dissenting.
Judge Garippo’s participation at DelVee-ehio’s trial did not deny DelVecehio a fair trial. I also disagree with the court that before Illinois can use DelVecchio’s confession against him, the state must prove at a hearing that the confession was voluntary. Therefore, I respectfully dissent.
I.
George DelVecehio comes before us having twice been convicted of murder. In 1965, during the course of a two-day crime spree, DelVecehio, then sixteen years old, pumped at least nine bullets into Fred Christiansen, an elderly man who had the misfortune of being alone in the same area where DelVec-chio and two companions were looking for somebody to rob. DelVecehio fired several of the shots into Christiansen while an accomplice was kicking the dying man in an attempt to silence his cries for help. After *520finally silencing Christiansen, DelVecchio and his compatriots made off with Christiansen’s wallet, which contained $11. DelVecchio confessed to killing Christiansen and pleaded guilty to the murder..
Twelve years later, DelVecchio broke into the apartment of Karen Canzoneri, apparently for the purpose of raping her. While in her apartment and before the rape, DelVec-chio slit the throat of her six-year old son, Tony. This court’s opinion graphically describes the damage DelVecchio’s knife did to Tony; to put it more simply, DelVecchio nearly decapitated the boy.
At his trial for Tony’s murder, DelVecchio was represented by competent counsel and had full opportunity to confront the witnesses against him. He did not contest the fact that he killed Tony; his defense, instead, was that he was too intoxicated by drugs to have formed the intent to commit murder. A jury heard all the evidence and found DelVecchio guilty beyond a reasonable doubt of murder, burglary, and deviate sexual conduct, and at a separate hearing sentenced him to death. The Illinois Supreme Court twice has upheld DelVecchio’s conviction and sentence. See People v. DelVecchio, 105 Ill.2d 414, 86 Ill. Dec. 461, 475 N.E.2d 840 (1985) (direct appeal); People v. DelVecchio, 129 Ill.2d 265, 135 Ill.Dec. 816, 544 N.E.2d 312 (1989) (appeal from denial of post-conviction relief).
Yet, the court holds today that DelVec-chio’s conviction and sentence must be set aside because DelVecchio did not receive a fair trial. This is so, says the court, because Judge Garippo, who presided at the Canzon-eri trial in 1979, had been involved in DelVec-chio’s prosecution in 1965. Judge Garippo was not the prosecutor of record in the Christiansen case. But Judge Garippo was chief of the criminal division of the Chicago State’s Attorney’s Office, and in 1965 he made the decisions to try DelVecchio as an adult rather than a juvenile and to expedite DelVecchio’s indictment so that DelVecchio could plead guilty and be sentenced before his seventeenth birthday. Judge Garippo also assigned his former trial partner in the State’s Attorney’s office to the Christiansen case and attended DelVecchio’s guilty plea hearing in 1965 as a spectator.
The court holds that Judge Garippo’s participation in the Christiansen prosecution violated due process in two ways. First, the court holds that any finding of actual bias or prejudice is unnecessary. Instead, as I understand the court’s opinion, Judge Garippo’s participation in the Christiansen prosecution created what the court characterizes as an “extreme conflict of interest” and presented him with a temptation to be biased against DelVecchio; this temptation, in turn, made it appear improper for Judge Garippo to preside at the trial. The court concludes that because of this appearance, Judge Garippo was required to disqualify himself from the Canzoneri case. His failure to do so violated DelVecchio’s right to due process, even if Judge Garippo was scrupulously fair in conducting the trial. Second, the court holds that even if a showing of actual bias is required, Judge Garippo’s participation in the Canzoneri trial violated due process.
I disagree with the court’s conclusions. An appearance of bias is not a sufficient ground to require disqualification of a judge absent an actual, substantial incentive to be biased. Judge Garippo’s participation in the Christiansen prosecution was not a sufficiently substantial biasing influence to require Judge Garippo’s disqualification in the Can-zoneri trial. Nor is the evidence in this case sufficient to show that Judge Garippo was actually biased against DelVecchio.
Certainly the appearance of justice is important in our system and the due process clause sometimes requires a judge to recuse himself from a case without a showing of actual bias, where a sufficient motive to be biased exists. A long line of Supreme Court cases compel these general conclusions. In Tumey v. Ohio, 273 U.S. 510, 532, 47 S.Ct. 437, 444, 71 L.Ed. 749 (1927), the Court stated that “[ejvery procedure which would offer a possible temptation to the average man as a judge ... not to hold the balance nice, clear, and true between the state and the accused denies the latter due process of law.” Where such temptations exist, the due process clause “may sometimes bar trial by judges who have no actual bias and who would do their very best to weigh the scales of justice equally between contending par*521ties.... [T]o perform its high function in the best way, ‘justice must satisfy the appearance of justice.’ ” In re Murchison, 349 U.S. 133, 136, 75 S.Ct. 623, 625, 99 L.Ed. 942 (1955); Aetna Life Ins. Co. v. Lavoie, 475 U.S. 813, 825, 106 S.Ct. 1580, 1587, 89 L.Ed.2d 823 (1986).
