¶ 1. Lucien Agnello appeals a judgment of conviction for first-degree intentional homicide, party to a crime. The issues are: (1) whether Agnello's confession was voluntary and therefore admissible; and (2) if we affirm the trial court's ruling that the confession was voluntary, whether Agnello was *264entitled to a trial because of the Wisconsin Supreme Court's vacation of his original conviction. We conclude the trial court correctly decided his confession was voluntary. We also conclude that Agnello is entitled to a trial. Accordingly, we affirm the order of the trial court denying Agnello's motion to suppress his confession, vacate the judgment of conviction, and remand for trial.
PROCEDURAL HISTORY
¶ 2. On February 18, 1996, the Milwaukee Police Department arrested Agnello in connection with the murder of his foster father, Theodore Agnello. Near the end of several interrogation sessions taking place between 4:30 a.m. to 3:20 p.m. the day after his arrest, Agnello confessed to the murder. After the trial court determined in a Goodchild1 hearing that Agnello's confession was voluntary, Agnello pleaded guilty. Ag-nello appealed the judgment of conviction, contending that the trial court committed error by allowing the prosecutor to cross-examine him at the Goodchild hearing regarding the truthfulness of his confession, and that the trial court erred in deciding his confession was voluntary. This court affirmed the judgment of conviction.
¶ 3. The supreme court granted Agnello's petition for review and concluded that the prosecutor had improperly questioned Agnello at the Goodchild hearing about the truthfulness of his confession. Because the trial court had based its ruling that Agnello's confession was voluntary in part on his answers to the *265improper questions, the supreme court reversed that ruling, vacated the conviction and remanded the case to the trial court for a new Goodchild hearing. State v. Agnello, 226 Wis. 2d 164, 182, 593 N.W.2d 427 (1999).
¶ 4. After hearing testimony at the second Good-child hearing, the newly assigned judge2 determined that Agnello's confession was voluntary. The trial court also reinstated the judgment of conviction. The court reasoned that, after the second hearing, Agnello was returned to the same position he was in after the original judge denied his motion to suppress. Agnello appealed to this court, contending the trial court erred in determining his confession was voluntary and the supreme court's vacation of his conviction meant he was entitled to a new trial. This court certified these issues to the supreme court under Wis. Stat. Rule 809.61, and the supreme court granted certification. However, the supreme court ultimately vacated its order granting certification and remanded the case to this court.3
FACTUAL BACKGROUND
¶ 5. The trial court at the second Goodchild hearing found the following facts regarding Agnello's confession. Agnello was arrested at approximately 11:20 p.m. on February 18, 1996. He was transported to the police department at about 2:00 a.m. the following day, February 19. After getting to the station Agnello was placed in a standard interrogation cell estimated to be *266between eight-by-ten to ten-by-ten feet in size. The room had a table, some chairs, and a ring on the wall at about the level of the table. Agnello was questioned in the room from approximately 4:30 a.m. to 3:20 p.m. in three sessions involving different teams of officers, with breaks during all sessions. The first session took place from approximately 4:30 to 6:00 a.m. The second session took place from approximately 6:00 to 7:30 a.m. The third session began at approximately 10:15 a.m. and ended at approximately 3:20 p.m., with a break from 11:40 a.m. to 1:15 p.m. and a later twenty-minute break. Agnello was advised of his Miranda rights and waived them three times: once at the beginning of the first session and twice more over the course of the interrogation. Agnello began to admit culpability at about 11:40 a.m. By the end of the third session, Agnello had completed his confession that he was involved in the killing of his foster father.
¶ 6. The court found that throughout the interrogation, Agnello was coherent, gave appropriate answers to questions, and was not seen sleeping. There was no indication Agnello was under the influence of alcohol or drugs. The detectives were not armed. Agnello was given food, coffee, water, and cigarettes over the course of the interrogation, and he was allowed to use the bathroom at least once. He was asked if he needed sleep, if it was okay to continue, and he indicated he wanted to go on. The officers did not make promises to Agnello in return for the confession nor did they threaten him. They did not scream or raise their voices.
¶ 7. The trial court found that Agnello was not handcuffed to the ring on the wall when he was being interrogated, but was likely handcuffed to the ring during at least some of the breaks. When handcuffed to the ring, the court found, Agnello would not have been *267able to move around in the room, but would have been able to rest his head on the desk. The undisputed testimony of the detectives who interrogated Agnello was that he was not handcuffed during the interrogation sessions themselves. Some of the detectives could not recall whether Agnello was handcuffed during breaks in the questioning; others did not believe he was, although one detective testified that Agnello was handcuffed during a break from approximately 11:40 a.m. to 1:15 p.m. Two detectives stated it was department policy to handcuff a suspect if the suspect would be alone in a room for a substantial amount of time. The undisputed testimony on the manner of handcuffing a person to the ring was that one wrist was handcuffed and the other handcuff was on the ring. The location of the ring in relation to the table was such that when a person was handcuffed to the ring, he or she could rest the arm of the handcuffed wrist on the table when sitting down, could stand up, and could move a step or two.
