Although I ultimately agree that Bates is not entitled to habeas relief on his claims, I write separately to emphasize my disagreement with much of the Majority’s analysis. First, with respect to Bates’s ineffective assistance of counsel claim, although I believe that trial counsel was ineffective, given the dearth of clearly established law on this point, I agree that the state court’s adjudication was not an unreasonable application of clearly established federal law. Similarly, with respect to Bates’s claim involving the resentencing jury’s lack of awareness about his consecutive life sentences, I disagree with the Majority’s conclusion that existing Supreme Court precedent forecloses his claim. Nevertheless, because I agree that the Florida Supreme Court’s adjudication was not contrary to clearly established law of the Supreme Court, I ultimately concur in the outcome of that claim.
A.
Bates’s murder trial began with a prayer in the presence of the jury, and the victim’s husband subsequently gave testimony informing the jury that the prayer was delivered by none other than the victim’s own minister. This testimony had no probative value, but it had great potential to prejudice the jury against Bates. The prayer inserted God into Bates’s trial, and the husband’s testimony made clear whose side God was on.
Bates argues that his trial counsel, Bowers, rendered ineffective assistance by failing to object to this highly prejudicial sequence of events, either when the trial judge asked the victim’s minister to pray or when the victim’s husband’s testimony linked the victim to the minister’s church. Specifically, Bates insists that in the racially charged context of this case, where a black defendant stood before an all-white jury, beginning the trial with a prayer by the victim’s minister was not generic and benign as the district court and the Majority concludes, particularly after the jury became aware of who delivered the prayer. Moreover, Bates insists that the Florida Supreme Court did not conduct the proper cumulative Strickland prejudice analysis because it failed to consider the totality of the circumstances within the context of Bates’s trial.1
The Florida Supreme Court concluded that this Strickland claim failed on the merits. See Bates v. Dugger, 604 So.2d 457, 459 n. 4 (Fla.1992). Although Bates insists that the Florida Supreme Court’s determination is not entitled to AEDPA *1308deference because it was insufficient and relied upon inaccurate facts, the Supreme Court has clarified that “[w]here a state court’s decision is unaccompanied by an explanation, the habeas petitioner’s burden still must be met by showing there was no reasonable basis for the state court to deny relief.” Harrington v. Richter, 562 U.S. 86, 131 S.Ct. 770, 784, 178 L.Ed.2d 624 (2011); see Jones v. GDCP Warden, 2014 WL 1088312, *10 (11th Cir. March 20, 2014) (“AEDPA mandates deferential review of any claim that a state court ‘adjudicated on the merits,’ 28 U.S.C. § 2254(d), and does not impose any specific requirements on how a state court should announce its decision.”). Therefore, we owe the Florida Supreme Court’s rejection of Bates’s ineffective assistance of counsel claim AEDPA deference, and given that conclusion, I agree with the Majority that Bates is not entitled to habeas relief. I am not aware of any clearly established federal law, nor has Bates cited any, which indicates that the Florida Supreme Court’s determination is an unreasonable determination under Strickland.
This is not to say that I agree with the Florida Supreme Court’s decision. I do not. I concur and do not dissent only because that court’s decision is not necessarily an unreasonable one. Unlike the Majority and that court, however, I believe that Bates has presented a persuasive Strickland claim. One would expect reasonably competent counsel, following a prayer by a murder victim’s minister and subsequent testimony linking the victim to the minister’s church, to request permission to approach the bench, to object, and to ask for a mistrial. This is especially true given the context of Bates’s original trial: it was a high profile, racially charged murder case in a small community.2
I agree with the Majority that interrupting either the prayer or the husband’s testimony with an immediate objection would have been imprudent. I also agree that if Bowers had waited too long to object, his objection would have been barred by the contemporaneous-objection rule. See, e.g., United States v. Turner, 474 F.3d 1265, 1267 (11th Cir.2007) (refusing to consider an objection to testimony where the “objection was made by defense counsel only the next day ” (emphasis added)). I disagree, however, that competent counsel would not find an opportunity to object in the rather large window of time between the moment the husband gave the objectionable testimony (when an objection would have been imprudent) and “the next day” (when an objection would have been untimely). Had Bowers asked to approach the bench immediately after the husband’s testimony concluded but before the next witness was called, the objection would have been timely. The contemporaneous-objection rule does not foreclose objections raised after a witness’s testimony, particularly if counsel explains the reason for any delay.
I also fail to see why Bowers would have been conflicted about approaching the bench to request a mistrial. The upside, securing a mistrial, could not have been *1309more beneficial to Bates given the jury’s mounting prejudice against him. In this context, the prospects for securing a mistrial need not be very great to make it incompetent not to even ask for one.3 The Majority suggests that there is a significant downside to asking for a mistrial in the manner just described, but I do not see it. The Majority explains that approaching the bench would have invited the jury to speculate and that if the objection had been overruled, the jury would have been left bewildered. Even if this is true, is it really better to leave the jury prejudiced, instead? I cannot take seriously the notion that prosecutors would voice no objection to beginning a murder trial with a prayer by the defendant’s minister, even though all the reasons The Majority discusses for not objecting would be equally applicable in that context. When God is inserted into a trial on the opponent’s side, whatever slight reservations competent attorneys have about raising objections (not wanting to be quarrelsome, not wanting to exhaust the court’s patience, and not wanting to “bewilder” the jury) pale in comparison to such obvious and significant prejudice.
