Opinion of the Court by
Justice ROACH.I. INTRODUCTION
Appellant was convicted of being a second-degree persistent felony offender and his sentences for attempted rape and kidnapping were enhanced. On appeal, he contends that the trial court erred (1) by refusing to allow mitigation evidence and (2) by requiring that he remain in leg shackles in front of the jury. Finding no error, we affirm Appellant’s convictions.
II. BACKGROUND
Appellant, Ricky Barbour, was convicted of first-degree attempted rape, kidnapping, and fourth-degree assault in November 1994. The jury recommended sentences of ten years for attempted rape, twenty years for kidnapping, and twelve months and a $500 fine for fourth-degree assault, all to run consecutively. The jury also found that Appellant was a second-degree persistent felony offender (“PFO II”), and recommended enhanced sentences of twenty years for the attempted rape and two hundred years for the kidnapping.
Appellant appealed his conviction to this Court as a matter of right, contending that the trial court erred in admitting evidence of two out-of-state felony convictions, which were used to prove his status as a PFO II. In May 1996, we reversed the PFO II conviction and sentence enhancement and remanded the matter to the Hart County Circuit Court for retrial of the PFO II charge. Barbour v. Commonwealth, NO.1995-SC-000078-MR (Ky. June 13,1996).
On remand, Appellant moved the trial court to bar retrial of the PFO II charge on the grounds of double jeopardy, collateral estoppel, res judicata, due process, and equal protection. The trial court denied Appellant’s motion. He then petitioned the Court of Appeals for a writ of prohibition, which was denied. We affirmed the decision of the Court of Appeals in 2001. Barbour v. Raikes, No.2000-SC-000249-MR (Ky. March 15, 2001).
After the failure of his writ petition, Appellant filed a pretrial motion pursuant to KRS 532.055(2)(b), the truth-in-sentencing statute, to allow him to introduce mitigation evidence at the retrial of the PFO II charge. Specifically, Appellant sought to introduce evidence of his post-conviction conduct, including proof of his completion of the Sexual Offender Treatment Program and other programs in prison, and evidence of his remorse. The trial court granted Appellant’s motion on November 12, 2002.
On July 30, 2004, Appellant filed a motion in limine to limit the retrial “to the PFO phase and not a full truth-in-sentencing proceeding.” On August 9, 2004, the trial court issued an order that granted the motion in limine but that also overruled the previous order allowing the introduction of mitigation evidence at the retrial. The trial court’s order stated in part: “Since the defense cannot introduce mitigation evidence, the Commonwealth is also barred from introducing victim impact evidence. The only issue on retrial is the PFO II charge.” On August 11, 2004, in a conference in the judge’s chambers immediately before the PFO retrial, Appellant mentioned the mitigation evidence issue again, and the judge confirmed that the *609PFO proceeding would not include mitigation or victim impact evidence.
Also at this in-chambers meeting on August 11, 2004, Appellant objected to wearing leg shackles during the PFO proceeding. The Commonwealth responded by requesting that the shackles remain on Appellant. The judge ruled that the Appellant would remain in shackles, but offered to give the jury an admonition regarding the procedure. Appellant expressed concern that such an admonition would draw more attention to the shackles. Appellant then declined the judge’s offer to give the instruction, and no such instruction was given at the PFO proceeding.
The jury found Appellant guilty of being a PFO II and enhanced his sentences accordingly: the sentence for first-degree attempted rape conviction was enhanced from ten years to twenty years and the sentence for kidnapping was enhanced from twenty years to fifty years. The sentences were set to run consecutively for a total of seventy years imprisonment. Appellant appeals to this Court as a matter of right. Ky. Const. § 110(2)(b).
III. ANALYSIS
We address the issues in the order in which they appear in Appellant’s brief.
A. Mitigation Evidence
Appellant first claims that his motion to introduce mitigation evidence during the remanded PFO proceeding was denied improperly. Appellant contends such a ruling contravenes KRS 532.055, which provides for the introduction of mitigation evidence during sentencing, and Boone v. Commonwealth, 821 S.W.2d 813 (Ky. 1992), which held that “[wjhere ... a subsequent separate punishment phase is necessary because of reversible error committed only in that phase at the original trial, common sense dictates that the second jury must be told something about what transpired during the earlier guilt phase if they indeed are not ‘to sentence in a vacuum without any knowledge of the defendant’s past criminal record or other matters that might be 'pertinent to consider in the assessment of an appropriate penalty.’ ” Id. at 813 (quoting Commonwealth v. Reneer, 734 S.W.2d 794, 797 (Ky.1987)).
