Full Opinion

NOT FINAL UNTIL TIME EXPIRES TO FILE REHEARING MOTION AND, IF FILED, DETERMINED IN THE DISTRICT COURT OF APPEAL OF FLORIDA SECOND DISTRICT LEO SCHOFIELD, JR., ) ) Appellant, ) ) v. ) Case No. 2D18-2175 ) STATE OF FLORIDA, ) ) Appellee. ) ) Opinion filed May 27, 2020. Appeal from the Circuit Court for Polk County; J. Kevin Abdoney, Judge. J. Andrew Crawford of J. Andrew Crawford, P.A., St. Petersburg and Seth Miller, Executive Director, Innocence Project of Florida, Tallahassee, for Appellant. Ashley Moody, Attorney General, Tallahassee, and Blain A. Goff, Assistant Attorney General, Tampa, for Appellee. SALARIO, Judge. Leo Schofield appeals from a final order denying, after an evidentiary hearing, his motion for postconviction relief under rule 3.850 of the Florida Rules of Criminal Procedure.1 He has filed two prior motions under rule 3.850, both of which were denied after evidentiary hearings, with the denials being affirmed by this court. See Schofield v. State, 67 So. 3d 1066 (Fla. 2d DCA 2011); Schofield v. State, 681 So. 2d 736 (Fla. 2d DCA 1996). We affirm in this appeal as well. In 1989, Mr. Schofield was convicted of the first-degree murder of his wife and sentenced to life in prison. The victim's body was found in a canal near the intersection of State Road 33 and Interstate 4 in Lakeland. Her car was abandoned on the interstate. When the case was tried, Mr. Schofield knew that several unidentified fingerprints had been found inside his wife's car. Part of his unsuccessful defense at trial was that the person who left the fingerprints in the car murdered his wife. Mr. Schofield filed the second of his two prior motions for postconviction relief—the first motion is not relevant here—in 2009. In it, he alleged that the fingerprints from his wife's abandoned car had been matched to one Jeremy Scott, who was then and is now serving a life sentence for an unrelated murder. His motion alleged that the identification of Mr. Scott as the person who left the fingerprints was newly discovered evidence that would probably produce an acquittal on retrial. The postconviction court summarily denied the motion. After we reversed, it conducted an evidentiary hearing. During the hearing, Mr. Scott explained the presence of his fingerprints in the car by testifying that he used to drive up and down I-4 in Lakeland looking for abandoned cars, from which he would steal stereo equipment. That testimony was corroborated by evidence that he told law enforcement the same thing before he knew 1The motion is styled "Defendant's Second Amended Motion for Postconviction Relief Newly Discovered Evidence – Jeremy Scott's Confession." -2- that his fingerprints had been found inside the victim's car and before law enforcement told him that speakers and an amplifier were missing from the car. The postconviction court found Mr. Scott's testimony credible and determined that, in light of Mr. Scott's credible explanation for his fingerprints being in the car, the newly discovered evidence was unlikely to establish that he was the murderer and therefore unlikely to produce an acquittal on retrial. It rendered an order denying relief. We affirmed the postconviction court's order, relying largely on its finding that Mr. Scott's explanation for his fingerprints in the car was credible. Schofield, 67 So. 3d at 1070, 1072. We explained that even though Mr. "Scott's testimony was impeached to some extent, we must defer to the credibility determination made by the postconviction court." Id. at 1070. Because "the postconviction court's . . . credibility determinations [were] supported by competent, substantial evidence and . . . the postconviction court properly applied the law to the facts," there was "no legal basis upon which to reverse the postconviction court's ruling." Id. at 1072. Mr. Schofield filed the postconviction motion at issue in this appeal in January 2017. In it, he alleged two claims seeking a new trial based on newly discovered evidence. The first claim alleged that Mr. Scott had confessed to the murder both orally and in writing. The second claim alleged that, over a decade earlier, Mr. Scott told an inmate that he roomed with in prison that he had committed the murder. The postconviction court conducted an evidentiary hearing, during which Mr. Scott, the prison inmate, an investigator for the State Attorney's office, and a lawyer who listened to a telephone call during which Mr. Scott confessed to the murder all testified. Mr. Scott's testimony at the evidentiary hearing was, to put it mildly, bizarre. Mr. Schofield's counsel had to have Mr. Scott