The Majority opinion further supposes that Bowers may have refrained from objecting because he feared having his objection overruled, but that can hardly justify an attorney’s decision not to raise an objection in the first place. If that were an adequate justification, a lawyer could never be faulted for failing to raise an objection because it is always true that an objection might be overruled. This logic depends on the fallacy that having an objection overruled is highly counterproductive, such that a lawyer who hears prejudicial testimony is placed in a lose-lose situation: objecting is bad because the objection might be overruled, but remaining silent is also bad because the prosecution may inject even more prejudice into the trial. But here, at worst, Bowers faced a win-no lose situation. Had he raised the objection, he might have secured a mistrial and spared his client from prejudice, but at worst, his objection would have been overruled out of the jury’s hearing, leading to a momentary break in the proceedings at an already natural breaking point (between the testimony of two witnesses). In short, Bowers passed up an opportunity for a significant upside in order to avoid a virtually non-existent downside.4
*1310Further, the fact that Bowers thought “nothing of the prayer” and that he never even considered moving for a mistrial is all the more reason to believe that his failure to object was incompetent and was not the product of thoughtful consideration. He did not balance the pros of objecting to this highly prejudicial evidence against the cons of potentially drawing attention to the prejudice, as the Majority would have it. Instead, Bowers incompetently failed to see the prejudice at all and sat idly by as the court and the prosecution stacked the deck against his client.5 As I have already said, however, while I believe Bates established a Strickland claim, we cannot grant habeas on this basis because the Florida Supreme Court’s contrary conclusion was not unreasonable.6
*1311It is bad enough that, in the course of denying habeas relief on Bates’s Strickland claim, the Majority countenances a defense attorney’s failure to object to highly prejudicial proceedings that have no probative value in a trial. However, I believe that the manner in which the Majority relies upon Establishment Clause cases to conclude that Bates’s Strickland claim fails is even more problematic and unnecessary. Because the Florida Supreme Court addressed Bates’s Strickland claim on the merits, the first question we must address is whether that court unreasonably applied Strickland in holding that Bowers’s failure to object to a prayer and subsequent testimony was not ineffective. I agree that it did not, and we could leave it at that.
To be fair, I discussed how I would have addressed the Strickland issue were we to analyze the issue de novo, so I cannot quarrel with the Majority’s decision to do the same. I must, however, raise a few issues with the Majority’s analysis, which is in deep tension with Strickland itself. To begin with, I would frame the question differently than has the Majority. The Majority asks only whether Bowers was incompetent for failing to object to the prayer itself. Had Bowers objected only to the prayer, I agree that his objection might have been best framed as an Establishment Clause challenge. That is not precisely what Bates claims Bowers should have objected to, however. Instead, Bates also insists that Bowers was incompetent for failing to object to the prayer after the husband’s testimony linked the prayer and the minister to the victim. In other words, we need not address whether the Establishment Clause prevented the trial court from inviting God into the courtroom; rather, we must address whether competent attorneys would object to testimony placing God on the victim’s side. That changes the objection from an Establishment Clause challenge to a run-of-the-mill objection to highly prejudicial testimony that has no probative value. The Establishment Clause cases cited by the Majority are irrelevant to this inquiry.
Even assuming that we must address whether Bowers was incompetent for failing to raise an initial objection to the prayer itself on Establishment Clause grounds, we have no reason to address the Establishment Clause cases. As the Majority makes clear, it believes the prayer itself was entirely innocuous, unobjectionable, and did not prejudice the jury against Bates in any way. The Strickland analysis should thus be quite simple: competent attorneys in Bowers’s position would not have objected because nothing about the prayer harmed Bates. The Majority’s Strickland analysis should stop there. After all, regardless of the odds of success, why object or ask for a mistrial if there is no reason to think that the next jury will be any more sympathetic (or less prejudiced) than the current one?
To justify discussing the Establishment Clause, the Majority explains that it is only analyzing what a competent attorney in Bowers’s position would have done. In order to do this, the Majority claims we need to know whether Bowers’s objection to the prayer itself would have been sustained under the Establishment Clause because that, in turn, informs our analysis of whether a competent attorney in Bowers’s position would have objected. I disagree with both propositions.
*1312In analyzing Bates’s Strickland claim, our goal is to recreate the situation confronting Bowers when he failed to object. The Majority states that it is reasonable to conclude that counsel’s failure to object was not deficient because the “United States Supreme Court has never held that it is a violation of either the Due Process Clause or the Establishment Clause to begin a criminal trial with a prayer, let alone a violation of the Establishment Clause that would require reversal, a mistrial, or any other form of relief.” Maj. Op. at 1290. The Majority also asserts that “we are deciding ... that given the state of the law at the time of trial (and now), it was (and still is) not clearly established that the opening prayer violated the Establishment Clause____ As a result, a reasonably competent attorney could conclude that objecting ... would not benefit his client.” Maj. Op. at 1298, n. 9 (emphasis added).
