that the "W" on Petitioner's RV was brown, not blue. (RT 5949-50.)

f. Investigators' Recollection of Sandoval's Statements

(1) Officer Pimentel

At trial, Pimentel testified he raised with Sandoval the possibility the vehicle was a Dodge Caravan. (RT 7181-82.) He could not remember who first mentioned a Dodge van, but agreed that it would have been unlikely for the fire investigator to have noted what Pimentel had said. (Id.) He did not know why his report did not mention a Dodge RV or van. (RT 7184.) Later in the examination, and notwithstanding Sandoval's testimony at both trials that she had used the term "Dodge Caravan," Pimentel said that he, not Sandoval, had used the words "Dodge Caravan." (RT 7208.)

At the second trial, Pimentel testified that Sandoval called the vehicle a Winnebago from the start. (RT 7189-91.) When questioned why the word Winnebago does not appear in his report or why he previously had testified Sandoval did not call it a Winnebego, Pimentel said that he remembered this fact after he testified in the first trial.45 (RT 7189-91.) He denied that the word "Winnebago" could have first been used during Sandoval's discussion with Officers Lee and Ridenour. (RT 7211-12.)

(2) Captain Reuscher

At trial, Captain Reuscher agreed his handwritten notes suggest Sandoval may have described the vehicle as a Dodge Caravan. (RT 6699.)

Even though he used his contemporaneous interview notes to prepare his report, the report does not include reference to a beige RV or a Dodge Caravan. (RT 6732-33.) Reuscher admitted drafting the report after Sandoval had identified the vehicle and Petitioner had been arrested. (Id.) Reuscher thought it likely Sandoval had used the words "beige RV Dodge van type" and that he included that description in his report by way of his reference to her observing a white Winnebago.46 (RT 6734-35.)

(3) Detective Buehler

Detective Buehler escorted Sandoval to the RV sales lot and drove her back to Petitioner's house where she identified the RV.

Buehler testified that there were Winnebagos at the RV sales lot and that some had "W's" and stripes down the side. (RT 8188-89.) Buehler did not recall Sandoval referring to any such features as she toured the lot with him (even though, according to her testimony, those were the features she recalled about the suspect vehicle). (CT 140; RT 5945.) Instead, Buehler testified that Sandoval focused on the slanted shape of the frond end of the vehicle and, in so doing, eliminated motor homes with a sleeper compartment over the driver's cab.47 (RT 8190-91.) Buehler made no notes of any characteristics of the vehicle Sandoval said she had observed; there was a discussion about a spare tire and ladder, but he could not recall whether Sandoval said the vehicle had either. (RT 8193.) He recalled that the "big thing" Sandoval focused on was the slanted front end. (RT 8193.)

Even though his report mentioned nothing about a Winnebago, the "W," or a wide stripe, and despite his initial trial testimony mentioned above, Buehler testified for the first time at the second trial that Sandoval had indeed told him the RV had a side stripe and a "W." (RT 8294-95; Buehler Rpt., Ex. J-XV.)

(4) Detective Trogdon

Detective Trogdon did not interview Sandoval. He was assigned to canvas the neighborhood on the morning of the fire and ask if anyone had seen a class "C" motorhome. (RT 3217, 3227-28.) Trogdon described class "C" motorhomes as a type of RV where the nose of the vehicle sticks out and usually has a sleeper compartment above the driver's compartment. (RT 3217-18.) Trogdon did not include information regarding a Winnebago or a "W" despite Sandoval's alleged reference to them.

g. Description of the Back of the Vehicle

Sandoval exhibited consistent difficulty describing the back of the RV. She did not identify it when she was first driven by Petitioner's house. When she was shown a picture of the rear of Petitioner's RV at the first trial, she said it was different than the one she had seen driving around before the fire. (RT 1506-07.) She testified that she had been able to see the back and roof of the vehicle as it passed back and forth in front of her apartment multiple times. (RT 5896.) She felt that the features on the back of the vehicle were some of its most distinctive. Still she could not identify the vehicle from the rear, and those features she thought most distinctive on the suspect vehicle were not present on Petitioner's vehicle.

During the preliminary hearing she described the rear ladder and window as two of the most distinctive attributes of the suspect's RV. (RT 5941.) Petitioner's RV did not have a ladder. Despite looking at the rear window of Petioner's RV, she did not identify it as the suspect RV. No information regarding a ladder or back window was recorded in the police reports. During the preliminary hearing, almost immediately after stating that she could not identify the vehicle from the back, she stated that the back of the vehicle was one of the features that reminded her of the vehicle she had seen that night.48 (CT 137, 139.)

h. Testimony of Thomas Streed

Petitioner called Thomas Streed, a former homicide detective and a behavioral scientist with a Master's Degree and Ph.D. in Human Behavior, to examine and discuss issues relating to the reasonableness of police procedures employed in connection with Sandoval's eyewitness identification. (Streed Rpt., ECF No. 102-8.)

(1) Questionable Police Practices

Streed notes that Sandoval was interviewed shortly after the fire, very early in the morning of January 15, 1997. The interviewing officers' reports, made after Sandoval identified the vehicle and Petitioner was arrested, state that Sandoval identified a white or off-white RV with a blue "W." However, he noted, Reuscher's handwritten notes made during his interview with Sandoval describe a "beige RV Dodge Van type" and "Dodge Caravan." At trial, Reuscher admitted that he did not refer to a Dodge vehicle in his report because Sandoval had by then identified the Winnebago. (Streed Rpt. at 12-13.)

According to Streed, California peace officers are taught that the first description obtained from an eyewitness tends to be the most accurate. (Street Rpt. At 13.) The original description provided by a witness is particularly important when testimony is inconsistent and changes over the course of the investigation. (Id.) Moreover, reasonable peace officers are aware that audio or video recordings are the most reliable form of documenting statements. (Id.) Here, the interview either was not recorded or the tape of the recording had been destroyed. Both Officer Lee and Sandoval remember the interview being taped.49 (Id.)

(2) Suggestive Second Showup

Streed was of the opinion that Sandoval's testimony regarding the RV was influenced by subsequent exposure to the vehicle in the second of the two police showups. (Streed Rpt. at 10.) California peace officers are taught that the process of repeatedly showing an eyewitness the same suspect or object can create a suggestion to the eyewitness that the suspect or object is the correct choice. (Streed Rpt. at 14.) Here, Sandoval was shown Petitioner's RV twice in less than hour. The first time, Sandoval could not identify it. The second time, she did. At this second viewing, Sandoval observed five or six police cars at Petitioner's residence; it appeared that something was happening and possibly the police were focused on the RV.50 (RT 1546, 5947.)

E. Circumstantial Evidence

A significant amount of circumstantial evidence was presented at trial; some probative of Petitioner's guilt and some supporting his claim of innocence. To make the determination required by Schlup, the Court will review all such evidence having any perceived probative value.

1. Unlawful Detainer Action

Daniel Jones and his family rented the home on 1319 Ronald Avenue from Petitioner. On November 5, 1996, Jones notified Petitioner that he planned to move his family out by the first of December. Jones testified that Petitioner did not show any sign of dissatisfaction with this news, and in fact seemed to understand Jones' economic reason for moving. (RT 5749.)

The parties disagreed however on whether Jones had pre-paid his last month's rent. Petitioner presented Jones with a three-day "pay rent or quit" notice on the same day that Jones told Petitioner he was moving. (RT 5748-49.)

Jones testified that he intended to move out by December 1, but was unable to do so. (RT 5749-50.)

A hearing was held on the unlawful detainer action on December 20, 1996. (RT 3307.) The court found that Jones did owe rent for the months of November and December, awarded Petitioner $1147.50 as unpaid rent and $405 in fees and costs, and restored possession of the property to Petitioner. (RT 3307-10.) At the hearing, however, Petitioner agreed to stay the writ of execution until January 1, 1997, so the Jones family could remain in the house for the holidays. (RT 3311-12, 3319-20.)

Petitioner's unlawful detainer counsel stated that Petitioner was not particularly happy about not getting paid rent, but that his reaction was unremarkable when compared to other landlords the attorney had assisted in evicting tenants. (RT 3313.) Petitioner's counsel thought that, in any event, Petitioner likely would ultimately collect the judgment because Jones was employed. (RT 3311.)

The tenants were scheduled to be locked out of the property on January 16, 1997. (RT 2396.) However, on January 13, 1997, the sheriff's office advised Petitioner's eviction agency that the writ of execution had to be amended to reflect Ronald "Avenue" rather than Ronald "Street." (RT 2396.) When told this would likely take a week to correct, Petitioner was upset, but not extraordinarily so. (RT 2404, 7737.) Ruth Seere, who worked at the eviction agency, also was frustrated with the Sheriff's refusal to overlook the technical error. (RT 2407-08.)