The court’s presentation of the facts surrounding Judge Garippo’s contacts with Del-Vecehio succeeds in leaving the impression that Judge Garippo had some “temptation” to be biased and that it did not look good for Judge Garippo to preside at DelVecchio’s trial. Indeed, given that it is possible to paint the facts in as bad a light as the court has done here, it probably would have been prudent for Judge Garippo to have disqualified himself from presiding at DelVecchio’s trial. But the question in this case is not about my notion of prudence or whether any judge on this panel would have disqualified himself if faced with the circumstances Judge Garippo faced. See Whitacre v. Davey, 890 F.2d 1168, 1171 (D.C.Cir.1989). The question, rather, is whether Judge Garippo denied DelVeechio a fair trial simply by presiding at DelVecehio’s trial.
Despite the Supreme Court’s broad pronouncements about “the appearance of justice” and “possible temptations,” we cannot answer the fair trial question simply by concluding that it looked bad for Judge Garippo to preside at trial. If the question truly is whether a defendant received a fair trial, bad appearances alone should not require disqualification to prevent an unfair trial. What may appear bad to an observer, especially in hindsight, may not have influenced — or, more importantly, may not have had any real possibility to influence — the judge in his decision-making process. Suppose a judge does not know a close relative has a financial interest in a case he tries. To the outside observer aware of the interest but unaware of the judge’s lack of knowledge, it would look bad for the judge to try that case. But if the judge does not even know about the relative’s financial interest, how could he be tempted to throw the case? And if no actual incentive exists for the judge to be biased — if the judge does not have reason to be partial — how could the judge’s presiding over the trial deprive a party of his right to a fair trial before an impartial judge? See Bradshaw v. McCotter, 796 F.2d 100, 101-03 (5th Cir.1986) (Gee, J., concurring.).
Bad appearances alone do not require disqualification. Reality controls . over uninformed perception. This court cites no case in which the Supreme Court has overturned a verdict on due process grounds based on a mere appearance of bias. Our court was correct in holding that “a litigant is not denied due process by either the ‘appearance’ of partiality or by circumstances which might lead one to speculate as to a judge’s impartiality.” Margoles v. Johns, 660 F.2d 291, 296 (7th Cir.1981). When the Supreme Court talks about the “appearance of bias,” it is not saying that bad appearances alone require disqualification; rather, it is saying that when a judge is faced with circumstances that present “some [actual] incentive to find one way or the other” or “a real possibility of bias,” a court need not examine whether the judge actually was biased. Id. at 297 (quoting Howell v. Jones, 516 F.2d 53, 57-58 (5th Cir.1975)); Bradshaw, 796 F.2d at 102 (Gee, J., concurring). Absent the incentive for bias, however, disqualification is not required despite bad appearance.
Moreover, while it might be fair to say that Judge Garippo faced some “possible temptation” to be biased against DelVeechio, not every “possible temptation” to be biased presents a sufficient possibility of bias to require disqualification. This is evident from other language in the same eases in which the court talks about “possible temptations.” Thus, in Aetna and Tumey, the Court noted that “ ‘[n]ot all questions of judicial qualification ... involve constitutional validity-[Mjatters of kinship, personal bias, state policy, remoteness of interest, would generally be matters of legislative discretion’” — even though any of these matters, particularly kinship and personal bias, could offer a “possible temptation” to be biased. Aetna, 475 U.S. at 820, 106 S.Ct. at 1584; Tumey, 273 U.S. at 523, 47 S.Ct. at 441. At some point, a “ ‘biasing influence ... will be too remote and insubstantial to violate the constitutional constraints.’ ” Aetna, 475 U.S. at 826, 106 S.Ct. at 1588 (quoting Marshall v. Jerrico, 446 *522U.S. 238, 243, 100 S.Ct. 1610, 1614, 64 L.Ed.2d 182 (1980)).
This merely recognizes, at least implicitly, that in the real world, “possible temptations” to be biased abound. Judges are human; like all humans, their outlooks are shaped by their lives’ experiences. It would be unrealistic to suppose that judges do not bring to the bench those experiences and the attendant biases they may create. A person could find something in the background of most judges which in many cases would lead that person to conclude that the judge has a “possible temptation” to be biased. But not all temptations are created equal. We expect — even demand — that judges rise above these potential biasing influences, and in most cases we presume judges do. Compare Withrow v. Larkin, 421 U.S. 35, 47, 95 S.Ct. 1456, 1464, 43 L.Ed.2d 712 (1975), in which the Supreme Court held that the contention that the combination of investigative and adjudicatory functions necessarily disqualifies an administrative adjudicator “must overcome a presumption of honesty and integrity in those serving as adjudicators.”
The common law recognized this reality. At the time the American court system was established, the common law of disqualification “was clear and simple: a judge was disqualified for direct pecuniary interest and for nothing else.” John P. Frank, Disqualification of Judges, 56 Yale L.J. 605, 609 (1947). No other interest would suffice to require, or even permit, a judge to disqualify himself, including evidence of bias against a litigant. Id. at 611-12; see also Aetna, 475 U.S. at 820, 106 S.Ct. at 1585. As Blackstone put it, “the law will not suppose a possibility of bias or favour in a judge, who is already sworn to administer impartial justice, and whose authority greatly depends upon that presumption and idea.” 3 W. Blackstone, Commentaries *361 (quoted in Aetna, 475 U.S. at 820, 106 S.Ct. at 1585).