DISCUSSION
I. Voluntariness of Agnello's Confession
¶ 8. Under the due process clause of the Fourteenth Amendment, confessions that are not voluntary are not admissible. Rogers v. Richmond, 365 U.S. 534, 540 (1961). When we review a trial court's determination on the voluntariness of a defendant's confession, we affirm the trial court's findings of historical facts unless they are clearly erroneous. See State v. Kieffer, 217 Wis. 2d 531, 541, 577 N.W.2d 352 (1998); Wis. Stat. § 805.17(2) (2001-02). However, the application of the constitutional standard to historical facts is a question *268of law, which we review de novo. Kieffer, 217 Wis. 2d at 541. The State has the burden of proving voluntariness by a preponderance of the evidence. Agnello, 226 Wis. 2d at 182.
¶ 9. In deciding whether a confession is voluntary, we inquire whether the confession was procured by coercive means or was the product of improper pressures exercised by the police; this is the focus of our inquiry because it is determinative of whether the inculpatory statement was the product of "free and unconstrained will, reflecting deliberateness of choice." State v. Clappes, 136 Wis. 2d 222, 235-36, 401 N.W.2d 759 (1987). In our inquiry we must determine whether the defendant was "the victim of a conspicuously unequal confrontation in which the pressures brought to bear on [the defendant] .. . exceeded the defendant's ability to resist." Id. at 236. We make this determination after looking at the totality of the circumstances surrounding the confession and balancing the personal characteristics of the defendant against the pressures imposed by the police to induce the defendant to respond to the questioning. Id. The personal characteristics to be considered may include the defendant's age, education and intelligence, physical and emotional condition, and prior experience with police. Id. These must be balanced against police pressures and tactics used to induce admission, such as the duration of the questioning, the general conditions under which the confession took place, any excessive physical or psychological pressure brought to bear on the declarant, any inducements, threats, or other methods used to compel a response, and whether the defendant was informed of his right to counsel and right against self-*269incrimination. Id. at 236-37. See also State v. Hoppe, 2003 WI 43, ¶¶ 39-40, 261 Wis. 2d 294, 661 N.W.2d 407.
¶ 10. Police conduct does not need to be egregious or outrageous in order to be coercive; subtle pressures are coercive if they exceed the defendant's ability to resist. Hoppe, 261 Wis. 2d 294, ¶ 46. If a defendant's condition renders him or her uncommonly susceptible to police pressures, those pressures may be coercive even though under another set of circumstances, they might not be coercive. Id. "[A]s interrogators have turned to more subtle forms of psychological persuasion, courts have found the mental condition of the defendant a more significant factor in the voluntariness calculus." Colorado v. Connelly, 479 U.S. 157, 164, (1986). A confession's truth or falsity has no direct bearing on the determination of voluntariness. Agnello, 226 Wis. 2d at 177.
¶ 11. Agnello points to several factors that, he asserts, made his confession coerced: (1) being handcuffed to the ring in the interrogation room; (2) the deprivation of sleep; (3) psychological coercion concerning the discussion of his foster mother; and (4) the length of isolation and interrogation, including questioning by "relay teams" of detectives. Considering these factors as part of the totality of the circumstances and balancing Agnello's personal characteristics against the pressures imposed by the officers, we conclude his confession was voluntary.
¶ 12. We will assume for purposes of this decision that Agnello was handcuffed to the ring on the wall for all breaks between interrogations. Agnello states in his brief, "The practice of chaining the accused to the wall has a strikingly medieval flavor that is offensive to basic *270notions of civilized and humane treatment.... The practice is coercive in and of itself." To support the proposition that this physical restraint, by itself, made his confession involuntary, Agnello cites three cases in which confessions have been ruled involuntary. In United States ex rel. Adams v. Benzinger, 507 F.2d 390, 393-95 (7th Cir. 1974), the recitation of facts included that the defendant was "locked to a radiator" for some period of time; however, that was not one of the three elements the court viewed as rendering the confession involuntary. The three elements were: (1) the defendant was not informed of his right to assistance of counsel and right to remain silent; (2) he was awakened from sleep by the police for questioning; and (3) he was of "low intellect and limited education." Id. at 394-95. In Knight v. State, 518 So. 2d 799, 800 (Ala. Crim. App. 1987), the defendant, who was "essentially illiterate," was handcuffed to a chair for most of a six-hour interrogation by two detectives, one armed, and both detectives used a "rapid-fire" questioning technique. That defendant also testified that he asked for food but was given nothing to eat, asked for a lawyer twice, and was threatened with a whipping. Id. Finally, in State v. Tom, 613 P.2d 842, 844 (Ariz. App. 1980), the defendant was handcuffed with his hands behind him and his face covered in a towel while being interrogated by the same officers who had beaten him and threatened to kill him during his "violent arrest" earlier.