declared an adverse witness and -3- ask leading questions. Mr. Scott did ultimately confess to the murder of Ms. Schofield. But then he also confessed to murdering every other person who was murdered in Polk County between 1987 and 1988. He admitted that he told Mr. Schofield's defense team that he would confess to Ms. Schofield's murder for $1000. He acknowledged that his confession was inconsistent with his prior testimony that although his fingerprints were in the car, he was not involved in the victim's murder. He testified that the oath he swore to tell the truth in court does not mean much to him. Mr. Scott explained that he was confessing to the murder because he was tired of being hauled from prison to county jail for interviews and court proceedings and just wanted to go back to prison and stay there. In that connection, he admitted that he had engaged in acts of self-mutilation and smeared excrement on jailhouse walls for the purpose of forcing authorities to move him to and from various places of confinement. That testimony was corroborated, without objection, by the investigator for the State Attorney's office, who had interviewed Mr. Scott after the filing of the motion for postconviction relief at issue in this appeal. The investigator also testified, again without objection, that during that interview Mr. Scott denied murdering the victim and instead repeated his explanation that he had broken into the victim's car to steal her stereo equipment. The investigator stated that Mr. Scott told him that he had told Mr. Schofield's counsel that he would confess for $1000 and that he often confessed to crimes or bad acts in exchange for money, property, or things of benefit within the prison system. The inmate with whom Mr. Scott previously lived testified that immediately after Mr. Scott arrived at the correctional facility, Mr. Scott asked him for help—the inmate has paralegal training—because he had just been contacted by law enforcement -4- about the murder of Mr. Schofield's wife. During that conversation, Mr. Scott confessed to murdering her. The inmate was uncooperative during cross-examination. The State's questioning revealed, however, (1) that Mr. Scott actually arrived at the facility well before law enforcement contacted him about the murder, (2) that the facts of the murder varied in important respects from the facts that the inmate testified Mr. Scott gave him, and (3) that there had been prior incidents of animosity between Mr. Scott and the inmate. For his part, during his testimony, Mr. Scott denied ever discussing the murder with the inmate. The postconviction court denied relief. In a written order, it explained that to grant a new trial on the basis of newly discovered evidence, it must find both that the evidence is newly discovered—in that it was not previously known—and that the evidence would probably produce an acquittal on retrial. With respect to each claim, the court concluded that the evidence was newly discovered but that it would not probably produce an acquittal in a new trial. As to the latter conclusion, the court summarized the evidence at Mr. Schofield's trial and the evidence presented in connection with Mr. Schofield's new trial motion and determined that the new evidence was unlikely to produce an acquittal on retrial because Mr. Scott's confession was not credible and the prison inmate's testimony was not credible. In this timely appeal, Mr. Schofield argues that the postconviction court erred in determining that the evidence was not likely to produce an acquittal on retrial. In assessing this argument, we review the postconviction court's factual findings and credibility determinations for competent substantial evidence. See Green v. State, 975 So. 2d 1090, 1100 (Fla. 2008). We review its application of the law to the facts de novo. See Preston v. State, 970 So. 2d 789, 798 (Fla. 2007). -5- To prevail on a postconviction claim based on newly discovered evidence, a defendant must prove two things: (1) that the evidence was not known to the trial court, the defendant, or counsel at the time of trial, and it could not have been known through diligence, and (2) that the evidence is of such a nature that it would probably produce an acquittal on retrial. See Preston, 970 So. 2d at 797 (citing Jones v. State, 709 So. 2d 512, 521 (Fla. 1998) ("Jones II")). This case hinges on the second prong of the test, which is satisfied when the new evidence "weakens the case against [the defendant] so as to give rise to a reasonable doubt as to his culpability." Jones II, 709 So. 2d at 526 (quoting Jones v. State, 678 So. 2d 