770 So.2d 1167 (2000)
Ceasar ROBINSON, Petitioner,
v.
STATE of Florida, Respondent.
No. SC93210.
Supreme Court of Florida.
October 5, 2000.
James Marion Moorman, Public Defender, and Kevin Briggs, Assistant Public Defender, Tenth Judicial Circuit, Bartow, Florida, for Petitioner.
Robert A. Butterworth, Attorney General, Robert J. Krauss, Senior Assistant Attorney General, Chief of Criminal Law, and Patricia A. McCarthy, Assistant Attorney General, Tampa, Florida, for Respondent.
WELLS, C.J.
We have for review State v. Robinson, 711 So.2d 619 (Fla. 2d DCA 1998). We accepted this case based on express and direct conflict with the opinions in Tibbs v. State, 397 So.2d 1120 (Fla.1981), Johnson v. State, 442 So.2d 185 (Fla.1983), Evans v. State, 692 So.2d 966 (Fla. 5th DCA 1997), and Borgess v. State, 455 So.2d 488 (Fla. 1st DCA 1984). We have jurisdiction. See art. V, § 3(b)(3), Fla. Const. Although we approve the Second District's reversal of the trial court's order granting postconviction relief in this case, we reach our conclusion through reasoning that differs from that of the district court, and we remand for proceedings consistent with this opinion.
The facts in this case, as set forth in the decision of the Second District below, are as follows:
In the early morning hours of December 5, 1981, Avil Francis was shot to death while sleeping at a residence in Winter Haven, Florida. His newlywed wife, Bernadette Francis, was shot in the head, but survived the shooting. The residence was the former marital abode of Ms. Francis and Robinson, who were divorced five months prior to the shootings. At a trial in August 1982, the jury found Robinson guilty of the first-degree murder of Mr. Francis and the attempted first-degree murder of Ms. Francis. The trial court sentenced Robinson to life in prison with a minimum of twenty-five years, plus a consecutive prison term of twenty years with a three-year minimum.[[1]]
Approximately fourteen years later, in July 1996, Robinson filed a motion for postconviction relief asserting newly discovered evidence. Robinson alleged that Ms. Francis and her daughter lied at his trial concerning their eyewitness accounts of the shooting. In support of his motion, he filed the affidavit of a fellow inmate, Wilbert Hollins, who purportedly obtained such information during a conversation with Ms. Francis in New York, in 1989. Robinson claimed that he learned of this information only a few months prior to filing his 1996 motion for postconviction relief.
In response to Robinson's motion, the State agreed to an evidentiary hearing. Following the evidentiary hearing, the trial court granted postconviction relief by written order in May 1997.
711 So.2d at 620 (footnote omitted). The State appealed, and the Second District reversed, finding that the trial court had abused its discretion in that the court had not recognized that the circumstantial evidence presented at trial, coupled with direct evidence in the form of testimony of Ms. Francis and her daughter Shantel, "so clearly outweighs the newly discovered, impeachment evidence that no reasonable person would have reached the conclusion that the trial court did." Id. at 623.
In analyzing Robinson's alleged newly discovered evidence, the district court applied the two-prong test this Court provided in Jones v. State, 591 So.2d 911 (Fla. 1991) (Jones I), and further explained in Jones v. State, 709 So.2d 512 (Fla.1998) (Jones II). Id. at 622. As to the first prong, the court found that Hollins' testimony constituted newly discovered evidence in that it was unknown at the time of the trial and that neither Robinson nor his counsel could have known of it by the use of diligence. See id.
Concerning the second prong of the Jones test, which assesses the probable effect of the newly discovered evidence at retrial, the court noted that Hollins' testimony would be admissible on retrial as impeachment evidence under section 90.608, Florida Statutes (1997). See id. After recognizing recent case law holding that impeachment evidence alone may in some circumstances warrant a new trial, the district court found that the trial court erred in failing to balance the weight of Hollins' impeachment testimony "against all of the other evidence stacked against Robinson." Id. at 623. The district court concluded that the trial court abused its discretion in determining that Hollins' testimony "would probably produce an acquittal on retrial" and reversed the trial court's grant of postconviction relief. Id.
