State v. Grim

854 S.W.2d 403 (1993) STATE of Missouri, Respondent, v. Robert GRIM, Appellant. No. 74892. Supreme Court of Missouri, En Banc. May 25, 1993. Rehearing Denied June 29, 1993. *405 Deborah Wafer, James S. McKay, St. Louis, for appellant. William L. Webster, Atty. Gen., Joan F. Gummels, Asst. Atty. Gen., Jefferson City, for respondent. THOMAS, Judge. Robert Grim was convicted of second degree murder, armed criminal action, and first degree burglary. He appealed his convictions, and the Court of Appeals, Eastern District, reversed, finding the evidence insufficient to support the jury's verdicts. We granted transfer and now decide the case as if on original appeal. Mr. Grim offers four reasons why his convictions should be reversed; we treat each argument in turn and affirm the judgment. I. SUFFICIENCY OF THE EVIDENCE Mr. Grim first argues that the evidence does not support his convictions. The standard for appellate review of the sufficiency of the evidence to support a criminal conviction was stated by this Court in State v. Dulany, 781 S.W.2d 52, 55 (Mo. banc 1989): On review, the Court accepts as true all of the evidence favorable to the state, including all favorable inferences drawn from the evidence and disregards all evidence and inferences to the contrary. [Citation omitted.] In reviewing a challenge to the sufficiency of the evidence, appellate review is limited to a determination of whether there is sufficient evidence from which a reasonable juror might have found the defendant guilty beyond a reasonable doubt. Under this standard, the Court must review the evidence adduced at trial and examine the inferences reasonably supported by that evidence to determine whether the jury's verdict is proper. The Dulany standard echoes the due process standard announced by the United States Supreme Court in Jackson v. Virginia, 443 U.S. 307, 99 S.Ct. 2781, 61 L.Ed.2d 560 (1979). Missouri might choose to require more evidence to support convictions on appellate review, but due process mandates that all convictions be supported at least to this extent. For the most part, we have not required anything more than the constitutionally required minimum and employ the Dulany standard. One exception, however, has been in cases based upon circumstantial evidence. The State now questions the necessity of this different treatment and asks us to re-examine and reject the circumstantial evidence rule. A. The Circumstantial Evidence Rule The circumstantial evidence rule originated as a higher standard to which circumstantial evidence cases were held. Because of a basic distrust of criminal convictions based upon circumstantial evidence and nothing more, we required the prosecution in such cases to meet a different burden. The distrust took form in the rule, which was used both as a standard of appellate review and as a jury instruction. In its appellate form the rule was stated: "Where the conviction rests on circumstantial evidence, the facts and circumstances to establish guilt must be consistent with each other, consistent with the guilt of the defendant, and inconsistent with any reasonable theory of his innocence." State v. Pritchett, 327 Mo. 1143, 39 S.W.2d 794, 796-97 (1931). Although this statement of the rule is quite similar to our modern rule, the older cases reveal the distrust of circumstantial evidence and the resulting higher standard. Where a chain of circumstances leads up to and establishes a state of fact inconsistent with any theory other than the guilt of the accused, such evidence is entitled to as much weight as any other kind of evidence; but the chain, as it were, must be unbroken, and the facts and circumstances disclosed and relied upon must be irreconcilable with the innocence of the accused in order to justify his conviction. Pritchett, 39 S.W.2d at 797 (quoting State v. Morney, 196 Mo. 43, 50, 93 S.W. 1117, 1119 (1906)) (emphasis added in Pritchett).