delivered the opinion of the Court.
Four hours after learning that police suspected him of raping his girlfriend’s ten-year-old daughter, defendant sought to clear his name by agreeing to take a lie-detector test. Though polygraph evidence is generally inadmissible, defendant, without counsel, signed a stipulation before submitting to the exam in which he not only agreed to take the test but also (1) agreed that the polygraph examiner was an expert at his craft, (2) waived any objection to the admissibility of the expert’s testimony, (3) waived the right to call another expert or witness about that evidence, and (4) agreed that the results of the polygraph exam would be admissible at trial. Those decisions would ordinarily have been made with the help of a lawyer.
*74Defendant faded the exam. As a result, otherwise inadmissible polygraph evidence was introduced against him at trial based on consent he alone provided. In a case that rested on the testimony of the accuser, with no medical or physical evidence to corroborate her words, defendant was convicted.
The sweeping waiver of trial rights in this case, without the assistance of a lawyer, compromises the integrity of the criminal trial process. Our obligation to advance the sound administration of justice, and our continuing doubts about the reliability of polygraph evidence, warrant the exercise of the Court’s supervisory authority to bar the admission of polygraph evidence based on stipulations entered into without counsel. We therefore affirm and modify the judgment of the Appellate Division, which reversed the conviction on this ground for different reasons.
In addition, we address State v. Guenther, 181 N.J. 129, 854 A.2d 308 (2004), which permits a defendant, in carefully limited circumstances, to impeach the credibility of a victim-witness by inquiring into prior false criminal allegations the witness made. We now hold that false allegations made after the underlying accusation may also be relevant for impeachment purposes.
Defendant sought to impeach the victim with evidence that she falsely accused an employee of the Division of Youth and Family Service (DYFS) of sexual abuse shortly after making allegations against the defendant. Because Guenther was decided after defendant’s trial, the trial court did not have the benefit of that ruling in evaluating defendant’s argument. We agree with the Appellate Division that, on remand, the trial court should conduct an N.J.R.E. 104 hearing in the manner outlined in Guenther to evaluate whether the proposed impeachment evidence may be used at a new trial.
I.
A.
Defendant A.O. is an immigrant from Nigeria who lived with his girlfriend, J.I., and her daughter, C.I., since 1998. Late in the *75afternoon on April 27, 2001, J.I. called defendant and relayed that there was a problem with C.I. At the time, the girl was ten years old. J.I. asked defendant to come to the Child Advocacy Center of the Union County Prosecutor’s Office. On arrival, he found three police officers waiting to speak with him.
For the next four hours, the detectives questioned defendant about allegations the girl had made in school earlier that day. According to a classmate and a school official, the girl had said that defendant was “doing sex to her,” “abusing her,” and forcing her to engage in sexual acts.
During the course of the interview, the police advised defendant of his Miranda rights three separate times, the first at 6:40 p.m. At around 11 p.m., after repeatedly denying the accusations, defendant asked the officers how he could prove his innocence. A detective proposed that he take a polygraph examination, and defendant agreed. An assistant prosecutor then met with defendant and gave him a standard, five-page polygraph stipulation form used by the Union County Prosecutor’s Office. The prosecutor reviewed the form with defendant for about half an hour. The prosecutor read the form aloud to defendant, stopping after each paragraph to ask defendant whether he understood it. Defendant initialed each paragraph in turn, and asked no questions.
The stipulation informed defendant that he had an absolute right to refuse to be tested. Should he agree to take the test, though, the document made clear that the results could be used against him at trial. Among other provisions, the stipulation provided that the results generated by the test were admissible as evidence, that the examiner was an expert in administering and analyzing polygraph examinations, that neither party would object to the admissibility of the expert’s testimony, and that neither party would call any other experts to dispute the results. Specifically, the form stated:
[IJrrespective of the outcome of the examination, the results of the examination shall be admissible as evidence in the Grand Jury, or on behalf of either the prosecution on either the State’s direct or rebuttal case, or the defense (the subject) in any criminal trial(s) of the subject for the subject for the above-*76mentioned crime(s). This means that the polygraph examiner may give his opinion as to whether or not the subject exhibited evidence of attempts at deception during the examination.
(3) The polygraph examiner is acknowledged by both parties to be an expert in all phases of both administering polygraph examinations and in the analysis of polygraph chart recordings____
The opposing party hereby expressly waives any and all objections to the admissibility of such expert testimony as to his qualifications, the manner in which he conducted the examination, the test questions asked the subject, his expert opinion as to whether or not the subject exhibited evidence of attempts at deception, and the possibility of error.
(4) Neither party shall have the right to introduce another polygraph expert or any other person, other than the subject himself, in reference to the original polygraph expert’s testimony.
(6) ... The results of any other polygraph examination shall not be admissible unless covered by a separate stipulation agreement.
Defendant signed the stipulation at 11:38 p.m. At the time, he was not represented by counsel, not under arrest, and had not been charged with a crime.
Minutes before midnight, Detective Sergeant John Kaminskas began the polygraph examination. Kaminskas was a nineteen-year veteran of the force who had previously administered hundreds of polygraph exams. Kaminskas asked defendant four pertinent questions about whether he had sexually abused C.I., and defendant answered “no” each time. According to Kamins-kas, the polygraph revealed “indications of deception” for each response. Defendant was arrested soon afterward.
