concurring in part and dissenting in part. In the main I agree with the disposition the majority makes with respect to the errors assigned by the defendant-appellant. However, I strongly disagree with the affirmation of the death sentence. I dissent from the imposition of the death sentence because the evidence of guilt in this case, while sufficient to meet the various standards which an appellate court must use to measure legal error, is far from overwhelming.
The majority recognizes: “A review of the circumstantial evidence presented by the state leads us to conclude that reasonable minds could reach different conclusions as to whether each material element was proved beyond a reasonable doubt.” (Emphasis added.) The majority further says: “Given the state of the record, we hold that the court of appeals correctly determined that while the evidence was not conclusive of guilt, it was sufficient to withstand a motion for acquittal under Crim. R. 29.” (Emphasis added.)
Our duty, as imposed by R.C. 2929.05, does not stop with the review of legal error. Before approving a sentence of death, we are required by that statute to make two additional determinations. First, we must “weigh all of the facts and other evidence disclosed in the record in the case and consider the offense and the offender to determine whether the aggravating circumstances the offender was found guilty of committing outweigh the mitigating factors in the case.” Second, we must determine “whether the sentence of death is appropriate.”
With regard to our second obligation the statutory mandate includes this language: “The court of appeals or the supreme court shall affirm a sentence of death only if the particular court is persuaded from the record that the aggravating circumstances the offender was found guilty of committing outweigh the mitigating factors present in the case and that the sentence of death is the appropriate sentence in the case.” (Emphasis added.)
This mandate goes beyond a bare sufficiency of the evidence test and it is not discharged by a mere proportionality review. The legislature has made our responsibility in reviewing the sentence in a capital case different from that in any other criminal case. In essence, we are constituted as a super jury to review the record and to decide whether the death sentence is appropriate. We are not bound, as in other cases, by the findings of fact made by the trier of fact. We must be “persuaded. ”
Can anyone quibble with the idea that lack of certainty as to a defendant’s guilt (even if the evidence is sufficient as a matter of law) should be a consideration in deciding whether the death penalty is appropriate? In this case, there is a substantial possibility that the defendant may not be guilty. I emphasize the word “may,” because, as indicated, I believe the record does pass muster as to legal sufficiency.
*30I turn now to the evidence. The first consideration is the state’s contention that the defendant has the same blood type as that of the assailant. This was an important part of the case presented by the state at trial. The state called a medical technologist who had no sample of seminal fluid from the victim’s vagina but did have vaginal and oral swabs from which she found the presence of spematozoa. From her testing, she was able to determine that the swabs came from a person in blood group A. The defendant is in blood group A. What troubles me is that the pathological report indicates that the victim was also in blood group A.
It is recognized that in order to determine the blood type of an assailant in a rape case, where oral and vaginal swabs have been taken from the victim’s body, medical evidence must identify an antigen which was known not to have been produced by the victim herself See Gaensslen & Camp, Forensic Serology: Analysis of Bloodstains and Body Fluid Stains, 2 Forensic Sciences (Wecht Ed. 1987) 29-83. If the victim was a secretor,3 the recovery of a type A antigen from the swab obtained from the victim (who was herself a type A) offers no information concerning the blood type of the assailant, because the recovered antigens could have as easily originated from the victim as from the assailant. As is stated by Gaensslen & Camp, supra:
“In vaginal swabs or washings from sexual assault victims, there is always a mixture of the vaginal fluids of the victim and the seminal fluid of the perpetrator (assuming that semen is present). For this reason, the victim’s blood type and secretor status must be determined in order to know what substances found in the samples might be from the semen. In many cases, we cannot tell anything about the blood group of the perpetrator because the seminal blood group substances are masked by those of the victim’s secretions.” (Emphasis added.) Id.