As the common law recognized, and as experience teaches, the lure of lucre is a particularly strong motivation, and therefore judges ought to be prohibited from presiding over cases in whose outcomes they have a direct financial interest. Of course, the Supreme Court has held the due process clause requires disqualification for interests besides pecuniary interests. But the constitutional standard the Supreme Court has applied in determining when disqualification is necessary recognizes the same reality the common law recognized: judges are subject to a myriad of biasing influences, judges for the most part are presumptively capable of overcoming those influences and rendering evenhanded justice, and only a strong, direct interest in the outcome of a case is sufficient to overcome that presumption of evenhandedness.
The Supreme Court’s disqualification cases illustrate this point. The cases requiring disqualification all involved “direct, personal [and] substantial” influences on the judges involved. Aetna, 475 U.S. at 822, 106 S.Ct. at 1585. In each of these cases, it is fair to say that the influences involved struck at the heart of human motivation, that an average man would find it difficult, if not impossible, to set the influence aside.
In both Tumey and Aetna, for example, judges were faced with “direct, personal, substantial, pecuniary interest[s].” See Aetna, 475 U.S. at 824, 106 S.Ct. at 1586. In Tumey, the judge in a criminal ease was paid only if the defendant was convicted. 273 U.S. at 520, 47 S.Ct. at 440. In Aetna, Justice Embry of the Alabama Supreme Court cast the deciding vote and wrote the majority opinion in a case establishing a legal proposition that “had the clear and immediate effect of enhancing both the legal status and settlement value” of two pending cases Justice Embry had filed as a plaintiff. 475 U.S. at 822-824, 106 S.Ct. at 1585-1587.
The judge in Ward v. City of Monroeville, 409 U.S. 57, 93 S.Ct. 80, 34 L.Ed.2d 267 (1972), faced a temptation similar to those faced by the judges in Tumey and Aetna. In Ward, the Court held that the mayor of Monroeville could not sit as a judge in traffic court. The mayor was responsible for the town’s finances and revenue production; Monroeville derived a “major part” of its income from fines and other costs imposed in that court. The Court sensibly held that the mayor’s responsibility for town finances “may make him partisan to maintain the high level of contribution from the mayor’s court.” *523Id. at 60, 93 S.Ct. at 83. That interest, like the judge’s personal financial interest in Tu-rney, provided the mayor sufficient incentive to find against the defendant that the mayor could not, consistently with the due process clause, sit as judge. See id.
The Court has also required disqualification in the face of a litigant’s direct personal insults to a judge. In Mayberry v. Pennsylvania, 400 U.S. 455, 91 S.Ct. 499, 27 L.Ed.2d 532 (1971), the defendant during the course of his trial caled the judge, among other epithets, a “dirty [S.O.B.j,” a “dirty tyrannical old dog,” a “stumbling dog,” and a “fool,” had charged the judge with running a “Spanish Inquisition,” and had told the judge to “Go to hell.” Id. at 466, 91 S.Ct. at 505. The Court held that the judge could not subsequently try the litigant for contempt in the face of this abuse. The litigant’s insults were “apt to strike ‘at the most vulnerable and human qualities of a judge’s temperament.’ ” Id. (quoting Bloom v. Illinois, 391 U.S. 194, 204, 88 S.Ct. 1477, 1483, 20 L.Ed.2d 522 (1968)). See also Taylor v. Hayes, 418 U.S. 488, 94 S.Ct. 2697, 41 L.Ed.2d 897 (1974) (judge who had become embroiled in a “running controversy” with an attorney that resulted in “marked personal feelings ... on both sides,” and during which the judge had displayed an “unfavorable personal attitude” toward the attorney, could not try the attorney for contempt).
In In re Murchison, the Court held that the combination of prosecutorial and adjudicatory functions in that case violated due process. In Murchison, a state court judge acted as a “one-man grand jury” investigating police corruption. Two witnesses before the judge in his role as grand juror answered questions in such a way as to, convince the judge that those witnesses had committed contempt. The judge charged the two witnesses with contempt and subsequently tried and convicted them. 349 U.S. at 134-135, 75 S.Ct. at 624-625. The Supreme Court held that the trial before the same judge who brought the contempt charges violated the defendants’ right to due process. “Having been part of [the accusatory process] a judge cannot be, in the very nature of things, wholly disinterested in the conviction or acquittal of those accused.” Id. at 137, 75 S.Ct. at 625-626.
But a comparison of the situations in which the Court required disqualification with situations in which the Court did not find disqualification was required makes it clear that not all “possible temptations” toward bias require a judge to disqualify himself. For example, in Aetna, the Court held that Justice Embry’s general antipathy toward and frustration with insurance companies did not require him to disqualify himself. 475 U.S. at 820-21, 106 S.Ct. at 1584-1585. “[O]nly in the most extreme of cases” would the Constitution require disqualification for this type of general bias. Id. at 821, 106 S.Ct. at 1585.
Likewise, not all contemptuous conduct by a person prevents a judge from trying the person for contempt. The Court refused to hold a judge disqualified from trying a witness for contempt in Ungar v. Sarafite, 376 U.S. 575, 84 S.Ct. 841, 11 L.Ed.2d 921 (1964), despite the witness’ continued criticism of the judge, disobedience to his orders during trial, and “disruptive, recalcitrant and disagreeable commentary.” Id. at 584, 84 S.Ct. at 847. One might suppose this disobedience and criticism offered a “possible temptation” to be biased against the defendant. But the judge had not become “personally embroiled” with the litigant, and the Court refused to “assume that judges are so irascible that they cannot fairly deal with resistance to their authority or with highly charged arguments about the soundness of their decisions.” Id.