[1] *406 In more recent years, courts have come to view circumstantial evidence as no different from direct evidence. See, e.g., Holland v. United States, 348 U.S. 121, 140, 75 S.Ct. 127, 137, 99 L.Ed. 150 (1954). Missouri courts now apply a more relaxed version of the rule. Specifically, we have retained the language of the traditional standard but continue in the next sentence with a limitation on the rule: Where the conviction rests on circumstantial evidence, the facts and circumstances to establish guilt must be consistent with each other, consistent with the guilt of the defendant, and inconsistent with any reasonable theory of his innocence. In such cases, the evidence need not be absolutely conclusive of guilt, nor must the evidence demonstrate the impossibility of innocence. E.g., State v. Livingston, 801 S.W.2d 344, 347 (Mo. banc 1990) (emphasis added). To withstand scrutiny under the Dulany standard, i.e., the due process minimum, a conviction must be supported by enough evidence that a reasonable juror, taking all evidence in the light most favorable to the State, would be convinced beyond a reasonable doubt. Under the circumstantial evidence rule, the inquiry would turn to whether there exists a reasonable theory of innocence that is consistent with all of the evidence. The theory need not be so likely that no reasonable person could discount it. At this point, the circumstantial evidence rule provides no more guidance and a reviewing court, or a juror, must simply decide whether the theory presented is reasonable. Confusion as to how likely a theory has to be before it is "reasonable" adds to the overall confusion regarding the rule. In a sense, the issue is whether the Dulany standard and the circumstantial evidence rule require a different quantum of evidence to support a conviction. If the two rules lead to identical results in all cases, there is no need to retain the circumstantial evidence rule since it could only lead to confusion to have two rules for one standard. If, on the other hand, the two rules sometimes yield different results in the same case, problems arise. Because the Dulany standard has a constitutional foundation, no conviction can be affirmed when the evidence fails under this test. If the circumstantial evidence rule is interpreted to require less evidence than the Dulany standard, then it will lead to the violation of defendants' constitutional rights and must be rejected. If the circumstantial evidence rule is used to require more evidence than the Dulany standard, we must consider what added quantum of evidence and convincing power is contemplated to overcome "a reasonable theory of innocence," how this differs from the Dulany standard for appellate review and the phrase "beyond a reasonable doubt," and whether the circumstantial evidence rule can be applied consistently and fairly by diverse judges and juries passing upon a wide spectrum of facts and circumstances. Even overcoming these hurdles, we must ask ourselves whether there is any need to hold circumstantial evidence cases to a higher standard. No reason remains to perpetuate this different rule. Any societal distrust of circumstantial evidence has long been abandoned. We no longer need to hold circumstantial evidence cases to a higher standard than direct evidence cases. If a jury is convinced beyond a reasonable doubt, so long as the evidence meets the minimal appellate standard required by due process, we need not disturb the result simply because the case depended wholly, mostly, or partially upon circumstantial proof. The State urges us to reject the rule because "the overwhelming majority" of jurisdictions have already done so. In fact, many jurisdictions have ceased to follow the rule, including the federal courts. Of course, we do not decide our cases based upon which rule wins favor in the most states. That other courts have reached the same conclusion reinforces our decision but *407 does not require it. Suffice it to say that we are not the first court to reach the issue, nor are we the first to reject the rule. Those interested in the reasoning employed by other courts may find profitable an examination of two recent decisions—one from Ohio, the other from Texas. State v. Jenks, 61 Oh.St.3d 259, 574 N.E.2d 492 (1991); Geesa v. State, 820 S.W.2d 154 (Tex.Crim.App.1991). In Jenks, the Ohio Supreme Court rejected the circumstantial evidence rule as an instruction to juries and as an appellate standard of sufficiency of the evidence. Jenks, 574 N.E.2d 492. Before doing so, the court reviewed the law of many states and the federal courts. Id. at 498-502. The court then overruled previous decisions that had examined the question and upheld the use of a circumstantial evidence instruction. Id. at 503. Almost as an aside, the court then disposed of the rule as a gauge of sufficiency of the evidence. Id. In reaching its conclusion, the Ohio court noted first that "[c]ircumstantial evidence and direct evidence inherently possess the same probative value.... Hence, we can discern no reason to continue the requirement that circumstantial evidence must be irreconcilable with any reasonable theory of an accused's innocence in order to support a finding