United States Court of Appeals
                              For the Eighth Circuit
                          ___________________________

                                  No. 16-3391
                          ___________________________

                               United States of America

                         lllllllllllllllllllll Plaintiff - Appellee

                                            v.

                                    Johnny Madison

                        lllllllllllllllllllll Defendant - Appellant
                                        ____________

                    Appeal from United States District Court
                  for the Northern District of Iowa - Sioux City
                                 ____________

                               Submitted: May 8, 2017
                                Filed: July 20, 2017
                                   ____________

Before RILEY, BEAM, and SHEPHERD, Circuit Judges.
                            ____________

RILEY, Circuit Judge.

     The district court1 sentenced Johnny Madison to 140 months in prison after
Madison pled guilty to two counts related to methamphetamine distribution. See
21 U.S.C. §§ 841(a)(1), (b)(1)(B)(viii), 846. Madison appeals his sentence and

      1
        The Honorable Leonard T. Strand, Chief Judge, United States District Court
for the Northern District of Iowa.
contends the district court erred in three ways by: (1) overstating the amount and
purity level of methamphetamine attributed to him; (2) crediting testimony suggesting
he possessed a firearm; and (3) imposing a substantively unreasonable sentence. We
affirm. See 28 U.S.C. § 1291 (appellate jurisdiction).

I.     BACKGROUND
       Madison entered into a conspiracy to distribute methamphetamine at some
point in 2014 or early 2015. Law enforcement became aware of this fact in March
2015, when a confidential source reported Madison was dealing methamphetamine
out of his house in Sioux City, Iowa. Officers then executed a controlled buy from
Madison on July 1, when the confidential source paid $1,200 for 24.5 grams of
methamphetamine that was later tested and shown to be 99.1% pure. The confidential
source contacted Madison again about two weeks later, this time looking to buy two
ounces (about 56.7 grams) of methamphetamine. Madison agreed to the sale, but said
he needed another day to acquire the full amount. The deal never took place,
however, and the investigation continued.

       Officers arrested Madison on state drug charges in late July. Madison’s wife
(now ex-wife), Kelli, spoke with officers on July 29, and named Josh Navrkal and
Daniel Suhr as two of Madison’s suppliers. Madison agreed to a proffer interview
in early September. The interview was not recorded and the parties disagree about
what Madison said.2 There is no question Madison admitted to purchasing a total of
141.75 grams of methamphetamine from Navrkal, and 56.7 grams from another
supplier, Seid Tilahun. According to the police, Madison also admitted to purchasing
a quarter pound of methamphetamine from Suhr on three separate occasions—340.2
grams in all. Madison denies saying any such thing. Suhr, who was also questioned


      2
       According to Sergeant Troy Hansen, the Sioux City Police Department often
did not record proffer interviews. No doubt there are good reasons for this practice,
though we note a video certainly would be helpful in cases like this.

                                         -2-
by the police at one point, said he only sold Madison about 42.5 grams of
methamphetamine, an estimate Madison “agrees with.” There is no indication
Madison said in the interview how pure any of this methamphetamine was, and these
specific quantities were never seized or tested for purity.

       Madison was indicted in federal court in November 2015 for distributing and
conspiring to distribute five or more grams of pure methamphetamine. See 21 U.S.C.
§§ 841(a)(1), (b)(1)(B)(viii), 846. He pled guilty to both counts. The United States
Probation Office prepared a Presentence Investigation Report (PSR), which attributed
519.1 grams of actual methamphetamine to Madison based upon his reported
admissions from the proffer interview and the lab tests of methamphetamine seized
from Madison’s suppliers. This put Madison slightly over the 500-gram threshold
and raised his base offense level to 34. See United States Sentencing Guidelines
(Guidelines or U.S.S.G.) § 2D1.1(c)(3). The PSR recommended a three-level
reduction for acceptance of responsibility, which resulted in a total offense level of
31 and an advisory Guidelines range of 168 to 210 months. Madison objected to the
PSR, arguing the gross quantity of methamphetamine was too much and the purity
rate too high. Madison also moved for a downward variance.