Even though the Jones family failed to move out on the first of January and Petitioner was forced to evict them, he reportedly showed no animosity toward them. (See, e.g., RT 3312-13 (Petitioner's unlawful detainer attorney testified Petitioner expressed no antipathy toward the Joneses), RT 3687-88 (Petitioner's son testified Petitioner expressed no anger toward the family)).

2. Argument with Mrs. Jones?

At trial, the prosecution presented Hope Warner, the manager of the mobile home park where the victims had been scheduled to move. She testified that a day or two before the fire she had seen an angry exchange between the driver of a Winnebago and Mrs. Jones at the mobile home park. (RT 8040-46, 8053-55, 8084, 8166-68, 8127, 8170.) She later identified the driver of the vehicle as Petitioner. (Id.) She said another witness, Apolis Crane, was present at the time of the argument. (RT 8057-58.)

There was no evidence presented as to why Petitioner would confront Mrs. Jones there rather than in Petitioner's rental home where she still lived. There was no evidence as to how Petitioner would know to visit the mobile home park at the same time Mrs. Jones visited.

At the first trial, Petitioner called Russell Downing, the owner of the mobile home park, to testify regarding Warner's credibility. He stated that he had been approached by a Federal Emergency Management Agency investigator about a possibly falsified receipt by Warner on a flood damage claim. (RT 3181-85.) He also stated that Warner's daughter sued him for injuries sustained in a fall in a fence-post hole on his property, even though the alleged fall occurred three weeks after the fence had been completed. (Id.)

Apolis Crain testified that he was at the mobile home park during the time that Warner described the exchange having taken place and did not witness any argument. (RT 3277.) He also testified that Warner called him before the trial and threatened to report his allegedly unlicensed business activity to the State if he did not "get [his] story straight." (RT 3277-78.)

3. Moved RV

On the night of January 14, one of Petitioner's neighbors saw him pull his Winnebago out onto the street around 8:00 p.m. (RT 8313-15.) Daniel and Georgina Treece, also neighbors, noticed the Winnebago was parked out on the street when they went to work at 11:30 that night. The Treeces had never before seen the RV parked on the street at night. The RV was returned to the side of Petitioner's house when the Treeces returned home from work about 8:30 the next morning. (RT 6126-28, 6143-44.)

4. Prior Rental Fire

On January 31, 1995, a fire occurred at a tenant-occupied property owned by Petitioner. (See Ex. R-C.) The next day, Petitioner served the tenant with a 30-day notice of eviction. (Ex. R-B.) Petitioner, a licensed contractor, received money from his insurer to repair the home himself. (See Ex. R-C.) The fire was electrical in nature and ruled accidental; no evidence of criminal activity was identified.51

Respondent, like the prosecution at trial, argued that evidence of this event was relevant to show Petitioner's motive insofar as it proved Petitioner had knowledge that a fire could generate an insurance payment and create a reason to remove tenants. This evidence was excluded by the judge at Petitioner's trial. (See RT at 338-42.)52

5. Financial Motive?

Bertha Love, an employee at Petitioner's bank, testified that Petitioner came into the bank a week before the fire and inquired about the bank taking back one of his three properties. (RT 7117-18, 7131-322.) Love claimed that Petitioner "was very upset" and "irate," and that he "wanted [her] to take the keys" to the Ronald house. (RT 7113-14.) However, Love's typewritten memorandum on the event says nothing about Petitioner being upset or irate. It simply reflects essential details of his request to voluntarily surrender a loan. (RT 2447-48, 7121-22, 7124, 7130-31.) Love pointedly refuted Detective Lee's claim that she had said that Petitioner had "thrown" his keys at her. (RT 3244, 7135.) Love acknowledged Petitioner's heavy Greek accent and general tendency to speak loudly in a boisterous and demonstrative manner, but stated she could differentiate between him being loud and being upset. (RT 7118-19, 7137-38.)

Petitioner's son, Demetrios Souliotes, testified that Petitioner's English was poor and that his frustration communicating, inability to know the correct English word to use, use of his hands in speech and strong enunciation of words often caused others to think his father was upset when he was not. (RT 3689-90.)

Banker Louis Bacigalupi, Love's supervisor, also met with Petitioner. He did not believe Petitioner was upset. He described Petitioner as "a little bit nervous" but otherwise engaged in a fairly routine conversation about loan options. (RT 7724-26.) Bacigalupi was unable to answer Petitioner's questions and referred Petitioner to the loan service department. (RT 7722.)

Real estate broker Larry Titus contacted Petitioner after he learned of the unlawful detainer action from court records. (RT 7155.) On December 4, 1996, Petitioner agreed to list the Ronald Avenue house for sale with Titus. Petitioner told Titus he wanted no more rentals. (RT 7157.) While not discussed, Titus felt Petitioner understood it would be difficult to show the house to potential buyers while the tenants were still there. (RT 7158.)

Titus discussed Petitioner's flexibility on sales price. Petitioner was willing to lower the asking price if he received no offers in thirty days. (RT 7162.)

At trial, Michael Marks, an auditor with the United States Department of Treasury's Bureau of Alcohol, Tobacco and Firearms, examined Petitioner's financial records to evaluate his financial status at the time of the fire. (RT 6979, 6981-6983.) According to him, as of January 15, 1997, Petitioner had assets worth about $394,000 and liabilities of about $209,000. (RT 7009.) Marks put Petitioner's monthly income at $2,613 and his monthly expenses at $2,528; he noted Petitioner had $16,579 in liquid savings in the bank.53 (RT 7008-09, 7018-7020.) Marks also testified that Petitioner had a history of paying all his bills on time and leaving no balances on his credit cards. (RT 7094-95.) Petitioner's monthly mortgages on his properties totaled $1,504, not including taxes and insurance. (RT 7018-7020.)

At the evidentiary hearing, Scott Spertzel, a certified public accountant, testified to his similar examination of Petitioner's financial position. (HT 346.) Spertzel explained that at the time of the fire Petitioner's net worth exceeded $210,000. (HT 347.) He testified that Petitioner's liquid cash savings were sufficient to maintain his current lifestyle for at least eight months even if his income ceased entirely. Petitioner had numerous options to secure additional cash or income if needed. (HT 358-59.)

Spertzel calculated Petitioner's estimated monthly expenses to be somewhat less than Marks, at $2,234 a month. (HT 351; ECF No. 102-5, Ex. B at 16 (Spertzel Rpt.).) He concluded that Petitioner's expected monthly income, comprised of workers' compensation payments, Social Security payments, and rental income,54 totaled approximately $3,550 per month. (See Spertzel Rpt. at 12-13.) Petitioner was also due, in February 1997, to begin collecting net interest payments of $287.50 on the $30,000 loan. (RT 7062; Spertzel Rpt. at 7.) Had Petitioner moved in with his girlfriend, Jill LeBlanc, as had been discussed, he could have rented his residence at about $600 per month. (RT 3347; Spertzel Rpt. at 12.) Including these additional sources of income, Spertzel calculated that Petitioner's potential monthly income in the beginning of 1997 was $4,450.50. (Id.)

Spertzel described multiple options available to Petitioner to generate additional income or cash. In addition to his house, rental properties, and bank accounts, Petitioner had a retirement account with over $33,000, mountain property at the Odd Fellows Sierra Camp, and several vehicles which he could have sold. (Spertzel Rpt. at 17; RT 7074:4-12.) He was also in the process of litigating a workers' compensation claim that was expected to yield a judgment, including attorney fees, of at least $100,000. (RT 3459, 3466.) (At the first trial, Gary Nelson, Petitioner's workers' compensation attorney, testified that Petitioner's workers' compensation case for injures sustained while employed at Montgomery Ward was projected to produce an estimated lump sum payment of at least $100,000. (RT 3445-46, 3454-55.))

Like Marks, Spertzel noted that Petitioner regularly paid his credit card bills in full each month and he was not in default or foreclosure on any properties. (RT 7014-15, 2973, 7078; HT 351.)

At the time of the fire, Petitioner had identified a potentially interested buyer of the Ronald Avenue house. Petitioner had listed the property for sale for $89,950. That same month Earl Linam approached him to negotiate a purchase. (RT 7983-86.) Linam did not think Petitioner seemed desperate for money; in fact, Petitioner offered to loan Linam the down payment on the house. (RT 7992-93.)