In essence, the Majority is asserting that counsel cannot be found incompetent for failing to raise an objection under Strickland unless the Supreme Court has clearly established that the unraised objection would have been sustained. That simply cannot be the case, as competent trial counsel’s goal is not to have 100% of his objections sustained; it is instead to secure the most favorable circumstances for his client. When assessing a trial court’s actions, the competent attorney’s first question is not, “What have federal courts clearly said on this subject, and if I object, how likely am I to be sustained?” Instead, the competent attorney’s first question is, “Do these actions harm my client’s interests, and if so, what are the best arguments I can make to remedy the prejudice that has just occurred?” If the best argument is one that has neither been explicitly accepted or rejected by federal courts, and counsel decides not to make the argument for that reason, I think that decision would very likely be incompetent.
Indeed, the question of how clearly established the grounds for an objection are in federal law is often largely irrelevant to a trial attorney in state court. For example, assuming that the Supreme Court had clearly established that Bates would have been entitled to a mistrial had Bowers objected to the prayer, competent counsel still may not have objected for reasons wholly unrelated to the clarity of federal law. If, as the Majority claims, the jury was not prejudiced against Bates either before or after the prayer, then even if Bowers could have secured a mistrial, why would he want to? Competent lawyers do not halt proceedings only to start them over again — even if the Supreme Court has clearly established that they can— when there is no reason to believe that starting over will be any better for their clients.
On the other hand, if Supreme Court precedent was not clearly established one way or the other, and a defendant’s trial began with a prayer asking the jury to bring the defendant to justice or to bring closure to the victim’s family, the prejudice would be extreme. In that case, it would likely be incompetent for counsel to do nothing regardless of whether the Supreme Court had clearly established that beginning a trial with a biased prayer is grounds for a mistrial, unless he had legitimate strategic reasons for doing nothing. That case law is not definitive on the issue is no excuse for the attorney to sit in silence and to do absolutely nothing to suggest that gaps in case law be filled in in his client’s favor.7
*1313In order to justify its discussion of the Establishment Clause in a case about a Strickland claim, the Majority explains that counsel cannot be deemed ineffective for -failing “to raise meritless arguments.” Diaz v. Sec’y, Dept, of Corr., 402 F.3d 1136, 1142 (11th Cir.2005) (emphasis added). As the Majority explains, however, the most that can be said of case law regarding any objection Bowers might have raised under the Establishment Clause is that federal law does not clearly establish that such an objection would prevail. I have no problem concluding that attorneys cannot be found incompetent for failing to raise meritless claims, but I think it is entirely different — and highly inappropriate — to suggest that attorneys could never be found incompetent for failing to raise a claim simply because it is not clearly established. But this is the interpretation of Strickland the Majority advances. Worse still, in a case where our only task is to assess the reasonableness of the Florida Supreme Court’s resolution of a Strickland claim, the Majority attempts to convert a potential Establishment Clause claim from “not clearly established” . to “meritless.” This ' obvious overreach makes a mess of Strickland and potentially forecloses a claim that we have absolutely no reason to address.
At least by explaining that “we are not deciding whether the opening prayer violated the Establishment Clause,” Maj. Op. at 1300, n. 9, the Majority recognizes that whether trials can begin with prayers is still an open question. It is important to point out that this remains true despite the Majority’s -claim that “federal courts of appeal have rejected such Establishment Clause challenges where the content of the prayer did not prejudice the defendant or substantially impair his right to a fair trial.” Maj. Op. at 1290 (emphasis added). In support of this proposition, the Majority cites Isaacs v. Head and Marsh v. Chambers, neither of which support the Majority’s conclusion. See Isaacs v. Head, 300 F.3d 1232, 1252-53 (11th Cir.2002) (rejecting a habeas claim that the state court had unreasonably applied the Supreme Court’s Establishment Clause precedent by not reversing a conviction where there had been a prayer to open a trial, reasoning that there was a lack of Supreme Court precedent supporting a reversal); Marsh, 463 U.S. 783, 103 S.Ct. 3330, 77 L.Ed.2d 1019 (1983) (holding that a prayer at the commencement of a legislative session did violate the Establishment clause). These cases tell us only that it is not clearly established that Bowers’s objection to the prayer would have been sustained. They do not tell us that it should have (or would have) been overruled had it been made, or that the objection would have been merit-less such that Bowers cannot be faulted for having failed to raise it. The same can be said of the Supreme Court’s recent decision in Town of Greece, N.Y. v. Galloway, — U.S.—, 134 S.Ct. 1811, 188 L.Ed.2d 835 (2014) (rejecting an Establishment Clause challenge to opening town board meetings with a sectarian Christian prayer).8
*1314Ironically, the Majority emphasizes that Bates cannot disguise an Establishment Clause claim as a Strickland claim because he was only granted a Certificate of Appealability (COA) on the Strickland issue. See Maj. Op. at 1283, n. 1. Given that Bates is precluded from making an Establishment Clause claim, it is inappropriate — particularly in an AEDPA case concerning a Strickland claim — for the Majority to imply that Establishment Clause cases foreclose objections to prayers at the beginning of a trial. Ultimately, however, because of the deference afforded state courts under AEDPA and the absence of clearly established federal law of the Supreme Court, I concur in concluding that the Florida Supreme Court’s adjudication of the Strickland claim was not unreasonable. See 28 U.S.C. § 2254(d)(1).
B.
