UNITED STATES of America, Plaintiff-Appellee,
v.
Sergei Paul LUDWIG, Defendant-Appellant.

*1245 Thomas A. Fleener, Fleener & Vang, LLC, Laramie, WY, for Defendant-Appellant.

Steven K. Sharpe, Assistant United States Attorney (Christopher A. Crofts, United States Attorney, with him on the brief), Cheyenne, WY, for Plaintiff-Appellee.

Before MURPHY, TYMKOVICH, and GORSUCH, Circuit Judges.

GORSUCH, Circuit Judge.

Driving through Wyoming one afternoon in August, Sergei Ludwig was stopped for speeding. The trooper became suspicious more might be afoot after Mr. Ludwig *1246 offered a strange story about his travel plans. Eventually, the trooper ran a certified drug dog around the car. The dog soon alerted and in a hidden compartment the trooper found 11 pounds of drugs. Before us, as before the district court, Mr. Ludwig argues that the trooper had no lawful basis to stop him, no reason to detain him, no authority for searching his car. He questions the dog's reliability, complaining that the dog successfully identifies "seizable" quantities of illegal drugs only 58% of the time. He challenges, as well, the indictment against him and the legality of the sentence he received. In the end, we affirm.

I

A

Patrolling Interstate 80, Trooper David Chatfield spotted a car traveling at what he estimated to be ten miles over the speed limit. After his radar gun confirmed this, the trooper turned on his emergency lights and gave chase. The car, driven by Mr. Ludwig, pulled onto the shoulder of the highway. But, oddly, the car didn't stop. Instead, it continued slowly along the shoulder. Only after almost a minute did it finally come to rest.

The trooper approached the car and Mr. Ludwig rolled down his window. A strong waft of cologne "hit [the trooper] in the face"—something the trooper testified is often used to mask the smell of illegal drugs. Aplt.App. Vol. II at 306. Mr. Ludwig appeared "very, very, very nervous." Aplt.App. Vol. I at 192. So nervous that his hands were trembling and he had difficulty retrieving his wallet from his pocket.

Eventually, Trooper Chatfield asked Mr. Ludwig to follow him to his patrol car. As the trooper wrote up the speeding ticket, he asked Mr. Ludwig about his travel plans. Mr. Ludwig described himself as an "IT administrator." Aplt.App. Vol. II at 314. He said he was returning to New Jersey from California. He said his employer had sent him to California to help a different company with a "server problem." Aplt.App. Vol. II at 314. He said he chose to drive, not fly, all the way out west. Yet, he said, he had stayed in California only four days before beginning his trek back to New Jersey. Asked where he had spent last night, Mr. Ludwig said he had slept in his car. During their talk, Trooper Chatfield noticed that Mr. Ludwig declined to make eye contact and remained very nervous. Trooper Chatfield also noticed that both the registration and proof of insurance for Mr. Ludwig's car were in someone else's name.

By this time the trooper's suspicions were aroused. After he finished writing the ticket, the trooper asked Mr. Ludwig if he would answer a few more questions. Mr. Ludwig said no. Believing he had reasonable suspicion to detain Mr. Ludwig for further investigation, the trooper told Mr. Ludwig that he thought there were drugs in the car. The trooper instructed Mr. Ludwig to stand aside while he released a drug detection dog from his patrol car. Once released, the dog quickly alerted to Mr. Ludwig's vehicle. Trooper Chatfield opened the trunk and the dog alerted again. Searching the car, Trooper Chatfield spotted a recently welded metal patch that seemed to conceal a compartment. At this point, Trooper Regina Schulmeister arrived at the scene, closely followed by Lieutenant Tom Adams. Eventually, the officers found 11.3 pounds of ecstasy hidden in the compartment.

B

Finding himself under indictment for violating 21 U.S.C. §§ 841(a)(1) and (b)(1)(C), Mr. Ludwig moved to suppress the drugs found in his car, arguing that his detention and the search of his car violated the Fourth Amendment. The district *1247 court disagreed and held evidence of the drugs admissible.