The line between interests that require disqualification and those that do not is not always clear. Consider the contrast between Murchison and Withrow v. Larkin, 421 U.S. 35, 95 S.Ct. 1456, 43 L.Ed.2d 712 (1975), a case involving similar circumstances. In Withroiv, the Wisconsin Medical Examining Board commenced an investigation to determine whether a doctor had committed certain illegal acts. The Board subsequently decided to hold a hearing to determine whether the doctor had committed the alleged illegal acts and whether to suspend the doctor’s license temporarily. 421 U.S. at 39-41, 95 S.Ct. at 1460-1461. The Supreme Court held that the Board could adjudicate the *524same charges it had investigated and decided to prosecute without violating the doctor’s due process rights. See id, at 47-55, 95 S.Ct. at 1464-1468.
One may wonder why the Court found a due process violation in Murchison but not Withrow. In both eases, the same person (or body) had served as both prosecutor and adjudicator. In both cases, the adjudicators seemed to have had a “possible temptation ... not to hold the balance nice, clear, and true.... ” But the Court distinguished Withrow from Murchison; the procedure in Murchison violated due process “not only because the judge in effect became part of the prosecution and assumed an adversary position, but also because as a judge, passing on guilt or innocence, he very likely relied on ‘his own personal knowledge and impression of what had occurred in the grand jury room’ an impression that ‘could not be tested by adequate cross-examination.’ ” 421 U.S. at 53, 95 S.Ct. at 1467 (quoting Murchison, 349 U.S. at 138, 75 S.Ct. at 626).
Not even all financial interests are disqualifying. In Aetna, for instance, six other judges were members of the plaintiff class in the two suits Justice Embry had filed. Although those six justices conceivably had a “slight pecuniary interest” in the outcome— and, therefore, a “possible temptation” to be biased — the Court held that their disqualification was not required. 475 U.S. at 825-826, 106 S.Ct. at 1587-1588. The justices’ interests, though pecuniary, were not “direct, personal, and substantial.... ” Id. Any gain to these justices from a favorable decision in Aetna was “speculative and contingent;” the class had not yet been certified, and no class relief had been awarded. Id.
The Court’s cases require that we go beyond generalizations about “possible temptations” in deciding whether Judge Garippo was required to disqualify himself. The question is not whether some possible temptation to be biased exists; instead, the question is, when does a biasing influence require disqualification absent a showing of actual bias? Consistent with the common law, we begin in answering this question by presuming “the honesty and integrity of those serving as adjudicators.” Withrow, 421 U.S. at 47, 95 S.Ct. at 1464; Dyas v. Lockhart, 705 F.2d 993, 997 (8th Cir.1983). Disqualification is required only when the biasing influence is strong enough to overcome that presumption, that is, when the influence is so strong that we may presume actual bias. See Dyas, 705 F.2d at 996-97. This occurs in “situations ... in which experience teaches that the possibility of actual bias is too high to be constitutionally tolerable.” Withrow, 421 U.S. at 47, 95 S.Ct. at 1464. A court must be convinced that a particular influence, “under a realistic appraisal of psychological tendencies and human weakness,” poses “such a risk of actual bias or prejudgment that the practice must be forbidden if the guarantee of due process is to be adequately implemented.” Id.
Judged by this standard, Judge Garippo was not required to disqualify himself unless he was actually biased against DelVeechio. Judge Garippo faced none of the biasing influences involved in the cases in which the Supreme Court required disqualification. Judge Garippo had no financial interest— direct or indirect — in the outcome of DelVec-chio’s trial. He had never been subject to any personally insulting, abusive, or even disrespectful remarks by DelVeechio or his attorneys. He did not serve the dual role of prosecutor and judge in the Canzoneri trial.
Judge Garippo had been involved in some aspects of DelVeechio’s prosecution for murder in 1965 — fourteen years before the Can-zoneri trial. But as both the Fifth and Sixth Circuits have held, no per se rule disqualifies a judge because he has prosecuted a defendant in the past. Corbett v. Bordenkircher, 615 F.2d 722, 723-24 (6th Cir.1980); Murphy v. Beto, 416 F.2d 98, 100 (5th Cir.1969). This rule is consistent with the Supreme Court’s cases. Prosecuting a defendant in one ease is not the kind of action from which we can presume bias or prejudgment in a future ease.
The court finds grounds for Garippo’s disqualification not in the mere fact that he was involved in the prosecution of DelVeechio in 1965, but rather in the intersection of the 1965 and 1979 cases. Specifically, despite the overall aura of impropriety the court attempts to create from Judge Garippo’s involvement in the 1965 prosecution, the court at bottom finds that the due process clause required Judge Garippo to recuse himself for *525two reasons: first, because Judge Garippo in 1979 “was forced to sit in judgment of his own decisions as a prosecutor” in 1965; and second, because “it would have been natural” for Judge Garippo “to hold a grudge” against DelVecchio because Judge Garippo had given DelVecchio a break in 1965 by allowing him to be sentenced as a juvenile rather than as an adult. Neither of these perceived influences is sufficient to require Judge Garippo to disqualify himself from the Canzoneri trial absent actual bias.