Bates makes three related claims with respect to revised Florida Statute § 775.082(1), which provides for a possible sentence of life without parole. - First, Bates argues that the resentencing judge erred by failing to allow Bates to waive his right against ex post facto application of laws in order to apply the newer version of § 775.082(1) to him, which provides a possible sentence of life without parole.9 Second, Bates maintains that the judge erred by failing to enter Bates’s soliloquy seeking retroactive application of the sentencing statute into evidence so that the jury would know that Bates was willing to fore-go any potential opportunity for parole. Third, Bates argues that the resentencing judge erred by failing to instruct the re-sentencing jury that Bates had been senténced to two life terms and one 15-year term for the other crimes which would run consecutively, facts which he believes would have made the jury less likely to recommend death because they would have known that he would be imprisoned for a long time.10 Bates asserts that these *1315failures denied him due process and a fundamentally fair capital sentencing under the Eighth and Fourteenth Amendments. See Stringer v. Black, 503 U.S. 222, 232, 112 S.Ct. 1130, 1137, 117 L.Ed.2d 367 (1992) (“[W]hen the sentencing body is told to weigh an invalid factor in its decision, a reviewing court may not assume it would have made no difference if the thumb had been removed from death’s side of the scale.”).
In its adjudication on the merits, the Florida Supreme Court held:
In Florida, without clear legislative intent to the contrary, a law is presumed to apply prospectively____ We find no unequivocal language that the Legislature intended this [1994] amendment to apply retroactively. We have previously held that this statute was not applicable to crimes committed before its effective date____
Our analysis of this issue causes us to reject appellant’s waiver arguments. Because the 1994 amendment can have no effect on appellant’s sentencing, we conclude that the waiver of an ex post facto claim in respect to the 1994 amendment to section 775.082 is of no consequence. The waiver of ex post facto rights would only be an issue if the-statute could have an effect on appellant’s sentence which, as we have stated, it cannot.
Appellant’s alternate contention, that the jury should have been advised that appellant would agree to waive the possibility of parole, is also unavailing under Florida’s capital sentencing scheme because, as the trial court ruled, “[a] defendant cannot by agreement confer on the court the authority to impose an illegal sentence.” Williams v. State, 500 So.2d 501, 503 (Fla.1986). At the time appellant committed this murder, the Legislature had not established life without the possibility of parole as punishment for this crime.
In his second issue, appellant argues that the State took advantage of the trial court’s failure to instruct the jury on a sentence of life without the possibility of parole during cross-examination of appellant’s witnesses and closing argument by making future dangerousness an issue for the jury. Appellant did not object to either the State’s cross-examination or closing argument on this ground, and the issue is therefore procedurally barred. Steinhorst v. State, 412 So.2d 332, 338 (Fla.1982). Moreover, after reviewing the record, we do not agree that the State’s cross-examination or argument raised the specter of appellant’s future dangerousness....
As part of his third issue, appellant contends that the fact that he was already sentenced to two life terms plus fifteen years and that those sentences were to run consecutively to the sentence for the murder was relevant mitigation “in the sense that [it] might serve as a basis for a sentence less than death.” We have rejected similar arguments in Franqui v. State, 699 So.2d 1312, 1326 (Fla.1997); Marquard, v. State, 641 So.2d 54 (Fla. 1994); and Nixon v. State, 572 So.2d 1336 (Fla.1990).
These other sentences are not relevant mitigation on the issue of whether appellant will actually remain in prison for the length of those sentences. The length of actual prison time is affected by many factors other than the length of *1316the sentence imposed by the sentencing court. The introduction of this evidence would open the door to conjecture and speculation as to how much time a prisoner serves of a sentence and distract jurors from the relevant issue of what is the appropriate sentence for the murder conviction. Regarding this issue appellant’s brief states “[T]he state argued that [appellant] would be eligible for parole after serving the mandatory minimum.” Appellant, however, makes no record reference to support that statement; nor has our independent review of the record revealed support for that statement. As we stated regarding the previous issue, our review of the record causes us to find that the State did not violate Hitchcock v. State, 673 So.2d 859, 860 (Fla.1996), by injecting appellant’s future dangerousness into its evidence or argument. We conclude that the trial court followed our precedent and did not abuse its discretion in respect to this issue.
Bates v. State, 750 So.2d 6, 10-11 (Fla. 1999) (footnote omitted). I note that in his concurrence, Justice Shaw of the Florida Supreme Court emphasized that the question posed by the resentencing jury indicated that they were clearly confused, and the court should have just answered with a simple “Yes” in response to whether the jury was limited to life with a minimum of 25 years or the death penalty, and “No” in response to whether they could recommend life without the possibility of parole. Id. at 20 (Shaw, J. concurring). There was also a vigorous dissent in which Florida Supreme Court Justice Anstead, along with two other Florida justices, held that the majority’s refusal to accept Bates’s waiver of his ex post facto rights was “unnecessarily harsh and inconsistent with ... prior case law.” Id. at 20 (Anstead, J., dissenting). In fact, Justice Anstead explained that such a waiver would be consistent with prevailing legislative policy, as indicated by the legislative amendment itself, and that the court has repeatedly recognized that a defendant can waive his constitutional protections. Id. at 21 (citing Bowles v. Singletary, 698 So.2d 1201 (Fla. 1997); Melvin v. State, 645 So.2d 448 (Fla. 1994)).