After this, Mr. Ludwig discovered that all three troopers involved in his detention had vehicles equipped with video cameras. Yet, to this point, the government had produced only two videos, from the cars of Trooper Chatfield and Lieutenant Adams. Mr. Ludwig had nothing from Trooper Schulmeister's car. When Mr. Ludwig complained about the missing video, the government replied that it had been automatically deleted by a program that erases videos after a certain period of time; the government asserted that its failure to save the video before deletion was accidental. Learning all this, Mr. Ludwig filed a motion to dismiss the indictment. He also filed a second motion to suppress.

After entertaining additional evidence and argument, the district court denied both motions. When it did, Mr. Ludwig entered a conditional guilty plea while preserving his right to appeal the disposition of his suppression motion, his motion to dismiss, as well as the sentence he eventually received. Those remain the three issues now before us on appeal. We discuss them by turn.

II

First is Mr. Ludwig's challenge to the disposition of his suppression motion. Because that motion independently challenges the constitutionality of his initial stop, his extended detention, and the search of his vehicle, we address each of these matters separately. In doing so, we review legal questions de novo but view the facts in the light most favorable to the government as the prevailing party. Watson v. United States, 485 F.3d 1100, 1103 (10th Cir.2007). We also accept the district court's specific factual findings unless clearly erroneous—no easy hurdle to clear, requiring the defendant to show that the findings are more than possibly or even probably wrong but pellucidly so. Id. at 1108.

A

"[T]he decision to stop an automobile is reasonable," and so consistent with the Fourth Amendment, "where the police have probable cause to believe that a traffic violation has occurred." Whren v. United States, 517 U.S. 806, 810, 116 S. Ct. 1769, 135 L. Ed. 2d 89 (1996). Mr. Ludwig says probable cause was lacking in his case because Trooper Chatfield's radar gun was unreliable. Unreliable because of the trooper's allegedly shoddy maintenance habits. The district court, for its part, found none of this persuasive and credited the radar gun's reading. But even overlooking this, Mr. Ludwig's stop was still lawful. It's long been the case that an officer's visual estimation can supply probable cause to support a traffic stop for speeding in appropriate circumstances. United States v. Vercher, 358 F.3d 1257, 1262-63 (10th Cir.2004); United States v. Bourassa, 411 F.2d 69, 71 (10th Cir.1969). In this case, the district court found the trooper's visual estimate credible and we are given no reason to believe otherwise. To be sure, as Mr. Ludwig well notes, the eyes can deceive and the trooper's training in speed estimation dates back to his time in the police academy. But Mr. Ludwig neglects to mention that the trooper enjoyed a fine view, watching Mr. Ludwig's car approach as he was parked in the highway's median. Mr. Ludwig also fails to note the day was crystal clear and the trooper possessed 15 years' experience as a highway patrolman watching cars and estimating speeds. And Mr. Ludwig offers us no affirmative reason to think that the trooper forgot his training or that his estimate should be discredited for any other reason. In these circumstances, the district court's factual finding about the reliability of the trooper's visual estimation *1248 remains untouched, must be affirmed, itself sufficient to support the traffic stop.

B

Even if his stop was lawful, Mr. Ludwig says his continued detention after the trooper finished writing the ticket was not. For its part, the government concedes that Trooper Chatfield detained Mr. Ludwig beyond the time it took to issue a traffic ticket. Neither does (nor can) the government argue that Mr. Ludwig consented to this continued detention. So in order to justify Mr. Ludwig's continued detention after Trooper Chatfield finished writing the ticket—for that detention to be "reasonable" for Fourth Amendment purposes—it falls to the government to show that it had reasonable suspicion to believe criminal activity may be afoot. Terry v. Ohio, 392 U.S. 1, 30, 88 S. Ct. 1868, 20 L. Ed. 2d 889 (1968). In deciding whether this standard is met, we are told we must examine the "totality of the circumstances." United States v. Arvizu, 534 U.S. 266, 273, 122 S. Ct. 744, 151 L. Ed. 2d 740 (2002). Several considerations, in combination, lead us to hold the reasonable suspicion standard satisfied here.