Thus, where semen itself was not specifically found, and tests were conducted by use of oral and vaginal swabs, it was essential to identify the blood type and secretor status of the victim.4 In sum, it appears to me that the state failed to prove that the defen*31dant’s blood type was the same as that of the assailant.5
Another piece of the evidence which disturbs me is the human hair which was found on the back of the victim’s hand. The victim’s hands had been covered with plastic bags in the coroner’s office in order to prevent disturbing potential evidence, which would be further scrutinized at the crime laboratory. When the plastic bags were removed from the victim’s hands, a human hair was discovered. Further analysis of the hair demonstrated that it did not originate from either the victim or the defendant. At trial, the state argued that it is not uncommon for police or crime scene personnel to lose a hair, which is later recovered and thought to be of evidentiary value. While this may have been the case, the better approach would have been to have the hair analyzed against all crime scene personnel who could have deposited it. Such an elimination procedure is not overly burdensome given the penalty sought to be extracted by the state.
The balance of the evidence presented by the state was circumstantial. It included the following: (1) the defendant had a scratch on the left side of his face consistent with that of a fingernail scratch, but no tissue was found under the victim’s fingernails; (2) the defendant had painted a portion of the victim’s duplex and a copy of the contract for painting was found on the kitchen table the day after the murder was discovered; (3) the defendant was familiar with the layout of the victim’s house; (4) the victim was apprehensive about the defendant; (5) the defendant spoke with the victim for about ten minutes in the afternoon preceding the murder about painting her windowsills, and a portion of one of the sills was used to stab the victim in the neck; and (6) the defendant offered several inconsistent stories about his whereabouts on the night of the murder between the hours of 9:15 p.m. and 12:45 a.m.
A conviction may be affirmed upon circumstantial evidence and even upon a record with as many holes as this one. But in this case, I am not persuaded that the death sentence is appropriate, even though the crime would merit that penalty if there were no doubt as to the defendant’s guilt. Accordingly, I would affirm the defendant’s conviction, but remand the case to the trial court for imposition of a life sentence. See State v. Penix (1987), 32 Ohio St. 3d 369, 513 N.E. 2d 744.
Sweeney and Locher, JJ., concur in the foregoing opinion.In this case the state did not determine whether the victim was a secretor. However, the medical technologist testified that eighty percent of all people are secretors, so there is an eighty percent chance that she was.
It is surprising, in view of scientific medical advances, that the state simply proceeded with an ABO bloodgrouping analysis after finding the presence of sperm from the swabs. As noted by Gaensslen & Camp, supra: “There is a wide range of genetic markers in human blood and body fluids * * * [which] can, in many cases, greatly limit the number of people who might have deposited the stains.” Id. at 29-5 to 29-6. For instance, aside from ABO testing, seminal-vaginal mixtures can be examined for phosphoglucomutase (PGM) isoenzymes to provide a degree of individualization as to the perpetrator. Although this procedure may not necessarily help identify the rapist, due to a masking of the seminal type by vaginal secretions, we do not know whether the state even attempted to utilize this procedure in the instant case.
Further, to address the problem of sex offender identification, a test known as the DNA comparison test is available. See Giusti, Baird, Pasquale, Balazs & Glass-berg, Application of Deoxyribonucleic Acid (DNÁ) Polymorphisms to the analysis of DNA Recovered from Sperm (1986), 31 J. of Forensic Sciences 409. In this test, the masking problems inherent in ABO blood grouping and HLA testing are virtually nonexistent, since the analysis is conducted *31upon the sperm itself, a quantity of which was recovered in the instant case. See, also, Kanter, Baird, Shaler & Balazs, Analysis of Restriction Fragment Length Polymorphisms in Deoxyribonucleic Acid (DNA) Recovered from Dried Bloodstains (1986), 31 J. of Forensic Sciences 403-408. A further advantage of the DNA comparison test is that the procedure can, as of this date, still.be attempted by the state in the case sub judice given the stability of the material for a period of four or more years. See Butzel, Genetics in the Courts (1987) 611. It is unfortunate that the state failed to utilize these procedures, which could have made the issue of this defendant’s guilt or innocence far less murky.
Even if the proof established the assailant as being in blood group A, such is of only moderate weight as circumstantial evidence. According to the state’s medical technologist, that group would include approximately thirty-two to thirty-five percent of the male population in Cuyahoga County. (See, also, Gaensslen & Camp, supra, at 29-83 to 29-84.)