BROWN, J.

Section 805.17(3), STATS., states in pertinent part: "Upon its own motion . . . the court may amend its findings or conclusions or make additional findings or conclusions and may amend the judgment accordingly." Jon R. Olsen, the defendant in this operating while intoxicated case, maintains that before the trial court can exercise its authority to reconsider on its own, either the parties or the court must have previously considered the legal issue that the court now wishes to comment upon. We have two responses. First, there are no prerequisites established by the statute except perhaps timeliness. Second, even if there were, the subject matter of the reconsideration here was factual in nature, not legal. The court was simply changing its mind about what it believed to be the important factual issue to be decided. This the court has a right to do under the statute. We affirm.

The trial to the court took place on March 27, 1998. The Village of Thiensville's case against Olsen was largely mine-run. The Village police arrived after a call from a citizen. They found Olsen's car in a drift of snow, resting partially down an incline. The arresting officer detected the odor of alcohol on Olsen's breath and observed that he was "slow in responding." Olsen performed field sobriety tests and, in the officer's opinion, he failed. Olsen was arrested and taken to the police station. There, an intoxilyzer test was performed and a breath alcohol reading of 0.25% was registered. The citizen who called the police testified that Olsen had an odor of alcohol on his breath five minutes after the car was disabled. Also, the officer found an empty half-pint bottle of vodka in the car.

Olsen's defense was that he was not intoxicated when he drove the car. Rather, he was watching the Packer game at his sister's apartment where he had consumed three or four cans of beer. He decided to take a break and go to a liquor store to purchase cigarettes and, while there, also bought one-half pint of vodka to take back. On the way back, he took a short detour to observe a house for sale in which he had an interest. He got stuck while trying to turn around. He testified that it was not until after trying to get his car out of the drift that he decided to drink the entire one-half bottle of vodka because his afternoon was "kind of ruined." While not stated to the trial court, Olsen's claim apparently was that the high test score was due to the vodka and that the vodka was consumed when he was no longer operating his vehicle.

Immediately after the parties had rested, the trial court announced that "the only issue in the case is when did he drink the bottle of vodka." The trial court said that it had no problem with the Village's case-in-chief, that the arresting officer believed Olsen to be under the influence and the breath test confirmed it. However, the trial court observed that the citizen who called the police did not testify that Olsen was under the influence of alcohol. The court believed Olsen's story that he drank the vodka sometime after the car went off the road. The Village objected to the trial court's analysis, arguing that Olsen's story was incredible. But the court was convinced that because Olsen testified that he purchased the vodka shortly before the car was disabled, and because this evidence was not contradicted, Olsen's testimony was credible. The trial court found Olsen not guilty.

Just days later, the trial court, on its own motion, set the matter on for a hearing to reconsider its judgment. At the hearing on the motion, the court wasted no time in explaining why it moved to reconsider on its own motion. To paraphrase the court, it held that it had not given sufficient weight to the presumption established by the breath test. This presumption holds "not withstanding the fact that he may have consumed an alcoholic beverage in the interim." The trial court concluded that the "burden of proof then shifts to the defendant to establish that the consumption of the one-half pint of Vodka would have materially affected the breath test to the extent that it would put it over the legal limit."

Following the trial court's holding, it allowed colloquy regarding the decision. We read Olsen's objection to have argued the following: The issue at trial was "when did Olsen drink the vodka" and not "assuming the vodka was consumed after the car was disabled, what effect would the vodka have on the breath test." Thus, Olsen contended that the trial court was changing the issue which was the subject of debate at the trial. Olsen complained that the court was, in reality, changing the rules by now requiring him to obtain an expert to show the effect of the vodka upon the breath test results.

The trial court responded by reiterating its reason for reconsideration — that it had not accorded to the Village the proper weight of the statutory presumption at the time of trial. Once accorded, Olsen's defense was insufficient to rebut the presumption. The trial court changed its finding to guilty, assessed a forfeiture and Olsen appeals.

Construction of a statute is a question of law. See O'Neill v. Buchanan, 186 Wis.2d 229, 233, 519 N.W.2d 750, 752 Ct. App. 1994). We review statutory construction issues de novo, without deference to the trial court. See id.

Olsen's theory is that reconsideration assumes a question which has been previously considered. He reads O'Neill to so hold and contends that O'Neill governs this case. Olsen then asserts that neither prior to nor during the course of the trial did either party argue or present testimony relating to the statutory presumptions of intoxication or the respective burdens of proof. Thus, the trial court did not have the occasion to consider such arguments.