¶ 13. We do not agree that these cases support Agnello's position. In these cases the handcuffing occurred in the context of egregious police conduct that did not occur here, such as beatings, armed officers, failure to honor defendants' constitutional rights and the placement of a towel over the suspect's face. These cases do not stand for the proposition that handcuffing *271in itself is coercive. In addition, unlike in those cases, Agnello was not handcuffed during the interrogation sessions themselves, but only during breaks when the officers left the room. Also, Agnello's mental capacity, literacy, and education4 are not factors making it more likely that police conduct was coercive.
¶ 14. Agnello's characterization of this handcuffing during breaks as "medieval" is not supported by the record. The police testimony indicated it was standard procedure to handcuff suspects in this manner when they would be alone for long periods of time, in order to prevent suicide attempts, escape, or property damage to the room. Agnello was able to sit on a chair while handcuffed and he could lay his head on the desk.
¶ 15. We next address Agnello's argument that he was deprived of sleep. Agnello points out that he was not allowed to lie down during his time overnight in custody, and he argues it is common sense that someone detained overnight would require sleep. Clearly, when Agnello was handcuffed to the wall, he was precluded from lying down. However, the trial court found this did not preclude sleep. The court stated:
it would not have been the most comfortable situation in which to sleep, but... it would have been possible certainly for the defendant to catch some sleep, but the ultimate point here is that the uncontroverted testimony was that the defendant never indicated that he wanted to sleep, that he needed to sleep, and that to the contrary, he told the police that he wanted to go on and even appeared impatient to continue with the interrogation.
*272¶ 16. These findings by the trial court are supported by the record. There was testimony that Agnello did not say he was tired or ask for time to rest, that he "appeared impatient" and that he said he felt well enough to continue with questioning when asked.
¶ 17. Agnello also argues that the detectives put undue psychological pressure on him by discussing his foster mother, Mary Agnello, during the interrogation. Mary had also been taken into custody the night of Agnello's arrest. One of the detectives testified that in the course of the interrogation "any time Mary's name came up... his [Agnello's] outward appearance changed. He seemed to be worried about her." The detective further testified that, at about 11:30 a.m., he told Agnello he was going to take a break with Agnello so he could speak to Mary. At that point Agnello said Mary was not involved and he began to break down and sob. The trial court found that very shortly thereafter, at approximately 11:40 a.m., Agnello began to admit culpability.
¶ 18. We do not agree with Agnello that the references to Mary constituted excessive psychological pressure that coerced his confession. The references to Mary did not in any sense resemble the types of threats and inducements that courts have held to be coercive, such as threatening that a suspect's children would be taken away if she did not cooperate, Lynumn v. Illinois, 372 U.S. 528, 534 (1963), or conditioning outside contact upon accession to police demands. Haynes v. Washington, 373 U.S. 503, 507 (1963). It is true that more subtle psychological pressure was considered coercive in Hoppe, 261 Wis. 2d 294, ¶¶ 55, 60, but only because of the particular vulnerability of the defendant. There the court concluded that the pressures of using emotional topics such as the death of the defendant's *273parents, the feelings of the victim's family, and the defendant's military service in Vietnam exceeded the defendant's ability to resist because of his "severely debilitated mental and physical condition." Id. Assuming that the police here used the references to Mary to pressure Agnello psychologically, it was not excessive pressure and Agnello was not unusually vulnerable because of any mental or physical condition. In addition, Agnello's prior police contacts made him less vulnerable than someone without that experience.