DelVecchio did file motions to exclude use of his 1965 confession and conviction in the Canzoneri trial, and Judge Garippo did have to rale on these motions. But as the Court in Murchison stated, the interest necessary to disqualify a judge “cannot be defined with precision. Circumstances and relationships must be considered”. 349 U.S. at 136, 75 S.Ct. at 625. The circumstances in this case — the actual contents of the motions and the issues they raised — do not warrant a finding that Judge Garippo was so likely to have prejudged the motions’ merits that we can presume he could not impartially decide the motions.
DelVecchio challenged the use of his 1965 conviction at the 1979 trial because the conviction was irrelevant to the issues at the 1979 trial. In another motion, DelVecchio also alleged that his 1965 conviction was illegal because the statute that allowed juveniles to be charged as adults violated due process in two ways: the statute did not require a hearing to determine if a juvenile should be charged as an adult; and the statute left the decision to charge a juvenile as an adult to the “absolute and totally standardless discretion” of the State’s Attorney and circuit court judge. Finally, DelVecchio challenged the use of his 1965 confession in the 1979 trial because the confession was involuntary and obtained in violation of Miranda.
The court would have us presume that Judge Garippo was unable to consider these motions impartially. But certainly there is no reason to think, let alone presume, that Judge Garippo could not impartially consider the relevance of the 1965 conviction. That question in no way required Judge Garippo to judge any decision he had made in 1965. Judge Garippo had no personal stake in finding the 1965 conviction relevant.
Nor did Judge Garippo have any real personal stake in the decisions of the other motions. The court finds it “inexcusable” for Judge Garippo to have considered DelVec-ehio’s motion challenging the constitutionality of the law in 1965 allowing juveniles to be charged as adults. It is true that Judge Garippo made the decision to charge DelVec-chio as an adult (a decision that Judge Garip-po in his 1989 deposition described as a “no-brainer” — 16-year olds accused of murder were almost always charged as adults). But DelVecchio’s motion did not focus on Judge Garippo’s decision; it focused on the statute under which he made that decision. The motion did not allege that Judge Garippo himself violated the Constitution except to the extent that he followed the allegedly unconstitutional statute. There is no reason to presume that a judge in Judge Garippo’s position could not impartially decide DelVec-chio’s motion, unless perhaps he had some professional or intellectual stake in the statute’s constitutionality. But the record reveals no such stake. There is no indication, for example, that Judge Garippo had anything to do with drafting or passing the statute, or that he had ever defended the statute’s constitutionality. The motion in no way required Judge Garippo to “judge his own decision as a prosecutor.”
Neither did DelVeechio’s motion regarding the admissibility of his confessions require Judge Garippo in any real sense to “judge his own decision.” Judge Garippo was at best only tangentially involved in taking DelVee-chio’s confession. DelVecchio gave two confessions, one to police and one to Assistant State’s Attorney Patrick Tuite, Judge Garip-po’s subordinate. Judge Garippo read Tuite’s report about the confession. But Judge Garippo himself took no confession from DelVecchio. In fact, Judge Garippo’s deposition indicates that he first read about the crime in the newspaper and did not even know of the case when the police and Tuite took DelVecchio’s confessions. The motion does not allege, nor does the record indicate, that Judge Garippo improperly trained his subordinates, or that he required or allowed subordinates to extract involuntary confessions from suspects, either systematically or in this case. Again, any personal stake *526Judge Garippo would have had in upholding the confession was minimal. There is no reason to presume that Judge Garippo could not impartially evaluate the work of the police or his former subordinate.
Our decision in Barry v. United States, 528 F.2d 1094 (7th Cir.1976), is instructive on whether DelVecchio’s pretrial motions required Judge Garippo to disqualify himself. In Barry, the defendant was indicted under a then-novel (but subsequently upheld) construction of the Hobbs Act. Barry’s trial judge had been a United States Attorney. As United States Attorney, the judge had personally decided to use the Hobbs Act to prosecute activity such as that Barry was involved in. Id. at 1097. One of the questions the judge had to decide at trial was whether the Hobbs Act applied to Barry’s activities. Id. at 1100.
The judge in Barry had at least as much professional and intellectual stake in the outcome of the Hobbs Act question as Judge Garippo had in the outcomes of DelVecchio’s motions to exclude his 1965 conviction and confession. He himself had made the decision to apply the Hobbs Act the way it was applied in Barry’s case. But after examining Supreme Court precedent and the practices of Supreme Court justices in deciding whether to disqualify themselves, this court held that the judge’s interest was insufficient to require disqualification under the due process clause. See id. at 1099-1100 & 1100 n. 19.
