dissenting. In a substantial number of cases in this Court, the real legal issue is to determine the proper standard of review. That is, there has been some error in the trial court proceedings, and we must decide whether the error warrants appellate intervention. At one extreme, we can hold that any error, however small and technical, must have affected the outcome, and reverse and remand for the trial court to redo the case without the error. At the other extreme, we can affirm if it is possible to conclude that the trial court might have reached the result it did if there had been no error, even if we have no idea what the trial court would have done in the absence of the error. Not surprisingly, up until now we have generally avoided the extremes and intervened only when it is clear that the error may have affected the result and we cannot say what the trial court would have done if there had been no error.
Although the majority opinion does not acknowledge it, this is a case where the only real issue is to determine the applicable standard of review; all else follows from this choice. Again, without acknowledging what it has done, or that the matter is debatable, the majority uses the extreme pro-affirmance standard of review and affirms because it was possible for the trial court to have revoked defendant’s probation on the admissible evidence presented, without considering the inadmissible evidence. Of course, the majority cannot, and did not, *446conclude that the trial court ignored the inadmissible evidence or would have revoked defendant’s probation based solely on the admissible evidence. The only way to find out what the trial court would have done is to ask it; this is exactly the step the majority refuses to take.
Standards of review are important. The review standard is often more likely to determine the result of an appeal than the substantive law about which the parties are arguing. Thus, we should take as much care in developing a review standard as we do in developing the substantive law in issue in the appeal. Unfortunately, this review standard is chosen and applied with no stated rationale and without the benefit of briefing and argument.1 No one, and particularly not the defendant, will know why the majority chose to apply a standard of review that made this appeal illusory; and we will have no idea to what cases this standard of review will apply in the future.
Certainly, the majority did not adopt this standard of review because our precedents required it. In fact, this new standard of review is inconsistent with the vast majority of our decisions, and the majority has made no attempt to explain or reconcile this inconsistency.
A probation-revocation proceeding is a hybrid criminal/civil proceeding. As a result, we should look to the standard of review for evidentiary errors in both types of cases. In civil cases, the standard of review is normally governed by V.R.E. 103. Under that standard, errors in the admission of evidence are grounds for reversal if “a substantial right of the party is affected.” In applying this standard of review, we do not look to whether the trial court could have reached the same result without relying on the offending evidence. We reverse when the court could have reached a different result relying only on admissible evidence.
For example, in Jakab v. Jakab, 163 Vt. 575, 581, 664 A.2d 261, 263-64 (1995), a trial to court, we affirmed the trial court judgment despite its admission of inadmissible evidence because “[w]e can say with sufficient certainty that the conclusion would have been the same even if the [inadmissible evidence] had not been considered.” On the *447other hand, in Gilbert v. Gilbert, 163 Vt. 549, 556, 664 A.2d 239, 242 (1995), we reversed because of the admission of inadmissible evidence, saying, “If we could be sure that the [inadmissible evidence]' played no role in the court’s decision, we would not reverse.”
In criminal cases, our standard of review is more rigorous because of the elevated standard of proof. Thus, where evidence is admitted in violation of a defendant’s constitutional rights, we can find the error harmless and affirm only if the error was harmless beyond a reasonable doubt. See State v. Lynds, 158 Vt. 37, 42, 605 A.2d 501,503 (1991). This standard applies in trials to court. See State v. Percy, 149 Vt. 623, 629, 548 A.2d 408, 411 (1988). We have recently held that it applies to nonconstitutional errors. See State v. Carter, 164 Vt. 545, 555, 674 A.2d 1258, 1265 (1996).
No doubt because of the liberty interest at stake, our one case announcing a standard of review for an evidentiary error in a probation-revocation case chose a standard similar to the criminal standard. Thus, in State v. Emery, 156 Vt. 364, 371, 593 A.2d 77, 81 (1991), we affirmed despite the admission of inadmissible evidence because “we are convinced that introduction of the evidence did not have the slightest effect in this case and was therefore harmless.” The majority has ignored the standard of review announced in Emery.2
The only exception to the uniform rejection of the standard of review used by the majority has been in juvenile cases. In re B.S., 163 Vt. 445, 659 A.2d 1137 (1995), the case cited by the majority, charts the development of this limited standard of review in juvenile cases. See id. at 454, 659 A.2d at 1143. I believe this exception is justified in juvenile cases because of the primary focus on-the interests of the children and the adverse effect on children caused by the uncertainty of lengthy litigation. The exception is particularly inappropriate where the interest at stake is the defendant’s liberty. In any event, the majority makes no attempt to justify why it has relied on a limited exception for juvenile cases rather than the general standard of review for evidentiary errors.
*448Other courts reviewing probation revocations where the trial court has admitted hearsay in violation of the probationer’s confrontation rights have used review standards similar to our criminal or general civil standards. For example, in United States v. Bell, 785 F.2d 640 (8th Cir. 1986), the defendant was charged with breaching a probation condition that required obedience to all local, state and federal laws. The affidavit stated that the defendant had three times tested positive for marijuana in his blood and once was arrested for DUI and possession of narcotic paraphernalia and marijuana. During the hearing, evidence was admitted through defendant’s former probation officer that the defendant was once investigated for cocaine trafficking. The Court of Appeals concluded that the evidence supporting the original charges was proper, but the hearsay evidence of the cocaine investigation violated the defendant’s confrontation rights. Because it found that the trial court “seems to have relied on this evidence in deciding to revoke Bell’s probation,” it remanded for reconsideration of the sanction without consideration of the inadmissible evidence. Id. at 645; see also United States v. Lacey, 648 F.2d 441, 444-45 (5th Cir. 1981) (where trial court’s decision to revoke defendant’s probation must be remanded so that judge can state factual findings and reasons relied upon for revocation).3 We adopted a similar approach in State v. Styles, 166 Vt. 615, 616, 693 A.2d 734, 735 (1997), although in that case the State alleged violations of more than one condition and the court did not specify which of the violations warranted revocation.