Approximately one week later, C.I. recanted her allegations against defendant on the day she was to undergo a medical examination. As a result, no exam was performed. Shortly after, DYFS removed C.I. from her home after concluding that the girl's mother had “failed to protect her.” During the course of the next year, DYFS placed C.I. first in a shelter, then a foster home, and finally with her aunt. While living in the shelter, C.I. also accused a DYFS worker of sexually assaulting her, but she recanted several days after prosecutors interviewed her about the incident. *77In December 2002, while living with her aunt, C.I. reaffirmed her initial abuse allegations against defendant.
On December 12, 2001, a Union County Grand Jury returned an indictment against defendant charging him with first-degree aggravated sexual assault (N.J.S.A. 2C:14-2(a)(l)); second-degree attempted aggravated sexual assault (N.J.S.A. 2C:14-2(a)(l)); two counts of second-degree sexual assault (N.J.S.A. 2C:14-2(b), (c)(1)); and second-degree endangering the welfare of a child (N.J.SA 2C:24-4(a)). The allegations of sexual abuse spanned the period from February 4, 2000 to April 17, 2001.
At trial, C.I. testified that defendant committed a number of sexual acts against her, including rape. No physical or medical evidence was introduced to corroborate her testimony.
The State also introduced the polygraph results and highlighted that evidence throughout the trial. In its opening, the State explained it would counteract defendant’s denial with proof that he failed a polygraph exam. As promised, Detective Sergeant Ka-minskas testified about his expertise as a polygrapher and his examination of the defendant. After explaining to the jury that defendant’s test results showed deception, Kaminskas testified that he had a 100 percent accuracy rate in the prior 302 tests he had administered. He remarked that he “never had a confirmed mistake.” The State underscored that testimony in closing argument by noting “[t]his particular defendant failed. And [Kamins-kas is] 100 percent accurate____[Defendant] denies these allegations and he thinks [he can] fool the machine. Guess what? He can’t. The machine said he lied.”
In addition, the State presented testimony in an effort to explain why C.I. recanted, and subsequently reaffirmed, her allegations against defendant. The prosecution offered evidence that C.I. was not supported by her family and was pressured to recant by her grandmother and mother. The State also introduced expert testimony on Child Sexual Assault Accommodation Syndrome to explain that sexual assault victims are more likely to recant if they do not receive family support.
*78The defense argued that C.I. fabricated the allegations against defendant because she disagreed with his strict parenting style. In an effort to attack C.I.’s credibility, defense counsel sought to make use of a DYFS report that revealed that C.I. had accused a DYFS worker of sexually assaulting her and later recanted. The State countered with proof that the abuse had actually occurred. After reviewing certain DYFS records and a videotape of C.I. in camera, the trial court concluded the evidence was not relevant and could not be used. The court noted C.I.’s allegations against the DYFS worker had been investigated and substantiated, notwithstanding C.I.’s recantation. As a result, the trial judge reasoned, this was not a case of a child “lying about two people.” The trial court did not conduct a hearing under N.J.R.E. 104.
On April 8, 2004, the jury found defendant guilty of first-degree aggravated sexual assault, second-degree sexual assault, and second-degree endangering the welfare of a child. After merging the latter two counts into the first, the sentencing court imposed a term of imprisonment of eighteen years, with a nine-year period of parole ineligibility. Defendant was also sentenced to community supervision for life.
B.
The Appellate Division reversed defendant’s conviction and remanded for a new trial. State v. A.O., 397 N.J.Super. 8, 27, 935 A2d 1202 (App.Div.2007). Rejecting a contrary ruling in State v. Reyes, 237 N.J.Super. 250, 567 A2d 287 (App.Div.1989), the panel held that admitting polygraph results in evidence based on the stipulation defendant signed without a lawyer “impinged upon defendant’s Sixth Amendment right to counsel.” A.O., supra, 397 N.J.Super. at 27, 935 A2d 1202. The panel recognized that although “the Sixth Amendment right to counsel does not attach in the investigative stage, this does not mean that an uncounseled defendant may ... be induced to make binding decisions concerning trial strategy which ordinarily would be made in consultation with trial counsel.” Id. at 22, 935 A.2d 1202.
*79The panel distinguished State v. McDavitt, 62 N.J. 36, 297 A2d 849 (1972), which upheld the admission of polygraph results where a defendant told the jury mid-trial that he would take a polygraph test and, after consulting with his attorney, agreed to do so. The panel also noted the ongoing debate about the reliability of polygraph tests and concluded “it is fundamentally unfair to permit an uncounseled defendant to stake his fate on what may be the equivalent of a coin toss.” Id. at 25, 935 A.2d 1202.
The Appellate Division found that admitting the polygraph evidence in this close case amounted to plain error. Absent that evidence, which the prosecution emphasized was “100 percent” accurate, the State’s ease hinged on the “uncorroborated testimony of a child witness who had recanted and then withdrawn her recantation.” Id. at 27, 935 A.2d 1202.
