HEFFERNAN, J.

Robert M. Cook was found not guilty of first-degree murder by reason of a mental disease or defect which deprived him of the substantial capacity either to appreciate the wrongfulness of his conduct or to conform his conduct to the requirements of society. As a consequence of this verdict, on December 21, 1971, the trial judge ordered Cook committed to the Department of Health and Social Services, to be placed in institutional custody, care, and treatment until discharged under the provisions of sec. 971.17, Stats. Cook was committed to the Central State Hospital.

Cook's 1972 application for release was denied; and in September, 1973, he petitioned for re-examination, contending that his mental disease or condition had disappeared and that he could be safely discharged without danger to himself or others. After a hearing, the trial court issued an order denying the discharge. This appeal is from that order.

Only one expert witness, a well-recognized psychiatrist, testified. At the hearing, he concluded that the mental condition which occasioned the outburst of violence and the murder of 1971 had been cured and that Cook could now be released with safety to himself and to others. Despite the fact that the psychiatrist was the only expert witness, the trial judge denied the petition for release. The only question presented to the court on this appeal is whether the trial judge could reject the recommendations and conclusions of the only expert witness, which were uncontradicted by any other testimony. We conclude that the judge's decision in respect to a release under the provisions of sec. 971.17, Stats., is of a discretionary nature and will not be reversed or set aside by this court unless the record reveals an abuse of discretion. We affirm the order.

Judge JACKMAN considered the entire factual background to these proceedings. He took into consideration his own knowledge of the defendant's past conduct. He carefully considered the expert's findings and conclusions and compared them to the conclusions which had been reached by the same psychiatrist a year earlier. The evidence and the facts were rationally and carefully considered. Under those facts and in the light of the entire history of Robert Cook's case, the conclusion reached by the judge was a reasonable one, based on all the factors necessary for his decision. We cannot conclude that the decision constituted an abuse of judicial discretion.

The attorney for Cook argued that the test to be used on review is whether the finding reached by the judge is contrary to the great weight and clear preponderance of the evidence. Since the only expert testimony was that of Dr. Leigh M. Roberts, a psychiatrist, and he stated that the defendant was no longer mentally ill and could be released without danger, defendant argues that the trial judge's conclusion was contrary to the only evidence. On the other hand, the state takes no position on the burden of proof and argues that, irrespective of the standard of review, a court need not give unquestioning credence to any witness, whether or not that witness qualifies as an expert. Sec. 971.17(2), Stats., provides:

"The burden shall be on the defendant to prove that he may safely be discharged or released without danger to himself or others. If the court is so satisfied, it shall order the discharge of the defendant . . . ."

It should be noted that the statute does not ask the trial judge to hinge his decision on the question of whether the defendant remains mentally ill. Rather, the judge must consider whether the defendant will be a danger to himself or to others. This essentially is the standard which a trial judge uses when he sentences defendant in an ordinary criminal proceeding. For example, under the American Law Institute, Model Penal Code, sec. 7.03(3), an extended term of imprisonment is appropriate if the defendant "is a dangerous, mentally abnormal person whose commitment . . . is necessary for protection of the public."

Under the American Bar Association, Standards Relating to Sentencing Alternatives and Procedures, sec. 2.5(c)(i), total confinement is appropriate if it is "necessary in order to protect the public from further criminal activity by the defendant." In a sense then, a trial judge who is faced with the question of whether a defendant should be released after a period of confinement under sec. 971.17, Stats., is faced with the same problem that frequently faces a sentencing judge, i.e., the determination of dangerousness. This court established in McCleary v. State (1971), 49 Wis.2d 263, 182 N.W.2d 512, that a sentence should be imposed only in the exercise of judicial discretion. We believe that the sentencing decision and the release decision under sec. 971.17 are analogous. Each is to be exercised on the basis of judicial discretion, after the rational consideration of all relevant factors. Moreover, sec. 971.17 provides that a defendant will be discharged when the defendant has proved that he may be released with safety to himself and to others. Only if the court is "so satisfied . . . shall [it] order the discharge of the defendant."

The use of the verb, "satisfy," in this context is indicative that the judge's decision is of a discretionary nature and must be subjectively determined by the process of rationally applying the facts to the particular circumstances. The Oxford English Dictionary, Oxford University Press, Letter "s," page 123, gives as the definition of "satisfy" appropriate to the context of this statute: "7. To furnish with sufficient proof or information; to assure or set free from doubt or uncertainty; to convince."