Nevertheless, upon review of the relevant Supreme Court precedent, I agree with the Majority here that the Florida Supreme Court’s adjudication was not contrary to, or an unreasonable application of, clearly established federal law. With respect to Bates’s first two claims, the state court had previously held that this statute was not applicable to crimes committed before its effective date, and there was nothing in its legislative history to indicate that defendants could choose which sentencing statute would apply. Bates, 750 So.2d at 10; see Hudson v. State, 708 So.2d 256, 262 (Fla.1998). Further, Bates has cited no federal law requiring or even allowing a defendant to waive the laws applicable to his sentencing, regardless of whether such a waiver would be favorable to him or not.
However, Bates’s third claim, regarding the jury’s knowledge about his other convictions, gives me much pause. The Florida Supreme Court explicitly rejected Bates’s argument that the fact that he was already sentenced to two life terms plus 15 years and that those sentences were to run consecutively to the sentence for murder was relevant mitigation “in the sense that [it] might serve as a basis for a sentence less than death.” Bates, 750 So.2d at 11. The Florida Supreme Court continued to say:
These other sentences are not relevant mitigation on the issue of whether appellant will actually remain in prison for the length of those sentences. The length of actual prison time is affected *1317by many factors other than the length of the sentence imposed by the sentencing court. The introduction of this evidence would open the door to conjecture and speculation as to how much time a prisoner serves of a sentence and distract jurors from the relevant issue of what is the appropriate sentence for the murder conviction.
Id. Cutting against the Florida Supreme Court’s finding that such evidence would be irrelevant, longstanding Supreme Court precedent explicitly holds that evidence which may call for a penalty less severe than death is relevant in a capital sentencing. See Mills v. Maryland, 486 U.S. 367, 367-77, 108 S.Ct. 1860, 1867, 100 L.Ed.2d 384 (1988) (holding that “the risk that the death penalty will be imposed in spite of factors which may call for a less severe penalty is unacceptable and incompatible with the commands of the Eighth and Fourteenth Amendments”); Lockett v. Ohio, 438 U.S. 586, 604, 98 S.Ct. 2954, 2964-65, 57 L.Ed.2d 973 (1978) (holding that “the Eighth and Fourteenth Amendments require that the sentencer, in all but the rarest kind of capital case, not be precluded from considering, as a mitigating factor, any aspect of a defendant’s character or record and any of the circumstances of the offense that the defendant proffers as a basis for a sentence less than death” (emphasis and footnote omitted)).
One of the primary factors that any sentencing body must consider is the need for incapacitation of the defendant in order to protect the public. See, e.g., 18 U.S.C. § 3553(a)(2)(c) (stating that one of the primary needs of a sentence is to “protect the public from further crimes of the defendant”). In this case, the resentencing jury was undoubtedly interested in incapacitating Bates because it asked if it could impose a life sentence instead of the death penalty, indicating that some jurors wanted to incapacitate Bates, without the possibility of parole, for longer than the 12 years remaining on the 25-years-to-life sentence that Bates would have received if the jury had voted against death. Admittedly, guaranteed life without parole was not an option even if Bates’s other sentences were considered because it appears that under Florida’s previous system of parole, Bates would have been eligible for review for parole, at an absolute minimum, 12 years after the end of his sentence for the first degree murder conviction.11 However, given the jury’s question, we do *1318not see any limiting principle in Lockett that would render it inapplicable. Indeed, the fact that Bates would be, at a mini-mum, incarcerated for an additional 12 years following his sentence for first degree murder, bringing total incarceration if the jury voted against death to at least another 24 years, is a “mitigating factor ... that the defendant proffer[ed] as a basis for a sentence less than death.” Lockett, 438 U.S. at 604, 98 S.Ct. at 2964. It is clear from the jurors’ question that incapacitation was important to them, but on this critical issue, the jury was kept in the dark.12
We recognize that Simmons v. South Carolina made a jury instruction on parole ineligibility mandatory only in a context where lifetime ineligibility was at issue, only in the context of rebutting aggravating evidence, and only where the jury itself had a life without parole option. 512 U.S, 154, 175, 114 S.Ct. 2187, 2200, 129 L.Ed.2d 133 (1994) (O’Connor, J., concurring) (holding that when the state raised the specter of a defendant’s future dangerousness, the court violated his due process rights by refusing to instruct the jury that, as an alternative to a capital sentence, the sentence of life imprisonment included absolutely no possibility of parole). Simmons did not, however, either foreclose or explicitly extend that mandatory instruction to eligibility for parole for a term of years rather than for a term of life or to sentences rendered for other convictions not before the jury. Id. Subsequently, Ramdass v. Angelone held that the instruction on ineligibility for parole is only required when ineligibility is established with certainty as a matter of state law. 530 U.S. 156, 166, 120 S.Ct. 2113, 2120, 147 L.Ed.2d 125 (2000). Thus, Ramdass also does not foreclose applying the rule from Simmons in a case like this, where there is certainty under state law that Bates would be ineligible for parole, at a bare minimum, for a term of 12 additional years following the term served for the first degree murder. Nor does Ramdass foreclose application of Lockett’s rule that relevant mitigating evidence cannot be kept from a capital sentencing jury. Indeed, the plurality in Ramdass permitted Virginia to insist on certainty before instructing the jury on ineligibility for parole precisely because of how relevant ineligibility for parole is to a sentencing jury. Id. at 180-SI, 120 S.Ct. at 2127-28.13
Furthermore, Lockett’s rule that defendants are entitled to present evidence that may tend to prove to a jury that they deserve a sentence less than death may be sufficiently clear to control this case, even under deferential AEDPA review of state *1319court decisions. See Panetti v. Quarterman, 551 U.S. 980, 958, 127 S.Ct. 2842, 2858, 168 L.Ed.2d 662 (2007) (“AEDPA does not require state and federal courts to wait for some nearly identical factual pattern before a legal rule must be applied. Nor does AEDPA prohibit a federal court from finding an application of a principle unreasonable when it involves a set of facts different from those of the case in which the principle was announced. The statute recognizes, to the contrary, that even a general standard may be applied in an unreasonable manner.” (internal citations omitted)). Thus, according to the Supreme Court’s holding in Panetti, the evidentiary rules announced in Lockett and Mills need not specifically address the instant factual scenario in order to be applied to grant habeas relief.