In addition, the panel ruled that on remand the trial court should conduct the N.J.R.E. 104 hearing that Guenther requires to determine whether defendant may question witnesses about C.I.’s later abuse allegation against the DYFS worker. Id. at 27-28, 935 A.2d 1202. Guenther, decided after defendant’s trial, created a narrow exception to N.J.R.E. 608 and allows a defendant to challenge a witness’s credibility with evidence that the witness made a prior false accusation of a crime. 181 N.J. at 154, 854 A.2d 308; N.J.R.E. 608(b). The Appellate Division saw “nothing in the Guenther opinion that would necessarily limit the holding to only a prior false accusation, as opposed to a false accusation made a short time after the accusation against the defendant.” Id. at 28, 935 A.2d 1202 (emphasis added).
The panel also ruled that the trial court’s in camera review of records from DYFS and the Prosecutor’s Office was not an adequate substitute for an N.J.R.E. 104 hearing. Although the trial judge found that C.I.’s allegation of abuse was substantiated, it did “not set forth on the record any facts concerning the thoroughness of the investigations or the evidentiary basis for those agencies’ conclusions.” Id. at 29, 935 A.2d 1202. The panel noted that “[t]he second recantation may be a critical issue for the *80defense,” and “it is the judge and not the agencies who must make the decision as to its admissibility.” Ibid.
In a concurring opinion, Judge Weissbard urged this Court to bar the use of polygraph evidence altogether, even when offered pursuant to a stipulation signed by a defendant with counsel. Id. at 30, 935 A.2d 1202 (Weissbard, J.A.D., concurring). Judge Weissbard noted that the “foundation of our evidence rules ... is to provide the fact-finder "with only reliable and probative evidence.” Ibid. Because polygraph evidence is widely considered both unreliable and prejudicial, he maintained it “does not make sense to permit the introduction of ... [such] evidence simply because the parties agree to its admissibility.” Id. at 31, 935 A.2d 1202.
We granted the State’s petition for certification. 194 N.J. 446, 945 A.2d 1289 (2008).
II.
The County Prosecutor and Attorney General argue that defendant had no Sixth Amendment right to counsel at the time he was interviewed because adversary proceedings had not yet begun. As a result, they contend that the Appellate Division erred in holding that admission of defendant’s polygraph test results violated his Sixth Amendment rights. They also maintain that the polygraph results were properly admitted at trial under McDavitt. In addition, the County Prosecutor emphasizes that defendant only waived the right to challenge the admissibility of the polygraph evidence, not its reliability.
With regard to C.I.’s later accusations against the DYFS worker, the Attorney General conceded at oral argument that Guenther could apply to false allegations made before and after the underlying incident. The County Prosecutor and the Attorney General argue that no N.J.R.E. 104 hearing is needed, however, because the trial court in effect complied with Guenther’s mandate by reviewing relevant evidence and determining that C.I.’s allegations had been substantiated and were therefore not false.
*81Defendant argues that the Appellate Division’s decision to bar admission of polygraph results can be affirmed on different grounds. He asserts that polygraph results are scientifically unreliable and generally inadmissible at trial. He also argues that the narrow exception in McDavitt does not apply to the different facts of this case. Defendant urges that McDavitt be overruled or limited to its facts.
Defendant also argues that the Appellate Division properly held that Guenther should apply to subsequent false criminal accusations. He contends that a false accusation after the primary event, if closer in time, may be even more probative than a prior false allegation. Defendant argues that he is entitled to a remand for an N.J.R.E. 104 hearing consistent with Guenther on the admissibility of C.I.’s allegation against the DYFS worker.
III.
We first consider whether the admission of the stipulated polygraph results violated defendant’s Sixth Amendment right to counsel. The Sixth Amendment to the United States Constitution provides that “in all criminal prosecutions, the accused shall enjoy the right ... to have the Assistance of Counsel for his defence.” U.S. Const, amend. VI. The New Jersey Constitution guarantees the same right in nearly identical language. See N.J. Const, art. I, H10.
Under federal law, an accused’s Sixth Amendment right to counsel does not attach until “ ‘adversary judicial proceedings have been initiated against him.’ ” Rothgery v. Gillespie County, —. U.S. -,-, 128 S.Ct, 2578, 2583, 171 L.Ed.2d 366, 375 (2008) (quoting Kirby v. Illinois, 406 U.S. 682, 688, 92 S.Ct. 1877, 1881, 32 L.Ed.2d 411, 417 (1972)); see also United States v. Gouveia, 467 U.S. 180, 187-88, 104 S.Ct. 2292, 2297-98, 81 L.Ed.2d 146, 154 (1984). As the Supreme Court has explained, that interpretation “is consistent not only with the literal language of the Amendment, which requires the existence of both a criminal [prosecution] and an accused,” but also with the “core purpose” of *82the right to counsel, which is “to assure aid at trial.” Gouveia, supra, 467 U.S. at 188, 104 S.Ct. at 2297-98, 81 L.Ed.2d at 154 (citation and internal quotation marks omitted).
Once “adversary judicial proceedings” have begun, the right applies not only at trial, but to any “critical stage” of the prosecution, that is, any stage in which the substantial rights of the accused may be affected. See Estelle v. Smith, 451 U.S. 454, 470-71, 101 S.Ct. 1866, 1877, 68 L.Ed.2d 359, 373-74 (1981).