309, 315 (Fla. 1996)). In determining whether that standard is met, the postconviction court must consider the newly discovered evidence together with all other "admissible evidence that could be introduced at a new trial." See Hildwin v. State, 141 So. 3d 1178, 1184 (Fla. 2014) (citing Swafford v. State, 125 So. 3d 760, 775-76 (Fla. 2013)). The credibility of the newly discovered evidence can be critical in determining whether it gives rise to a reasonable doubt as to the defendant's culpability. See, e.g., Hitchcock v. State, 991 So. 2d 337, 350 (Fla. 2008) (explaining that the credibility of the witnesses was "critical to the newly discovered evidence analysis"); Robinson v. State, 865 So. 2d 1259, 1262 (Fla. 2004) (describing the "critical credibility issue that arises with [newly discovered] recantation testimony"). That only makes sense. If new evidence is believable and bears on some important issue in the case, it may be likely, when considered with all of the other admissible evidence, to produce an acquittal. If the new evidence is unworthy of belief, however, it is hard to say the evidence, even if exculpatory, would probably change the result in a new trial. -6- Indeed, our supreme court has routinely approved the denial of newly discovered evidence motions when the postconviction court determines that the evidence is not credible, the credibility finding is supported by competent substantial evidence, and the court correctly applies the law to the facts. In Hitchcock, for example, a defendant convicted of first-degree murder produced testimony from a witness who said the defendant's brother confessed to her that he, and not the defendant, had committed the murder. 991 So. 2d at 349-50. Two additional witnesses also gave newly discovered evidence in the form of testimony about statements the defendant's brother made to them that implied that he was responsible for the murder. Id. at 349- 50. The postconviction court found that the testimony suffered "from an inherent lack of credibility" and was therefore inadmissible under section 90.804(2)(c), Florida Statutes (2003). Id. at 350. The supreme court affirmed the denial of relief, but without reaching the question of admissibility upon which the postconviction court decided the case.2 Id. at 350, 364. It held instead that the motion failed on the merits because the testimony of the three witnesses was not credible. Id. at 350. The court stated: Assuming without deciding that the newly discovered evidence would be admissible . . ., Hitchcock has not demonstrated that the newly discovered evidence would probably produce an acquittal or life sentence on retrial because the witnesses were not convincing. The credibility of [the witnesses] is critical to the newly discovered evidence analysis . . ., and the circuit court's finding that these witnesses were not credible is supported by competent, substantial evidence. 2The State has raised a similar argument about the admissibility of the hearsay evidence of Mr. Scott's confessions in this case. Like the supreme court in Hitchcock, we need not and do not reach that issue here. 991 So. 2d at 354 n.14. -7- Id. at 350. The court considered the testimony in light of the other evidence and concluded that "the testimony of these three witnesses, which lacked credibility and merely partially inculpated [the defendant's brother] because he expressed personal responsibility for the murder in only one of the comments, is not evidence that so 'weakens the case against [the defendant] so as to give rise to a reasonable doubt as to his culpability.' " Id. at 351 (second alteration in original). More recently, in Sweet v. State, 248 So. 3d 1060-61, 1065 (Fla. 2018), a defendant convicted of first-degree murder and related offenses filed a postconviction motion based on (1) newly discovered testimony from one of the victims who recanted her trial testimony identifying the defendant and (2) newly discovered testimony from an alleged witness stating that the defendant was not involved. After an evidentiary hearing, the postconviction court found the recantation testimony not to be credible because the victim had poor recollection, was a regular drug user, and was on friendly terms with the defendant. Id. at 1067. It found the witness incredible because he had delayed coming forward for twenty-four years, had no explanation for the delay, and gave testimony that was inconsistent. Id. at 1067-68. It denied relief on the basis that the testimony would not probably produce an acquittal on retrial. Id. at 1065, 1068. The supreme court affirmed. Id. at 1069. It