O'Neill does not support Olsen's position. In that case, the appellant, an interested party to a will construction, did not appear at the hearing construing the will. He later moved for relief under § 806.07, STATS., claiming excusable neglect. Alternatively, he asked for reconsideration of the order construing the will. The trial court denied both motions and we affirmed. In doing so, we held that "[t]his course of events is akin to default by a party." O'Neill, 186 Wis.2d at 234, 519 N.W.2d at 752. We opined that "[i]f a party has not yet appeared and presented arguments in the litigation, the court has not considered that party's arguments in the first instance." Id. Although we did observe how "reconsideration assumes that the question has previously been considered," this must be read in the context of the discussion concerning the effect of a default upon a person's standing to move to reconsider. See id. The statement certainly cannot be read as a blanket holding construing § 805.17(3), STATS., as requiring a condition precedent. Moreover, Olsen's reliance upon O'Neill completely ignores footnote 3 of our opinion which stated, "We note that our opinion does not address the authority of the court to reconsider a decision upon its own motion." See O'Neill, 186 Wis.2d at 235 n. 3, 519 N.W.2d at 753. O'Neill does not help Olsen.

We consider § 805.17(3), STATS., to be unambiguous. The statute plainly allows a court "[u]pon its own motion" to " amend its findings or conclusions." See § 805.17(3) (emphasis added). There is no condition precedent listed except perhaps that the court must act within twenty days of its original decision. There is no requirement, therefore, that the reason for the reconsideration must have been a subject of discussion, testimony or debate at the original trial or hearing. Good reason exists for such a rule. A judge's job is to do justice. A judge endeavors to come to the right result. The law gives a judge the right to change his or her mind, so long as it is done in a timely fashion and the parties are given a fair chance to be heard. As was the case here, the state also allows a judge to engage in critical analytical thinking that was not at all considered by the court in the first instance. A judge should not have to live with the consequences of a decision that he or she, upon reflection, believes to be wrong. We reject Olsen's argument for a contrary construction.

Moreover, Olsen misconstrues what the trial court did in this case. The trial court did not surprise Olsen after judgment. The presumption of intoxication was present at the trial. In fact, it was stipulated to by the parties. It never went away. After testimony closed and before either of the parties had an opportunity to make arguments to the court, the court announced that the issue, as the court saw it, was "when did [Olsen] drink the bottle of vodka." This was not an issue brought to the court's attention by the parties. It was an issue seized upon by the court and the court alone. After settling on the question, the court answered it. Olsen drank the vodka after operating the vehicle. At that time, the court was obviously convinced that answering the question it had decided was the key to the case. So, the trial court found Olsen not guilty.

But upon reflection, the trial court decided that it had asked the wrong question or at least had not asked enough of the right question. The issue was not simply whether Olsen drank the vodka after operating the vehicle, but also whether an affirmative factual finding would be sufficient to overcome the presumption created by the .25 reading. What the court was doing was reconsidering its earlier belief that the question should be solely whether Olsen drank the vodka after operating the vehicle. The trial court changed the question that it had sua sponte asked initially in favor of a new sua sponte question. After the court was satisfied that it had now framed the question correctly, it found that Olsen had not produced the quantum of evidence sufficient to convince it that the presumption of intoxication had been rebutted.

Olsen claims that the trial court's decision effectively forces persons in Olsen's position to prove not only that the alcohol was consumed after operation of the vehicle ceased, but also that the postoperation alcohol was the reason for the high breath test result. This, he claims, is an error of law because the law does not require, as a condition precedent to rebutting the presumption of intoxication, the production of an expert to prove the effect of postoperation alcohol on a chemical test for intoxication In Olsen's words, a particular "type" of evidence is not a necessary prerequisite to overcoming the presumption of intoxication.

But a close look at the trial court's decision shows that it did not change its mind based on a belief that, as a matter of law, Olsen needed an expert to rebut the presumption of intoxication. Rather, the trial court simply decided that the question the court itself had framed at the conclusion of evidence was insufficient. In sum, the trial court's decision to reconsider was fact intensive, not law intensive. Perhaps had the testimony been that Olsen had not one drop of alcohol while watching the Packers, and had there been no testimony of alcohol on Olsen's breath five minutes after he was seen driving his car, the trial court may never have felt the need to reconsider its original decision. In that case, had the trial court answered the original question the same way that it did in this case, the obvious inference would then have been that, of course, the test reading was solely the result of the vodka. But those were not the facts. The trial court changed the question because the facts in the case compelled it. That finding is affirmed.

By the Court. — Judgment affirmed.