Although Appellant asks us to consider whether mitigation evidence is admissible during a separate, remanded PFO phase, this issue simply was not preserved for our review. While Appellant had at one time moved the trial court to allow the introduction of mitigation evidence under the truth-in-sentencing statute, we believe that request was effectively withdrawn when he subsequently filed a motion in limine requesting “that the retrial be limited to the PFO phase and not a full truth-in-sentencing proceeding.” The trial court granted Appellant’s motion in limine, thus overruling its previous order granting the motion to allow introduction of mitigation evidence and stating that neither Appellant nor the Commonwealth could introduce evidence under the truth-in-sentencing statute. Appellant’s current challenge to this ruling cannot be justified, given that it was the direct result of his own motion. Otherwise, we would be forced to allow Appellant to pursue inconsistent, if not mutually exclusive, positions, first by seeking to take advantage of the truth-in-sentencing statute, and later moving for its application to be barred. The trial court gave Appellant exactly what he requested in the later motion in limine: a retrial limited to the evidence related to the PFO charge, not a full truth-in-sentencing hearing. The effect of this ruling was that all other evidence, including mitigation evidence, could not be introduced.
*610Appellant claims to have again raised the issue of the introduction of mitigation evidence subsequent to the trial court’s ruling that such evidence was not to be presented. Specifically, he cites statements by his attorney during the in-chambers conference immediately before the PFO proceeding. He claims that his attorney’s statements during that conference amounted to a reassertion of his earlier request to present mitigation evidence. A review of the record, however, reveals that this is a mischaracterization of what occurred at the conference. Appellant’s counsel raised the issue in the conference by saying: “I believe your most recent ruling was that it would be a bare PFO proceeding and that my understanding was there would be neither mitigation nor victim impact.” The judge then described the limited nature of the PFO proceeding and concluded that the truth-in-sentencing only applied to the penalty phase, and not the PFO proceeding. Rather than raising an objection at this point, Appellant’s counsel merely reiterated the judge’s decision to restrict content of the PFO phase by asking: “So it’s my understanding that the jury will be shown the evidence only; no opening, no closing, and not the truth-in-sentencing?” The judge responded affirmatively, and Appellant’s counsel then replied, “OK.” This exchange hardly constitutes an objection. Moreover, this discussion reveals, at the very least, Appellant’s attorney’s complacence with precluding the application of the truth-in-sentencing statute, which contained the provision for the introduction of mitigation evidence. And when viewed in light Appellant’s motion to restrict the evidence to be introduced at the PFO proceeding, it is clear that the conversation during the conference immediately before the proceeding was merely an attempt by Appellant’s attorney to ensure that the constraints on evidence were clear to everyone involved, not a reassertion of Appellant’s rights under the truth-in-sentencing statute.
We cannot consider Appellant’s alleged objection to the exclusion of mitigation evidence as grounds for reversing Appellant’s conviction when the exclusion was prompted by his own motion and was not properly preserved by any subsequent objection.
B. Shackles
Appellant next claims that the trial court improperly ordered that he remain in leg shackles during the PFO proceeding. Under the common law, shackling a defendant during trial, absent exceptional circumstances, was heavily disfavored. See, e.g., Deck v. Missouri, 544 U.S. 622, 626, 125 S.Ct. 2007, 2010, 161 L.Ed.2d 953 (2005) (noting that “[t]his rule has deep roots in the common law” and discussing in some detail the history of the rule). But see id. at 638, 125 S.Ct. at 2017 (Thomas, J., dissenting) (arguing that the common law rule against shackling was motivated by the physical pain caused by the type of restraints employed in the 17th and 18th centuries and their interference with the defendant’s ability to defend himself — not concerns about the presumption of innocence, etc. — and that the rule is therefore not applicable to modern restraints, which are much lighter and less painful). The United States Supreme Court has recognized that the rule is part of the constitutional guarantee of a fair trial. Id. at 629, 125 S.Ct. at 2012 (“[T]he Fifth and Fourteenth Amendments prohibit the use of physical restraints visible to the jury absent a trial court determination, in the exercise of its discretion, that they are justified by a state interest specific to a particular trial.”); see also Illinois v. Allen, 397 U.S. 337, 344, 90 S.Ct. 1057, 1061, 25 L.Ed.2d 353 (1970) (“But even to contemplate such a technique, much less see *611it, arouses a feeling that no person should be tried while shackled and gagged except as a last resort”). Thus, it is clear that this general rule against shackling is an elemental aspect of modern trial practice. Kentucky has gone so far as to codify the rule as part of the Rules of Criminal Procedure: “Except for good cause shown the judge shall not permit the defendant to be seen by the jury in shackles or other devices for physical restraint.” RCr 8.28(5).
