OPINION
SINGLETON, Judge.James Houston appeals his sentence of fifteen-years imprisonment, with three years suspended, imposed following his nolo contendere plea to manslaughter, former AS 11.15.040. Houston was originally charged with and convicted of second-degree murder. He previously appealed that conviction and the supreme court reversed, finding error in the lower court’s failure to grant a bifurcated trial on the separate defenses of insanity and self-defense and in *1026the court’s compelling a psychiatrist hired by the defense to testify against the defendant. Houston v. State, 602 P.2d 784 (Alaska 1979). After the remand, the district attorney changed the charge to manslaughter. Houston pleaded no contest to that charge on May 6, 1980. After hearing argument from the parties and the testimony of a correctional officer, the trial court imposed the sentence which forms the basis of this appeal.
Houston’s crime is described in the opinion of the supreme court, 602 P.2d at 785-86. After drinking heavily throughout the evening, Houston went into the men’s room of an Anchorage bar and shot and killed a man. Houston testified that the victim had made a derogatory, racial remark to him and had reached toward his pocket, as if going for a gun. One witness testified he had not heard Houston or the victim say anything and that only two to four seconds had elapsed between the time the victim entered the men’s room and the first shot.
Houston was a twenty-six-year-old army sergeant at the time of the offense, serving an unaccompanied tour of duty at Fort Richardson during which his wife and child resided in Germany, their native homeland. Houston had served honorably in the army for seven years, including 11 months in Vietnam. He had no prior criminal record. Following his service in Vietnam, he began to drink heavily. The drinking intensified after he came to Alaska. Eight days before the shooting, he checked himself into El-mendorf Hospital for alcohol abuse treatment. He was diagnosed a chronic alcoholic and released to resume his military duties two days before the crime. He had been detoxified but had received little or no psychological counseling. Upon his release Houston began drinking heavily again, and he subsequently had little recollection of the events during the time between his discharge and the shooting.
Three psychiatrists diagnosed Houston as a chronic alcoholic. Although they disagreed over whether he was legally responsible for his actions on the night of the shooting, they agreed that the shooting appeared to have occurred as a reflex reaction — a response to a perceived threat such as one that he had been trained to react to by the military.
Houston had been incarcerated for almost four years at the time of his resentencing. He had been a model prisoner who had received no disciplinary reports and had been praised for his institutional adjustment and job performance in prison. He had completed over a year of college with high grades. The initial sentencing order recommended that he receive psychological and alcoholism counseling, and the first classification report by the Division of Corrections states “his alcoholism and emotional instability should be considered.” The prison records do not show that Houston received any significant counseling during his first four years in prison.
Houston was originally sentenced to fifteen years’ imprisonment after his conviction for second-degree murder. Resentenc-ing on the manslaughter plea occurred in May, 1980, before the same judge who conducted the original trial, and sentencing. Houston was given a sentence of fifteen years’ imprisonment, with three years suspended. The reasons for the imposed sentence were briefly set out by the judge.1
*1027Houston lodges four separate attacks on his sentence. First, Houston claims that the sentencing judge failed to discuss adequately the Chaney factors.2 Although the court “need not recite the goals of sentencing as long as it is clear that it has considered those goals” Evans v. State, 574 P.2d 24, 26 (Alaska 1978),3 full explanation of a sentencing decision contributes to the rationality of the sentence, facilitates the reviewing court’s evaluation of the propriety of the sentence, and fosters public confidence in the criminal justice system. A full explanation may also aid the correctional authorities and have therapeutic value in assisting the defendant to accept his sentence without bitterness. Alpiak v. State, 581 P.2d at 665 n.2; Perrin v. State, 543 P.2d 413, 418 (Alaska 1975); State v. Chaney, 477 P.2d 441, 447 n.26 (Alaska 1970).
We believe the trial court adequately addressed the facts in a manner consistent with the cases cited. It is clear from the judge’s remarks that he had assumed that a high potential for rehabilitation existed when he imposed a relatively lenient sentence for second-degree murder and that he viewed Houston’s good intervening institutional record as justifying that original determination. We therefore conclude that the court adequately addressed rehabilitation. Less adequate is the treatment of isolation, deterrence, and affirmation of community norms. Where a trial judge rejects probation as a suitable sentence and concludes that a period of incarceration is in order, the record should reflect the basis for that conclusion. Incarceration and isolation are not synonymous. A judge may feel that a period of incarceration is necessary for rehabilitation or deterrence to emphasize to the defendant the seriousness of his offense and the likely consequences of recidivism. However, as a sentencing goal, isolation is reserved for those who can be neither rehabilitated nor deterred; that is, those who are reasonably likely to commit further criminal activity unless incarcerated.