As Barry indicates, Supreme Court disqualification practice is also instructive. Justice Black, Murchison’s author, was one of the principal authors of the Fair Labor Standards Act and helped shepherd the Act through Congress. Yet, Justice Black sat in United States v. Darby, 312 U.S. 100, 61 S.Ct. 451, 85 L.Ed. 609 (1941), the case deciding the Act’s constitutionality. Justice Black’s is not an isolated case. For example, Justice Frankfurter, as a law professor, had written extensively in the field of labor law and had co-authored a book, The Labor Injunction, that had severely criticized the use of injunctions against labor unions. Justice Frankfurter had also played an important part in drafting the Norris-LaGuardia Act, an act designed to correct what was considered the abusive use of injunctions by federal courts in labor disputes. Yet, Justice Frankfurter not only sat in, but wrote the opinion in United States v. Hutcheson, 312 U.S. 219, 61 S.Ct. 463, 85 L.Ed. 788 (1941), a leading case interpreting the Act’s scope. See Laird v. Tatum, 409 U.S. 824, 831-832, 93 S.Ct. 7, 11-12, 34 L.Ed.2d 50 (1972) (memorandum of Justice Rehnquist). Justice Jackson participated in McGrath v. Kristensen, 340 U.S. 162, 71 S.Ct. 224, 95 L.Ed. 173 (1950), a case raising the same issue he had decided as Attorney General. Laird, 409 U.S. at 833, 93 S.Ct. at 13. And on the Supreme Court, Justice Holmes sat in several cases reviewing decisions by the Supreme Judicial Court of Massachusetts that he had participated in. See Laird, 409 U.S. at 836, 93 S.Ct. at 14 (citing cases).
While the practices of Supreme Court justices regarding disqualification may not be dispositive, they are persuasive. Certainly, it would be presumptuous for us to hold that some of the most distinguished jurists in our nation’s history regularly violated the Constitution by sitting in cases in which we ourselves might conclude that disqualification is prudent. Even in the light cast by the court, any interest Judge Garippo had in the outcome of DelVecchio’s motions seems to be no greater than — and probably less than — the interests of Justice Black in Darby, Justice Frankfurther in Hutcheson, Justice Jackson in McGrath, or Justice Holmes in any of the cases in which he sat on review from the Massachusetts Supreme Judicial Court. The stake those Justices had in the outcomes of those cases was much more than that of a judge who merely “expressed views on an abstract matter of law.” (Op. at 517 n. 7) All these Justices had a real professional and intellectual stake in the decisions in those cases, the type of stake that Judge Garippo supposedly should have had in the outcome of DelVecchio’s motions. Yet they decided they could put that stake aside and rule fairly. There is no reason to presume that Judge Garippo was not likewise able to put aside the minimal interest he might have had in the decision of DelVecchio’s motions.
The court’s second reason for holding that Judge Garippo was required to disqualify himself is also insufficient. The court states *527that it would have been “natural” for Judge Garippo to have held a “grudge” against DelVecchio. This is so, the court posits, because Judge Garippo “arguably bore some personal responsibility” for DelVecchio being on the street in 1977 because Judge Garippo agreed to expedite DelVecchio’s 1965 indictment so that DelVecchio could be sentenced as a juvenile. This arguably resulted in a lesser sentence than DelVecchio otherwise might have received. The upshot of all this appears to be that Judge Garippo might have felt so guilty about letting DelVecchio off easy in 1965 that he would have been predisposed against DelVecchio in 1979.
This excuse for reversing is pure conjecture. Any biasing influence involved here is far too remote to require Judge Garippo to have disqualified himself. Judge Garippo’s decision to expedite the indictment might have made it a bit more likely that DelVec-chio would receive a shorter sentence than he otherwise would have received. But even when DelVecchio was sentenced as an adult when he reached age 21, a judge in the circuit court apparently gave him a break by sentencing him to only fourteen years, which with good time credits left him with only two or three years to serve. The judge based his decision in large part on DelVecchio’s exemplary prison record, a record Garippo could not have foreseen in 1965. And even if Del-Vecchio had been sentenced as an adult in 1965, a lenient judge in the Circuit Court might just as well have decided to give Del-Vecchio a break. While this requires speculation, we do know that Judge Garippo had nothing to do with sentencing DelVecchio; once Garippo made the decision to expedite DelVecchio’s indictment, DelVecchio’s sentence was out of his hands. There was no way in 1965 for Judge Garippo to have known how long a sentence DelVecchio was likely to receive. It is highly unlikely that Judge Garippo would have felt sufficiently guilty about a decision he had so little to do with that he could not have put this guilt aside at DelVecchio’s trial. The due process standard for disqualification requires an influence or interest which we can conclusively presume would cause the average judge to be biased. It is most unreasonable to presume the average judge would weigh the scales against a defendant to whom he indirectly gave a break fourteen years earlier.
But even if Judge Garippo had undoubtedly given DelVecchio a clear break — for example, if Judge Garippo had agreed to imposition of the minimum sentence in exchange for a guilty plea — the risk of bias is still insufficient to hold that Judge Garippo was required to disqualify himself. Compare a prosecutor to a sentencing judge. In a system of indeterminate sentencing, such as the federal system before the Sentencing Guidelines, judges make thousands of decisions on sentence length, based in large part on the judges’ personal evaluations of the defendants before them. Any one of those decisions could turn out to be wrong. But does a judge’s incorrect decision to give a defendant a sentencing break in one case necessarily prevent the judge from presiding at that defendant’s trial in a case fourteen years later? Would the Circuit Court judge who actually gave DelVecchio the minimum sentence have been required to disqualify himself from the 1979 trial? Can we presume that the average judge in this situation would violate his oath to be impartial? Most trial judges would rightly reject that suggestion. To suggest partiality in such situations is inconsistent with the “presumption of honesty and integrity of those serving as adjudicators.” Withrow, 421 U.S. at 47, 95 S.Ct. at 1464.