The majority offers no policy reason to change the standard of review for probation-revocation cases, and I know of no policy reason to overrule Emery or to adopt a more limited standard of review for probation-revocation cases. There are, however, strong reasons to reject the majority’s standard. The district court is the “sole trier of fact” in probation-revocation proceedings. State v. Bushey, 149 Vt. 378, 382, 543 A.2d 1327, 1329 (1988). The majority’s standard inevitably involves us in weighing the evidence, although we have not seen the witnesses. Thus, based solely on the transcript, the majority details the cross-examination of the mother of the contacted child, because it finds it effective in “undermining] her claim that alterna*449tive care arrangements minimized any contact between defendant and S.L.” This is a judgment for the trial court, not this Court.
The real defect in the majority’s standard of review, however, is that there is no trier of fact. Despite statements like that above, the majority has not found the facts; it has concluded only that the trial court could have found them in a certain way. Although the trial court found facts, it did so based in part on inadmissible evidence. No court has found the facts based on the admissible evidence.
There are serious due process concerns about a system in which no court finds facts based on proper evidence. In Gagnon v. Scarpelli, 411 U.S. 778, 786 (1973), the United States Supreme Court held that due process requires “a written statement by the factfinders as to the evidence relied on and the reasons for revoking probation.” See also Black v. Romano, 471 U.S. 606, 613-14 (1985). Under the majority’s analysis, a probation revocation is valid even if the evidence relied upon is inadmissible in whole or in part, as long as there was some evidence, not relied upon, to justify the revocation. This standard of review eviscerates the due process requirement that the trial court state the evidence it relied upon.
The danger of the new standard of review is apparent from the majority’s analysis of the facts of this case. The trial court revoked probation because defendant’s contacts with children under sixteen years of age were “numerous” and not “inadvertent.” The only evidence that directly supports that contacts were numerous and not inadvertent was the improperly admitted hearsay. The majority’s answer is that positive evidence of frequency of contacts could be supplied by the trial court’s disbelief of the testimony of defendant and the mother of the child that contacts were rare and inadvertent. If the possibility of the trial court’s disbelief can prevent reversal in this case, then it can in virtually any case.4 The Austin rule that the court must make an explicit finding of good cause to admit hearsay in violation of a probationer’s confrontation rights will be nullified by the standard of review.
This new standard of review lacks even-handedness in using evidence not explicitly relied upon by the' trial court. Defendant *450argued to this Court that the district court erroneously admitted and relied upon a treatment summary that contained hearsay. The majority rejects the argument because “the record is devoid of any indication that the trial court relied upon the summary.” I believe that this analysis is correct, but it is abandoned when the majority finds evidence that supports affirmance. Whether or not the trial court relied upon particular pro-affirmance evidence is irrelevant to its review.
Finally, I think it significant that the extreme standard of review has been chosen to avoid what is likely to be an easy and expeditious error correction. As we held in Styles, the only consequence of reversal is that the trial court must make findings and conclusions based solely on the nonhearsay evidence, a task that would have taken far less time than the opinions in this appeal. If the nonhearsay evidence is as strong as the majority believes, the outcome would be the same, but we would have dealt fairly with a defendant who faces a lengthy jail sentence as a result of this case.
Under Emery, we must ask whether the error in admitting evidence played no role in the court’s decision to revoke probation. Because the trial court revoked probation based on its conclusion that numerous violations had occurred, we cannot conclude that the inadmissible evidence played no role in the trial court’s decision. As in Styles, I would remand for the trial court to issue a revocation decision solely on the admissible evidence, and I dissent from the majority’s failure to do so. I am authorized to say that Justice Johnson joins in this dissent.
Defendant argued that we should reverse the trial court because it relied on evidence inadmissible under State v. Austin, 165 Vt. 389, 390, 685 A.2d 1076, 1077 (1996). The State argued that Austin should not be applied retroactively. It never argued that even if Austin applied, the case should be affirmed on the rationale adopted by the majority and.never addressed the standard of review at all. Thus, the standard of review was never addressed by either party.
I do not understand the majority’s argument that Emery does not state a standard of review inconsistent with that employed in this ease. In that case, defendant argued that two hearsay statements should not have been admitted, and this was cause for reversal. This Court first noted that the statements were “cumulative.” 156 Vt. at 371, 593 A.2d at 81. Under the standard of review used by the majority in this case, the decision should have gone no further since the trial court decision was, therefore, supported by admissible, untainted evidence. Instead, the Court went on to hold that any error in admission was not cause for reversal because the “evidence did not have the slightest effect in this case.” Id.
United States v. Frazier, 26 F.3d 110, 114 (11th Cir. 1994), the case cited by the majority, is not inconsistent with the approach of the other federal courts because the court analyzed the issue as we did in Emery. In Frazier, the court found no reversible error because the nonhearsay evidence in support of revocation was overwhelming. Id.
The majority suggests that there is more than a possibility of disbelief in this case, but it bases this conclusion on its own evaluation of the weight of the evidence rather than any indication that the trial court used its disbelief of the testimony of the mother as affirmative evidence of numerous and intentional contacts. In doing this, it is usurping the role of the trial court to determine credibility of witnesses and weight of the evidence.