First, Mr. Ludwig didn't stop after the trooper signaled him to pull over. Instead, he moved onto the shoulder of the highway but then, for no reason compelled by traffic conditions, continued to drive for about "a quarter mile to a half a mile." Aplt.App. Vol. II at 303. In all, Mr. Ludwig took 44 seconds to stop after crossing the shoulder. The trooper testified that he thought this behavior "unusual." Id. Recognizing as much ourselves, this court has repeatedly held that a driver's failure to stop his vehicle promptly is a factor that can contribute to reasonable suspicion of criminal activity. See United States v. Villa-Chaparro, 115 F.3d 797, 799, 802 (10th Cir.1997); United States v. Elkins, 70 F.3d 81, 83 (10th Cir.1995); United States v. Walraven, 892 F.2d 972, 973, 975 (10th Cir.1989).

Mr. Ludwig responds that he stopped faster than any of the defendants in these other cases. And that may be so. But he overlooks this court's decision in United States v. Hunnicutt, where we factored into the reasonable suspicion analysis the fact that it took "ten to twelve seconds over the course of a half of a mile" for the defendant to stop his car in response to the officer's lights—a shorter period of time than it took Mr. Ludwig to stop. 135 F.3d 1345, 1347, 1349-50 (10th Cir.1998). Given this precedent, and even assuming that Mr. Ludwig's behavior wasn't enough by itself to justify his prolonged detention, we can't say it contributes absolutely nothing to a reasonable suspicion analysis, one that commands us always to consider the totality of the circumstances.

Second, Trooper Chatfield noticed an overpowering smell of cologne when Mr. Ludwig rolled down his window. The trooper testified, based on his experience, that strong masking odors are often used by drug couriers to hide illegal drugs. And, though the use of cologne or perfume is certainly consistent with lawful activity, our cases have acknowledged that it is "commonly used to mask the odor of drugs" and so can contribute to a reasonable suspicion calculus. United States v. Ortiz-Ortiz, 57 F.3d 892, 895 (10th Cir. 1995); see also United States v. Salzano, 158 F.3d 1107, 1114 (10th Cir.1998); United States v. Stone, 866 F.2d 359, 362 (10th Cir.1989).

Mr. Ludwig contends that Trooper Chatfield simply lied about what he smelled. As proof, Mr. Ludwig points us to Lieutenant Adams's statement that he didn't smell anything. But this "proof" conflates two very different things from two very different times. As the district court noted, Lieutenant Adams's statement *1249 refers to his failure to detect the smell of marijuana later when he was shown the secret compartment in Mr. Ludwig's vehicle. It does not in any way refer to, much less contradict, Trooper Chatfield's testimony that he smelled cologne inside the passenger compartment of Mr. Ludwig's vehicle when he first arrived on the scene.^[1]

Third, Mr. Ludwig was driving a vehicle registered to a third party who wasn't present. This is a factor we "have often held" may "indicat[e] a stolen vehicle or drug trafficking." United States v. Olivares-Campos, 276 Fed.Appx. 816, 821 (10th Cir.2008) (unpublished); see also United States v. Turner, 928 F.2d 956, 959 (10th Cir.1991). Mr. Ludwig asks us to hold that this factor becomes relevant only when the driver cannot provide details about the owner; and here, he says, he did supply such details. We decline to—and cannot—adopt Mr. Ludwig's rule. In United States v. Turner, we found it enough to contribute to reasonable suspicion that the "defendant was driving a car that was not registered to him or to his passenger." 928 F.2d at 959. That precedent requires us to hold that this factor is at least relevant to the reasonable suspicion analysis.

Fourth, and likewise, Mr. Ludwig's account of his travel was suspect. He claimed that he was called out to San Jose—a hub of the computer industry—all the way from New Jersey to fix a computer server problem. At the least, this suggested the server problem was no mere glitch but something critical, requiring specialization. Yet if this were true, it was surely curious that the San Jose company would be willing to wait for Mr. Ludwig to drive cross-country. And curiouser still that the company would refuse to pay for lodging or a rental car. Considering these oddities together, we believe, as the district court did, that Trooper Chatfield reasonably regarded Mr. Ludwig's claimed travel plans as suggesting that all was not as it seemed.

