dissenting.
Although I agree with the court on all other issues, I cannot subscribe to the majority’s view that the admissible evidence against Billingsley was “not overwhelming to the point where we would be comfortable finding [that] admission of Billingsley’s detailed confession was harmless beyond a reasonable doubt.” My assessment of the admissible evidence differs significantly from that of the majority of the court and leaves me with no discomfort in finding harmless error.
Quite apart from the unrecorded, post-Miranda statements improperly admitted against Billingsley, the state established that Billingsley called the police and turned himself in for committing the Quick Stop robbery. Before any interrogation occurred, and without any prompting, Bill-ingsley confessed that he had committed the robbery. Suppression of the details provided by Billingsley during the ensuing custodial interrogation, in no respect precluded the state from establishing the basic and undisputed fact that Billingsley turned himself in and confessed to the offense for which he was convicted.
Billingsley’s confession was supported by evidence that verged on being overwhelming in its own right. When Billings-ley turned himself in less than two days after the offense, he was still dressed in the clothing that he wore when he robbed the Quick Stop store: dirty jeans, a dirty and worn button-up blue jean jacket, and scuffed, brown, square-toed “dingo” boots. This clothing matched precisely the description given by the robbery victim, James Oswald, immediately after the robbery. While no single item of Billingsley’s clothing was unique, the combination was certainly striking and singular — sufficiently so to preclude dismissal as mere coincidence.
Billingsley’s physical appearance, moreover, precisely matched the description given by Oswald, who had obtained an excellent and unobscured view of the perpetrator. Oswald was able to provide an unusually detailed description of the robber. He described a twenty-eight-year-old white male, with a ruddy complexion, about six feet tall, medium build, dark hair, a mustache, and a full untrimmed beard. Bill-ingsley fit virtually every detail of this description save one: he was twenty-nine years old instead of twenty-eight.
Shortly after Billingsley turned himself in and confessed, Oswald was shown a photographic lineup that included a snapshot of Billingsley as well as five other individuals who were similar in appearance. Despite the similarities, Oswald was able to pick out two photographs that resembled his assailant; one of them was Billingsley.
At trial, when he confronted Billingsley personally, Oswald identified him without any equivocation; Oswald testified that he had “no doubt at all” that Billingsley was the man who had held him at gunpoint. Nor was Billingsley able to suggest any plausible reason why Oswald should have had a doubt.
Finally, undisputed evidence at trial established that shortly after the Quick Stop robbery the police attempted to track the robber with a dog. The dog led them across the street and into a trailer park *1108before losing the robber’s scent. The place where the dog lost the scent was less than fifty feet short of the trailer where Bill-ingsley resided.
In opposition to this evidence, Billingsley attempted to establish an alibi: that he was in his trailer, passed out from drinking, at the time of the offense. To bolster this claim, Billingsley produced Gregory Dahl, a close friend who shared the trailer with Billingsley at the time of the robbery. Dahl claimed that he was awake at 3 a.m., the time the robbery occurred, playing video games in his trailer. He said that Bill-ingsley remained passed out next to him on the couch during that time. The state, however, thoroughly impeached Dahl’s testimony. Apart from establishing that Dahl had made prior inconsistent statements to the police, the state proved that he was intoxicated on the night of the robbery and that his estimates of the time at which various events occurred that night were substantially inaccurate.
The totality of the properly admissible evidence against Billingsley leaves little room for reasonable doubt. Of course, no matter how strong the evidence, a resourceful mind can always conjure exotic reasons for acquittal. But speculation about implausible and irrational hypotheses does not advance the harmless error analysis, for in considering whether error is harmless we are bound to presume that the jury acted reasonably. See, e.g., Hill v. Lockhart, 474 U.S. 52, 61, 106 S.Ct. 366, 371-72, 88 L.Ed.2d 203 (1985).
Here, Oswald’s unequivocal identification of Billingsley, taken in conjunction with the physical evidence, was a strong and sufficient showing of guilt, albeit not necessarily overwhelming. Defendants are routinely, though perhaps not invariably, convicted of robbery on such evidence. Add to this evidence the fact that Billingsley turned himself in and confessed his commission of the robbery, and it becomes difficult to conceive how a rational jury— one that applied its reason and common sense to the law — could have returned a verdict of acquittal.
Judge Buckalew, who had the opportunity to see and hear all of the witnesses at trial (and whose view of the evidence I believe deserves at least a modicum of deference) did not hesitate an instant before declaring that the properly admitted evidence established Billingsley’s guilt, and that the failure to record Billingsley’s custodial interrogation was harmless beyond a reasonable doubt. I would join in Judge Buckalew’s conclusion. In my view, the suppressible statements obtained by the police during Billingsley’s custodial interrogation amounted to little more than superfluous decoration on a cake that was already fully baked and well iced.
I therefore dissent.