It is clear in this case that the trial court considered Houston dangerous despite his rehabilitative potential. Given the nature of the crime and the circumstances surrounding its commission, as well as the absence of a reasonable explanation for its occurrence, we are unable to conclude that the sentence imposed reflects improper consideration or balancing of isolation, deterrence, and affirmation of community norms.4
*1028Houston’s second and third claims are that impermissible considerations entered into the sentencing decision. He argues that certain remarks of the sentencing judge, directed to him just prior to the announcement of the sentence, indicate that the judge improperly relied on speculation that the shooting was racially motivated.5 The state responds that there was evidence in the record to suggest that the killing was racially motivated and that the comments reflect only the judge’s inability to ascertain the true reason for the shooting. We agree with the state that the judge’s comments reflect his perplexity over a killing which was difficult to understand. The judge’s remarks followed Houston’s own statement concerning his inability to explain the shooting. Under these circumstances, we find no error. The trial court must endeavor to ascertain the reasons for a criminal episode in order to determine the likelihood of a reoccurrence and the consequent danger, if any, presented by the defendant to the community. Other things being equal, an unexplained killing may warrant a longer period of incarceration to protect the public than one where the likely explanation suggests causes that can be cured or controlled. Cf. Nelson v. State, 619 P.2d 480, 481 n.2 (Alaska App.1980) (Inexplicable nature of crime cited as one of the factors justifying the trial court’s conclusion that the prospect of the defendant’s rehabilitation was unlikely).
Houston argues next that other remarks by the sentencing judge indicate that the judge improperly relied on the prosecutor’s argument that Houston’s offense was really a first degree-murder.6 We are satis*1029fied that the trial judge did not sentence Houston as a convicted murderer.
The judge observed the demeanor of the witnesses and heard their examination and cross-examination. Testimony under oath and subject to cross-examination is clearly appropriate for consideration by the trial court in making the necessary fact findings in preparation for imposition of sentence. The court’s remarks indicate that it rejected the state’s theory of felony murder, i.e., a murder committed in the course of a robbery, and finally discounted racial motivation as the basis of the crime. The court remained in doubt regarding the genesis of the crime, a doubt shared by the defendant himself. We find no prejudicial error in the comments made by the trial judge in his dialogue with counsel and the defendant.
Houston’s final argument is that a sentence of fifteen years’ imprisonment with three years suspended is excessive. The state argues that the goals of isolation, deterrence, and community condemnation support the present sentence even if the record shows that Houston does not require further incarceration for his rehabilitation.
In light of the finding, which is supported by the record, that Houston committed an unexplained homicide requiring isolation to protect the community, we conclude that the sentence is not excessive. In appropriate circumstances, sentences of ten years or more for manslaughter are not clearly mistaken even where the defendant is a first offender with a good background. See, e.g., Notaro v. State, 608 P.2d 769, 770 (Alaska 1980); Padie v. State, 594 P.2d 50, 62-63 (Alaska 1979).
Because of the uncertainty surrounding Houston’s motivation for the killing and the extent to which his mental and emotional states present the risk of further recidivism, we cannot conclude that the length of the sentence is “clearly mistaken.” Although the trial court found Houston’s conduct while he was in custody commendable, this finding does not provide the assurance that Houston could function effectively outside a prison environment.
We assume that the trial court’s failure to include a provision for psychological and alcohol counseling in the final judgment was an oversight since it was included in the original judgment. Therefore, we direct the amendment of the present judgment to reflect that recommendation. We recognize, however, that this is a recommendation, not an order, and that the final determination regarding appropriate rehabilitative services is administrative, not judicial. Finally, we express no opinion as to whether Alaska Rule of Criminal Procedure 35(b) would warrant a trial judge in releasing or shortening the sentence of a convicted felon whom the trial judge believes dangerous to the community solely because of a disagreement between the judge and the Division of Corrections over the convict’s need for alcohol counseling or psychological treatment.