I think it likely, given the highly relevant nature of incapacitation to jurors when deciding whether to impose a capital sentence, that the Supreme Court would conclude that due process requires including such relevant mitigation evidence in a capital sentencing. Moreover, the trial court here, unlike the state trial court in Ramdass, could have, without “conjecture and speculation,” Bates, 750 So.2d at 11, told the jurors of the 12 years of guaranteed incapacitation that Bates would have to serve in addition to what they imposed for the murder conviction.14 Cf. Ramdass, 530 U.S. at 167, 120 S.Ct. at 2120 (holding that a Simmons instruction was not warranted because defendant’s third conviction under Virginia’s three-strike rule was not final under Virginia law at the time the jury considered the murder sentence). Thus, I find the Florida Supreme Court’s reasons for refusing to permit the jury instruction unpersuasive.
Nevertheless, a competing principle announced by the Supreme Court in California v. Ramos, that state courts are entitled to deference in determining what evidence may go before a sentencing jury, combined with AEDPA’s deferential standard of review, precludes us from applying Mills and Lockett to this case. California v. Ramos, 463 U.S. 992, 1001, 103 S.Ct. 3446, 3453, 77 L.Ed.2d 1171 (1983) (stating that it is ordinarily proper to “deferí] to the State’s choice of substantive factors relevant to the penalty determination”). The Supreme Court further reiterated in Ramdass that “States are entitled to some latitude [as] the admissibility of evidence at capital sentencing ... remains ... an issue left to the States.” Ramdass, 530 U.S. at 169, 120 S.Ct. at 2122-23. Even the Ramdass dissent recognizes the discretion typically afforded to state supreme courts: “This is not to say ... that the Constitution compels States to tell the jury every single piece of information that may be relevant to its deliberations. Indeed, in California v. Ramos, we held it ordinarily proper to defer to the State’s choice of substantive factors relevant to the penalty determination.” Id. at 194-95, 120 S.Ct. at 2135 (Stevens, J., dissenting).
Therefore, despite the Supreme Court’s holding in Panetti that “AEDPA does not require state and federal courts to wait for some nearly identical factual pattern before a legal rule must be applied,” 551 U.S. at 953, 127 S.Ct. at 2858, I conclude that Supreme Court’s emphasis on the deference afforded to state courts’ evidentiary *1320rulings under AEDPA, in light of the ambiguity in the law created by Simmons and Ramdass, precludes us from granting Bates habeas on this claim. This highlights a troubling consequence of AEDPA case law: where a precedent cannot obviously be extended to the case we have before us, and where the Supreme Court has spoken in a tangentially related way to the situation at issue without explicitly covering it, we figuratively throw up our hands, repeat the refrain that AEDPA requires deference to state courts, and deny habeas relief. To be clear, I believe that the Florida Supreme Court’s determination on this issue was contrary to the rule articulated in Lockett and reiterated in Mills, but that the more recent precedent in Simmons and Ramdass, although distinguishable, generates sufficient ambiguity as to preclude relief in an AEDPA context unless or until the Supreme Court tells us otherwise. For these reasons alone, I concur in affirming the district court’s decision denying Bates habeas relief.
. Part of Bates’s theory may be, as the Majority says, “that an all-white jury cannot give a black defendant charged with the murder of a white woman a fair trial.” Maj. Op. at 1288. Had Bowers secured a mistrial by requesting one after the prayer or at least after the husband’s testimony, Bates believes he may have benefitted from a racially diverse jury. That, however, is not Bates’s only argument, or his best one. To be perfectly clear, I am focusing here on Bates’s claim that in the context of a racially charged environment, which included an all-white jury, a white Christian victim whose religion was made evident by her minister’s prayer, and a black defendant, his counsel’s failure to object to the minister’s prayer after the husband's testimony was objectively unreasonable. The prejudice Bates suffered as a result of this unobjected-to sequence of events was not facing an all-white jury, which perhaps may have been able to give Bates a fair trial before they listened to the minister’s opening prayer. Rather, Bates claims, and I agree, that the jury was far less likely to be able to give Bates a fair trial after the prayer and the husband’s testimony, and that any competent counsel would have objected. This testimony linked the minister’s plea for God’s guidance to the victim herself, turning a potentially innocuous prayer into a not so subtle reminder that Bates stood accused of murdering a Christian woman and that her minister was interested in the trial.