       The government presented evidence to support the PSR’s calculations at the
sentencing hearing. Sergeant Troy Hansen testified about the procedure for proffer
interviews, generally—officers present for the interview take notes, and then one
officer is responsible for converting those notes into a report that must be drafted
within five days of the interview. Here, the post-proffer report supported the
challenged paragraphs in the PSR regarding gross quantity. Although Sergeant
Hansen did not write this particular report, he was present for the interview and
confirmed the summary was consistent with the notes he and his two fellow officers
had taken. Sergeant Hansen also stated law enforcement seized and tested various
methamphetamine samples from Tilahun, Navrkal, and Suhr, which served as the
basis for the high average purity rates used in the PSR.

                                         -3-
      Kelli testified for the government as well. She began by describing multiple
instances of abuse during her tumultuous relationship with Madison. After one such
incident, Kelli called the police when she found a rifle between the mattresses in
Madison’s bedroom. Apparently, the police seized the weapon, and Madison later
admitted to Kelli the gun was his. Madison was never charged despite being a felon
who could not legally possess a gun. Kelli then shared what she knew about
Madison’s brief history as a drug dealer. In addition to her general awareness that
Madison bought, sold, and used methamphetamine, Kelli was present for two
purchases Madison made from Suhr. On both occasions Madison paid “around a
thousand” dollars for a quantity that was about the size of a “tennis ball,” figures
consistent with an ounce (about 28.3 grams) of methamphetamine. That said, Kelli
made clear she was not around every time Madison met with Suhr.

       The district court found the government “pretty easily establishe[d]” its burden
to prove Madison was responsible for 519.1 grams of pure methamphetamine,3 and
considered the method used to calculate the purity level to be “perfectly appropriate.”
Thus, the district court adopted the PSR and agreed the advisory Guidelines range
was 168-210 months. As for any variance, the district court recognized there were
“a lot of moving parts in this case.” There were mitigating factors like Madison’s
drug and gambling addictions, “history of some mental illness,” and supportive
family. But there were also aggravating factors like Madison’s gun possession, “a
long career of committing crimes,” and the credible allegations of domestic assault
made against him. To the district court, these factors “balance[d] each other out.”
Yet the district court varied down anyway, reasoning the overall quantity calculation
was based on “very imprecise” (but sufficient) evidence and fearing a within-
Guidelines sentence would be “too severe” given how close Madison was to the 500-
gram threshold that had raised his base offense level. The district court gave Madison


      3
      This amount does not include the 24.5 grams from the controlled buy or the
two “ball size” quantities Kelli said she saw Madison purchase from Suhr.

                                         -4-
“the benefit of the doubt” and sentenced him to 140 months in prison—28 months
below the bottom of his advisory Guidelines range. Madison appeals.

II.    DISCUSSION
       A.     Drug Quantity and Purity
       Madison argues the government did not meet its burden to hold him
responsible for 519.1 grams of pure methamphetamine and the district court erred in
concluding otherwise. See United States v. Houston, 

, 878 (8th Cir.
2003) (“The government bears the burden of proving drug quantity by a
preponderance of the evidence.”). The district court’s conclusion as to drug quantity
is a factual one we review for clear error, and “we will only reverse when the entire
record definitely and firmly illustrates that the lower court made a mistake.” United
States v. Marshall, 

, 894 (8th Cir. 2005). For quantity calculations, a
clear error exists when there is insufficient evidence or a misapplication of the law.
See United States v. Walker, 

, 420-21 (8th Cir. 2012). Conversely, clear
error does not exist just because the district court relied on imprecise information,
provided the evidence “has sufficient indicia of reliability to support its probable
accuracy.” U.S.S.G. § 6A1.3(a); accord 

(“A sentencing
court may ‘determine drug quantity using imprecise evidence, so long as the record
reflects a basis for the court’s decision.’” (quoting United States v. Bradley, 

, 1126-27 (8th Cir. 2011))).