Streed concluded Petitioner had no financial incentive to destroy the house. He likley would have received far less than the $97,000 insurance policy limit. The policy gave Allstate the option to pay either the cost of replacement or its actual cash value. (HT 356; Streed Rpt. at 9-10.) Allstate determined that the actual cash value of the Ronald Avenue house just prior to the fire was only $55,000, well below both the policy limit and the house's listing price. (HT 357; Exs. J-XXIX, J-XXX.) That amount would not have covered the $75,591 balance remaining on Petitioner's mortgage for the house (HT 357), and was less than the listing price of $89,950.

6. Testimony of Girlfriend Jill LeBlanc

Petitioner's girlfriend, Jill LeBlanc, testified at the first trial that Petitioner had been scheduled to drive the RV to her house on the day of the fire. (RT 3341-42.) The two were planning to go dancing that night in the Bay Area and then spend several days traveling along the California coast, camping in the RV along the way, before attending a swing dancing convention in Monterey on January 23, 1997. (Id.)

LeBlanc also testified that at about the time of the fire she and Petitioner had been making plans for Petitioner to move into her home in Pleasanton, California, about 50 miles west of Modesto. (RT 3326-29.) She explained that Petitioner was tired of taking care of multiple rentals and was thinking about selling 1319 Ronald Avenue. (RT 3329.) LeBlanc was independently secure financially. She was not aware of any financial problems on Petitioner's part. (RT 3329, 3333-34.) Two days before the fire, she went with Petitioner to the Greek Consulate in San Francisco in preparation for a two month trip to Greece together. (RT 3336-3341.) Petitioner did not ask LeBlanc to contribute financially to pay for the trip. (Id.)

7. Tunnel Vision/Confirmatory Bias

a. Overview

At the evidentiary hearing Petitioner's expert, Thomas Streed, testified regarding a criminal investigation phenomenon he called "tunnel vision" or "confirmatory bias." He described this as something that occurs when investigators become convinced that a particular individual is the suspect and then disregard or exclude potentially exculpatory information and focus solely on evidence that supports their existing conviction. (HT 281.) Confirmatory bias causes a loss of objectivity that may subconsciously steer the investigator to a certain conclusion. (HT 282, 319.) Streed saw evidence of confirmatory bias in the investigation of Petitioner. (HT 282-83.) He referred, for example, to investigators' failure to follow up on a complaint of a particularly suspicious RV lingering near the fire on the morning of the fire and the report of a car speeding away from the scene just as another witness arrived. He also noted investigators' disregard of the fact that the motor of Petitioner's RV was cold on the morning of the fire and the fact they could not find in Petitioner's house the shirt Sandoval described the suspect wearing. (HT 283-85.)

Streed also expressed concern with inconsistencies in investigative reports (such as Captain Reuscher's failure to include in his report reference to Sandoval's description of a Dodge van or Caravan.) (RT 287-29.) Streed noted too that in the statement of probable cause used to secure a search warrant, Detective Lee stated that Sandoval, when shown the rear of Petitioner's RV, was "not sure" if the RV was the one she had seen. However, both his and Ridenour's report described her as saying it "did not look like the vehicle" she had seen. (HT 295-96; Exs. J-XIII, J-XIV.)

It is also of concern that Modesto Police Department investigators focused on Petitioner so quickly after the fire. Detective Owen testified that by about 9:30 a.m. he had been directed to seek a search warrant for Petitioner's home, and he immediately went to Petitioner's home to get information for the warrant from Detective Buehler. (RT 7239-40.) The police claimed that Sandoval's identification of Petitioner's Winnebago was the impetus for the search warrant of Petitioner's residence. (RT 7276-77.) However, Sandoval did not identify Petitioner's Winnebago until at least 10:37 a.m. (Exh. J-XV at 3.) It thus appears police began the warrant application process before any identification of Petitioner's RV implicated him in the crime. (RT 8186-87.) Petitioner was arrested the day of the fire. There is little evidence other leads were investigated prior to the arrest.

Unpursued leads favorable to Petitioner are discussed below.

b. Cold RV Engine

The police, once alerted to Petitioner's RV, attempted to determine if its engine was still warm. At approximately 10:25 a.m. the morning of the fire, Petitioner, at the request of investigators, lifted the hood of his Winnebago and Detective Ridenour, expecting it to be warm if it had been driven as described by Sandoval the night before, touched the engine. He found it cold to the touch. (RT 7799-7800; Ex. J-XIV.)

At the first trial, defense expert Dr. Donald Myronuk testified regarding an experiment he conducted on a similar RV to show that the engine would have retained detectible warmth had it been driven in the manner described by Sandoval. (RT 3565-66.)

c. No Blue and White Shirt

The police disregarded their inability to recover a blue and white checkered Pendleton-type shirt when they searched Petitioner's house pursuant to a warrant even though the warrant instructed officers to search for precisely this item.55 (Ex. P-12; HT 285-87.)

d. Ford Taurus

Steve Hamilton, a passer-by, was perhaps the first to arrive at the scene (and, incidentally, was injured attempting to ensure no one was still on the premises as it burned). He testified that when he arrived at the fire, a Ford Taurus-like automobile was parked in front of 1319 Ronald Avenue with its lights out and the driver seemingly watching the fire. (RT 1604, 6219-20, 6238-39, 6241-42.) As Hamilton approached, the driver sped away, running the stop sign at Ronald's intersection with Tully. (RT 6240-41.) Hamilton described the driver as a 25-to-35-year-old Caucasian male with a dark skin tone, "longish light hair," a mustache, and glasses. (RT 1604, 6242-43.) Hamilton thought the driver appeared to act like a pyromaniac watching his work. (RT 6241-42.) Two other witnesses also heard squealing tires around the time the fire started. (RT 6030-31, 6034-35, 7144-46.)

There is no record of any investigation into these facts. Captain Reuscher acknowledged that the information was important and that he definitely would have been interested in it, but no one shared it with him. (RT 6737-38.) Streed agreed, and testified that this information should have been further investigated. (HT 324-25.)

e. Other RVs

Another Winnebago was seen being driven suspiciously in the vicinity of 1319 Ronald Avenue on the morning of the fire.

At the first trial, Linda Flores, a witness otherwise uninvolved with the case, testified that at some time before 7:00 a.m. on the morning of the fire she saw a beige or off-white, older model, Winnebago with "an olive green" or "brownish green stripe" and a "W" on the side drive around the parking lot of the unopened shopping center where she worked. (RT 3287-88.) An opening in the center exposed the alley leading to 1319 Ronald Avenue. (Id.) Flores testified that the driver suspiciously circled the parking lot, made approximately four U-turns, and repeatedly looked down the alley towards Ronald Avenue. (RT 3287-88, 3291.) The man she saw driving this Winnebago was not Petitioner. She was positive Petitioner's Winnebago was not the Winnebago she saw driving around the parking lot. (RT 3289-90, 3301-02.)

Flores called the police to report this when she got off work that night at around 6 p.m. She received no follow-up call or visit. (RT 3290-91.)

Another witness, Tim Campbell, testified that on two separate occasions during Petitioner's trial he witnessed a Winnebago he thought was "identical" to Petitioner's being driven around in Modesto. (HT 341-44.) On one occasion, he saw this Winnebago only a couple blocks from the site of the fire. (HT 341.)

During trial, a Winnebago employee was called to testify regarding how very few RVs like Petitioner's were produced. (See RT 2148-49, 6837.) He said Winnebago produced a total of 15,000 RVs of that model, of which 1,068 were shipped to California. (Id.)

f. Michelle Jones' Psychological Issues

At the first trial, Petitioner presented an expert psychiatrist, Dr. Fred Rosenthal, to testify regarding victim Michelle Jones' psychiatric history. (RT 3384-87.) Based on his review of medical, psychiatric, and autopsy records, Rosenthal described Jones as a troubled individual with a long history of emotional problems starting in childhood. (RT 3390-94.) She had a difficult childhood: she had been diagnosed with cancer at an early age, had a below average IQ, was placed into foster care, ran away from home, lived on the streets for some time, and had a history of drug use. (RT 3394-95.) Rosenthal testified that children with this type of history usually become seriously depressed adults. (RT 3395.)

In 1994, Jones attempted suicide by cutting her wrists. (RT 3397.) She had sought psychiatric treatment several years before the fire, but had difficulty maintaining appointments. (RT 3398.)

No drugs were found in her system at the time of death. However, during a period several years before the fire, she had been a fairly heavy occasional user of cocaine, amphetamines, marijuana, and nicotine. (RT 3399-3400.) Rosenthal thought this history reflected poor judgment and simplistic thinking. (RT 3401-02.)

F. Evaluation of the Evidence

1. New Fire Science Evidence

a. Arson?