¶ 19. Finally, we address Agnello's argument that the length of isolation and detention, along with "relay-team" questioning, led to his confessing involuntarily. The trial court estimated the length of Agnello's detention before he began confessing from two different events: twelve hours when measured from Agnello's arrest and nine and one-half hours when measured from his arrival at the police station, because that was, in the trial court's words, when Agnello "probably had a sense of just how serious this situation was." Regardless of which estimate is used, we conclude the length of detention was not impermissibly long. Agnello cites to several cases in which courts have ruled confessions involuntary because they were procured only after long periods of detention and interrogation.5 The suspects in *274all but one of these cases, Spano v. New York, 360 U.S. 315 (1959), were detained for longer periods of time, ranging from twenty hours to seven days. In addition, in many of the cases the suspects were denied their right to counsel, whereas the trial court here found that Agnello waived his Miranda rights three times.
¶ 20. Agnello emphasizes that in Spano, the length of detention was only eight hours and the court nonetheless ruled the confession coerced. However, the other relevant circumstances in Spano were significantly different than those in this case. In Spano, the suspect was a twenty-five-year-old man who had "no past history of law violation or of subjection to official interrogation," only one-half year of high school, and a history of "emotional instability"; he was denied his right to counsel and interrogated virtually nonstop for eight hours. Id. at 321-23. In addition, the police had a young officer who was a childhood friend of the suspect interrogate him and falsely state that a phone call the suspect had made to him could cause him (the friend) to lose his job, which would harm the friend's wife and child. Id. at 323. In contrast, Agnello was given breaks during much shorter interrogation sessions and he was not denied counsel. He also had had previous contacts *275with police, and his mental capacity and education were not factors making it more likely that police conduct was coercive.
¶ 21. We do not agree with Agnello that the questioning methods of the police in this case were "relay-team" tactics. "Relay" questioning implies that different interrogators relieve each other in an effort to put unremitting pressure on a suspect. See 3 William E. Ringel, Searches & Seizures, Arrests & Confessions § 25.2(b) (2d ed. 1979). The trial court here found there were breaks both during and between interrogation sessions and at least one of the changes in interrogation teams was due to a shift change. These findings are supported by the record and do not constitute "relay-team" tactics that put unremitting pressure on Agnello.
¶ 22. We agree with the trial court's statement that "the most troubling thing here with respect to voluntariness is the lapse of time" between arrest and confession. However, considering this factor along with all the other circumstances, we are satisfied the confession was voluntary. We have already concluded that the physical restraint and psychological tactics used were not coercive, and the trial court found that Agnello was not deprived of sleep, food, or water during the interrogation sessions. The trial court also found that the detectives were not armed during the sessions, they did not make promises to Agnello in return for the confession, nor did they threaten him. In addition, the trial court found that Agnello had had multiple contacts with police as a juvenile, which is a factor that makes it less likely he would succumb involuntarily to police questioning. There is no evidence that Agnello was vulnerable because of any personal characteristic. Accordingly, we conclude the police did not coerce Agnello *276into giving an involuntary confession. The trial court therefore properly denied Agnello's motion to suppress his confession.
II. Interpretation of the Supreme Court's Ruling Vacating Agnello's Conviction
¶ 23. Because we conclude that Agnello's confession is admissible, we next consider whether Agnello was entitled to a trial following the supreme court's vacation of his original conviction. In its decision concerning the first Goodchild hearing, the supreme court stated, "[W]e reverse the court of appeals, vacate Agnello's conviction, and remand the cause to the circuit court for a new Goodchild hearing." 226 Wis. 2d 164 at 182. Agnello asserts that vacation of his conviction entitles him to a new trial. The State argues that the scope of the remand was limited to a new Goodchild hearing. In the State's view, the supreme court intended that, if Agnello's confession were ruled inadmissible, he would be allowed to withdraw his plea; but if his confession were again found voluntary, the error of the first Goodchild hearing would be corrected and the trial court was to reinstate the conviction and sentence.
¶ 24. The State urged us to certify this question to the supreme court if we had doubts about the procedural correctness of the reinstatement, saying, "There is no need for speculation when the supreme court may be called upon to clarify its intent." However, because the supreme court has vacated its acceptance of certification and remanded the case to this court, it is now our task to ascertain the supreme court's intent in *277using the language it did. We view this as a question of law and therefore decide it de novo.6
¶ 25. We conclude that both Agnello's and the State's readings of the supreme court's language are reasonable. The supreme court expressly vacated the conviction and did not direct the trial court to reinstate it if it found the confession voluntary. As Agnello explains, there is a rational basis for not reinstating the conviction even if the trial court decided the confession was voluntary. When Agnello entered his plea, he did so after a Goodchild hearing at which he answered questions about the underlying crime, answers that might have been admissible at trial under some conditions. The second Goodchild hearing corrected the errors of the first hearing, but also arguably changed the circumstances relevant to the decision to enter a plea or go to trial. On the other hand, the State's view is also reasonable: the supreme court may have assumed the trial court would reinstate the conviction if it found the confession voluntary after a proper Goodchild hearing even though the supreme court did not direct the trial court to do so.