The judgment of the superior court is AFFIRMED as modified.7
. The full explanation given by the sentencing judge is as follows:
[W]hen I sentenced you to 15 years for the second degree murder I took into consideration your age and your background and your problem with alcohol and your substantial time in Vietnam. Whether that had anything to do with this, I don’t know, but I listened to the psychiatrist which would suggest that this might have been an automatic reaction; I’m not persuaded that that’s really all that meaningful in this situation .... When I look at these facts and think in terms of manslaughter, this would be the most aggravated type of manslaughter. But my legal training and ideas of justice demonstrate to me that on resentencing, some adjustment has to be made. I’m going to sentence you to 15 years and I’ll suspend 3 of it and you’ll be placed on probation then for a period of time when you’re released. I think that having some contact with a probation officer is probably going to be helpful in this case .... And Mr. Houston, I am pleased with your *1027progress while in confinement and I considered that in my sentence.
. State v. Chaney, 477 P.2d 441 (Alaska 1970).
. See also Padie v. State, 594 P.2d 50, 62 n.40 (Alaska 1979); Chappell v. State, 592 P.2d 1218, 1220 (Alaska 1979); Alpiak v. State, 581 P.2d 664, 665 n.2 (Alaska 1978).
. We would prefer that trial judges specifically address each of the Chaney criteria and expressly relate each criterion to the facts the court considers relevant in the specific case before it. Such remarks should clearly identify the Chaney criteria accorded preference, explain why preference was given, and should be presented in concluding remarks after the parties’ final arguments and the defendant’s allo-cution. To enforce these guidelines strictly would, however, result in remands for resen-tencing in a substantial number of cases with little likelihood that the ultimate results would change. In order to determine why the trial judge imposed a specific sentence, we will review the record of the entire hearing to consider the evidence presented, the issues perceived to be significant by the parties, and the trial court’s reaction to the arguments presented. If based upon that determination we can conclude that the trial court’s decision is supported by substantial evidence and is consistent with the Chaney criteria, we will affirm the sentence even though the judge’s concluding remarks, viewed in isolation, are incomplete. Here it is clear that the trial judge considered Houston dangerous and believed that despite Houston’s good record in prison and previous good record in the military, a long period of imprisonment was necessary to protect the public. We would, however, remind the trial bench of the comment voiced by the supreme court in Amidon v. State, 604 P.2d 575, 578 n.7 (Alaska 1979) (citations omitted), where the court said:
[W]e caution judges that the practice of ‘readopting’ previous sentence remarks at resen-tencing is inadvisable. Due care must be taken to demonstrate a thorough and thoughtful sentencing decision.
We also note that the new criminal code may require a more detailed statement of reasons *1028than that mandated by present case law. See AS 12.55.025(a). Interpretation of this section must, however, await cases decided under the new code.
. Appellant points to the following remarks:
THE COURT: But Mr. Houston, I’ve often speculated as to why you — I have some frightening conclusions. You know, you bought the gun that day and you loaded it and you were — indicated some fear of the people that were surrounding you. It’s a thing that’s always really terrorized me was that I hope that you didn’t decide on that occasion that you were going to kill a white man. That’s — I often wondered about that.
MR. HOUSTON: No, it didn’t come down like that.
THE COURT: Didn’t come down like that.
MR. HOUSTON: No.
THE COURT: Because it doesn’t make any sense. But that’s the only thought that ran through my mind. I don’t — I wasn’t too impressed with the idea that it was a robbery, I don’t know why it didn’t — but just from your remarks that you made in the other bar which would suggest to me that you felt some hostility and I assume that it would be from a so-called dominant race, but I don’t know. But for whatever it’s worth, that thought ran through my mind and I’m not persuaded that that’s the reason but that thought ran through my mind so you can reflect on that ....
. The comments in question include the following remarks:
MS. FABE: ... [Sumabat ] was an unusual kind of manslaughter. And I would submit that this one is as well. And the ....
THE COURT: Of course if you look — if you take a hard look at the facts, it looks more like murder one than anything else but I realize it’s ....
MS. FABE: Well, Your Honor, I think — it couldn’t be murder one just because there’s
THE COURT: Well, I mean, just_
MS. FABE: ... no premeditation. I mean it was — it happened so quickly that I think that there’s no ....
THE COURT: Course we don’t ....
MS. FABE: ... question ....
THE COURT: We don't know how quickly it lasts.
MS. FABE: ... he couldn’t have premeditated ....