New Jersey law follows the same approach. Article I, Paragraph 10 of the New Jersey Constitution is “consonant with the Federal Constitution on the issue of when the right to counsel is triggered.” State v. P.Z., 152 N.J. 86, 110, 703 A2d 901 (1997) (citation omitted). Accordingly, this Court has held that Article I, Paragraph 10’s right to counsel attaches “upon the return of an indictment or like process because, prior to that point in time, ‘the State’s investigative effort ... is at a preliminary stage, ... the police may still be attempting ... to solve the erime[,] ... [and] the State’s decision to prosecute has not solidified.’ ” Ibid, (alterations in original) (quoting State v. Tucker, 137 N.J. 259, 290, 645 A 2d 111 (1994), cert. denied, 513 U.S. 1090, 115 S.Ct. 751, 130 L.Ed.2d 651 (1995)); see also Reyes, supra, 237 N.J.Super. at 263-64, 567 A 2d 287 (finding right to counsel not triggered when criminal defendant entered into polygraph stipulation before being formally charged).
In this case, defendant had not been formally charged at the time he signed the polygraph stipulation. He consented to a lie-detector test on the same day he was first accused of sexual assault and before any arrest or charge. Because the State’s investigation was at a preliminary stage when defendant agreed to the stipulation, the Sixth Amendment does not afford him a basis for relief. P.Z., supra, 152 N.J. at 110, 703 A2d 901.
IV.
We next consider the enforceability of the stipulation in light of the law regarding polygraph evidence and the facts of this case.
*83A.
As a general rule, polygraph results are not admissible in evidence in New Jersey. State v. Domicz, 188 N.J. 285, 312-13, 907 A2d 395 (2006); McDavitt, supra, 62 N.J. at 44, 297 A2d 849; State v. Driver, 38 N.J. 255, 261, 183 A2d 655 (1962). In 1972, this Court held in McDavitt that “to date ... lie detector testing has not yet attained scientific acceptance as a reliable and accurate means of ascertaining truth or deception.” 62 N.J. at 44, 297 A2d 849. We reaffirmed that view recently and noted that “[i]n the more than thirty years since McDavitt, serious questions about the reliability of polygraph evidence remain.” Domicz, supra, 188 N.J. at 313, 907 A2d 395.
There remains a “lack of scientific consensus concerning the reliability of polygraph evidence, which in turn is reflected in the disagreement among state and federal courts concerning the admissibility of such evidence.” Id. at 312, 907 A2d 395 (citing United States v. Scheffer, 523 U.S. 303, 309-12, 118 S.Ct. 1261, 1265-66,140 L.Ed.2d 413, 419-21 (1998) (reviewing scientific studies showing that accuracy of polygraph tests ranges from 50 to more than 90 percent)). Some studies suggest that the accuracy rate is “little better than could be obtained by the toss of a coin.” Scheffer, supra,, 523 U.S. at 310, 118 S.Ct. at 1265, 140 L.Ed.2d at 419 (citing Iacono & Lykken, The Scientific Status of Research on Polygraph Techniques: The Case Against Polygraph Tests, in 1 Modem Scientific Evidence § 14-5.3).
Nonetheless, to many citizens who serve on juries, polygraph evidence—presented by experts and arrayed in scientific language—has an aura of infallibility.1 That impression “can lead jurors to abandon their duty to assess credibility and guilt” and rely instead on the examiner’s expert opinion. Scheffer, supra, *84523 U.S. at 314,118 S.Ct. at 1267,140 L.EdM at 422. As a result, “the vast majority of states either ban polygraph evidence altogether or do not admit such evidence absent a stipulation between the State and defendant.” Domicz, supra, 188 N.J. at 312-13, 907 A 2d 395.