It is axiomatic that one cannot be convicted of arson unless there is proof, beyond a reasonable doubt, that arson in fact occurred.

As noted, the parties agree that the cause and origin of the Ronald Avenue fire cannot be determined from the evidence presented. Original expert conclusions that flammable liquids were used as an accelerant and that the fire was caused by arson no longer have probative value.

Only Shilling testified at the evidentiary hearing that the fire was intentionally set. However, his conclusions were so clearly based on outdated and now universally discredited theories that they cannot be considered to have true evidentiary value. His reported examination of potential accidental sources is similarly found to be of little probative value. The appliances he believed he had examined were buried in a large pile of rubble by the time he arrived at the scene. His report provided no more than a brief reference to his examination of these appliances even though then-existing standards called for documentation of their elimination as potential causes of the fire. Carman, an obviously well-qualified fire science expert, and the one who presented as perhaps the most objective and credible of those who testified at the evidentiary hearing, opined that the original investigator's visual inspection of electrical outlets was inadequate. (HT 74-78.) Other experts at the evidentiary hearing agreed that the original investigation was poorly documented. Accidental causes, while considered, were not thoroughly investigated.

The only significant difference of opinion between the experts at the evidentiary hearing was whether the fire had one or multiple points of origin. However, even DeHaan, who thought there was some evidence of multiple points of origin, acknowledged that neither that theory nor the cause of the fire could be established with any measure of certainty.

The Court recognizes that science has not ruled out arson as the cause of this fire. Respondent notes, correctly, that the inability to determine the cause of the fire leaves open the possibility it was intentionally set. (Resp. Brief at 26-27.) Of course, something more than a possibility is needed for a juror to find one guilty of arson. This Court must make a determination as to whether any reasonable juror could find beyond a reasonable doubt on the basis of the remaining evidence that the fire was intentionally set and set by Petitioner.

The comparative effect of the absence of the expert evidence of arson is difficult to measure. One cannot simply say that one of a series of equally-weighted factors leading to Petitioner's conviction has now been removed. Rather, as the prosecutor argued, it must be acknowledged that the now-excluded evidence provided a very substantial foundation upon which virtually all the other evidence rested. Surely it was easier for the jury to focus on identifying an arsonist once unrebutted experts had confirmed that the fire was caused by arson.

The sheer weight of the scientifically unrefuted evidence of arson itself likely had an added effect. At trial, the fire investigators expressed great confidence that the fire was intentionally set using an accelerant. They told the jury that certain identified factors were conclusive proof of an arson caused by an ignitable liquid and that there could be "no doubt" this was the case.56 (RT 6957.) Even though these experts were subjected to rigorous cross-examination, no competing expert evidence was proffered at the second trial. The prosecution's closing argument relied heavily on the expert evidence as a basis for finding Petitioner guilty:

(RT 9050.) The jury had no comparable evidentiary basis upon which to reject that testimony from two very experienced, presumably impartial, retired area fire captains.

Now, without expert findings of arson, the other evidence against Petitioner must be viewed in a more critical light. The Court must determine how reasonable and contentious jurors would react when all acknowledge that from a scientific standpoint there is an equal likelihood this fire was accidental and not a criminal act. To make that determination, the Court shall analyze and weigh the probative value of the remaining evidence.

b. Missing Physical Link

It is now known that the MPDs from Petitioner's shoes are not the same as the MPDs found at the scene. The absence of this physical evidence linking Petitioner directly to the fire scene is similar in import and effect to the removal of the fire causation evidence. Petitioner's shoes no longer tell a tale; they do not point "the finger of guilt" at Petitioner. There is no link between Petitioner's shoes and the fire. There is no MPD evidence placing Petitioner at the scene of the fire.

Respondent argues that even if the MPD evidence does not tie Petitioner's shoes to the fire scene, the presence of MPDs at the scene was unusual. (Resp. Post-Hr'g Brief at 28.)This assertion, however, runs contrary to the stipulation that "[d]etectable MPDs are commonly found on many household products and consumer goods, including the solvents in glues and adhesives used in floor coverings and footwear, residues of dry cleaning solvents, insecticides and cleaning agents." (Undisputed Fact 14.) The presence of detectible MPDs at the fire scene no longer lead the experts to attribute the cause of the fire to MPDs; both Petitioner and Respondent's experts concluded that the cause of the fire could not be determined.

With regard to the presence of MPDs, Carman explains, "For instance, the mere presence of a medium petroleum distillate in a scene has little bearing on the start of a fire if that material had normally been present in the form of lamp oil or other similar materials. It is incumbent on investigators to eliminate the possibility such items were present in an area prior to a fire. With the presence of various chemicals and synthetic materials in houses and other buildings, such a task can be time consuming and difficult." (Carman Rpt. 16-17.) Given expert opinions and the stipulation of the parties that MPDs are commonly found in household products, the Court concludes that the mere presence of an MPD at the fire scene is of no probative value in evaluating Petitioner's guilt.

This new evidence that MPDs found on Petitioner's shoes were not the same as those found at the scene must be considered to be of tremendous significance in any evaluation of guilt versus innocence in this case. At trial, the jury was presented with scientific evidence that not only placed Petitioner in the crime scene but also strongly implicated that the materials found on his shoes was precisely that which had been used to start the fire. The prosecution relied heavily on such evidence. Now, all agree that qualified experts cannot determine if the fire was arson and that the chemicals on Petitioner's shoes are distinct from those found in the fire. The parties' stipulations regarding the fallibility of the fire science used in Petitioner's trial drastically and fundamentally changes the evaluation of Petitioner's guilt. The deletion of this scientific evidence leaves a large gap which can only be filled with evidence that the Court finds to be far less compelling.57

Here, the principal remaining evidence of Petitioner's guilt is Sandoval's disputed testimony that identified Petitioner as the driver of the motor home from which someone exited and visited the scene minutes before the fire started. Other circumstantial evidence of guilt, mostly focused on Petitioner's alleged motives for committing the arson, remains. This Court shall determine if a reasonable juror would convict Petitioner in light of the present state of the evidence.

2. Monica Sandoval

a. Identification of Petitioner

The evidence viewed in the light most favorable to Respondent establishes that Monica Sandoval identified Petitioner as the driver of the RV which stopped while the driver exited the vehicle and walked towards 1319 Ronald Avenue with a sack. Shortly afterwards, the driver returned to the RV without the sack, drove away and then passed nearby again to look suspiciously toward 1319 Ronald Avenue just before the house became engulfed in flames. If credible, her testimony is very strong evidence of Petitioner's guilt. However, Petitioner argues strenuously that the discrepancies, inconsistencies and contradictions in Sandoval's testimony and official reports of her statements leave the former wholly unworthy of credit.

The Court must determine the weight to be given to this eyewitness evidence.

At the outset, the Court cannot disregard the very peculiar circumstances that made Sandoval a witness that morning. She acknowledged being of a state of mind that studies show negatively impacts ones ability to observe and record recoverable information. It is, at the least, very odd to think that anyone would stand outside on a balcony for two and one half hours on a cold, wet night when there was a view of the street from a window inside. (RT 5893-94.) Only once, after an hour and a half, did Sandoval go inside to get a jacket and turn off lights. She denied doing anything else to occupy or amuse herself during that entire time.58 (RT 5984.) Curiously, she did nothing when she saw an RV drive up and down her street time after time in the middle of the night and then stop while its driver surreptitiously delivered a sack to 1319 Ronald Avenue.

Sandoval's view of the driver at that point did not aid in identifying him. She could barely conclude it was a man (though, paradoxically, she noted relatively small details of his shirt). This admittedly extremely distraught witness later briefly saw a portion of the driver's face at a great distance, under poor lighting conditions and through a windshield.

The reenactment video, though far from determinative on the issue, is enlightening in illustrating the difficulty anyone would have in trying to see, record and recall the face of an RV driver in the circumstances described by Sandoval. In this regard the Court recognizes the clear discrepancy between the amount of light shown in the still photograph compared to that shown in the video. Because the Court believes it common knowledge that photographic equipment settings can affect the amount and quality of light reflected in a photo or film, the Court is not persuaded the video accurately portrays the exact same lighting conditions Sandoval encountered on January 15, 1997. Nevertheless, the reenactment, along with Balletto's testimony regarding her observations during the filming, demonstrate the scene generally and the difficulty one would have in observing a very briefly-exposed part of a face through a windshield at that distance. The Court does find credible Balletto's testimony that she was unable to observe in any meaningful way the face of the driver of the RV during the reenactment.