There is some question, however, whether the constitutional rule against routine shackling of criminal defendants in the presence of the jury applies beyond the guilt phase of a trial. The United States Supreme Court has recently addressed the issue in Deck v. Missouri, 544 U.S. 622, 626, 125 S.Ct. 2007, 2010, 161 L.Ed.2d 958 (2005). In the first sentence of that opinion, the Court formulated the issue specifically as “whether shackling a convicted offender during the penalty phase of a capital case violates the Federal Constitution.” Id. at 624, 125 S.Ct. at 2009. The Court’s answer to that question is found in the next sentence: “We hold that the Constitution forbids the use of visible shackles during the penalty phase, as it forbids their use during the guilt phase, unless that use is ‘justified by an essential state interest’ — such as the interest in courtroom security — specific to the defendant on trial.” Id. at 624, 125 S.Ct. at 2009. Considering that statement alone, with its lack of limiting language, it is unclear whether the Court intended the rule to apply only in capital cases or to apply broadly in all penalty phase jury proceedings. An examination of the Court’s reasoning for extending the rule in Deck, however, indicates that it is limited and applies only to the penalty phase of capital cases.
In building up to its analysis regarding the use of shackles during the penalty phase of a capital case, the Court in Deck again described the issue under consideration, noting specifically that it was “considering] shackling not during the guilt phase of an ordinary criminal trial, but during the punishment phase of a capital case.” Id. at 630, 125 S.Ct. at 2012. The Court also noted that the modern rule against shackling enjoys a tripartite rationale, namely that the practice might undermine (1) the presumption of innocence and the related fairness of the factfinding process, (2) the exercise of the right to counsel to secure a meaningful defense, and (3) the dignity of the judicial process. Id. at 630-31, 125 S.Ct. at 2013. The Court recognized that the latter two considerations remain during the penalty phase of a capital case, but that the first consideration was less obviously applicable because the presumption of innocence no longer exists. The Court then noted: “Nonetheless, shackles at the penalty phase threaten related concerns. Although the jury is no longer deciding between guilt and innocence, it is deciding between life and death. That decision, given the severity and finality of the sanction, is no less important than the decision about guilt.” Id. at 632, 125 S.Ct. at 2014 (citations and quotation marks omitted). The Court also noted that accurate and reliable decision making is the paramount concern in death penalty cases, but that the implication of dangerousness inherent in the shackling of a defendant could undermine that concern in various ways. Concerned that “[i]n these ways, the use of shackles can be a thumb on death’s side of the scale,” id. at 632-33, 125 S.Ct. at 2014 (internal citations, quotation marks, and brackets omitted), the Court concluded: “Given the presence of similarly weighty considerations, we must conclude that courts cannot routinely place defendants in shackles or other physical restraints visible to the jury during the penalty phase of a capital proceeding.” Id. at 633,125 S.Ct. at 2014.
*612The Court’s reasoning amounts to recognition of the intersection of the modern rationale for the rule against shackling with the Court’s insistence that death penalty cases are different than other criminal prosecutions. This is likely why we have been unable to find a case that applies Deck outside the context of a capital case, and why those courts that have applied, or at least discussed, the case have consistently noted that its essential holding was to extend the rule against routine shackling in the presence of juries in capital proceedings. See, e.g., Lakin v. Stine, 431 F.3d 959, 963 (6th Cir.2005) (“Although Deck was only recently decided, the Supreme Court made clear that the guilt phase rule was clearly established law pri- or to the date of the Deck decision. Deck addressed whether the guilt phase rule is applicable in the capital sentencing context.”); United States v. Howard, 429 F.3d 843, 856 (9th Cir.2005); In re A.H., 359 Ill.App.3d 173, 295 Ill.Dec. 709, 833 N.E.2d 915, 922 (2005); United States v. Honken, 381 F.Supp.2d 936, 980 (N.D.Iowa 2005). "While none of the courts we have reviewed have noted that Deck was limited only to capital proceedings, it is telling that none have said otherwise, even in dicta.