Twenty eight states bar the admission of polygraph evidence outright. Pulakis v. State, 476 P.2d 474, 479 (Alaska 1970); Bloom v. People, 185 P.3d 797, 807 (Colo.2008); State v. Porter, 241 Conn. 57, 698 A2d 739, 742 (1997); State v. Okumura, 78 Hawai'i 383, 894 P.2d 80, 94 (1995); People v. Jackson, 202 Ill.2d 261, 269 Ill.Dec. 481, 781 N.E.2d 278, 282 (2002); Morton v. Commonwealth, 817 S.W.2d 218, 221-22 (Ky.1991); State v. Legrand, 864 So.2d 89, 98 (La.2003), cert. denied, 544 U.S. 947, 125 S.Ct. 1692,161 L.Ed.2d 523 (2005); State v. Harnish, 560 A.2d 5, 8 (Me.1989); State v. Hawkins, 326 Md. 270, 604 A2d 489, 492 (1992); Commonwealth v. Martinez, 437 Mass. 84, 769 N.E.2d 273, 278 (2002); People v. Jones, 468 Mich. 345, 662 N.W.2d 376, 382 (2003); State v. Jones, 753 N.W.2d 677, 690 (Minn.2008); State ex rel. Kemper v. Vincent, 191 S.W.3d 45, 49 (Mo.2006); State v. Hameline, 344 Mont. 461, 188 P.3d 1052, 1055-56 (2008); Mathes v. City of Omaha, 254 Neb. 269, 576 N.W.2d 181, 184 (1998); Petition of Grimm, 138 N.H. 42, 635 A2d 456, 464 (1993); State v. Fleming, 350 N.C. 109, 512 S.E.2d 720, 738 (1999); Thornburg v. State, 985 P.2d 1234,1241-42 (Okla.Crim.App.1999), superseded by statute on other grounds by Okla. Stat. tit. 12 § 2403 (2008); Commonwealth v. Brockington, 500 Pa. 216, 455 A.2d 627, 629 (1983); State v. Werner, 851 A.2d 1093, 1104 (R.I.2004); Sabag v. Continental South Dakota, 374 N.W.2d 349, 352 (S.D.1985); State v. Damron, 151 S.W.3d 510, 515-16 (Tenn.2004); Nesbit v. State, 227 S.W.3d 64, 66 n. 4 (Tex.Crim.App.2007); State v. Hamlin, 146 Vt. 97, 499 A.2d 45, 53-54 (1985); Elliott v. Commonwealth, 267 Va. 396, 593 S.E.2d 270, 282-83 (2004); State v. Lewis, 207 W.Va. 544, 534 S.E.2d 740, 744 (2000); State v. Dean, 103 Wis.2d 228, 307 N.W.2d 628, 653 (1981); see also People v. Angelo, 88 N.Y.2d 217, 644 N.YS.2d 460, 666 N.E.2d 1333, 1335 (1996) (polygraph evidence properly excluded where there continues to be no showing *85that such evidence is generally accepted as reliable by scientific community).
Virtually all the other states to consider the issue—eighteen in total—limit the admission of polygraph evidence to cases where both parties stipulate to its use. Wynn v. State, 423 So.2d 294, 297 (Ala.Crim.App.1982); State v. Hoskins, 199 Ariz. 127, 14 P.3d 997, 1014 (2000); Ramaker v. State, 345 Ark. 225, 46 S.W.3d 519, 525 (2001); People v. Wilkinson, 33 Cal. 4th 821, 16 Cal.Rptr.3d 420, 94 P.3d 551, 564-67 (2004); Melvin v. State, 606 A2d 69, 71 (Del.1992); Davis v. State, 520 So.2d 572, 573-74 (Fla.1988); Thornton v. State, 279 Ga. 676, 620 S.E.2A 356, 360 (2005); State v. Perry, 139 Idaho 520, 81 P.3d 1230, 1235 (2003); Jackson v. State, 735 N.E.2A 1146, 1152 (Ind.2000); State v. Countryman, 573 N.W.2d 265, 266 (Iowa 1998); Wilkins v. State, 286 Kan. 971, 190 P.3d 957, 970 (2008); Rose v. State, 123 Nev. 24, 163 P.3d 408, 417 (2007); State v. Weatherspoon, 583 N.W.2d 391, 393 (N.D.1998); State v. Souel, 53 Ohio St.2d 123, 372 N.E.2d 1318, 1323-24 (1978) (only for corroboration and impeachment); State v. Brown, 297 Or. 404, 687 P.2d 751, 776 n. 35 (1984) (citing State v. Bennett, 17 Or.App. 197, 521 P.2d 31, 33 (1974)); State v. Crosby, 927 P.2d 638, 642 (Utah 1996); State v. Thomas, 150 Wash.2d 821, 83 P.3d 970, 989-90 (2004); Schmunk v. State, 714 P.2d 724, 731 (Wyo. 1986).
Only New Mexico allows the admission of polygraph exam results without stipulation. Lee v. Martinez, 136 N.M. 166, 96 P.3d 291, 306-07 (2004).
Underscoring the widespread skepticism about the polygraph’s reliability, four states—Massachusetts, Wisconsin, North Carolina, and Oklahoma—have experimented with allowing the admission of polygraph evidence for a number of years, only to reject the practice and reinstate the traditional rule of inadmissibility. See Commonwealth v. Mendes, 406 Mass. 201, 547 N.E.2A 35, 41 (1989), overruling Commonwealth v. A Juvenile, 365 Mass. 421, 313 N.E.2d 120, 123-24 (1974); Dean, supra, 307 N.W.2d at 653, overruling State v. Stanislawski, 62 Wis.2d 730, 216 N.W.2d 8, 14 *86(1974); State v. Grier, 307 N.C. 628, 300 S.E.2d 351, 359-60 (1983), overruling State v. Milano, 297 N.C. 485, 256 S.E.2d 154, 162 (1979); Fulton v. State, 541 P.2d 871, 872 (Okla.Crim.App.1975), overruling Jones v. State, 527 P.2d 169, 174-75 (Okla.Crim.App. 1974); see also Porter, supra, 698 A.2d at 775-76.
Our view remains unchanged. This Court has not sanctioned and does not now entertain the admission of polygraph results. Nor does our holding in McDavitt offer support for the admission of the stipulated polygraph results in this case. That holding addressed very different facts, and we once again decline to “ ‘widen the small aperture of ... McDavitt.’ ” State v. Baskerville, 73 N.J. 230, 236, 374 A.2d 441 (1977) (quoting State v. Cole, 131 N.J.Super. 470, 471, 330 A.2d 594 (App.Div.1974)); see also Domicz, supra, 188 N.J. at 313, 907 A.2d 395 (“[W]e are not prepared to extend McDavitt to unstipulated polygraph examinations, even in a suppression hearing presided over by a judge.”).