To this Mr. Ludwig answers that, so long as a traveler doesn't give contradictory or contradicted descriptions of his travel plans, those plans should not be factored into the reasonable suspicion calculus. Again, we disagree. Although contradictory or contradicted travel plans themselves very well may add something all their own to the reasonable suspicion analysis, we have never suggested that exposed lies are required. Bizarre travel plans may, by themselves, contribute to reasonable suspicion that criminal rather than innocent activity is under way. See United States v. White, 584 F.3d 935, 943, 951 (10th Cir.2009); United States v. Sokolow, 490 U.S. 1, 9, 109 S. Ct. 1581, 104 L. Ed. 2d 1 (1989) ("While a trip from Honolulu to Miami, standing alone, is not a cause for any sort of suspicion, ... surely few residents of Honolulu travel from that city for 20 hours to spend 48 hours in Miami during the month of July."); United States v. Simpson, 609 F.3d 1140, 1151-52 (10th Cir.2010) ("That [defendant] chose to drive [two and a half days] to spend a single night in Reno ... contributes to a finding of suspiciousness.").

Fifth, Mr. Ludwig was exceptionally nervous throughout the encounter. Of course, everyone gets nervous when stopped by a police officer. And for this reason we have emphasized—repeatedly— *1250 that "nervousness is of limited significance in determining reasonable suspicion." United States v. Wood, 106 F.3d 942, 948 (10th Cir.1997) (quoting United States v. Fernandez, 18 F.3d 874, 879 (10th Cir. 1994)). At the same time, this court has just as consistently held that the fact we all get nervous when stopped supplies "no reason ... to ignore" entirely evidence of unusually extreme or prolonged nervousness. United States v. Williams, 271 F.3d 1262, 1269 (10th Cir.2001). And that's what the record here suggests. Mr. Ludwig's hands trembled so much that he had difficulty reaching his wallet in his front pocket. He refused to make eye contact when answering the trooper's questions. He took long pauses before each response he offered the trooper. He never settled down but, as the district court found, remained exceptionally nervous throughout the stop.^[2]

Considering all of the above factors together and in light of our precedent, we hold that a particularized and objective basis existed for suspecting Mr. Ludwig of drug trafficking—thus permitting officers to detain him while a drug dog was deployed around the outside of his car. Neither do we think we might faithfully reach any other conclusion. In United States v. Sanchez-Valderuten, this court said that an officer had reasonable suspicion to detain the defendant for questioning based on the masking smell of air freshener in his vehicle, the defendant's apparent evasion of the officer's questions about his point of departure, and the defendant's seemingly unusual choice to travel I-70 when driving from New York to Washington state. 11 F.3d 985, 989 (10th Cir. 1993). In United States v. Turner, this court affirmed a district court's finding of reasonable suspicion based on the defendant's nervousness, the implausibility of the defendant's story to the officer, and the fact that the vehicle was not registered to the defendant or his passenger. 928 F.2d at 959. We fail to see any convincing way in which we might distinguish Mr. Ludwig's case from these or many other cases where this court has found reasonable suspicion present.

C

So the trooper had reasonable suspicion to detain Mr. Ludwig. The question remains whether and when, during this detention, the trooper gained probable cause to permit his search of the car consistent with the Fourth Amendment. It is well settled that a drug dog's sniff of the outside of a car is not itself a search for Fourth Amendment purposes and so doesn't require a showing of probable cause to justify it. See Illinois v. Caballes, 543 U.S. 405, 409, 125 S. Ct. 834, 160 L. Ed. 2d 842 (2005). But it is equally well settled that a trooper's opening of a car trunk is a search and does require probable cause to make it constitutional. By the time the trooper opened Mr. Ludwig's car, of course, the drug dog had alerted to the presence of narcotics. And a positive alert by a certified drug dog is *1251 generally enough, by itself, to give officers probable cause to search a vehicle. See United States v. Parada, 577 F.3d 1275, 1282 (10th Cir.2009).