Sandoval's initial description of the driver's age and facial characteristics did not match Petitioner and it changed over time. No reason has been given to question the accuracy of original investigation reports attributing to Sandoval an estimate of the suspect's age as being in the "thirties" or in the "35 to 45" age range. As noted, the age shifted upward as matters progressed. Sandoval's belated descriptions of an older man suggests she, consciously or subconsciously, altered the description to better conform to Petitioner's likeness rather than describe what she saw on the night of the fire.

Of greatest significance to the Court is Sandoval's belated identification of Petitioner after describing a much younger man, after telling investigators she had not seen the driver well enough to identify him, and after being unable to identify him in a photograph on the morning of the fire. Her newfound ability to identify him some six months later came only after she had seen his face in the photographic lineup, she had seen at least one newspaper photo of him identified as the suspect arsonist, and she saw him sitting in court in front of her alongside defense counsel and wearing a red jail jumpsuit. She identified him after a break in court proceedings during which she might have talked with the prosecutor who then unexpectedly reopened his questioning of her to ask her if she could then make the never-before-possible identification. Neither she nor anyone else has offered any explanation, or even reasonable supposition, as to how she might have improved her ability to identify the RV driver during that intervening six months. (A reasonable juror might well wonder whether the dismissal of criminal charges for assault with a deadly weapon influenced her testimony, despite her assertions that there was no "deal" with the prosecutor. (RT 5981.))

It is possible that Sandoval observed a portion of the driver's face. It cannot be said with absolute certainty that, perhaps on seeing Petitioner's face in court, she did not suddenly recollect what she had seen six months earlier. However, there are so many factors militating against such conclusions, such bizarre circumstances surrounding the alleged sighting of the driver, and so many uncertainties, contradictions, and inconsistences in her testimony and in the various reports of it, the Court concludes that no reasonable juror would credit this identification. Dysart, who the Court found to be a very well-qualified and objective59 expert on such matters, concluded, after considering the totality of the circumstances, that Sandoval's identification of Petitioner was "wholly unreliable." (HT 157-58.)

The Court agrees.

b. Identification of RV

The Court now undertakes to determine the weight a reasonable juror would be expected to give to Sandoval's identification of Petitioner's RV as the one she saw being driven that night. The identification was made on the morning of the fire, but in other respects it suffers from the same types of inconsistencies and infirmities as her identification of Petitioner. Various factors supporting and impeaching the RV identification are reviewed below.

(1) Dodge Van or RV

The confusion over whether Sandoval first described a Dodge Caravan or an RV is neither easily explained nor insignificant. Both Sandoval and Reuscher testified, and Reusher's notes reflect, that Sandoval likely used the term Dodge Caravan when initially interviewed. That is difficult to reconcile with Sandoval's assertion that she knew all along that she had observed a Winnebago because of its blue "W."

If Sandoval had initially described an RV with a blue "W" and stripe, one would have expected her and Officer Buehler to have focused on those details as they toured the RV sales lot looking for a vehicle similar to the one she had seen. Buehler remembered that several RV's on the sales lot had stripes and a "W." However, he reported that the slanted front end was the critical feature searched for; he did not mention a "W" or stripe in his report.

The references to a Dodge Caravan conflict with Sandoval's later identification of an RV, with later-drafted police reports, and with her preliminary hearing and trial testimony. Moreover, Sandoval's inability to remember if she actually described a "W", who was the first to mention a "W", where the "W" was located, or why "W" stood for Winnebego or RV raises additional concern regarding the reliability of her identification.

(2) Wrong Features

Many of the other features Sandoval described as having been the most distinctive on the suspect vehicle and those which helped her identify Petitioner's vehicle were not in fact present on Petitioner's RV: the ladder, the rear window with a white curtain, and a white license plate. Conversely, she maintained that the suspect vehicle had a spare tire and nothing on the roof. Petitioner's had a spare tire and an air conditioner on the roof. Despite claiming she identified the RV based on its rear window, she could not identify the RV from the rear either on the day of the fire or, from a photograph, at trial. It is recalled that according to notes of her early contact with officials, she described a beige vehicle. Petitioner's RV is white.

(3) Suggestiveness

As noted, Sandoval was first taken by Petitioner's house and asked to look at his RV at a time when he was at least considered a potential suspect. More significantly, she was brought back a second time and this time saw several police vehicles at Petitioner's house and activity potentially suggesting the RV was a focus of the investigation. The suggestive potential in visiting the same place twice and there seeing police activity cannot be ignored. Streed's well-qualified and credible opinions that such police behavior is discouraged, and why, is testament to its potential corrupting influence.

(4) Conclusion

It is the Court's opinion that the reliability of Sandoval's identification of Petitioner's RV suffers from some of the same infirmities as her unreliable identification of Petitioner as its driver. Clearly, those infirmities are not of the same number or magnitude as those in the case of the personal identification. There are fewer contradictions. There is no inexplicably regained recall months after the event. The suggestive events preceding the identification were less numerous and, to this Court, less toxic. In short, the Court concludes that the credibility of Sandoval's identification of the RV is much greater than the credibility of her identification of Petitioner. The former is entitled to some weight; the latter, none.

The weight ultimately given cannot be determined in a vacuum. Were there substantial other evidence against Petitioner and the RV identification the last piece of the puzzle, it would likely be enough to complete the puzzle. The challenge here is to weigh that evidence in light of what remains of other evidence and in light of Sandoval's overall credibility. The Court shall undertake such an evaluation after first examining some of the evaluative criteria.

c. The Evaluative Framework

Sandoval was vigorously cross-examined at trial. The jury was given several reasons to doubt the reliability of her identification of Petitioner and the RV. Schlup instructs that a Court must presume that a reasonable juror would conscientiously obey the instructions of the trial court. Schlup, 513 U.S. at 329. The jury was given instructions regarding the reliability of witness testimony before it deliberated on Petitioner's guilt. There is no reason to believe the jurors ignored the many facts and circumstances which detracted from Sandoval's credibility. It is then reasonable to argue, as Respondent in effect did, that reasonable, qualified jurors, presumably following contentiously the instructions given to them and essentially made aware of the several reasons to question Sandoval's testimony, nevertheless found Petitioner guilty beyond a reasonable doubt. Why, one may ask, should the Court second-guess that result?

This Court is not undertaking to second-guess the jury. It is attempting to divine what that jury, or any reasonable juror, would do if faced with having to evaluate the reliability of Sandoval's identification testimony without the persuasive, essentially unchallenged, evidence that this fire was started by an arsonist and also put Petitioner at the scene of the arson. That jury was not asked to determine whether Sandoval's testimony and other evidence would produce the same finding if there were no scientific proof of arson and no scientific evidence indicating Petitioner had been at the scene. Rather, at closing, the prosecution encouraged the jury to evaluate Sandoval's testimony in the light of the substantial evidence of arson, MPDs, and other circumstantial evidence:

It is the burden of this Court to determine what a reasonable juror would do without that scientific evidence. Specifically, this Court must determine in light of "all the evidence, old and new, incriminating and exculpatory" without regard to rules of admissibility, is it "more likely than not that no reasonable juror would have convicted [Petitioner] in the light of the new evidence." Schlup, 513 U.S. at 327-32; House, 547 U.S. at 537-538. Without fire investigation reports or MPD evidence implicating Petitioner, the import of Sandoval's testimony increases exponentially. Irrespective of the effectiveness of defense counsel's cross-examination at trial, Petitioner has properly challenged Sandoval's testimony in an attempt to show his actual innocence under Schlup. Respondent's is incorrect in asserting that evidence presented at the evidentiary hearing regarding the credibility of Sandoval's testimony is not relevant to a showing of actual innocence under Schlup.

Respondent asserts that a habeas court may not revisit the jury's credibility determinations. See Marshall v. Lonberger, 459 U.S. 422, 434 (1983) ("28 U. S. C. § 2254(d) gives federal habeas courts no license to redetermine credibility of witnesses whose demeanor has been observed by the state trial court, but not by them."). This broad proposition however stands in direct contravention to relevant Supreme Court pronouncements on the Schlup actual innocence gateway. "Because a Schlup claim involves evidence the trial jury did not have before it, the inquiry requires the federal court to assess how reasonable jurors would react to the overall, newly supplemented record. If new evidence so requires, this may include consideration of the credibility of the witnesses presented at trial." House, 547 U.S. at 538 (citing Schlup, 513 U.S. at 330).60 Like the court in Nickerson, under the Supreme Court's instruction in Schlup and House, this Court shall examine Sandoval's overall credibility.