concluded first that the postconviction court's determinations that the victim and witness were not credible was supported by competent substantial evidence. Id. at 1066-68. From there, it concluded that the postconviction court's ultimate decision was correct. Id. at 1068-69. It reasoned as follows: After finding that [the witnesses] were not credible, the postconviction court turned to consider the cumulative effect of all evidence that could be -8- presented at a new trial. The court considered the additional evidence presented at the evidentiary hearing . . . . After noting its previous findings that [the recantation testimony was not] credible, the postconviction court concluded that "[i]n light of all the evidence presented at the evidentiary hearing and all other evidence available," [the] newly discovered evidence and . . . recantation "would not produce a reasonable probability of a different outcome at a new trial." Thus, it is clear that the postconviction court properly applied the [newly discovered evidence] standard to Sweet's claim. Id. at 1068-69. This case is not different in any consequential respect.3 The postconviction court's findings that Mr. Scott and the inmate are not to be believed are 3The special concurrence argues that cases involving trial witness recantations are inapplicable here because when a jury renders a verdict consistent with a trial witness's testimony "we can assume that the jury has found that witness credible." Granting a new trial based on recanted trial testimony thus requires setting aside the trial jury's implicit credibility finding which, in turn, trenches on its status as sole arbiter of the facts, which is not a concern in this case because Mr. Scott did not testify at the trial. There are three problems with this argument. The first is that it has no support in the supreme court precedents regarding recanted testimony. On the contrary, the supreme court precedents regard recanted testimony with suspicion not for fear of invading the province of the trial jury, but because recanted testimony is "exceedingly unreliable"—in other words, highly incredible. Consalvo v. State, 937 So. 2d 555, 561 (Fla. 2006) (quoting Armstrong v. State, 642 So. 2d 730, 735 (Fla. 1994)); see also Armstrong v. State, 642 So. 2d 730, 735 (Fla. 1994) ("Moreover, recanting testimony is exceedingly unreliable. . . ." (quoting Bell v. State, 90 So. 2d 704, 705 (Fla. 1956))). Although Mr. Scott did not recant testimony given at trial, he did recant testimony under oath at a hearing on a prior postconviction motion, thus raising the exact same concerns of reliability. See Armstrong, 642 So. 2d at 735 (noting that reliability concerns are especially present "where the recantation involves a confession of perjury" (quoting Bell, 90 So. 2d at 705)). The second problem is that the special concurrence's theory is profoundly overbroad. It effectively reads any trial jury's general verdict of "guilty" as an implicit finding that every material statement by every State witness was believed by the jury. That sweeping assumption does not comport with the real world in which a jury, in rendering a verdict that is consistent with the evidence, may nevertheless have accepted or rejected various statements made by any witness. The third problem is that, as shown in the text, supreme court precedents that do not involve trial witness recantations analyze newly discovered evidence questions involving credibility the same way we do—i.e., holding that if a postconviction court's -9- amply supported by the evidence as described above. And although perhaps not as precisely drafted as the order in Sweet, the postconviction court's order here correctly recites the legal standard applicable to newly discovered evidence motions, describes the new evidence in detail, and, by way of block quotation from an order on an earlier postconviction motion, describes at length the evidence at trial. The court looked at the new evidence, determined that it was not credible, and determined that based on that, in the context of the evidence as a whole, the new evidence would not probably produce an acquittal on retrial. We have examined the order and the evidence under the standards applicable to appellate review of such orders and, as in Hitchcock and Sweet, have found no basis for reversal. See also State v. Riechmann, 777 So. 2d 342, 360 (Fla. 2000) (affirming denial of newly discovered evidence claim where the postconviction court determined that witness testimony was "less than credible and 'rife with inconsistencies' " and thus "would probably not have created a