The court attempts to buttress its analysis by pointing to what it considers to be “circumstantial evidence that Garippo was predisposed against DelVecchio_” (Op. at 516). In fact, the court holds that “Judge Garippo’s actions in the Canzoneri case were so egregious that even if proof of actual bias were required, we would still hold that the judge’s conflict of interest violated the defendant’s due process rights.” But the evidence of actual bias the court musters is hardly sufficient to prove actual bias.
The court points to Garippo’s deposition testimony that he feels personally betrayed by defendants who commit crimes after he has been lenient with them. But most sentencing judges probably feel this; it is a natural emotion, though hardly a disqualifying interest. Taken to its logical conclusion, *528this “predisposition” theory would prevent a trial judge from trying and sentencing someone a second time. Judge Garippo’s general antipathy toward defendants who abuse his lenience is akin to Justice Embry’s general antipathy toward insurance companies that the Supreme Court found to be insufficient to disqualify the justice in Aetna.
The court points to two other pieces of circumstantial evidence. The court notes that Judge Garippo’s courtroom clerk had checked out the Christiansen case files before the Canzoneri trial started “and thus before the judge had any reason to examine the old files.” The implication is that Judge Garippo took an unusual interest in DelVec-chio’s case, and that this unusual interest points to bias. But this is like saying that an appellate judge who checks out a record before oral argument has “no reason” to examine the record, and therefore is taking an “unusual” interest in that case. In fact, the judge has a very good reason to examine the record — that reason is called preparation. Trial preparation does not start at the beginning of trial. In any event, this “evidence” shows not so much bias as curiosity; and curiosity about an earlier case is not sufficient reason for us to presume — or even infer — bias on the part of the average judge.
Finally, the court notes Judge Garippo’s statement that “DelVecchio was more deserving of the death penalty than John Wayne Gacy.” (Op. at 517) But the court’s rendition of Judge Garippo’s words mischar-acterizes his actual statement and takes the statement out of context. Judge Garippo made an off-the-record comment to DelVec-chio’s attorney comparing three death penalty cases before him, including DelVecchio’s and Gacy’s. According to Judge Garippo, the statement was directed not at which of the three defendants was most worthy of death, but rather at which defendant was a more likely “candidate” for the death sentence — that is, which defendant was most likely to be sentenced to death — based on Judge Garippo’s knowledge of the facts of these cases.
Judge Garippo’s explanation makes sense. If Judge Garippo was trying to railroad Del-Vecchio, why would he tell DelVecchio’s lawyer? In any event, DelVecchio has not presented any evidence inconsistent with Judge Garippo’s testimony. In an affidavit, Del-Veechio’s trial attorney stated only that Judge Garippo “indicated his ranking of the cases with respect to their relative merits as death cases.... Garippo indicated that of the three defendants, DelVecchio in his view was the most appropriate candidate for imposition of the death penalty. He referred to Gacy as a ‘distant third.’ ” The affidavit does not state that Judge Garippo said DelVecchio was more “deserving” of the death penalty than Gacy. In context, then, Judge Garip-po’s statement about Gacy provides no evidence of any predisposition on Judge Garip-po’s part.
In sum, DelVecchio has given us no reason to presume that the average judge in Judge Garippo’s position would have been biased against DelVecchio. Therefore, Judge Gar-ippo was not required to disqualify himself unless he actually was biased against DelVee-chio. Although the court does hold that Gar-ippo actually was biased, the “circumstantial evidence” of bias the majority points to is not sufficient to prove actual bias. If that evidence is sufficient even to raise an inference that Judge Garippo was biased (and I believe it was not), DelVecchio is entitled at most to a remand for an evidentiary hearing on the subject, since the question on appeal is whether summary judgment is appropriate in this case. But there is no need for a remand. DelVecchio does not argue that Judge Garip-po actually was biased against him; instead, DelVecchio pinned his hopes on his argument that Judge Garippo should have disqualified himself because of the appearance of bias resulting from the alleged biasing influences that Judge Garippo faced. Because that argument fails, DelVeeehio’s conviction should remain intact.
II.
I also disagree that DelVecchio was entitled to a hearing to determine the voluntariness of his 1965 confessions. The question here is whether Illinois violated the Constitution by not allowing DelVecchio to challenge his 1965 confessions in 1979 when he bypassed the opportunity to challenge those confessions in 1965. In Jackson v. Denno, 378 U.S. 368, 84 S.Ct. 1774, 12 L.Ed.2d 908 *529(1964), the Supreme Court held that when a defendant challenges a confession as involuntary, due process requires that the state hold a hearing to determine the confession’s vol-untariness. But Jackson does not answer the question before us. In Jackson, the defendant did not pass up an opportunity to challenge his confession before challenging it years later in another case. Unlike in this case, the defendant in Jackson challenged his confession in the case in which the confession was given.