The instant case illustrates the Supreme Court's foresight in these regards. We do not know whether the jurors found Sandoval to be credible or incredible. It is possible that they convicted Petitioner despite finding Sandoval to be unworthy of belief. The strength of the arson and MPD evidence may have been so strong that Sandoval's testimony, if of any value, was little more than filler. Now, the universe of relevant factors to be considered by the jury has changed drastically, perhaps beyond recognition. Scientific evidence has been stipulated away and Sandoval's testimony is key. The case stands or falls in large part on the credibility of her identification testimony. This Court must evaluate it in the present context, and not defer to the jury's earlier evaluation in a wholly different context.

Respondent, citing Sistrunk v. Armenakis, 292 F.3d 669, 674-675 (9th Cir. 2002), also notes that some inaccuracies in portions of a witness's testimony do not necessarily prove the witness untruthful. The Court agrees. However, Sistrunk is distinguishable. In Sistrunk, the Petitioner attempted to show an expert witness had lied. In explaining how she had he arrived at the conclusion, the expert stated that young children "never" make false allegations. This was not true. Six of 1,293 allegations at issue were falsely made by pre-teen children. Sistrunk, 292 F.3d at 674-675. The court in Sistrunk found that this minor overstatement did not call into question the expert's testimony and ultimate conclusion, and further there was other sufficient evidence to support Petitioner's guilt. Id. at 673-77.

Sandoval's testimony suffers from more than a small inaccuracy. On its face it repeatedly lends itself to incredibility. It may not be unreliable, but it has many earmarks of unreliability, and it certainly is much more unreliable than the testimony in Sistrunk.

It is remembered, too, that the Schlup standard does not require absolute certainty about the petitioner's guilt or innocence. It seeks only to determine if a petitioner's case is so extraordinary that a meaningful avenue must be provided to avoid a manifest injustice. Schlup, 513 U.S. at 327; House, 547 U.S. at 538. "[W]here post-conviction evidence casts doubt on the conviction by undercutting the reliability of the proof of guilt, but not by affirmatively proving innocence, that can be enough to pass through the Schlup gateway to allow consideration of otherwise barred claims." Lee, 653 F.3d at 938; Sistrunk v. Armenakis, 292 F.3d 669, 673 (9th Cir. 2002) (en banc).

Lisker v. Knowles illustrates that neither conclusive proof of innocence or elimination of all evidence of guilt is necessary to pass through the Schlup gateway. 463 F.Supp.2d 1008, 1009 (C.D. Cal. 2006). Lisker was found guilty in the stabbing and beating death of his mother. Id. At trial, the prosecution presented critical evidence that Lisker lied and could not, as he claimed, have discovered his mother's body when he looked through the window of her house. Id. at 1013. The prosecution argued that this evidence showed Lisker must have known the location of his mother's body and lied about his discovery of it only because he was guilty. Id. However, reconstruction of the crime scene during an evidentiary hearing demonstrated that the victim's body was visible from the window. Id. at 1019-22. Given this exculpatory evidence, the Court allowed Lisker to pass through the actual innocence gateway despite evidence that he confessed to a jailhouse informant, had conditionally plead guilty to gain placement in the California Youth Authority, and later admitted guilt at parole hearings. Lisker, 463 F.Supp.2d 1025-27; 1029-32). The court found the jailhouse informant's testimony not credible as he had offered prosecution testimony in two other murder cases and his statement of facts did not match those of the crime. Id. Further, the court noted that petitioner's confessions lacked detail and were motivated by the hope of a shorter sentence or earlier release. Despite his confessions, the court found that petitioner had made a sufficient showing of innocence to pass through the Schlup gateway. Here, the Court is also asked to analyze a claim of innocence despite the presence of testimony implicating guilt.

d. Conclusion Regarding Sandoval's Credibility Generally

This Court concludes that no reasonable juror would find that Sandoval's identification of Petitioner or his vehicle could support a finding of Petitioner's guilt beyond a reasonable doubt.

Sandoval contradicts her own earlier testimony. She has difficulty explaining what she observed. Her brief, impaired and limited view of the driver of the vehicle at a time when she was very agitated and her failure to identify Petitioner in a lineup not only renders her later identification unexplained, it leaves her a generally incredible witness. Similarly, the fact that she apparently originally identified the white RV as a beige Dodge van or Caravan, could not identify it from the back, incorrectly described its features, and demonstrated many inconsistencies in her identification testimony generates similar, albeit lesser, skepticism. The two identifications cannot be viewed independently of one another. While the Court reviewed Sandoval's identification of the driver and the identification of the RV separately, Sandoval was the source of all of the eyewitness testimony. Her testimony for both identifications derived from the same circumstances and was influenced by the same factors. Sandoval's testimony regarding the identification of Petitioner is so unbelievable, even bizarre, that her testimony must be viewed with focused skepticism when she goes on to describe other events and objects she witnessed that night. Perhaps if Sandoval had not provided such an incredible chain of events leading to her identification of Petitioner, her identification of the RV might be worthy of more consideration. That is not the case here.

Other than the inference that Sandoval may have had a motive to cooperate with the police given criminal charges pending against her, there is no reason to think she had a reason to implicate Petitioner or that she is an inherently dishonest person. Rather she presents as a witness whose various observations and the circumstances under which she formed them simply run afoul of normal human experience. Perhaps her intense agitation that night, the extreme conditions to which she was exposed as the night wore on, the constant questioning from law enforcement, the clearly suggestive circumstances, and a desire to validate that which everyone seemed to believe all combined to influence negatively her ability to retain and describe what she actually saw. Perhaps she ended up with such a garbled memory of mixed images that she lost track of the truth. The Court will never know for certain. It knows it cannot and does not credit her testimony in any measurable way.

Thus, for all the reasons set out above, it is the conclusion of this Court that reasonable jurors, carefully listening to Sandoval's testimony, the testimony of experts on eyewitness identification, and conscientiously following relevant jury instructions, would not find Sandoval's identification of Petitioner or his vehicle reliable evidence of Petitioner's guilt.

The Court will examine and evaluate the probative value of the remainder of the evidence.

3. Circumstantial Evidence

The circumstantial evidence of Petitioner's guilt is flawed and provides little support for a finding of guilt.

Respondent suggests that the eviction delay produced a man so upset with being a landlord that he was willing to, and did, destroy his own property and murder his tenants, two of whom were young innocent children. Clearly, it made no sense to do either. The tenants were soon to be gone. Significant time and money had been devoted to legal action which was about to come to fruition (after having once been delayed voluntarily by Petitioner so as not to disrupt the tenants' holidays). There appears to have been no financial rationale for destroying the house. People, especially criminals, may act bizarrely, but there is no evidence Petitioner was insane, unintelligent, or irrational. His money management practices suggest the opposite. The evidence shows he was relatively comfortable financially. He apparently would not have benefitted financially from destroying the house and collecting on his insurance policy. He would have done better selling the house on the open market. He had a prospective buyer lined up at the time of the fire. The Court does not see how a reasonable juror could conclude Petitioner had a financial motive to burn down his property.

The allegation of an argument between Mrs. Jones and Petitioner at the mobile home park is found unworthy of serious consideration. Warner's credibility was put at issue. Her claim was uncorroborated and refuted by the very person she said would corroborate it. It is difficult to imagine any circumstances that would put Petitioner and Mrs. Jones at the mobile home park at the same time. The Joneses remained as holdover tenants at the Ronald property because flooding destroyed their plans to move to the mobile home park. Why then would Mrs. Jones be at the mobile home park? Why would Petitioner expect to find her there?

The fact that Petitioner's RV was parked on the street the night before the fire is a peculiar coincidence, but reason suggests it was no more than that. Crimes are committed by individuals who think differently than law abiding people; the crime itself reflects abnormal judgment. However, one must ask why Petitioner would draw such attention to a reportedly unique vehicle if he was going to use it to commit arson-murder. (Indeed, one might ask why anyone would choose to use as a getaway vehicle a unique Winnebago and drive back and forth as many as twenty times in front of the scene of his crime.) As noted, Petitioner's girlfriend's testimony about their RV camping trip scheduled to begin the day of the fire presents a reasonable explanation for the RV being moved from its ususal parking spot.