reasonable doubt in the minds of the jury"); Melendez v. State, 718 So. 2d 746, 748 (Fla. 1998) (affirming denial of newly discovered evidence claim where the postconviction court held that "[f]our of the five [new witnesses] were not credible witnesses and their testimony . . . falls short of the standard required to grant a retrial" because the "court properly applied the law, and its findings are supported by competent substantial evidence"); Jones II, 709 So. 2d at 521-23 (affirming postconviction court's decision that there was no reasonable probability, given the lack of the witnesses' credibility, that a new trial would have resulted in the defendant's acquittal). finding that the testimony is not credible, and is in effect unreliable, is supported by competent substantial evidence and the court correctly applies the law to the facts, then an order denying relief will be affirmed. - 10 - Mr. Schofield argues that this case is controlled by the supreme court's decision in Aguirre-Jarquin v. State, 202 So. 3d 785 (Fla. 2016), which he characterizes as requiring a new trial notwithstanding a postconviction court's determination that new testimony by a person identifying herself as the perpetrator is not credible. In Aguirre- Jarquin, a defendant convicted of first-degree murder and burglary with an assault or battery moved for a new trial based on newly discovered testimony from four people that the daughter of one of the victims of the crime confessed to them that she, and not the defendant, had committed the murders. Id. at 792-93. The postconviction court heard testimony from the four witnesses to the purported confessions. Id. It denied relief on the basis that the confessions appeared to it to be expressions of survivor's guilt or attempts to frighten people and thus that the testimony of third persons about them would not probably produce an acquittal on retrial. Id. at 790. The supreme court reversed, holding that the testimony, together with new DNA evidence, warranted a new trial and explained that the trial court's theories as to why the victim confessed could be considered by the jury at that trial. Id. at 795. Aguirre-Jarquin does not have any bearing on this case. Initially, Aguirre- Jarquin did not involve a postconviction court's determination as to the credibility of the person making the confession; the motion there was based on the testimony of alleged witnesses to confessions, not the testimony of the confessing person herself. See id. at 790. Here, in contrast, Mr. Scott testified at the evidentiary hearing and confessed on the witness stand to the murder of Mr. Schofield's wife. The postconviction court heard Mr. Scott give that confession and found, with more than enough support in the record, that it was not believable. Moreover, the postconviction court in Aguirre-Jarquin did not determine that the third persons who related the daughter's confessions were not - 11 - credible; it simply tried to explain the confessions away. Id. Here, in contrast, the postconviction court found, again with sufficient support in the record, that the inmate to whom Mr. Scott purportedly confessed more than a decade earlier was not believable. In sum, the postconviction court in this case expressly found, after hearing them testify, that both the person who confessed to the murder and the person who would relate an out-of-court confession at a new trial were not credible. Those circumstances place this case snugly within decisions like Hitchcock and Scott, which involved postconviction courts that expressly found the testimony of the relevant witnesses to be incredible. Aguirre-Jarquin does not involve such circumstances and has nothing to say about them. Based on its findings, it is clear that the postconviction court concluded that Mr. Scott’s purported confession and the inmate’s testimony as to Mr. Scott’s confession and other statements were entirely incredible and unreliable. As a result, the postconviction court determined that this evidence was not of such a nature that it would probably produce an acquittal at a retrial of this more than thirty-year-old case. The special concurrence is quite right that credibility is usually an issue for the factfinder at trial, but the analysis changes when a trial has ended in a verdict and a defendant seeks a second trial based on testimony he did not have before. As we have shown, a critical issue in that context is whether the newly discovered testimony has sufficient credibility or reliability such that it would probably produce an acquittal on retrial. See also Jones II, 709 So. 2d at 522-23 (holding