. During the postconviction evidentiary hearing, Bowers testified that the jurors could have drawn a prejudicial conclusion from the prayer and subsequent testimony given the racial tension in the case. In his brief, Bates explains that "[n]o black defendant in a death penalty case in Panama City had ever been acquitted where the victim was white. Of the five death sentences rendered in Bay County, all five were black defendants with white victims. Three of the sentences were handed down to Mr. Bates since 1983. The other two death sentences were imposed on Carl Jackson and Eric Turner, whose cases were overturned and reduced to life sentences by the Florida Supreme Court. See Jackson v. State, 359 So.2d 1190 (Fla.1978); Turner v. State, 645 So.2d 444 (1994).”
. There is no certainty that Bowers’s objection, had it been raised, would have been sustained, but by remaining silent, there was a certainty that a mistrial would not be granted. Depriving a client of a significant, even if not certain, opportunity for a mistrial under these circumstances is incompetent.
. The Majority opinion insists that we cannot expect competent attorneys to raise objections that are meritless. On that point, I agree. But a potentially losing claim is different than a meritless one. We might expect competent counsel to object to highly prejudicial evidence even if there is only a 49% chance that the objection will be sustained, but of course if the odds of success are only 1%, competent counsel’s calculus may change. This consideration assumes that the objectionable evidence is prejudicial, however, and as the Majority recognizes, in addition to considering the odds of having an objection sustained, competent counsel must also assess whether the evidence is bad for the client. All else being equal, as evidence becomes more prejudicial, competent attorneys will be more willing to raise objections that may not be sustained. And at the other extreme, when an event transpires that does nothing to harm a client’s interests, even if it is 100% certain that an objection to the event will be sustained, competent attorneys will not necessarily object because there is nothing to gain by doing so. Relying on this imminently logical proposition, the Majority suggests that if Bates could not ultimately show prejudice for purposes of Strickland, then we cannot conclude that his attorney was incompetent for failing to object. To be clear, I believe that Bates can show prejudice for Strickland pur*1310poses. But even if he could not, I disagree with the Majority opinion’s conclusion. The fact that an appellate court may ultimately conclude, looking back, that an error was not prejudicial for Strickland purposes does not mean that the potential error was not sufficiently prejudicial to mandate action by a competent attorney on the spot. We assess prejudice for Strickland purposes under the totality of the circumstances, with the benefit of all the evidence and with knowledge of all subsequent events. Prejudicial evidence to which an attorney might have objected is often outweighed by subsequent, overwhelming proof of guilt, and we often find that a Strickland claim fails on the second prong of the analysis for this reason. But it would be patently incompetent for an attorney listening to prejudicial evidence during a trial to analyze the effect on his client in the same way. For one thing, he obviously does not know what the totality of the circumstances will be because he cannot predict the future. Further, the fact that subsequent evidence might render earlier prejudice harmless does not mean that a lawyer should simply throw up his hands and allow the prosecution to pile prejudice on top of prejudice. Indeed, if that were the case, a lawyer representing a man confronted with overwhelming evidence of his guilt could never be incompetent under the first prong of Strickland. Surely that cannot be. Competent lawyers object to prejudicial testimony even if — perhaps particularly if— the cases against their clients are overwhelming.
. The Majority correctly notes that competent attorneys do not make all objections that could be made, even if there is nothing to lose by objecting. To be clear, I do not find Bates’s Strickland claim persuasive because there would have been no downside to raising an objection. Instead, I find Bates’s Strickland claim persuasive because, in addition to having almost nothing to lose by objecting, Bowers had a lot to gain. His client was prejudiced by proceedings at trial, and at that point, Bowers had two options: allow the trial to proceed, despite the obvious downside that the trial was infected with prejudice against his client, or object and request a mistrial, which has no perceptible downside and which would have given Bates at least a shot at a new trial uninfected by prejudice. Faced with these options, competent attorneys do not opt for silence.
. The Majority explains that Bowers admitted only that the jury could have drawn a prejudicial inference against Bates based on the prayer and subsequent testimony. From this, the Majority concludes that we would merely be speculating about whether or not prejudice occurred, which is not enough to establish a Strickland claim. This argument misses the point. Bates’s entire argument is that his counsel’s barometer for measuring prejudice was not functioning properly, so it does not do much good for us to rely heavily on that barometer now. While Bowers perceived no prejudice at the time and merely speculated after the fact that there might have been prejudice to Bates, Bates claims there was prejudice at trial but only realized later that the potential was there. More to the point, we are not here to evaluate what Bowers believed; instead, we are here to evaluate what objectively competent counsel would have believed and done. I think competent counsel would have assessed, on the spot, that the prejudice was real, not speculative, and I think competent counsel would have done something about it.
That the record does not reveal much about how prejudicial the testimony in question truly was does not, as the Majority suggests, support the government’s position. Instead, the silence in the record is a consequence of Bowers’s failure to perceive the prejudice; it *1311is not a sign that there was no prejudice. We know that testimony linked the minister, and thus the trial's opening prayer, to the victim, and we know that Bowers did nothing about it, leaving the record on this point underdeveloped. All this tells us is that Bowers failed to react, not that he had no reason to react.
. Further, from the standpoint of a criminal defense attorney operating in state court, that the Supreme Court has no clearly established precedent does not necessarily suggest that *1313there is a gap in case law at all. If the Florida Supreme Court had clearly established precedent interpreting the Establishment Clause to forbid prayers- at the beginning of a criminal trial and explaining that a mistrial is the only adequate remedy, then the absence of clearly established federal law would be entirely irrelevant to our Strickland analysis.