However, we need not wade too deeply into this debate because our Rule of Criminal Procedure on this point does not distinguish between the guilt and penalty phases of a trial. Rather, RCr 8.28(5) bars the routine shackling of a defendant, absent a showing of good cause, whenever he will be seen by the jury. This amounts to an extension of the prohibition of routine shackling to all jury-observed aspects of a criminal trial. Thus, in accordance with our rule, even though Appellant is challenging his shackling during the unique setting of a remanded PFO proceeding, we must still determine whether there was sufficient good cause in this case to justify shackling Appellant during the PFO proceeding.
Shackling of a defendant in a jury trial is allowed only in “the presence of extraordinary circumstances.” Peterson v. Commonwealth, 160 S.W.3d 730, 733 (Ky. 2005). Our long-standing practice has been to limit shackling to specific types of “exceptional cases, .... cases where the trial courts appeared to have encountered some good grounds for believing such defendants might attempt to do violence or to escape during their trials.” Tunget v. Commonwealth, 303 Ky. 834, 836, 198 S.W.2d 785, 786 (1947).1
There have been a few such exceptional cases in which we have upheld the practice of shackling, and in each case the trial court based its decision on specific findings of extraordinary circumstances. In Tun-get, our predecessor court upheld the trial court’s decision to shackle the defendant because of his history of violent escape attempts. The defendant, who was serving a life sentence for a murder conviction, obtained a gun while in jail and used it to trap four prison guards in his cell. The defendant then shot and killed an associate warden. 303 Ky. at 836, 198 S.W.2d at 786. Applying the Tunget standard, we have allowed a defendant to be shackled throughout trial because he had fled the courtroom and courthouse during arraignment, thus “the trial judge had good rea*613son to believe that [the defendant] was a man of sufficiently ‘demonstrated desperation’ that he might make a similar attempt during trial_” Commonwealth v. Conley, 959 S.W.2d 77, 78 (Ky.1997). Similarly, we have held that the trial court did not exceed its discretion in ordering a defendant to remain in leg shackles throughout the trial where his “belligerent conduct prior to trial certainly raised a serious issue of courtroom security,” and he refused “to assure the trial court that he would not engage in any physical or violent outbursts during trial.” Peterson v. Commonwealth, 160 S.W.Bd 730, 734 (Ky.2005). We have also held that shackling was justified where a defendant, who was skilled in martial arts, had successfully escaped once before, and had planned several escape attempts in the past. Hill v. Commonwealth, 125 S.W.3d 221, 235-36 (Ky.2004). These cases illustrate the sort of limited circumstances, complete with specific trial court findings, that have justified allowing a defendant to remain shackled before the jury-
In this case, the shackling of Appellant was not based on any specific finding that the he was violent or a flight risk. In fact, the prosecutor supported his shackling request only with nonspecific concerns about escape risk and safety, stating to the court:
Your Honor, I would ask that the Defendant remain in shackles. He’s dressed in street clothes today. As Your Honor has pointed out, Your Honor will inform the jury that he is already under sentence of the court for thirty years. He would not be prejudiced by the jury knowing or seeing any evidence that he is in fact in custody. As far as a flight risk, his circumstances speak for themselves. He is under sentence, the court has pointed out earlier that he’s already been flopped once. He knows the likelihood of serving out thirty years is very, very real. He has every reason to take advantage of being unshackled in a courthouse, to flee from the courthouse. And for the safety of the courtroom I ask that the shackles remain.
The Commonwealth now points to these statements as proof that the trial court’s decision was justified. Under the Commonwealth’s reasoning, a trial court would be free to predict a defendant’s behavior solely from his status as a convicted felon, without making any specific findings that he posed a risk of violence in or escape from the courtroom. If that were a correct application of the rule, shackling would be justifiable in almost any serious felony case and would likely become a routine practice. But as our predecessor court noted of the practice, “[a] court would hardly be justified in permitting this to be done in one murder case out of an average hundred coming to trial.” Tunget, 303 Ky. at 836,198 S.W.2d at 786.