that a postconviction court "properly proceeded to evaluate the credibility of [the newly discovered] testimony" and properly found the testimony of new witnesses "lacking in credibility" and, as a result, insufficient to warrant a new trial). Where, as here, competent substantial evidence supports a - 12 - postconviction court's conclusions based on the evidence before it that the new testimony would not probably produce an acquittal on retrial, affirmance of the postconviction court's order is the correct result. Affirmed. SILBERMAN, J., Concurs. ROTHSTEIN-YOUAKIM, J., Concurs specially with opinion. ROTHSTEIN-YOUAKIM, Judge, Concurring specially. I concur in the majority's conclusion that Mr. Scott's confessions do not warrant a new trial for Mr. Schofield. I am troubled, however, by the postconviction court's order to the extent that it appears to have been based on the court's personal view of the credibility of Mr. Scott and of Mr. Scott's fellow prisoner, Paul Kline. For a new trial to be warranted based on newly discovered evidence, "the newly discovered evidence must be of such nature that it would probably produce an acquittal on retrial." Jones v. State, 709 So. 2d 512, 521 (Fla. 1998) (Jones II) (citing Jones v. State, 591 So. 2d 911, 915 (Fla. 1991) (Jones I). "To reach this conclusion the trial court is required to 'consider all newly discovered evidence which would be admissible' at trial and then evaluate the 'weight of both the newly discovered evidence and the evidence which was introduced at the trial.' " Id. (quoting Jones I, 591 So. 2d at 916). [A]n evaluation of the weight to be accorded the evidence includes whether the evidence goes to the merits of the case or whether it constitutes impeachment evidence. . . . The trial court should also determine whether the evidence is - 13 - cumulative to other evidence in the case. . . . The trial court should further consider the materiality and relevance of the evidence and any inconsistencies in the newly discovered evidence. Id. (first citing Williamson v. Dugger, 651 So. 2d 84, 89 (Fla. 1994); then citing State v. Spaziano, 692 So. 2d 174, 177 (Fla. 1997); and then citing Williamson, 651 So. 2d at 89). Whether the postconviction court believes the truth of the newly discovered evidence is not a factor for it to consider in its weighing of all the evidence. See Jones II, 709 So. 2d at 521. Moreover, when the newly discovered evidence is a confession by someone other than the defendant, the court is expressly prohibited from considering the truth of the newly discovered evidence in connection with its admissibility determination. See Bearden v. State, 161 So. 3d 1257, 1263-64 (Fla. 2015) ("[U]nder Florida law, the credibility of an in-court witness who is testifying as to an out-of-court declaration against penal interest is not a matter for the trial court's consideration in determining whether to admit the testimony." (citing Carpenter v. State, 785 So. 2d 1182, 1203 (Fla. 2001))); Masaka v. State, 4 So. 3d 1274, 1283 (Fla. 2d DCA 2009) ("Once that admissibility threshold was met, the credibility of Panzo's [confession] and Masaka's defense was for the jury, not the trial court, to assess."). When the newly discovered evidence involves a witness's recantation of his or her testimony at the initial trial, however, "it is the duty of the court to deny a new trial where it is not satisfied that such testimony is true." Armstrong v. State, 642 So. 2d 730, 735 (Fla. 1994) (quoting Bell v. State, 90 So. 2d 704, 705 (Fla. 1956)). And that makes sense. When a witness testifies at trial, the jury, as the factfinder, assesses that witness's credibility. When the witness testifies a certain way and the jury's verdict is - 14 - consistent with that testimony, we can assume that the jury has found that witness credible. A motion for a new trial on the basis that a witness has recanted his or her testimony necessarily implicates the jury's credibility finding—if the witness's testimony was false, then the jury was mistaken in crediting it. Granting a new trial on that basis necessarily requires the trial court to set aside the jury's credibility finding. And because "it is elemental in our system of jurisprudence that the jury is the sole arbiter of the credibility of witnesses," Boatwright v. State, 452 So. 2d 666, 668 (Fla. 4th DCA 1984) (citing Barnes v. State, 93 So. 2d 863, 864 (Fla. 1957)), it makes sense that to set aside such a finding, the court at a minimum must itself find the witness's recantation credible. If it doesn't, the original jury's finding stands. If it does, then a second