THE COURT: I mean he bought the gun the same day. It only takes one second. You take a gun out and you point it at somebody and you pull the trigger. How much time is required for premeditation. You point a loaded weapon at him, you bought it the same day you know, he knows what a gun will do, it’ll blow you apart but ....
MS. FABE: Well, I that [sic] though the testimony at trial indicated that he was suffering from — all right, and again, the jury did not apparently — I mean it’s hard to say what weight this can be given since it’s not the theory that the jury apparently subscribed to but ....
THE COURT: I don’t — I agree with you, he was convicted of murder two, I’ve got a manslaughter, I realize that I have to reduce the sentence, you don’t have any argument from me because I gave him 15 years for what looked like a murder one. The reason I gave him 15 years because of the things that *1029you’re talking about. He was a good soldier, he had problems with alcohol, he was intoxicated, but I took all those ....
(Emphasis added).
. The dissent argues that the trial judge could and probably should have said more in explanation of his reasons at resentencing. We agree. We do not, however, agree that a remand for further trial court comments would serve any useful purpose in light of the existing record and we fear that such a remand might cruelly raise Houston’s hopes for a modification of sentence; hopes which, on this record, would most likely not be realized.
In those cases where inadequate sentencing findings have triggered remands one or more of four possibilities existed. The first possibility is that the trial court, given a more thorough, thoughtful, and sensitive evaluation of the existing record, might, however unlikely, reduce or modify the existing sentence favorably for the defendant. The second possibility is that the appellate court, uncertain as to the trial court’s reasons for the sentence imposed, might, if those reasons were expressed more fully, find them deficient and as a result reduce or modify the sentence favorably to the defendant. The third possibility is that the appellate court’s opinion found certain legal errors in the trial court’s sentencing which might on remand *1030result in a reduced or modified sentence. The fourth possibility is that the appellate court perceived some evidence in the record, apparently overlooked by the trial court, which given the trial court’s remarks might influence that court to deal more leniently with the defendant. Aside from an unenforceable recommendation that psychological or alcohol counseling be provided Houston, a recommendation included in the first judgment but inexplicably omitted from the second, we do not understand the dissent to suggest that any of the four possibilities mentioned exist here. A sentence of 12 years to serve would appear appropriate for someone convicted of any degree of homicide where the trial judge could reasonably conclude, as this one did, based upon substantial evidence in the trial record, that the defendant intentionally shot and killed a total stranger without an apparent explanation or reason. Houston’s hope for a reduced sentence was based upon the possibility that this court would interpret Avery v. State, 514 P.2d 637 (Alaska 1973) to require a trial court to ignore at sentencing any facts in the trial record relevant to a Chaney evaluation which were not either implicit in the trial jury’s verdict or the defendant’s plea. We have rejected that interpretation of Avery in Huckaby v. State, 632 P.2d 975, 977 n.2 (Alaska App.1981), in which we held that a trial court may consider any facts verified in the record which are material to the defendant’s rehabilitation, or the protection of the community, so long as the resulting sentence is within the range permitted for the crime to which the defendant was convicted and so long as the sentence is not clearly mistaken. McClain v. State, 519 P.2d 811 (Alaska 1974). Given that decision, it is highly unlikely that a remand for further sentencing comments would result in a modification favorable to Mr. Houston.
Finally, while we agree that courts should, as a matter of course, recommend alcohol, drug, or psychiatric treatment wherever such treatment might be helpful, we believe that the dissent’s treatment of the relation between Houston’s drinking and his crime includes two errors. The first, is the logical error of assuming post hoc ergo proper hoc; the fallacy of believing that because one event follows another the two are causally related. Houston’s crime clearly followed a period of substantial drinking. He has been diagnosed as an alcoholic. It does not necessarily follow that alcohol caused this crime or that successful alcohol or psychiatric treatment would eliminate the risk Houston poses to the community. We have concluded, in the light of this record, that the trial court could find that Houston’s killing is largely unexplained. Our careful review of the testimony concerning Houston’s alcohol use and psychiatric problems, if any, has not changed this view. The dissent’s second error is to assume that alcohol treatment, if provided, will necessarily succeed in eliminating whatever alcohol problems Houston has. We are less optimistic in this regard but agree that it cannot hurt and might help. The availability of alcohol or psychiatric treatment to Houston would not, however, in our view mandate or even warrant a substantially reduced sentence.