McDavitt created a very narrow exception to the rule barring polygraph evidence. In that ease, the defendant’s conduct before the jury provoked the defensive use of polygraph evidence. During his criminal trial, the defendant testified that, after his arrest, he had offered to take a polygraph test to prove his innocence. McDavitt, supra, 62 N.J. at 41, 297 A.2d 849. The prosecutor objected and was mistakenly overruled. Id. at 41, 43, 297 A.2d 849. With the door thus opened, the prosecutor asked on cross-examination if the defendant would be willing to take a polygraph that day. He was. Id. at 41, 297 A.2d 849. After further discussion outside of the jury’s presence, the tidal court granted a recess to allow the defendant time to confer with his lawyer. Id. at 42, 297 A.2d 849. Afterward, with the court’s approval, the parties stipulated as follows: if the defendant passed the test, the State would not oppose a motion for acquittal; if he failed, the test results would be presented to the jury. Id. at 41-42, 297 A.2d 849.
Those unusual facts gave rise to the exception the Court framed: polygraph results may be admitted in evidence on agreement of the parties if their stipulation is “clear, unequivocal and *87complete, freely entered into with full knowledge of the right to refuse the test and the consequences involved in taking it.” Id. at 46, 297 A2d 849. In addition, the examiner must be qualified and the test administered in accordance with established techniques. Ibid.
McDavitt neither discussed nor sanctioned a polygraph stipulation agreed to by a suspect alone.2 McDavitt, therefore, does not offer support for the stipulation used in this case.
B.
We are troubled by more than the prosecution’s misplaced reliance on McDavitt and have concerns about certain matters defendant was asked to stipulate to on his own.
First, we question defendant’s ability to stipulate to the expert’s qualifications. Defendant acknowledges in the stipulation that the polygrapher was an “expert in all phases of both administering polygraph examinations and in the analysis of polygraph chart recordings.” How can a suspect, unschooled in the complexities of polygraphy or the credentials needed to administer a valid examination, stipulate to that statement? What factual basis does a suspect possess to form a view of the examiner’s expertise? Nothing in the record allays this concern. As the Appellate Division noted, “[i]f this were a consumer contract, we might deem it unconscionable.” AO., supra, 397 N.J.Super. at 23, 935 A.2d 1202.
Second, the stipulation waives all challenges to the admissibility of the polygraph expert’s testimony. Although defendant *88may cross-examine the expert about his or her qualifications, the manner in which the examination was conducted, the expert’s opinion, and the possibility of error, the stipulation nonetheless provides for the automatic right of the expert to testify. In other words, even if defense counsel can undermine basic foundational elements of the expert’s testimony and establish at trial that the polygrapher was wholly unqualified, the opinion voiced was not well-grounded, or that the possibility of error was great, the stipulation authorizes the expert to present his or her findings to the jury. That practice offends the core purpose of our evidentia-ry rules. See N.J.R.E. 403, 702.
Third, the stipulation limits defendant’s ability to attack the polygraph evidence. While he may cross-examine the State’s expert, defendant cannot call another witness on the subject. In other words, another expert, no matter how well qualified, cannot offer a contrary opinion about the test results. From accident reconstruction to blood-sample tests, it is common practice for a party to try to rebut the other side’s expert testimony with an expert of its own. To be sure, we have strong reservations about allowing dueling experts to testify about polygraph results because of doubts about the polygraph’s reliability in general. See Dom-icz, supra, 188 N.J. at 314, 907 A 2d 395. But in our adversary system of justice, that legal issue is best addressed by lawyers, not suspects.
Fourth, the stipulation collapses questions about a suspect’s voluntary consent with the legal issue of admissibility. In evaluating a waiver of rights, the focus at first is on whether a defendant knowingly and voluntarily entered into the waiver agreement. Next, the focus shifts to whether the results of that waiver may be admitted in evidence. For example, a defendant can knowingly consent to a search, but in doing so does not agree to the admissibility of everything found during the search. The State must still establish that the evidence taken is admissible in accordance with substantive and evidentiary rules. A seized document that would otherwise be inadmissible—whether because *89the material was irrelevant, prejudicial, privileged, or hearsay—is not cured of its inadmissibility simply because a citizen agreed to its seizure. See, e.g., N.J.R.E. 401, 403, 504, 702. Likewise, defendants may waive their Miranda rights, but they do not stipulate to the admission of all statements that follow. An irrelevant or highly prejudicial comment would still be subject to evidentiary rules that might bar such statements. The same is true for a polygraph exam. A defendant can voluntarily agree to take the test, but its admissibility is a distinctly separate question.
Once properly advised of his rights, defendant could agree to submit to a polygraph. But the ancillary decisions made beyond that choice bear on trial strategy. Defendants typically rely on counsel to object to otherwise inadmissible evidence, attack a witness’s expertise, and decide the most effective way to challenge evidence before a jury. See Rules of Professional Conduct 1.2 (allocating authority between lawyer and client). The stipulation here, though, operated to eliminate counsel’s role by relying on a suspect’s consent.