Respondent analogizes this case to Albrecht v. Horn, 485 F.3d 103, 125 (3d Cir. 2007) in arguing that motive and opportunity evidence can outweigh evidence of an accidental cause of arson and leave the Schlup actual innocence gateway closed. (Resp. Brief at 27-28.) As here, Albrecht had presented convincing evidence that fire investigators' conclusions were not reliable. Albrecht, 485 F.3d at 121 ("Albrecht had convincingly shown that the fire science presented by the Commonwealth at his trial has since been discredited."). However, the court in Albrecht found "[t]here was sufficient other evidence that the fire was not accidental. . . ." Id. The other evidence against Albrecht was powerful and persuasive. Albrecht was extremely abusive towards his wife. Id. at 110. He was having an extramarital affair. Id. Witnesses observed "Mrs. Albrecht with her hair torn out and burn marks on her face . . . a bruise on her chest about the size of a saucer . . . and he had banged her head against the refrigerator, and she said her head was numb so she didn't feel it when he burned her face with a cigarette." Id. Four different witnesses testified that in the days before the fire Mr. Albrecht had repeatedly stated that "he was going to burn down his house" and "kill his wife," and that "he had a good lawyer, he would get away with it, [and] nobody would prove it." Albrecht, 485 F.3d at 110-111. Police found a "can, which usually held hydraulic oil for Albrecht's paving machine" in the trunk of Albrecht's car; the can "had soot on it, and tested positive for gasoline." Id. at 111. "[A] local gas station employee testified at the trial about Albrecht's attempt to purchase gasoline to put in a can the day before the fire." Id. The court found that the new fire science "did nothing to undermine the Commonwealth's damaging evidence" of his guilt. There was other ample evidence to make it more likely than not that no reasonable juror viewing the record as a whole would lack reasonable doubt." Albrecht, 485 F.3d at 125-26.

Similarly, in Ferranti v. United States, even though new scientific evidence challenged whether a clothing store fire had been intentionally set, petitioner, the store's owner, did not make a sufficient showing of innocence.61 2010 U.S. Dist. LEXIS 6116 (E.D.N.Y. 2010). The court found the weight of the other incriminating evidence against Ferranti to be "crushing." Id. at *38. Remaining evidence of guilt included: the store had not been profitable; petitioner renewed his fire insurance on this store shortly before the arson, but let insurance lapse on his other store; two witnesses testified that petitioner told them he would burn down the store if its financial condition did not improve; petitioner lied, and attempted to convince employees to lie, to investigators; and a witness testified that petitioner's co-conspirator indicated that he had set the fire for petitioner. Id. at *38-42. Given this evidence, the court concluded Petitioner had not made a showing of innocence despite the lack of scientific proof of arson.

Albrecht, Ferranti and Petitioner's cases involve alleged arson. The similarities end there. Petitioner made no overt threats or admissions, did not engage in ongoing criminally violent behavior, is not linked to the crime by any physical evidence, and is not implicated by the confession of a co-conspirator.

Other circumstantial evidence in this case is inconsistent with guilt. The engine of Petitioner's RV was cold to the touch on the morning of the fire. Police never recovered a shirt resembling the one described by Sandoval. Having so quickly identified Petitioner as the suspect, other leads were left unpursued. The Court finds these circumstances more meaningful than those argued as showing guilt.

On balance the Court is unable to find much in the way of motive or other circumstantial evidence that could measurably suggest Petitioner's guilt.

4. Conclusion: Actual Innocence Exception Met

Petitioner originally was convicted in large part because scientific evidence showed that an arsonist started the fire, scientific evidence put Petitioner at the scene of the fire, Sandoval identified Petitioner as the man repeatedly driving his RV suspiciously in front of the crime scene, and the prosecution presented a myriad of other evidence it argued showed Petitioner had a motive to commit the crime.

Petitioner has presented new, reliable, exculpatory, scientific evidence of his potential innocence. Schlup, 513 U.S. at 324. Respondent has stipulated to the accuracy of that new evidence. The parties now agree that no one can determine that the fire was intentionally set. They agree that substances on Petitioner's shoes were not the same as those found in the fire. We are left with Sandoval's identification testimony and a variety of circumstances, or inferences, some of which are claimed to be consistent with guilt, others the opposite.

This proceeding is to determine how reasonable jurors would have reacted if the prosecution had not presented the arson and MPD evidence and the defense had presented in rebuttal all the evidence available to it.62 The Court has undertaken the specific task of determining whether the remaining evidence would enable a reasonable and conscientious juror to find Petitioner guilty beyond a reasonable doubt.

As noted, the Court is unable to see how the purported motive evidence supports a finding of guilt. The Court has already concluded that Sandoval's identification of Petitioner as the driver of the RV is so incredible and so likely the result of suggestive events occurring after the crime as not to be deserving of any weight.

Ultimately then, Petitioner's guilt hinges largely on the credibility of Sandoval's identification of Petitioner's RV. The evidence might cause one to hesitate before concluding there was no possibility Petitioner committed this crime. It is, however, not necessary that the evidence conclusively exonerate him. The test is whether any reasonable juror properly instructed and conscientiously following these instruction would be more likely than not to find the Petitioner guilty beyond a reasonable doubt. Schlup, 513 U.S. at 327; House v. Bell, 547 U.S. at 537-538.

For all the reasons set forth above, the Court answers the latter question in the negative. The evidence remaining after the scientific evidence was removed is so weak it is insufficient to support a finding of Petitioner's guilt beyond a reasonable doubt. Petitioner's showing of innocence leaves this Court without confidence in the outcome of Petitioner's trial. Schlup, 513 U.S. at 316. The Court concludes Petitioner should pass though the Schlup gateway.

VII. STATUTORY TOLLING CLAIM

A. Introduction

Petitioner claims that he is entitled to statutory tolling with regard to his substantive actual innocence claim.63 The Court has determined that Petitioner need not show diligence to present his actual innocence gateway claim. However, should a diligence requirement exist, Petitioner may show diligence based on the same facts that he was diligent in presenting his substantive actual innocence claim.

B. Factual Summary; Issue

Petitioner's substantive actual innocence claim is based on newly discovered scientific evidence demonstrating that the MPD's from the fire scene and the MPD on Petitioner's shoes were not from the same source. It is undisputed that Petitioner did not discover this evidence, the factual predicate for this claim, until September 2005, when John Lentini, a scientist then employed with Applied Technical Services, for the first time used refined scientific techniques to analyze the underlying MPD data from Petitioner's case. (Decl. of John J. Lentini, Dec. 11, 2011, ECF No. 102-4.)

Lentini had been hired by Petitioner in 1997 to analyze chemical samples relating to the arson scene. (Id., ¶ 12.) At that time, Lentini tested the samples using then-applicable national standards, specifically American Society for Testing and Materials (ASTM) E1412 ("Standard Practice for Separation of Ignitable Liquid Residues from Fire Debris samples by Passive Headspace Concentration with Activated Charcoal") and ASTM E1618 ("Standard Test Method for Ignitable Liquid Residues in Extracts from Fire Debris Samples by Gas Chromatography-Mass Spectrometry (GC_MS)"). (Id., ¶ 13.) Upon testing, Lentini came to the same conclusion as Sarah Yoshida, the state chemist for the prosecution, i.e., that the MPD on Petitioner's shoes could not be differentiated from the MPD on the samples from the fire scene. (Id., ¶ 15.)

Then, in September 2005, Lentini was contacted by Petitioner's sister who, convinced of her brother's innocence, was exploring possible means of trying to establish that innocence. She asked Lentini again about the chemical analysis in Petitioner's case. (Id., ¶ 17.) Lentini reviewed his results and determined, for the first time, that there was a method he could use to make distinctions among the various compounds found within the class of chemicals comprising medium petroleum distillates. Using this method for the first time in November 2005, Lentini concluded that he could unequivocally distinguish the MPDs found at the fire scene from the MPD found on Petitioner's shoes. (Id., ¶ 18, 20.) Lentini provided a report of his results to Petitioner's counsel in December 2005.

Petitioner incorporated Lentini's new findings in an amended petition for writ of habeas corpus filed with the California Supreme Court on December 13, 2005.

Respondent agrees that Petitioner did not become aware of the new testing methods until Lentini was contacted in 2005. (Undisputed Fact 15.) Further, Respondent does not dispute that Lentini is correct or that the MPDs in the shoes and those from the fire did not originate from a common source. (Undisputed Fact 13.)

Respondent also acknowledges that Petitioner's substantive actual innocence claim would be deemed filed within the one year statute of limitations of AEDPA if the statue were tolled based on newly discovered evidence.

Accordingly, the only remaining question is whether Petitioner was reasonably diligent in discovering this new evidence, i.e., the factual predicate under 28 U.S.C. § 2244(d)(1)(D).