. The Majority also cites United States v. Walker, 696 F.2d 277, 282 (4th Cir. 1982). To be sure, that case advances the argument the Majority unnecessarily makes here that Bowers’s, objection, had he raised it, might not have been sustained. But the fact that the Majority has to resort to out of Circuit precedent to support its point shows that the question of what might have happened had Bow*1314ers objected is an open one in this Circuit. It also underscores how implausible it is to suggest that Bowers was competent based on the assumption that he did not raise an objection because he believed the state trial court would extend the law of our sister Circuit to overrule his objection. It would have been the government’s job to argue why the state court should adopt that precedent, and it is competent defense counsel’s job, if the objection would benefit his client, to explain why the Fourth Circuit's non-binding case was wrongly decided or is distinguishable. Under the Majority's view, the government’s job would become very easy, because as soon as an adverse ruling comes out in any federal court agáinst a defendant’s position, apparently competent defense counsel are no longer expected to raise the objection in any other jurisdiction.
. Alternatively, Bates maintains that the application of the statute as revised by the Florida Legislature in 1994 would not violate the prohibition against ex post facto application of laws. See Weaver v. Graham, 450 U.S. 24, 29, 101 S.Ct. 960, 964, 67 L.Ed.2d 17 (1981) (providing a two prong test to determine if a statute violates the ex post facto prohibition, asking (1) is the law retrospective, and if so, (2) if it is disadvantageous to the offender). Here, Bates argues that the amended § 775.082(1), if applied, would be advantageous, not disadvantageous, under the circumstances of his resentencing.
. Bates's claim regarding the relevance of his additional life sentences to the resentencing jury is within the broad scope of our COA, which authorized Bates to address ‘‘[wjhether the Florida Supreme Court’s rejection of Appellant’s claim that the trial court’s refusal to instruct the jury about Appellant’s parole eligibility, including the effect of consecutive sentences he had left to serve, was contrary to law established by the United States Supreme Court or objectively unreasonable in light of such precedent.” I disagree with the Majority’s argument that even if Bates’s claim is included in our COA, he has abandoned this claim by failing to "plainly and prominently” argue it on appeal. Maj. Op. at 1304, n. 13 (citing Sapuppo v. Allstate Floridian Ins. Co., *1315739 F.3d 678, 681 (11th Cir.2014)). Bates has cited the relevant Supreme Court precedents in his appellate brief and specifically argued that the resentencing jury’s lack of information about his additional consecutive sentences was unconstitutional. That is enough to place the issue before us.
. To understand Bates’s argument, it is critical to understand the options presented to the jury. The jury was given two choices: vote for death or for a term of life with the possibility of parole after 25 years. Bates had already served 13 years in prison, so the jury believed it had a choice between death or a life sentence with the possibility that Bates would be a free man in as few as 12 years. The actual consequences of the jury’s vote were quite different. Bates had been convicted of three other crimes, leading to two additional life sentences-and a 15-year sentence, all to run consecutively. As the Majority explains, each of these sentences carried the possibility of parole. For each of the two life sentences, Bates would serve a minimum of five years, and for the 15-year sentence, he would serve at least an additional two years. In total, these other sentences guaranteed that Bates would spend, at the very least, an additional 12 years in prison without eligibility for parole, on top of whatever sentence the capital sentencing jury selected. Thus, while the jury believed that if it did not vote for death, Bates might be free in 12 years, in reality, if the jury did not vote for death, Bates could not have been paroled for at least 24 years. In other words, a vote for life would have left Bates incapacitated for at least twice as long as the jury believed.
Misinformed as it was, the jury’s vote for death was still fairly close: 9-3. It seems reasonable to infer that some jurors who did not believe that 12 years of incapacitation was enough might have believed that 24 years was.
. Quite apart from the constitutional question presented here, it also seems to me that when we ask jurors to make morally difficult life-and-death decisions, we ought to fully inform them of the actual consequences of their choices.
. Further, I do not agree that this Court’s decision in Booker v. Secretary, Florida Department of Corrections, 684 F.3d 1121, 1126 (11th Cir.2012), forecloses Bates’s claim. In that case, which presented itself upon habeas review, we found that even if the Florida Supreme Court's resolution of the claim "clearly violates the spirit of ... Simmons, that does not mean that it constitutes an unreasonable application of clearly established federal law, which thus far has only addressed jury instructions in the circumstance of statutory parole ineligibility." Booker, 684 F.3d at 1126 (internal quotation marks omitted). At most, Booker says that violating the spirit of Simmons is not contrary to clearly established law. To be clear, I believe that Simmons and Ramdass do not foreclose relief on this claim, and that relief is supported by the Supreme Court's decision in Lockett and Mills, but I ultimately conclude that, given AEDPA deference, the Florida Supreme Court’s adjudication falls short of violating clearly established federal law of the Supreme Court.
. We do recognize, however, that while Bates’s counsel asked the resentencing judge for an instruction regarding his two additional life sentences and his 15-year sentence, to be served consecutively, the only absolutely definitive period of incarceration, under Florida law at the time, appears to be 12 years to run consecutively to Bates’s punishment for first degree murder. The trial judge, when confronted with either counsel’s request or the jury’s subsequent question, could have explained this to the jury.