To avoid that course, a number of other states allow polygraph results by stipulation only upon the approval of defendant’s counsel. See Wynn, supra, 423 So.2d at 299 (Alabama); State v. Valdez, 91 Ariz. 274, 371 P.2d 894, 900 (1962); Sanchez v. State, 675 N.E.2d 306, 308 (Ind.1996); Rose, supra, 163 P.3d at 417 (Nevada); Souel, supra, 372 N.E.2d at 1323 (Ohio); State v. Renfro, 96 Wash.2d 902, 639 P.2d 737, 739, cert. denied, 459 U.S. 842, 103 S.Ct. 94, 74 L.Ed.2d 86 (1982). Such an approach is consistent with the holding in McDavitt but was not followed here.
Our “overarching constitutional responsibility [is] to guarantee the proper administration of justice.” State v. Williams, 93 N.J. 39, 62, 459 A.2d 641 (1983). “When we perceive ... that more might be done to advance the reliability of our criminal justice system, our supervisory authority over the criminal courts enables us constitutionally to act.” State v. Romero, 191 N.J. 59, 74-75, 922 A.2d 693 (2006) (citing N.J. Const, art. VI, § 2, 113; State v. Delgado, 188 N.J. 48, 62, 902 A 2d 888 (2006)). *90We do so now to ensure greater fairness at trial and reliability of jury verdicts.
Relying on our supervisory authority, we bar the introduction of polygraph evidence based on stipulations entered into without counsel. We therefore affirm the Appellate Division’s decision to reverse defendant’s conviction. The conviction rested on the testimony of a young witness who recanted and then withdrew her recantation. No physical or medical evidence corroborated her testimony. To strengthen its ease, the State introduced and highlighted the polygraph evidence discussed above and presented it as “100 percent accurate.” We agree with the Appellate Division that “the polygraph evidence may well have made the difference between conviction and acquittal in this case.” AO., supra, 397 N.J.Super. at 33-34, 935 A.2d 1202. As a result, admission of the evidence was clearly capable of producing an unjust result, see Rule 2:10-2, and warrants reversal and a new trial.
C.
Judge Weissbard’s concurring opinion encourages us to take one more step: to reverse McDavitt and ban polygraph evidence altogether. He reminds us that the core concern of our evidence rules “is to provide the fact-finder with only reliable and probative evidence.” A.O., supra, 397 N.J.Super. at 30, 935 A.2d 1202 (Weissbard, J.A.D., concurring) (citing 1 Wigmore on Evidence § 7a (Tillers rev.1983)); see also Scheffer, supra, 523 U.S. at 309, 118 S.Ct. at 1264, 140 L.Ed.2d at 419. As was true in Domicz, however, we do not have an adequate record to make ultimate findings about the reliability of polygraph evidence at this time. See Domicz, supra, 188 N.J. at 312-13, 907 A.2d 395. Nonetheless, we harbor a number of concerns about McDavitt in light of developments since 1972.
McDavitt recognized that “lie detector testing has not yet attained scientific acceptance as a reliable and accurate means of ascertaining truth or deception,” but concluded that the “art of *91polygraph testing had developed to a point that its results were probative enough to warrant admissibility upon stipulation.” 62 N.J. at 44, 297 A2d 849. As support for that finding, McDavitt cited to two criminal trial courts that had conducted extensive hearings on the reliability of polygraph tests and found that the results were “now generally accepted by authorities in the field and ... capable of producing highly probative evidence in a court of law when properly used by competent, experienced examiners.” Id. at 45, 297 A2d 849. The Court cited specifically to United States v. Ridling, 350 F.Supp. 90 (E.D.Mich.1972), and United States v. Zeiger, 350 F.Supp. 685 (D.D.C.1972). Zeiger, however, was reversed summarily. See United States v. Zeiger, 475 F. 2d 1280 (D.C.Cir.1972). And Ridling was later criticized by its own and one other circuit court for its treatment of polygraph evidence. United States v. Alexander, 526 F. 2d, 161, 166 (8th Cir.1975); United States v. Frogge, 476 F.2d 969, 970 (5th Cir. 1973).
Furthermore, as discussed above, after 1972 four states allowed the admission of polygraph evidence for a number of years but reversed course because of questions about reliability among other reasons. See Commonwealth v. Mendes, supra, 547 N.E.2d at 41; Dean, supra, 307 N.W.2d at 653; Grier, supra, 300 S.E.2d at 359-60; Fulton, supra, 541 P.2d at 872; see also Porter, supra, 698 A.2d at 775-76.
Recent social science studies cast doubt on the reliability of polygraph evidence as well. See Scheffer, supra, 523 U.S. at 309-10, 118 S.Ct. at 1265, 140 L.Ed,.2d at 419-20 (reviewing social science evidence); Porter, supra, 698 A.2d at 759-68 (same); NRC Study, supra, at 323-53 (2003) (reviewing 194 separate studies of polygraph testing).