We conclude, therefore, that the statute envisages a sufficient quantum of proof to assure or convince a reasonable trial judge that a defendant may be released from a mental hospital without danger to himself or to others. In no case, however, can a judge ever conclude with absolute certainty that any person under all possible conditions will not be dangerous. Accordingly, we do not conclude that a trial judge's decision to discharge a person committed under sec. 971.17, Stats., shall be only on proof beyond a reasonable doubt that the release will entail no danger to the defendant or others. Rather, we will sustain a trial judge's exercise of discretion if, on the basis of the record and the rationale expressed, we can conclude that a judge could arrive at the decision in fact reached. To sustain the decision, we need not conclude that this court would agree with that conclusion.

In the instant case, the trial judge took into consideration that the psychiatrist who now testified that the defendant could safely be discharged, less than a year before had testified that he doubted that treatment of the defendant would be effective. In this frame of reference, it was not unreasonable for the trial judge to place less than absolute confidence in the psychiatrist's present conclusion, when only a year before he had stated that the mental illness would continue indefinitely. The trial judge, taking judicial notice of the witness' testimony a year before, recited the factors that would make the present testimony less than completely reliable.

There was also evidence to show that the defendant at the time of the original criminal trial was suffering from chronic paranoid schizophrenia, which had existed over a period of years, a condition that the same psychiatrist had testified would be long lasting and from which recovery could not be expected quickly. The trial judge took note of the prior testimony and concluded, in view of the long-standing history of the severe mental illness and the comparatively short period during which the defendant apparently was no longer mentally ill, that he was not satisfied that the condition was not temporary — a remission rather than a cure.

The trial judge also noted that the conduct stimulated by the mental illness resulted in a brutal murder. The judge pointed out that, in the event that the prognosis was not reliable, the risk to society was substantial indeed.

The trial judge also noted that alcohol had played a substantial part in triggering the defendant's past misconduct. He observed that, since alcohol was not readily available in the hospital, the defendant's reaction to its easy availability could not be forecast with reasonable certainty.

He also considered the psychiatrist's testimony that, although the defendant no longer was mentally ill, he nevertheless continued to suffer from a lesser form of mental disability. The psychiatrist characterized the present condition as a schizoid personality. Accordingly, an abnormal mental condition persisted even though there was evidence that the defendant was no longer acutely mentally ill.

The testimony of the psychiatrist also indicated that, although Cook had demonstrated on several occasions his ability to cope with aggravating situations, he still harbored an abnormal amount of anger within him and that such potential for anger could be triggered by the stresses of life outside of the institution and by the use of alcohol. The trial judge reasoned:

"[The] defendant has been institutionalized and has adjusted well to that state. Institutionalization has its frustrations and inconveniences, but it also is an ordered life without the sometimes insurmountable strains and stresses that the competition on the outside produce to the ordinary person."

The opinion of the trial judge shows that he carefully evaluated the expert medical testimony. In his opinion, the trial judge recited the circumstances under which the defendant would in all probability be obliged to live were he released. He carefully balanced the psychiatric conclusion that the defendant was no longer acutely mentally ill against the testimony that the defendant still had a mental disorder, that his personality contained an abnormal amount of sociopathic anger, and that he had not been completely tested in normal environment. On the basis of these conflicting considerations, the trial judge, after a thoughtful and articulated analysis, concluded that he was not "satisfied" that the defendant was not dangerous.

We believe that the trial judge did all that the statute asks of him. A trial judge when asked to determine at a sentencing procedure whether a defendant will be dangerous in the future is called upon to forecast the impossible. He must exercise his judgment on the basis of the facts as they are presented to him at the time the decision must be made. In considering the release of a patient who has committed a brutal killing for which he was not criminally responsible because of a mental defect, the trial judge faces a similar impossible task. He is required to consider a psychiatrist's testimony and his predictions, but the judge, not the psychiatrist, has been selected by the legislature as the officer of the state who must be "satisfied" that the release can be accomplished without danger to the defendant or to others. All that society can ask of a judge is that he give careful and rational consideration to the information available. If the conclusion he reaches is a reasonable one on the basis of the facts and the circumstances — if it is not the result of an abuse of discretion — this court will affirm the decision. The trial judge in the instant case considered all the relevant facts. He carefully analyzed the expert testimony and reached a conclusion that was reasonable under the circumstances.

By the Court. — Order affirmed.