Seeking to defeat this conclusion, Mr. Ludwig says this dog wasn't reliable. He says the dog's records— spanning some 200 pages in the record and covering 7 years' worth of data—reveal that its alerts have identified a seizable quantity of drugs only 58% of the time. This, in his view, is not enough to establish probable cause. And it surely goes without saying that a drug dog's alert establishes probable cause only if that dog is reliable. See id. at 1283. But none of this means we mount a full-scale statistical inquisition into each dog's history. Instead, courts typically rely on the dog's certification as proof of its reliability. See id.; United States v. Kennedy, 131 F.3d 1371, 1378 (10th Cir.1997) ("[W]ith a canine, the reliability should come from the fact that the dog is trained and annually certified to perform a physical skill.") (quotation omitted). After all, it is safe to assume that canine professionals are better equipped than judges to say whether an individual dog is up to snuff. And beyond this, a dog's credentials provide a bright-line rule for when officers may rely on the dog's alerts—a far improvement over requiring them to guess whether the dog's performance will survive judicial scrutiny after the fact. Of course, if a credentialing organization proved to be a sham, its certification would no longer serve as proof of reliability. But the judicial task, we hold, is so limited: to assessing the reliability of the credentialing organization, not individual dogs. And in this case there is no suggestion that the California Narcotic Canine Association, the organization that credentialed the drug dog in this case, is all smoke and mirrors.^[3]

Compelling the use of statistics to scrutinize the performance of individual dogs in every case would raise another problem. It would suggest that probable cause might be usefully defined by reference to some one-size-fits-all mathematical equation. And about this, too, we have our doubts. See Brinegar v. United States, 338 U.S. 160, 176, 69 S. Ct. 1302, 93 L. Ed. 1879 (1949) ("The rule of probable cause is a practical, nontechnical conception...."); Orin Kerr, Why Courts Should Not Quantify Probable Cause 14 (George Washington University Public Law and Legal Theory Paper No. 543, Mar. 27, 2011). Besides the difficulty of agreeing on a single number,^[4] such an enterprise would, among other things, risk diminishing the role of "judgment based on ... situation-sense," evolved over time and through case-by-case exposition. Kerr, Why Courts Should Not Quantify Probable Cause at 14. It would also ignore the fact that many factors relevant to a probable cause analysis (just like a reasonable suspicion analysis) aren't subject to easy quantification. How might one *1252 quantify the "suspiciousness" of a travel story, for example? Or the fact a defendant took an "unusual" amount of time to stop his car? More than percentages, precedent and common sense rules (like reliance on certification processes) ought to inform the probable cause analysis.

Mr. Ludwig's argument fails for still another reason. He admits that, based on historical performance, this dog's alert suggested a 58% chance of finding a seizable quantity of drugs.^[5] While we hesitate to get into the business of affixing figures on probable cause, if we were pushed to do so we would hold this to be enough. After all, probable cause doesn't require an officer's suspicion about the presence of contraband to be "more likely true than false." Texas v. Brown, 460 U.S. 730, 742, 103 S. Ct. 1535, 75 L. Ed. 2d 502 (1983); United States v. Padilla, 819 F.2d 952, 962 (10th Cir.1987); see also United States v. Garcia, 179 F.3d 265, 269 (5th Cir.1999) ("[T]he requisite `fair probability' is something more than a bare suspicion, but need not reach the fifty percent mark."); United States v. Limares, 269 F.3d 794, 798 (7th Cir.2001) ("`[P]robable cause' is something less than a preponderance."); United States v. Donnelly, 475 F.3d 946, 954 (8th Cir.2007) ("Probable cause ... does not require ... evidence demonstrating that it is more likely than not that the suspect committed a crime.") (quotations omitted). And several of our sister circuits have upheld searches involving dogs with track records on par with this one. See Limares, 269 F.3d at 798 (62% accuracy rate suffices to demonstrate probable cause); United States v. Anderson, 367 Fed.Appx. 30, 33 (11th Cir. 2010) (unpublished) (holding dog was reliable with a 55% accuracy rate); United States v. Koon Chung Wu, 217 Fed.Appx. 240, 246 (4th Cir.2007) (unpublished) ("[A]n accuracy rate of 60% is more than reliable enough for [the dog's] alert to have established probable cause").^[6]