¶ 26. To resolve this issue, we have attempted to find cases in which the supreme court vacated convictions in similar circumstances, and we have located State v. Jiles, 2003 WI 66, 262 Wis. 2d 457, 663 N.W.2d 798. Jiles involves a defendant very similarly situated to Agnello. Jiles moved to suppress his statements to police on the grounds that they were involuntary and *278his Miranda rights were violated. Id., ¶ 3. After the hearing, the court denied his motion. Id., ¶ 19. Jiles then pleaded guilty to one count of first-degree reckless injury by use of a dangerous weapon and one count of armed robbery with the use of force, both as a party to a crime. Id., ¶ 20. Jiles appealed his convictions, contending that he did not receive a full and fair hearing on the motion to suppress. Id., ¶ 28. The supreme court agreed that the Miranda-Goodchild hearing was defective, stating that the trial court judge had "intervened and assumed the State's burden of establishing the existence of proper Miranda warnings and voluntariness." Id., ¶¶ 28, 38. After deciding that Jiles Miranda-Goodchild hearing was inadequate, the court stated: "Jiles is entitled to withdraw his guilty plea and to be granted a new and sufficient Miranda-Goodchild hearing. After this hearing, he is entitled to a trial if he so desires." Id., ¶ 49 (emphasis added).
¶ 27. We find the decision in Jiles instructive on the supreme court's intent when it vacated Agnello's conviction. Jiles was decided just one month after the supreme court issued the order in this case vacating the certification. We therefore assume the supreme court had in mind the lack of specific directions in Agnello when it chose to specify that Jiles was entitled to a trial if he desired. We can discern no reason why the supreme court would treat such similarly situated parties differently. In each case the defendant entered a guilty plea after his motion to suppress his statements was denied. In each case the supreme court determined that the hearing on the motion was defective and the defendant was entitled to another hearing, and it reversed the conviction.
¶ 28. We therefore conclude the supreme court intended that Agnello would be entitled to a trial after *279the Goodchild hearing if he chose, just as the court granted Jiles that choice. Accordingly, we reverse the trial court's order denying Agnello's request for a trial and remand for trial.
By the Court. — Judgment and order affirmed in part; reversed in part and cause remanded with directions.
In State ex rel. Goodchild v. Burke, 27 Wis. 2d 244, 264-65, 133 N.W.2d 753 (1965), the court established the procedure for a pretrial hearing on the issue of the voluntariness of a confession.
The Hon. Elsa C. Lamelas was assigned to the case after the original judge, current Wisconsin Supreme Court Justice Diane Sykes, left the circuit court.
The supreme court was deadlocked 3-3 on whether to affirm or reverse the judgment of the trial court, with Justice Sykes not participating.
Agnello indicated on his guilty plea questionnaire that he has completed twelfth grade and obtained a GED.
Agnello cites to: Darwin v. Connecticut, 391 U.S. 346, 349 (1968) (suspect detained thirty to forty-eight hours); Spano v. New York, 360 U.S. 315, 322 (1959) (eight hours); Turner v. Pennsylvania, 338 U.S. 62, 63-64 (1949) (four days); Watts v. Indiana, 338 U.S. 49, 52-53 (1949) (six days); Ashcraft v. Tennessee, 322 U.S. 143, 149 (1944) (thirty-eight hours); Ward v. Texas, 316 U.S. 547, 551 (1942) (three days); Chambers v. Florida, 309 U.S. 227, 230 (1940) (seven days); Purpura v. United States, 262 F. 473, 477 (4th Cir. 1919) (twenty-four hours); Binns v. State, 344 S.W.2d 841, 842 (Ark. 1961) *274(fifty-two hours); Wright v. State, 9 A.2d 253, 256 (Md. Ct. App. 1939) (twenty hours). The suspects in Spano, Turner, Chambers, Purpura, and Wright were denied counsel. In addition, in Wright, the defendant was threatened with being hit, 9 A.2d at 256; in Watts, the defendant's "rudimentary" needs for sleep and adequate food were disregarded, 338 U.S. at 53; and in Ward, the defendant was moved several times from one town to another and told there were threats of mob violence. 316 U.S. at 551, 553.
When there is ambiguity in a trial court's judgment, we defer to that court's construction of its own language, as long as it is reasonable. Estate of Schultz v. Schultz, 194 Wis. 2d 799, 809, 535 N.W.2d 116 (Ct. App. 1995). However, in this case the trial court is in no better position than this court to construe the language of the supreme court.