In a case similar to this one, the Sixth Circuit refused to endorse such general reasons for shackling. Lakin v. Stine, 431 F.3d 959 (6th Cir.2005). In that case, the defendant was on trial for attempted escape, and the court held that it was erroneous to shackle him merely due to the nature of his charge: “Lakin was on trial for attempted escape, but that fact alone[ ] is not determinative. The nature of the charges against a particular defendant cannot themselves provide the entire justification for shackling; rather, all of the relevant factors must be considered, including alternative means of providing] a safe and fair trial.” Id. at 965. It was similarly unfair in this case for the trial court to impose shackles merely because Appellant has already been given a lengthy sentence.
In addition to the general justifications offered by the prosecutor immediately preceding the PFO retrial, the Commonwealth now cites an episode of verbal disobedience by Appellant as further justification for shackling him during the PFO *614proceeding. During a pretrial conference, Appellant complained about his attorney to the trial court. Appellant became upset at the conclusion of the hearing and engaged in the following exchange with the judge:
Judge: I don’t want to hear anything more.
Appellant: I’m gonna speak when I wanna speak.
Judge: No, you’re not, not in my court. Appellant: If it concerns me being locked up, I am.
The Commonwealth claims that Appellant’s tone of voice during this discussion showed his hostility. Nonetheless, little in this exchange raises a concern that Appellant would be violent or would try to escape. Indeed, a single display of displeasure or disrespect by a criminal defendant rarely rises to such a level as to justify shackling before the jury. Furthermore, we must also note that the pretrial conference during which this discussion occurred took place more than a year and a half before the PFO proceeding. More importantly, however, this episode was not offered at trial as part of the prosecutor’s purported justification for shackling Appellant in front of the jury. Rather, the incident has only been raised on appeal. That the incident, if it can even be called that, was not even mentioned at trial indicates it was hardly memorable and likely was not even considered by the trial judge in allowing Appellant to remain shackled.
All that being said, a trial court’s decision to keep a criminal defendant shackled before the jury is usually accorded a great deal of deference. See Tunget, 303 Ky. at 836, 198 S.W.2d at 786 (“It appears to be the generally prevailing rule in many jurisdictions of this country that trial courts will be upheld in their exercise of thoroughly sound and reasonable discretion in the matter of keeping certain defendants in shackles during their trials.”). However, in light of RCr 8.28(5)’s requirement of a showing of “good cause” for allowing the practice and the lack of any substantive evidence or finding by the trial court that Appellant was either violent or a flight risk, it is clear that the decision to require Appellant to appear at the PFO hearing in shackles was not justified.2 We conclude that the decision to do so in this case was an abuse of discretion.
This error, however, is subject to the harmless error rule. Under RCr 9.24, we “must disregard any error or defect in the proceeding that does not affect the substantial rights of the parties.” In light of this standard, we have held that “[t]he doctrine of nonprejudicial error, sometimes called ‘harmless error,’ is that in determining whether an error is prejudicial, an appellate court must consider whether on the whole case there is a substantial possibility that the result would have been any different.” Commonwealth v. McIntosh, 646 S.W.2d 43, 45 (Ky.1983).3
*615In weighing the impact of shackles being worn on the jury’s decision to find Appellant guilty of being a PFO II, we must consider that Appellant clearly met all the statutory requirements for being a PFO II under KRS 532.080. “Conviction as a Persistent Felony Offender is not a charge of an independent criminal offense but rather a particular criminal status.” White v. Commonwealth, 770 S.W.2d 222, 224 (Ky.1989). We have even gone so far as to state that “[i]n order to establish a persistent felony offender status, the Commonwealth merely needs to establish a simple check list of technical statutory requirements.” Pace v. Commonwealth, 636 S.W.2d 887, 890 (Ky.1982), overruled in part on other grounds by Commonwealth v. Harrell, 3 S.W.3d 349 (Ky.1999); see also White v. Commonwealth 611 S.W.2d 529, 531 (Ky.App.1980) (“The sole issue to be determined in a persistent felony offender phase of a trial is that of status; that is, does the accused meet the statutory definition of being a persistent felony offender”).