C. Petitioner's Diligence

1. Testimony of Lentini

Lentini testified at the evidentiary hearing that he was not able to differentiate the MPDs found on the shoes until he re-examined the data in 2005. Lentini acknowledged that the mechanics by which gas chromatography-mass spectrometry (GC-MS) could be used to distinguish MPDs did not change over the intervening years, but testified that prior to 2005 he had not devised the reasoning that enabled him to use those mechanics to make the distinction and, to his knowledge and belief, no else had either. Specifically, contpresented one emplating the fundamental principle of chemistry that small molecules evaporate more easily than larger ones, Lentini reasoned for the first time in 2005 that when exposed to heat or fire, lighter molecular weight MPD compounds would evaporate before heavier ones. (HT 106-09.) Accordingly, the MPD compounds remaining in liquid form at the fire scene should have had a greater proportion of heavier MPD compounds remaining, and thus a higher molecular weight, than those found in Petitioner's shoes which were not exposed to the heat of the fire. (Id.) Instead, he found the opposite; the MPD on the shoes were of a higher molecular weight than those from the fire scene. (Id.) Lentini retested his old samples using GC-MS and obtained the same results. (HT 108.)

Lentini had correctly followed the standard applicable to testing such samples in 1997. However, neither the standards nor any published papers described this new methodology. (HT 108-09.) Lentini intended to present a paper on the subject in early 2012 and to include a description of the methodology in a revision of one of his books64. (HT 110.)

In summary, Lentini was aware before 1997 that shoes could contain MPDs, but did not understand how to interpret the available data using the evaporation process to identify or eliminate a source of MPD at that time. (HT 113-14.) He admitted that the technology at the time of the fire enabled the differentiation, but neither he nor anyone else divined its potential significance until 2005. (HT 117.)

2. Testimony of DeHaan

DeHaan agreed that MPDs, being volatile in nature, would evaporate when exposed and that differences between types of MPDs could be identified by comparing the characteristics of the compounds through GC-MS. (HT 436.) He went on to say that MPDs could have been differentiated by visual inspection of GC-MS results at the time of testing in 1997 and then claimed that he, in fact, had personally done so before 1997.65 (HT 436-37, 440-43.)

However, on cross-examination, beyond remembering "doing cases like that" and doing so in classes he taught, DeHaan could not remember a specific instance where he had differentiated between MPDs. (HT 474-75, 477-78.) Later in his testimony he admitted that he could not recall ever having done so.66 (HT 474, 481.) DeHaan also acknowledged that, despite having searched, he could not refer to any published literature differentiating between sources of MPDs or any published protocols for doing so. (HT 477-81.)

DeHaan said it was unclear whether, at the time of Petitioner's trial, the question arose as to whether MPDs could have been differentiated. (DeHaan Decl., p.5, ECF No. 95-1.) However, he testified that Yoshida had taken his classes and thus would have learned how to visually compare and differentiate volatile compounds. He also felt that given her position, knowledge and skill level, she would have had a duty to attempt to distinguish the chemical compounds, if possible. (HT 520-22.) She did not, but instead testified she could only make a class-specific determination.67 (RT 8923-24.) This corroborates Lentini's belief that experts were not distinguishing between the compounds at that time. Regardless, at the time of trial DeHaan agreed with Yoshida that one could not determine if the MPDs had a common origin. (HT 519-20.) That is irreconcilable with Respondent's stipulation that the MPDs are chemically distinguishable and not from a common source. (Undisputed Fact 13.)

D. Conclusion as to Tolling

Petitioner asserts tolling under § 2244(d)(1)(D) based on his discovery of new MPD evidence. Petitioner must show that he was reasonably diligent in discovering the new MPD evidence to evoke such tolling under 28 U.S.C. § 2244(d)(1)(D) and present his substantive actual innocence claim under Herrera v. Collins, 506 U.S. 390 (1993).

In this regard, the Ninth Circuit stated with respect to Petitioner's case:

Souliotes, 622 F.3d at 1178 (emphasis in original).

Petitioner knew from trial that a state expert and his own defense expert, Lentini, had both concluded that the MPDs on the shoes and from the fire scene could not be distinguished. He knew that Lentini had been hired to try to refute the state's MPD evidence if possible, but he could not. Additionally, Yoshida, the prosecution's expert, had an ethical duty to fairly present the MPD evidence. (RT 522); see also, e.g., Code of Ethics, Cal. Assoc. Of Criminalists, §§ II(I); III(G). If either Lentini or Yoshida had been able to differentiate between the MPDs at the fire scene and Petitioner's shoes, it must be presumed they would have done so. They did not. Based on their results, Petitioner could reasonably have concluded that other criminalists would provide similar opinions.

Petitioner is not scientifically trained. He has been incarcerated at all relevant times. It is unclear what, if anything, he might have done to discover this information sooner. Both Lentini and DeHaan testified that there was no identifiable scientific literature differentiating between MPDs. Accordingly, up until the point that Petitioner's sister contacted Lentini, there was no reason for Petitioner to look further into challenging the results of the chemical testing. Even if Petitioner had been qualified to and had researched all available literature, none existed that would have caused him to question the scientific techniques used at trial. See Moore, 368 F.3d at 940 (explaining that a due diligence inquiry should take into account that prisoners are limited by their physical confinement). If DeHaan, a criminalist knowledgeable in the field, not subject to confinement and hired by Respondent, was not able to find any such publications or standards, it is unrealistic to believe that Petitioner could have fared better. Apparently it was only Petitioner's sister's conviction of his innocence that drove her to pursue again previously exhausted avenues.

DeHaan's claim that MPDs could have been distinguished at the time of Petitioner's trial by one properly trained is not credible. Despite having strong incentive to attempt to determine if the MPDs on the shoes and the fire scene matched, neither Petitioner's nor the prosecution's criminalist did so. If either could have done so, but did not, significant ethical concerns arise. Moreover, DeHaan was not able to identify any specific examples where MPD compounds had been differentiated. Inconsistences in DeHaan's testimony on this point erode the Court's confidence in his assertions. As previously noted, Dehaan's professional judgment has been questioned by others. (HT 503-04.)

Respondent asserts that the factual predicate of Petitioner's claim "that shoes could give off volatile residues as a result of components used in the manufacturing process" was known at the time of trial. (See Resp. Brief at 17.) This is true, and Petitioner's expert was aware of it at the time of trial. However, that is not the factual predicate of this claim. The actual predicate is the lack of a common origin in the MPD samples. These factual issues are distinct. Just because Petitioner was aware shoes could test positive for MPDs does not show that he was aware that the MPDs on the shoes could be distinguished from those at the fire scene. See Souliotes, 622 F.3d at 1179 (holding "an evidentiary hearing is necessary to determine when the scientific techniques used by Lentini in 2005 to discriminate among the MPD compounds were developed, and when such information would have become available to an inmate like Souliotes.").

Petitioner need only show that he exercised reasonable care in attempting to find the new information. Souliotes, 622 F.3d at 1178. From the testimony presented, Petitioner could not have discovered the new evidence any earlier that he did, and he did not delay in presenting it to the California Supreme Court. Petitioner was diligent in discovering the new evidence. He is entitled to statutory tolling with regard to his substantive actual innocence claim.

VIII. SUMMARY AND CONCLUSION

After weighing the trial evidence with that presented at the evidentiary hearing, this Court lacks confidence in the outcome of Petitioner's trial. The Court concludes that having had the opportunity to consider this same evidence, "no reasonable juror would [have found Petitioner] guilty beyond a reasonable doubt." House, 547 U.S. at 538. Petitioner is entitled to be equitably excepted from the untimely filing of his federal habeas petition.

It is recommended that Petitioner pass through the Schlup gateway and the Court proceed to consider the merits of his underlying claims.

Petitioner also is entitled to statutory tolling under 28 U.S.C. § 2244(d)(1)(D) with regard to his substantive actual innocence claim.

IX. RECOMMENDATION

Accordingly, IT IS HEREBY RECOMMENDED that the Court find that Petitioner has made a sufficient showing of actual innocence to serve as an equitable exception to the one year statute of limitations set forth by AEDPA.

It is FURTHER RECOMMENDED that the District Court allow Petitioner to proceed to the merits of his claims and refer the matter to the Magistrate Judge for further adjudication.

This Findings and Recommendation is submitted to the assigned District Judge, pursuant to the provisions of Title 28 U.S.C. § 636(b)(1) and Local Rule 304. Within fourteen (14) days after being served with the Findings and Recommendation, any party may file written objections with the Court and serve a copy on all parties. Such a document should be captioned "Objections to Magistrate Judge's Findings and Recommendation." Any reply to the objections shall be served and filed within fourteen (14) days after service of the objections. The parties are advised that failure to file objections within the specified time may waive the right to appeal the District Court's order. Martinez v. Ylst, 951 F.2d 1153 (9th Cir. 1991).

IT IS SO ORDERED.