Those studies explain that polygraphy relies on two assumptions: (1) that deception triggers certain emotional states; and (2) that those emotional states produce specific, measurable physiological changes in the body. Porter, supra, 698 A.2d at 759. As certain empirical evidence has shown, however, there is substan*92tial variation in how individuals respond physiologically when they are lying or telling the truth, and the responses that humans produce in such situations are not specific to either deception or truth-telling. Id. at 760 (citations omitted); NRC Study, supra, at 212-13. The inherent ambiguities in such responses, which arise from individual variations in the subject’s cardiovascular, electro-dermal and respiratory activity, often make it difficult for a test administrator to determine if the examinee is lying, nervous, tired, or simply trying to game the system. NRC Study, supra, at 4, 13-17, 216, 286-90.
As the Supreme Court observed, “there is simply no way to know in a particular case whether a polygraph examiner’s conclusion is accurate, because certain doubts and uncertainties plague even the best polygraph exams.” Scheffer, supra, 523 U.S. at 312, 118 S.Ct. at 1266, 140 L.Ed.2d at 421. Even more troubling, “to the extent that the polygraph errs, studies have repeatedly shown that the polygraph is more likely to find innocent people guilty than vice versa.” Id. at 333, 118 S.Ct. at 1276, 140 L.Ed.2d at 433-34 (Stevens, J., dissenting).
Compounding these questions about reliability is the fact that many lay people tend to view polygraph evidence as bordering on infallible. Id. at 314, 118 S.Ct. at 1267, 140 L.Ed.2d at 422 (majority opinion); A.O., supra, 397 N.J.Super. at 33, 935 A.2d 1202 (citations omitted). Thus, potentially unreliable polygraph evidence may receive undue weight and distract jurors from judging the credibility of witnesses directly.
Such concerns raise questions about the continuing wisdom of McDavitt. Because we lack a factual record, we cannot fully address those issues today. However, a proper record will have to be developed in the trial court the next time a party seeks to introduce stipulated polygraph evidence, agreed to by both sides. That evidence should be introduced only if the parties can first establish its reliability at an N.J.R.E. 104 hearing.
*93V.
We next review whether Guenther should be extended to cover false criminal allegations made after the underlying accusation. The Attorney General properly conceded at oral argument that it should. We agree.
Guenther created a narrow exception to N.J.R.E. 608 and held that “in limited circumstances and under very strict controls a defendant has the right to show that a victim-witness has made a prior false criminal accusation for the purpose of challenging that witness’s credibility.” 181 N.J. at 154, 854 A2d 308. The rule has since been codified at N.J.R.E. 608(b) and applies when the credibility of the victim-witness is the central issue in the case.
To evaluate the proffered impeachment evidence, trial courts must conduct a preliminary hearing under N.J.R.E. 104 and “determine by a preponderance of the evidence whether the defendant has proven that a prior accusation charging criminal conduct was made by the victim and whether that accusation was false.” Id. at 157, 854 A.2d 308. Guenther outlines five factors to consider at the hearing, including “the proximity of the prior false accusation to the allegation that is the basis of the crime charged.” Ibid.
Guenther recognizes that a witness’s prior false criminal allegations may be relevant to the witness’s credibility. That logic applies with equal force to false criminal allegations made soon after the primary allegation. As defendant aptly observes, a false accusation after an event, if closer in time, can be even more probative than a prior false allegation.
Guenther was decided four months after the trial court barred defendant from impeaching C.I. with evidence that she had accused a DYFS worker of sexually assaulting her and then recanted. As a result, the court did not have the benefit of the decision in addressing the later accusation. Before ruling, the trial court reviewed DYFS records and materials from the prosecutor’s office in camera and found that the initial accusation was *94“substantiated.” But the trial court’s review understandably did not comport with the dictates of Guenther.
On remand, the trial court should conduct the N.J.R.E. 104 hearing Guenther requires and make the necessary findings. We offer no opinion on the court’s prior ruling and simply remand for reevaluation in light of Guenther.
Guenther, as modified today, otherwise remains good law. Courts should continue to apply the factors it sets forth in deciding the question of admissibility and be mindful of its concerns to avoid distracting mini-trials. Only “in limited circumstances and under very strict controls,” id. at 154, 854 A.2d 308, does a defendant have the right to challenge a witness’s credibility regarding a later false criminal accusation.
VI.
For the reasons set forth above, we affirm and modify the judgment of the Appellate Division and remand for a new trial.
For affirmance/modification and remandment—Chief Justice RABNER and Justices LONG, LaVECCHIA, ALBIN, WALLACE, RIVERA-SOTO and HOENS—7.
Opposed—None.
Traditional polygraphs measure physiological responses of the person being questioned, typically respiration, heart rate, blood pressure, and galvanic skin responses. Those physiological changes are believed to reveal deception. Comm, to Review the Scientific Evidence on the Polygraph, Nat'l Research Council, The Polygraph and Lie Detection 12-13, 286-90 (2003) ("NRC Study").
In Reyes, supra, a majority of the Appellate Division panel concluded—over reservations expressed in a concurring opinion—that McDavitt did not "prohibit the waiver of the right to counsel at the signing of” a polygraph stipulation. 237 N.J.Super. at 264, 567 A.2d 287. In doing so, the decision could be read to expand McDavitt’s reach. Yet at the same time, the panel expressly invited this Court to address any "change or extension of” McDavitt. Ibid. We do not extend McDavitt in any way today.