Of course, Mr. Ludwig says that at least one problem still remains. In his view, the drug dog didn't actually alert on his own; Trooper Chatfield cued him. *1253 Aplt.App. Vol. I at 215. Before the district court, however, the government introduced the testimony of Kem Wallentine, an expert in the training and evaluation of narcotics dogs. He reviewed the traffic stop videos and found "no evidence of any clueing between [the trooper] and the dog." Aplt.App. Vol. II at 507. He further testified that the California Narcotic Canine Association had certified Trooper Chatfield and his dog twice in the twelve months before the traffic stop. And that organization, Mr. Wallentine explained, uses tests specifically designed to reveal whether a dog unconsciously cues to his handler rather than drugs. Of course, Mr. Ludwig introduced his own competing expert. And that expert testified that the trooper unintentionally prompted the dog to alert to Mr. Ludwig's vehicle. But at the end of this battle of the experts, the district court chose to credit Mr. Wallentine rather than Mr. Ludwig's expert. On appeal, we may not revisit the site of this battle, recreate it in our imaginations, and resolve it for ourselves anew. Neither is it enough for Mr. Ludwig to ask us (as he does) simply to credit his expert's conclusions rather than the government's. Instead, it is incumbent on Mr. Ludwig to show that the district court's resolution of the experts' credibility contest was not just wrong but clearly or pellucidly (and so reversibly) wrong. And this he has not done.^[7]

III

From Mr. Ludwig's motion to suppress we turn to his motion to dismiss. Mr. Ludwig argues that the government acted in bad faith when it destroyed the video footage from Trooper Schulmeister's patrol car. On this basis, he says, the Constitution required the district court to dismiss the indictment against him. Because the district court failed to do so, he says, we should.

To the extent the Constitution's due process guarantees impose a duty upon the government to preserve evidence, the Supreme Court has told us that the "duty must be limited to evidence that might be expected to play a significant role in the suspect's defense." California v. Trombetta, 467 U.S. 479, 488-89, 104 S. Ct. 2528, 81 L. Ed. 2d 413 (1984). To qualify as constitutionally material in this sense, the evidence must: (1) "possess an exculpatory value that was apparent [to the police] before the evidence was destroyed," and (2) "be of such a nature that the defendant would be unable to obtain comparable evidence by other reasonably available means." Id. at 489, 104 S. Ct. 2528. In addition, "if the exculpatory value of the evidence is indeterminate and all that can be confirmed is that the evidence was `potentially useful' for the defense," then the defendant must also show (3) "that the government acted in bad faith in destroying *1254 the evidence." United States v. Bohl, 25 F.3d 904, 910 (10th Cir.1994) (quoting Arizona v. Youngblood, 488 U.S. 51, 58, 109 S. Ct. 333, 102 L. Ed. 2d 281 (1988)).

Mr. Ludwig falters at the second step of this test. He cannot demonstrate that the deleted videotape is "of such a nature" that he "would be unable to obtain comparable evidence by other reasonably available means." Trombetta, 467 U.S. at 489, 104 S. Ct. 2528. He already possessed two other videos of the traffic stop, videos shot from the vehicles of Trooper Chatfield and Lieutenant Adams. And presumably the video from Trooper Chatfield's vehicle offered the most comprehensive account of the events that transpired, as he was the only officer present throughout the entire traffic stop.

To be sure, Mr. Ludwig complains that both recordings he possessed were inaudible. And we accept his premise that Trooper Schulmeister's video at least might have provided missing dialogue. But even assuming all this, it still isn't enough to make out a Trombetta claim. It isn't because Mr. Ludwig enjoyed at least one other means for obtaining comparable evidence—namely, by calling and questioning the witnesses to the event at his suppression hearings. In United States v. Parker, much as here, the defendant complained about a missing portion of a traffic stop video. 72 F.3d 1444, 1452 (10th Cir. 1995). We rejected the defendant's Trombetta claim, reasoning that the defendant "could have called" witnesses to the event "to adduce what the missing video tape evidence showed." Id. Exactly the same holds true here and in these circumstances we see no way we might allow Mr. Ludwig's Trombetta claim consistent with our precedent.^[8]

IV

Finally, Mr. Ludwig complains about his sentence. He says the district court should have held him eligible for the "safety valve," see 18 U.S.C. § 3553(f), or for a "minor role" adjustment, see U.S.S.G. § 3B1.2. Either way, he says, he would have and should have received a lesser sentence.