In this Court, Robinson contends that the Second District erred in that it assumed the trial court's role and erroneously reweighed the evidence rather than finding the "palpable" abuse of discretion which is required to overturn a trial court's granting of a new trial. Robinson contends that the trial court correctly weighed the credibility of the newly discovered evidence alongside the direct testimony and circumstantial evidence presented at trial. Robinson further argues that the testimony of Hollins seriously undermined the only direct evidence of guilt the State presented at trial, which was the testimony of Ms. Francis and her daughter.
As to our jurisdiction in this case, we find upon further consideration that there is no conflict with the alleged conflict cases, which concern the proper role of appellate courts in evaluating weight and sufficiency of evidence in various factually distinguishable contexts. However, we do find conflict between this case and our decisions in Jones I and Jones II as to the analysis for determining whether alleged newly discovered evidence warrants a new trial.
As to the merits of this case, we approve the Second District's conclusion that the trial court erred in granting postconviction relief. However, the basis for our holding as to trial court error is not that the trial court abused its discretion, as held by the district court. Rather, we find in the trial court's order a misapplication of law as set forth in the Jones cases.
In its decision below, the Second District did not deal with the trial court's legal error in the application of an incorrect test for determining whether newly discovered evidence warrants a new trial. In Jones I, we explained the test for determining whether to grant a new trial based on newly discovered evidence as first requiring a finding that the evidence was unknown and could not have been known at the time of trial through due diligence. See 591 So.2d at 916. Once past this threshold finding, a court must apply the second prong which requires a finding that the newly discovered evidence "would probably produce an acquittal on retrial." Id. at 915.
Here, the trial court properly analyzed the threshold step in the Jones test but then stated as to the second prong, "The Court finds that the newly discovered evidence could probably produce an acquittal on retrial." State v. Robinson, No. CF81-3683A1-XX, Order Granting Motion for Postconviction Relief at 6 (Fla. 10th Cir. Ct. order filed May 28, 1997) (emphasis added). The uncertainty of this statement is compounded by the trial court's earlier statement, "The Court finds the new evidence would have probably affected the outcome of the case." See Robinson order at 5. These statements are in conflict with the definitive standard we set forth in Jones II, in which we held, as stated, that newly discovered evidence must be of such nature that it "would probably produce an acquittal on retrial." 709 So.2d at 521 (emphasis added).
In Jones II, we further explained that to reach the conclusion as to the probable acquittal on retrial,
[T]he 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." [Jones I, 591 So.2d at 916.]
In considering the second prong, the trial court should initially consider whether the evidence would have been admissible at trial or whether there would have been any evidentiary bars to its admissibility. Once this is determined, an 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 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.
709 So.2d at 521 (citations omitted) (emphasis added). We also required in Jones II that the trial court undertake a cumulative analysis by evaluating the newly discovered evidence in conjunction with evidence presented at all prior evidentiary hearings and evidence presented at trial. See id. at 522. We agree with the district court that impeachment evidence could be part of this cumulative analysis, as this Court stated in Jones I, Jones II, and Williamson v. Dugger, 651 So.2d 84, 89 (Fla.1994).
Therefore, we determine that this case should be remanded to the trial court for the trial court to reconsider its ruling. Accordingly, we remand to the Second District Court of Appeal with directions for the trial court to conduct a reweighing of the newly discovered impeachment evidence and a cumulative analysis by taking into consideration all the evidence presented at trial and in the evidentiary hearing below.
It is so ordered.
SHAW, HARDING, PARIENTE, LEWIS and QUINCE, JJ., concur.
ANSTEAD, J., concurs specially with an opinion, in which PARIENTE, J., concurs.
ANSTEAD, J., specially concurring.
While I concur in the remand required by the majority, I write separately to address the difficulty trial courts (and the trial court in this case in particular), have encountered trying to cope with the confusing standards that have been articulated by appellate courts for various postconviction proceedings. It is critically important, of course, that the correct standard be understood and applied, since it will usually determine the outcome of the proceedings.
Perhaps the three most common grounds the courts see alleged for postconviction relief are claims of newly discovered evidence, ineffective assistance of counsel and the government's failure to disclose evidence. The United States Supreme Court and, in turn, this Court, have articulated separate standards of proof that a defendant must meet in order to prevail on these three distinct legal claims.