       Madison’s argument on this point is twofold. First, Madison claims he only
acquired about 42.5 grams of methamphetamine from Suhr, not 323.2 grams as found
by the district court. The primary piece of evidence supporting the district court’s
finding is the post-proffer report, which states Madison admitted to paying Suhr
between $3,200 and $3,600 for a quarter pound (about 113.4 grams) of
methamphetamine on at least three occasions. Sergeant Hansen corroborated the
report, noting it was consistent with the notes taken during the interview. According
to Madison, this evidence is “contradicted by several other more reliable pieces of

                                          -5-
information.” For instance, Suhr told officers he only sold Madison a total of 42.5
grams of methamphetamine, an “estimate” Madison later “adopted” as his own.
Other evidence showed Madison routinely dealt in quantities smaller than a quarter
pound.

       These conflicting accounts required the district court to decide which evidence
was more credible: the post-proffer report and Sergeant Hansen’s testimony, or Suhr’s
statement and such evidence that Madison often dealt in smaller quantities. The
district court credited the government’s evidence, and thus found Madison admitted
to the quarter-pound transactions with Suhr. We decline Madison’s invitation to
overrule that finding. To begin, a district court’s credibility determinations are
“virtually unreviewable on appeal.” United States v. Symonds, 

, 936
(8th Cir. 2001). We do not agree with Madison that his evidence is “more reliable”
or compelling. As for Suhr’s comment, a court is not obligated to disbelieve evidence
of a defendant’s prior admission in favor of a subsequent self-serving statement by
a third-party that the defendant then chooses to adopt. See, e.g., United States v.
Hicks, 

, 998 (8th Cir. 2005) (affirming reliance on a previous admission
where the district court “did not find [the defendant’s] changed account credible”);

(finding a quantity determination “well-supported”
when based on an officer’s testimony about the defendant’s prior admission, even
though the defendant later testified the amount was “overstated”). The evidence
Madison often dealt in smaller amounts fares no better, because it does not foreclose
the possibility Madison also purchased larger quantities from time to time. Once the
district court found Madison did indeed admit to the large transactions with Suhr, the
district court was then entitled to rely on that admission. See 

(“When a defendant makes admissions regarding drug quantity, a court may rely on
the admissions to establish the base offense level.”).

      Second, Madison contends the unrecovered drugs “should have been treated
as methamphetamine mixture,” in which case his base offense level would have been

                                         -6-
30, not 34. See U.S.S.G. § 2D1.1(c)(5). This contention is based on the fact law
enforcement never seized or tested any of the methamphetamine he admitted to
purchasing, i.e., there was no direct evidence of purity. However, officers did seize
and test quantities from each of Madison’s suppliers, with every stash having a purity
rate of 95% or higher, i.e., there was indirect evidence of purity. The PSR and district
court relied on these tests and applied the average purity rate for each dealer to the
amounts Madison obtained from Tilahun, Navrkal, and Suhr, respectively.

       The circumstances here are relatively common in drug-conspiracy cases—law
enforcement often is unable to recover all the methamphetamine a defendant is
deemed responsible for at sentencing. As a result, our rules on determining purity for
unrecovered quantities are well-established: “The government may prove the total
quantity of actual methamphetamine in a series of transactions by testing the purity
of a seized quantity and applying the percentage of actual methamphetamine in the
tested quantity to the unrecovered quantities.” 

; see also
United States v. Mesner, 

, 852-53 (8th Cir. 2004) (“To calculate the
actual methamphetamine quantity, the district court was required to apply the
percentage of actual methamphetamine found in the methamphetamine seized from
[the defendant] to the unrecovered . . . methamphetamine he acknowledged.”). Even
if there were no quantity recovered at all, the government can still prove purity
through circumstantial evidence, like “an expert’s testimony as to the normal purity
of methamphetamine produced in a lab.” 