The safety valve permits the district court to grant a two level reduction of a defendant's base offense level if the defendant meets five specified criteria. The only criterion at issue in this case specifies that "not later than the time of the sentencing hearing," the defendant must "truthfully provide[] to the Government all information and evidence the defendant has concerning the offense or offenses that were part of the same course of conduct or *1255 of a common scheme or plan." 18 U.S.C. § 3553(f)(5). Mr. Ludwig says he provided truthful and complete disclosure at his sentencing hearing, but the district court expressly disagreed with this assessment. The district court explained that, at the sentencing hearing, Mr. Ludwig testified he wasn't familiar with the last name of his drug source named "Alex." Aplt.App. Vol. III at 620. This even though he had known Alex for more than nine years. And even though he admitted that Alex had trusted him enough to transport $125,000 in cash. In light of these facts, the district court found Mr. Ludwig's testimony that he didn't know Alex's true identity "stretch[ed] credulity." Aplt.App. Vol. III at 648. For his part, Mr. Ludwig offers us no reason to hold this finding not just possibly or probably but clearly wrong.

Instead, Mr. Ludwig retreats to the claim that he was at least entitled to a two-level reduction for being a "minor participant." But a minor participant adjustment applies only to "a defendant who plays a part in committing the offense that makes him substantially less culpable than the average participant." USSG § 3B1.2 cmt. n. 3(A) (emphasis added). The district court said Mr. Ludwig did not meet this definition because, unlike "average" drug couriers, Mr. Ludwig was fully aware that he was transporting $125,000 in cash to California to purchase a large amount of illegal drugs. Once again, we are offered no reason or precedent suggesting that the court's factual assessments—about the knowledge possessed by "average" couriers and Mr. Ludwig—were clearly erroneous.

The judgment of the district court is

Affirmed.

NOTES

[1] As a fall back, Mr. Ludwig reads Trooper Chatfield's testimony as saying that it wasn't the mere smell of cologne that made him suspicious, but the fact that the smell had no obvious source. And this, Mr. Ludwig says, is a lie. A lie because a photograph of his car clearly shows an air freshener hanging from the rear view mirror. Whatever the source of the smell, however, it was there and we cannot say it is irrelevant under our precedents.

[2] The trooper said one additional sign of Mr. Ludwig's unease was the extra pressure he applied when signing the traffic citation. Mr. Ludwig contends that the record does not support this particular assertion. But even assuming Mr. Ludwig is right about this much, there remains plenty of other indications he was exceptionally nervous for a prolonged period. Along the same lines, Mr. Ludwig makes much of the fact that Trooper Chatfield's arrest report incorrectly stated that there was a shirt hanging in the rear window of his vehicle and almost nothing else in the car suggesting a cross-country trip. But the district court did not rely on any of this when conducting its Fourth Amendment analysis and neither do we. Nor do we see any reason why the district court was obliged to disregard all of the trooper's testimony on the basis of these claimed errors.

[3] This is not to say that a dog's alerts are necessarily unreliable just because the dog lacks an acceptable certification. An uncertified dog's accuracy could still, in theory at least, be established by examining its training history and record for reliability. Our point is that this is a needless exercise when, as here, the dog has been certified by an organization whose bona fides are unchallenged.

[4] "In one study, 166 federal judges were asked to quantify probable cause. Their answers ranged from 10% to 90% certainty, with an average of 44.52% certainty. See C.M.A. McCauliff, Burdens of Proof: Degrees of Belief, Quanta of Evidence, or Constitutional Guarantees?, 35 Vand. L.Rev. 1293, 1327-28 (1982). See also Christopher Slobogin, Let's Not Bury Terry: A Call for Rejuvenation of the Proportionality Principle, 72 St. John's L.Rev. 1053, 1082-85 (1998) (estimating probable cause at about 50%)." Kerr, Why Courts Should Not Quantify Probable Cause at 2 n. 6.