For example, we have held that in order to prevail on a claim of newly discovered evidence a defendant must show first that he has discovered evidence that was "unknown and could not have been known at the time of trial through due diligence." Majority op. at 1170. Secondly, it must be shown that the evidence "would probably produce an acquittal on retrial." Id. Of course, if the newly discovered evidence is asserted as grounds for a new penalty phase proceeding in a capital case, the second prong would require a showing that the evidence would probably produce a life sentence in a new penalty phase.
Importantly, this relatively high standard of proof and prejudice for newly discovered evidence claims has been explicitly rejected by the U.S. Supreme Court as a standard for demonstrating prejudice in claims based upon the ineffective assistance of counsel:
On the other hand, we believe that a defendant need not show that counsel's deficient conduct more likely than not altered the outcome in the case....
Even when the specified attorney error results in the omission of certain evidence, the newly discovered evidence standard is not an apt source from which to draw a prejudice standard for ineffectiveness claims. The high standard for newly discovered evidence claims presupposes that all the essential elements of a presumptively accurate and fair proceeding were present in the proceeding whose result is challenged. Cf. United States v. Johnson, 327 U.S. 106, 112, 66 S.Ct. 464, 90 L.Ed. 562 (1946). An ineffective assistance claim asserts the absence of one of the crucial assurances that the result of the proceeding is reliable, so finality concerns are somewhat weaker and the appropriate standard of prejudice should be somewhat lower. The result of a proceeding can be rendered unreliable, and hence the proceeding itself unfair, even if the errors of counsel cannot be shown by a preponderance of the evidence to have determined the outcome.
Strickland v. Washington, 466 U.S. 668, 693-94, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984) (emphasis supplied). In Strickland, the U.S. Supreme Court formulated a two-part test for ineffectiveness claims: "First, the defendant must show that counsel's performance was deficient. Second, the defendant must show that the deficient performance prejudiced the defense." Id. at 687.
In explaining the prejudice prong of the test the Strickland court explained:
Accordingly, the appropriate test for prejudice finds its roots in the test for materiality of exculpatory information not disclosed to the defense by the prosecution, United States v. Agurs, 427 U.S., at 104, 112-113, 96 S.Ct., at 2397, 2401-2402, and in the test for materiality of testimony made unavailable to the defense by Government deportation of a witness, United States v. Valenzuela-Bernal, supra, 458 U.S., at 872-874, 102 S.Ct. 3449-3450[, 73 L.Ed.2d 1193 (1982)] . The defendant must show that there is a reasonable probability that, but for counsel's unprofessional errors, the result of the proceeding would have been different. A reasonable probability is a probability sufficient to undermine confidence in the outcome.
. . . .
The governing legal standard plays a critical role in defining the question to be asked in assessing the prejudice from counsel's errors. When a defendant challenges a conviction, the question is whether there is a reasonable probability that, absent the errors, the factfinder would have had a reasonable doubt respecting guilt. When a defendant challenges a death sentence such as the one at issue in this case, the question is whether there is a reasonable probability that, absent the errors, the sentencerincluding an appellate court, to the extent it independently reweighs the evidence-would have concluded that the balance of aggravating and mitigating circumstances did not warrant death.
Id. at 694-95 (emphasis supplied). In applying Strickland, it is extremely important to note that the Court expressly rejected utilizing the newly discovered evidence standard for demonstrating prejudice. Hence, in ineffective assistance of counsel claims a defendant is not required to demonstrate that but for counsel's deficient conduct the outcome of the trial would have been different. See id. at 693 ("[A] defendant need not show that counsel's deficient conduct more likely than not altered the outcome in the case"). Rather, the Court ultimately focused on whether counsel's ineffectiveness was such as to "undermine confidence in the outcome." Id. at 694. The Fourth Circuit Court of Appeals has explained the application of the Strickland standard:
With respect to the prejudice requirement, the petitioner "must show that there is a reasonable probability that, but for counsel's unprofessional errors, the result of the proceeding would have been different." The level of certainty is something less than a preponderance; it need not be proved that counsel's performance more likely than not affected the outcome. Instead, the petitioner need only demonstrate "a probability sufficient to undermine confidence in the outcome."