; see also 

(“[W]e have consistently rejected arguments demanding direct
evidence of drug identity, quantity, or purity.”).

      The district court’s approach here is in line with these rules. The government
provided lab results for over 315 grams of methamphetamine seized from Madison,
Tilahun, Navrkal, and Suhr. These 315 grams came from ten separately seized
quantities, one from Madison and nine from his suppliers. None of the



                                          -7-
methamphetamine had a purity rate below 95%.4 We agree with the district court
these tests reflect “sufficient sample sizes” that could be used as the bases for
calculating and using average purity rates. See 

, 852-53;

; cf. United States v. Fairchild, 

, 778 (8th Cir.
1999) (acknowledging lab results of an admittedly “small portion” of drugs “standing
alone may be insufficient” to establish purity). The government also elicited expert
testimony that methamphetamine from dealers in Sioux City is often “97, 98, 99
percent pure,” a form of circumstantial evidence that bolsters the inferences drawn
from the tests. See 

. The district court did not clearly err
in finding Madison responsible for more than 500 grams of pure methamphetamine.

       B.    Firearm Possession
       Madison also claims the district court clearly erred in crediting Kelli’s
allegation that Madison possessed a rifle, and then considering that fact when
fashioning his sentence. See United States v. Hairy Chin, 

, 402 (8th Cir.
2017) (per curiam) (“A significant procedural error can occur if the district
court . . . selects a sentence based on clearly erroneous facts.”). As Madison
emphasizes, Kelli’s testimony was the only evidence regarding the gun—the
government did not offer any corroborative evidence, nor did Madison present any
contradictory evidence. That makes this a credibility issue, and the district court had
to decide whether it believed Kelli’s testimony. The district court expressly found her
testimony credible, both generally and in specific regard to the rifle. Given our oft-
quoted precedent on this issue—a district court’s credibility determination is
“‘quintessentially a judgment call and virtually unassailable on appeal,’” United
States v. Rodriguez, 

, 938 (8th Cir. 2013) (quoting United States v.
Quintana, 

, 702 (8th Cir. 2003))—the district court did not clearly err

      4
       The following is a breakdown of the amounts seized and average purity rates
used in the PSR: from Madison, one test of 24.5 grams, 99.1% pure; from Tilahun,
one test of 55.6 grams, 98% pure; from Navrkal, four tests of 141.2 grams total, 99%
pure; and from Suhr, four tests of 96.1 grams total, 95% pure.

                                         -8-
in crediting and considering Kelli’s testimony about the rifle when determining
Madison’s sentence.

       C.    Substantive Reasonableness
       Lastly, Madison posits his 140-month sentence is substantively unreasonable
for “a low-to-mid-level drug dealer,” especially given “relevant mitigating factors”
and “flaws” in the Guidelines for methamphetamine offenses. See 18 U.S.C.
§ 3553(a). We review a sentence’s reasonableness under a “highly deferential”
abuse-of-discretion standard. United States v. Roberts, 

, 992 (8th Cir.
2014). The district court imposed a sentence 28 months below the bottom of
Madison’s advisory Guidelines range. We have declared it “nearly inconceivable”
to imagine a case where the district court imposed a below-Guidelines sentence and
“abused its discretion in not varying downward still further.” United States v.
Lazarski, 

, 733 (8th Cir. 2009). This case is not an exception to that
general rule. The district court thoughtfully weighed the appropriate factors, and
even accepted some of Madison’s arguments for leniency (just not to the extent
Madison would have liked). Madison’s sentence is not substantively unreasonable.

III.   CONCLUSION
       The district court did not clearly err in any of its factual determinations, nor did
the district court abuse its discretion in sentencing Madison to 140 months in prison.
Affirmed.
                         ______________________________




                                           -9-