[5] The dog's accuracy rate may actually be higher than 58%. In fact, the district court found the number to be between 72% and 80%. Mr. Ludwig arrived at his 58% figure only by excluding alerts in which police found drug residue or paraphernalia, but not seizable quantities of drugs. Yet those were instances in which the dog accurately detected the presence of contraband, so we see no basis for treating them as errors. See United States v. Bertram, 307 Fed.Appx. 214, 217 (10th Cir.2009) (unpublished) ("[A]lerts in the field that ultimately reveal no discernible drugs are not necessarily false alerts."). After all, the Fourth Amendment's probable cause standard requires a showing that there is "a fair probability that contraband or evidence of a crime will be found." Illinois v. Gates, 462 U.S. 213, 238, 103 S. Ct. 2317, 76 L. Ed. 2d 527 (1983) (emphasis added). And Mr. Ludwig doesn't (and can't) dispute that drug residue and drug paraphernalia are contraband or evidence of a crime.

[6] Mr. Ludwig rejoins that the drug dog wasn't specifically trained to detect the drug ecstasy. But this makes no difference. The dog alerted to the presence of drugs. Once he did, probable cause existed to search Mr. Ludwig's vehicle for drugs and any drugs found in that search became legally admissible. See United States v. McCranie, 703 F.2d 1213, 1218 (10th Cir.1983) (alert by explosives-sniffing dog not trained to detect drugs created reasonable suspicion that defendant's suitcase contained contraband); United States v. Outlaw, 319 F.3d 701, 704 (5th Cir.2003) ("That the suitcase the canine alerted to later turned out to contain PCP, a drug the dog was not trained to detect, simply does not vitiate the agent's reasonable suspicion under these facts."); United States v. Robinson, 707 F.2d 811, 815 (4th Cir.1983) ("[The dog's] initial detection [] was sufficient to establish probable cause for a search for controlled substances—the fact that a different controlled substance was actually discovered does not vitiate the legality of the search.").

[7] Whatever the merits of his other arguments, Mr. Ludwig says that, at the very least, the district court unfairly deprived him of the chance to elicit testimony refuting the dog's reliability. But this argument, too, fails. During the first suppression hearing, the court permitted Mr. Ludwig's attorney to cross-examine the government's drug dog expert on the question of reliability. And he points to no evidence that he was precluded from doing so. Instead, Mr. Ludwig complains about the second suppression hearing, as his effort there to proffer an expert was rejected. But the district court didn't abuse its discretion by declining to revisit the reliability issue during that second hearing. See United States v. Wiseman, 172 F.3d 1196, 1207-08 (10th Cir.1999). While the court granted Mr. Ludwig's request for another suppression hearing, it did so on a limited basis, expressly excluding further testimony on the drug dog's reliability. We discern no abuse in that decision. After all, the court had already provided Mr. Ludwig with his chance to develop the record. And Mr. Ludwig offers no reason why he was entitled to have the court reopen and redo the record on this point.

[8] Mr. Ludwig answers this by saying that, at the second suppression hearing, the district court prevented him from adequately examining Troopers Chatfield and Schulmeister, as well as Lieutenant Adams, about the conversations missing from the tapes he received. The problem with this argument is that Mr. Ludwig did not raise this specific objection to the district court during the second hearing or present it as an issue for appeal in his opening brief. Both unpreserved before the district court and unpresented in an opening appellate brief, it is deemed waived. Alternatively, Mr. Ludwig complains that the district court should have at least held a third evidentiary hearing where, he says, he might have shown the government acted in bad faith when destroying the videotape. But this wouldn't have helped him. Bad faith is a necessary but not sufficient element of Mr. Ludwig's Trombetta claim. To prevail, Mr. Ludwig must also and separately show the videotape in question was irreplaceable. Finally, Mr. Ludwig says he should've been allowed to collect and introduce certain impeachment evidence at the second suppression hearing. This evidence, he suggests, would've shown that the officers' testimony was unreliable and so not an adequate replacement for the lost video. But this argument too fails, for it wasn't adequately developed in his opening brief. See Prost v. Anderson, 636 F.3d 578, 593-94 (10th Cir. 2011).