jury will have the opportunity to make its own credibility findings based on all of the evidence, including the recantation. But in either event, the ultimate credibility finding will have been made by a jury—the proper factfinder. Here, in contrast, Mr. Schofield's jury never had the opportunity to assess Mr. Scott's credibility or the credibility of his confessions. The same is true for Mr. Kline's credibility, and he has never recanted his testimony that Mr. Scott confessed to him. The question before the postconviction court was not whether a second jury should have the opportunity to reevaluate their credibility but whether a second jury should have the opportunity to evaluate their credibility in the first place. Consequently, to the extent that the postconviction court, in effect, sat as a "gatekeeper juror," I believe it exceeded its proper role in this analysis. Nonetheless, I agree that a new trial is unwarranted. First, I believe the confessions would be inadmissible in a new trial under either section 90.804(2)(c), Florida Statutes (2003), or Chambers v. Mississippi, 410 U.S 284 (1973), because they - 15 - lack the requisite indicia of trustworthiness for admission under either standard. See § 90.804(2)(c) ("A statement tending to expose the declarant to criminal liability and offered to exculpate the accused is inadmissible, unless corroborating circumstances show the trustworthiness of the statement."); Bearden, 161 So. 3d at 1265 (requiring that to be admissible pursuant to Chambers, the confession must be made "spontaneously to a close acquaintance shortly after the crime occurred"; that the confession be "corroborated by some other evidence in the case"; that the confession be "self-incriminatory and unquestionably against interest"; and that if there is any question about the truthfulness of the confession, "the declarant must be available for cross-examination"). None of Mr. Scott's confessions was made shortly after the crime occurred; rather, all were made after he had begun serving a life sentence for another murder. Not only was Mr. Scott's confession to Mr. Kline not corroborated by other evidence in the case, but the details Mr. Scott provided of the murder itself were flatly contradicted. Although Mr. Scott's later confessions were more consistent with the other evidence (although a striking discrepancy is his claim that he stabbed the victim to death in her car, but only a trace amount of blood was found in the car despite her having been stabbed twenty-six times), they were not made to close acquaintances but to investigators and attorneys for Mr. Schofield. In addition, the reliability of his later confessions was seriously undermined by his inquiries concerning "what was in it for him," his repeated denials of any involvement, and his outrageous admission to other crimes. See Dailey v. State, 279 So. 3d 1208, 1213 (Fla. 2019) ("Further, given that Pearcy claimed the affidavit was false and refused to testify about any of its substantive assertions, we agree with the circuit court's determination 'that Pearcy's affidavit is hearsay of an exceptionally unreliable nature and does not qualify as a statement - 16 - against interest.' "). And finally, although he has never testified in a trial against Mr. Schofield, Mr. Scott's confession is a recantation of his prior sworn testimony that all he did was steal the stereo out of the victim's abandoned car, and recantation testimony is "exceedingly unreliable." Armstrong, 642 So. 2d at 735; see also Lambrix v. State, 39 So. 3d 260, 272 (Fla. 2010) ("As this Court has noted repeatedly, recanted testimony is 'exceedingly unreliable.' " (quoting Heath v. State, 3 So. 3d 1017, 1024 (Fla. 2009))). Even if Mr. Scott's confessions met the threshold showing of reliability to warrant their admission, these same concerns prevent Mr. Schofield from establishing that they "would probably produce an acquittal on retrial." See Jones II, 709 So. 2d at 521. Moreover, evidence adduced at the evidentiary hearing that the postconviction court considered in finding Mr. Kline and Mr. Scott not credible, e.g., evidence suggesting that Mr. Scott and Mr. Kline had not even been housed together at the time Mr. Scott supposedly confessed to Mr. Kline, was certainly relevant to whether, in light of all the evidence that would be admissible in a new trial, the confessions would probably produce an acquittal. And, of course, overarching the entire analysis is Mr. Scott's remarkable inconsistency. Accordingly, while I cannot entirely join in the majority's reasoning, I concur in its decision to affirm. - 17 -