Young v. Catoe, 205 F.3d 750, 759 (4th Cir.2000) (citations omitted).
As noted in the Strickland quotes set out above, the prejudice prong of Strickland is derived from the same legal test for materiality developed in cases considering claims that the government withheld evidence from a defendant. These claims have come to be known as Brady claims after a Supreme Court case that recognized a constitutional violation in the government's improper withholding of evidence.[2] Because of the U.S. Supreme Court's express reliance on the preexisting standard set out in cases like United States v. Agurs, 427 U.S. 97, 104, 96 S.Ct. 2392, 49 L.Ed.2d 342 (1976), we can see that the prejudice standard for withheld evidence claims is closely related to the prejudice standard for ineffectiveness claims. In fact, in two recent cases the U.S. Supreme Court has reiterated this materiality standard when discussing the three-prong test for Brady claims.[3] In essence the Supreme Court has told us that the prejudice prong of a Brady claim is virtually identical to the prejudice prong of an ineffectiveness claim.
Unfortunately, our opinions have not been careful in preserving the bright-line distinction set out in Strickland between the prejudice standard in newly discovered evidence claims and the prejudice standard in ineffectiveness and Brady claims. In much of our case law involving Strickland claims, we restate the prejudice prong to require a showing of a probability of a different outcome as if the ineffectiveness standard is identical to the newly discovered evidence standard. Hence, we have contributed to the misunderstandings and confusion that exist as to the proper standards to be applied in postconviction proceedings. Of course, the Strickland opinion itself is somewhat confusing in its choice of words. In fact, Justice Souter has suggested that the Court use other language in describing the Strickland and Brady standards so as to avoid the very confusion discussed herein:
The Court speaks in terms of the familiar, and perhaps familiarly deceptive, formulation: whether there is a "reasonable probability" of a different outcome if the evidence withheld had been disclosed. The Court rightly cautions that the standard intended by these words does not require defendants to show that a different outcome would have been more likely than not with the suppressed evidence, let alone that without the materials withheld the evidence would have been insufficient to support the result reached. See ante, [ Strickler v. Greene, 527 U.S. 263,] at 289-290[, 119 S.Ct. 1936, 144 L.Ed.2d 286 (1999)]; Kyles v. Whitley, 514 U.S. 419, 434-435, 115 S.Ct. 1555, 131 L.Ed.2d 490 (1995). Instead, the Court restates the question (as I have done elsewhere) as whether "`the favorable evidence could reasonably be taken to put the whole case in such a different light as to undermine confidence'" in the outcome. Ante,[ Strickler v. Greene, 527 U.S. 263,] at 290 (quoting Kyles, supra, at 435, 115 S.Ct. 1555).
Despite our repeated explanation of the shorthand formulation in these words, the continued use of the term "probability" raises an unjustifiable risk of misleading courts into treating it as akin to the more demanding standard, "more likely than not." While any short phrases for what the cases are getting at will be "inevitably imprecise," United States v. Agurs, 427 U.S. 97, 108, 96 S.Ct. 2392, 49 L.Ed.2d 342 (1976), I think "significant possibility" would do better at capturing the degree to which the undisclosed evidence would place the actual result in question, sufficient to warrant overturning a conviction or sentence.
See Strickler v. Greene, 527 U.S. 263, 297-98, 119 S.Ct. 1936, 144 L.Ed.2d 286 (1999) (Souter, J., concurring in part and dissenting in part). For all these reasons, it should not be surprising that a trial court might not always get it exactly right in restating the appropriate standard. The trial court here obviously did a thorough and conscientious job in evaluating the claims presented and its work is to be commended.
PARIENTE, J., concurs.
NOTES
[1] The Second District Court of Appeal affirmed the conviction and sentence in Robinson v. State, 436 So.2d 112 (Fla. 2d DCA 1983).
[2] Brady v. Maryland, 373 U.S. 83, 83 S.Ct. 1194, 10 L.Ed.2d 215 (1963).
[3] See Strickler v. Greene, 527 U.S. 263, 119 S.Ct. 1936, 144 L.Ed.2d 286 (1999); Kyles v. Whitley, 514 U.S. 410, 115 S.Ct. 1555 (1995).