Given that the PFO proceeding is essentially a status determination, we have no doubt that the jury would have found Appellant guilty of being a PFO II whether the shackling error occurred or not. Second-degree-persistent-felony-offender status is defined in part as follows:
A persistent felony offender in the second degree is a person who is more than twenty-one (21) years of age and who stands convicted of a felony after having been convicted of one (1) previous felony. As used in this provision, a previous felony conviction is a conviction of a felony in this state or conviction of a crime in any other jurisdiction provided:
(a)That a sentence to a term of imprisonment of one (1) year or more or a sentence to death was imposed therefor; and
(b) That the offender was over the age of eighteen (18) years at the time the offense was committed; and
(c) That the offender:
[[Image here]]
2. Was on probation, parole, conditional discharge, conditional release, furlough, appeal bond, or any other form of legal release from any of the previous felony convictions at the time of commission of the felony for which he now stands convicted ....
KRS 532.080(2).
At the PFO proceeding, the prosecutor introduced evidence that Appellant satisfied each of these requirements. The prosecutor relied on two prior convictions, one committed in Arizona and one committed in Kentucky, to prove Appellant’s PFO II status. The judgment from Arizona, which was certified by the trial court, revealed that Appellant had committed an offense in January 1991 and was sentenced to three years for that offense in January 1993. The jury was also offered proof of a second offense that was committed in Kentucky in May 1994, while Appellant was on parole from his conviction in Arizona. Appellant’s counsel stipulated that Appellant’s age had been established by Appellant’s own statement that he was thirty-six years old as of May 1994. This statement was made in open court and was later shown to the PFO jury via videotape. Appellant’s statement established that he was over age eighteen when he committed his first offense in 1991 and over twenty-one when he committed his second offense in 1994. This evidence clearly established all of the requirements necessary to find Appellant guilty of being a PFO II, and Ap*616pellant did not introduce any contrary evidence.
In light of the overwhelming, unrebutted evidence that Appellant met the statutory requirements of being a PFO II, we conclude that the outcome would not have been different had Appellant appeared before the jury free of shackles. See Lakin v. Stine, 431 F.3d 959, 966 (6th Cir.2005) (“Despite the substantial risk of prejudice that shackles pose, we are compelled to conclude that the error was harmless in this case due to the overwhelming evidence against [the defendant].”). Moreover, given that a PFO proceeding is merely a status determination, it is unlikely that Appellant’s shackles contributed in any way to the jury’s finding. Therefore, given the circumstances of the remanded PFO proceeding and the overwhelming evidence of guilt, we hold that requiring Appellant to remain in shackles was harmless error, and did not affect Appellant’s substantial rights.
IV. CONCLUSION
For the foregoing reasons, we affirm the judgment of the Hart Circuit Court.
LAMBERT, C.J.; GRAVES, McANULTY, MINTON and WINTERSHEIMER, JJ., concur. SCOTT, J., concurs in result only.. We also note that the federal courts apply a similar standard when the constitutional preference against shackling is in play: “CT]he Constitution forbids the use of visible shackles ... unless that use is justified by an essential state interest — such as the interest in courtroom security — specific to the defendant on trial.” Deck, 544 U.S. at 624, 125 S.Ct. at 2009 (internal quotation marks omitted). Though, as noted above, the constitutional right is not in effect in this case, because our criminal rule simply extends the protection against shackling to all aspects of the jury trial, we note that the federal case law is persuasive authority as to whether the shackling was erroneous.
. We also note that when the trial judge offered to admonish the jurors about the effect of the shackling, he indicated he would tell them that the shackles were a policy of the sheriff's department. Such a policy belies the Commonwealth’s contention that Appellant’s shackling was not a routine undertaking. However, it is unclear from the record whether such a policy actually existed or the trial judge was simply attempting to reduce the prejudice of shackling to Appellant. As such, we do not rely heavily on this part of the record in reaching our decision.
. Though the error involved in this case was a violation of our Rules of Criminal Procedure, rather than a Constitutional right, we note that even violations of the constitutional rule against shackling are subject to harmless error analysis. See Deck, 544 U.S. at 635, 125 S.Ct at 2016 ("[Wjhere a court, without adequate justification, orders the defendant to wear shackles that will be seen by the juiy, the defendant need not demonstrate actual prejudice to make out a due process violation. The State must prove 'beyond a reasonable doubt that the [shackling] error complained *615of did not contribute to the verdict obtained.’ Chapman v. California, 386 U.S. 18, 24, 87 S.Ct. 824, 17 L.Ed.2d 705 (1967).” (alteration in original)).