Neither does Haring v. Prosise, 462 U.S. 306, 103 S.Ct. 2368, 76 L.Ed.2d 695 (1983), a ease the court cites to support its holding that DelVecchio is entitled to a hearing, answer the question before us. In Haring, the Supreme Court held that a guilty plea does not bar a convicted defendant from filing a § 1983 action to challenge the constitutionality of an illegal search in the case in which he pleaded guilty. To reach this conclusion, the Court had to grapple with two issues. The first was whether 28 U.S.C. § 1738, which requires that federal courts give full faith and credit to state court judgments, barred the plaintiffs § 1983 action. This depended on whether the law of issue preclusion in Virginia — the state in whose courts the defendant pleaded guilty — barred the plaintiffs § 1983 action. 462 U.S. at 313-317, 103 S.Ct. at 2373-2375. The Court held that § 1738 did not bar the plaintiffs action because Virginia’s law did not bar the action. Id. at 317, 103 S.Ct. at 2375. The second issue in Haring was whether the Court should “create a special rule of preclusion which nevertheless would bar litigation of his § 1983 claim.” Id. The Court decided not to create such a special rule.
Unlike in Haring, the question is not whether Illinois law would bar DelVecchio from challenging his confessions in his 1979 case because of his 1965 guilty plea and his failure to challenge those confessions in 1965. The Illinois Supreme Court has already held that Illinois law does bar such a challenge. Nor do we have to decide whether to create a federal rule of claim preclusion. Instead, the question before us is whether Illinois is constitutionally required to give DelVecchio a second opportunity to challenge his confessions. Haring does not answer this question because Haring did not purport to create a constitutional rule prohibiting the type of bar Illinois applied.
So the question remains: was Illinois required to give DelVecchio a second chance to challenge his confession? I think not. Illinois has a strong interest in seeing that legal issues based on factual circumstances, such as the voluntariness of a confession, are litigated as quickly as possible. As time passes, memories dim, witnesses die or become otherwise unavailable, and physical evidence deteriorates or is destroyed. Early litigation is thus more likely to ferret out the truth than litigation delayed by many years.
DelVecchio, it is true, has a significant interest in seeing that an involuntary confession is not used against him. But DelVecchio had that same interest in 1965 and had ample opportunity to challenge his confessions. In spite of his interest in attacking his confessions, DelVecchio made a reasoned decision — a decision that worked to his immediate advantage — to plead guilty and forego challenging his confessions. While it is true that we cannot presume DelVecchio’s plea was motivated by a decision that he could not win at a suppression hearing, see Haring, 462 U.S. at 318, 103 S.Ct. at 2375, the fact remains that DelVecchio did — for a very good reason — pass up the opportunity to challenge his confessions.
And that is not all DelVecchio did. Through his attorney, DelVecchio stated that he had no objection to the state’s introducing his confessions to establish the factual basis for his plea. DelVecchio’s attorney used portions of his confessions (specifically, the references to DelVeechio’s drug use on the night of the Christiansen murder) in arguing to the judge regarding DelVecchio’s possible sentence. DelVecchio may now regret those decisions, but he does not allege that his attorney’s performance in 1965 was deficient. Yet, DelVecchio demands a hearing many years after the fact to determine if his confessions were voluntary. But requiring Del-Vecchio to live with the decisions he made in 1965 does not violate the Constitution. Nothing in the Constitution, or in any cases *530interpreting the Constitution cited by Del-Vecchio or the court, requires a state to accept a confession when it benefits the accused but (much) later defend that same confession when the accused concludes that the benefit is gone. In the vernacular, the Constitution does not allow the accused to eat his cake and still have it.
Even if DelVecchio is entitled to a hearing, I disagree that the state must bear the burden of proving that DelVecchio’s confession was voluntary. The due process clause requires only that states use procedures that are not fundamentally unfair, not procedures that are most likely to produce favorable results for defendants. See Parke v. Raley, — U.S.—,—, 113 S.Ct. 517, 524, 121 L.Ed.2d 391 (1992); Medina v. California, — U.S.—, —, 112 S.Ct. 2572, 2580, 120 L.Ed.2d 353 (1992). Given the state’s interests in quickly deciding involuntary confession claims, the difficulties that arise from delaying a decision of those claims, and the opportunity for an early decision that Del-Vecchio passed up, it is not fundamentally unfair for Illinois at this late date to require DelVecchio to prove that his confession was involuntary.
III.
As the court notes, DelVecchio raises several other claims on appeal, all of which the district court rejected. Like the court, I will forego comment on those claims, other than to say I generally agree with the district court’s disposition of those issues. I would affirm all of the district court’s decision except its decision that DelVecchio is entitled to a hearing at which the state must prove that DelVecchio’s 1965 confessions were voluntary, and I would deny DelVecchio’s petition for habeas corpus.
Before: POSNER, Chief Judge, CUMMINGS, BAUER, CUDAHY, COFFEY, EASTERBROOK, RIPPLE, MANION, KANNE, and ILANA DIAMOND ROVNER, Circuit Judges.ORDER
Dec. 17, 1993.
On consideration of the petition for rehearing with suggestion for rehearing era banc filed by respondent-appellee, cross-appellant on November 9, 1993, the answer of petitioner-appellant, cross-appellee, and the reply of respondent-appellee, cross-appellant, a vote of the active members of the Court was requested* and a majority of the judges in active service have voted to rehear this case era banc.
IT IS ORDERED that rehearing era banc be, and the same is hereby, GRANTED.
IT IS FURTHER ORDERED that the opinion entered in this case on October 26, 1993, be, and is hereby, VACATED. This case will be reheard era banc at the convenience of the Court.
Judge Joel M. Flaum took no part in the consideration of the suggestion for rehearing en banc.