of guilt." Id. at 502. Thus, in the court's view, the instruction to the jury was no longer warranted. "Proceeding to consider the proper standard of appellate review, where the evidence is either circumstantial or direct, we conclude that the relevant inquiry on appeal is whether any reasonable trier of fact could have found the defendant guilty beyond a reasonable doubt." Id. at 503. Of similar import is the Texas Court of Criminal Appeals decision in Geesa, 820 S.W.2d 154. In Texas, the circumstantial evidence instruction had already been rejected in Hankins v. State, 646 S.W.2d 191 (Tex.Crim.App.1983). Thus, in Geesa, the Texas court had to determine whether the "reasonable-hypothesis-of-innocence analytical construct" retained any validity as a standard of review. Noting that the theoretical basis for the rule disappeared with the demise of the jury instruction, the court also pointed out a number of problems that the analytical construct gave rise to. In Texas, the rule led appellate courts to place themselves "in the posture of a `thirteenth juror'." Id. at 159. Also, the rule had served as a source of confusion in appellate courts as to what evidence should be considered in applying the rule, what light the evidence should be viewed in, and what the result should be in a given case. Id. at 160. Consequently, the Texas court rejected the rule, noting that many other jurisdictions had done the same. Id. at 161, n. 9. The State argues, and we recognize, that Missouri's circumstantial evidence rule is causing many of the same problems that Texas courts experienced. We might have chosen to try to solve these problems by redefining the phrase "reasonable theory of innocence" in such a way as to make it clear that the circumstantial evidence rule was equivalent to the Dulany rule. Such a redefinition would have disposed of the remnants of our distrust of circumstantial evidence and might have aided in eliminating confusion. However, maintenance of a different test, even one designed to produce identical results, could never entirely end the confusion surrounding the rule. Thus, rather than suffer an unnecessarily restrictive standard based upon an outdated view of circumstantial evidence or invite further confusion and error by adopting yet another version of the circumstantial evidence rule, we reject the circumstantial evidence rule as a standard for reviewing the sufficiency of the evidence. The State also asks us to end the practice of giving to the jury that portion of the circumstantial evidence instruction that mirrors the appellate circumstantial evidence rule. Because we conclude that the circumstantial evidence rule is confusing to judges and lawyers, who make a career out of shoveling the smoke of abstract ideas, we cannot ignore the fact that the rule has to confuse, and often mislead, lay jurors. Everything we have said concerning the appellate circumstantial evidence rule applies with equal force to the same language found in the jury instruction, *408 MAI-CR3d 310.02.[2] We have honed our reasonable doubt instruction with legal analysis brought on by decades of defendants' attacks coming from every point of the compass. We believe the reasonable doubt instruction fully and accurately instructs the jury on the risk of non-persuasion. The circumstantial evidence instruction no longer serves the same purpose it did when we reaffirmed its use in State v. Lasley, 583 S.W.2d 511, 515 (Mo. banc 1979) (explaining that abrogation of the rule was proper only in jurisdictions that had an instruction defining reasonable doubt). The second paragraph of MAI-CR3d 310.02 shall no longer be given. The rule might be seen as requiring less evidence than, the same amount of evidence as, or more evidence than the Dulany standard. No conviction can be affirmed on less evidence than Dulany requires. A different rule stating the same standard is confusing and redundant. The previous justifications for requiring more evidence have been abandoned. Under any interpretation of the quantum of evidence required by the circumstantial evidence rule, the rule is no longer valid. It should be, and is, rejected. Having rejected the circumstantial evidence rule, we must now determine whether the state's case is adequate under the Dulany standard. A brief narration of the relevant evidence is in order. B. The Evidence at Trial The crime took place in the City of St. Louis. The evening of July 28, 1988, Cora Bradford's niece became worried when her aunt failed to answer repeated telephone calls. The niece called the police and asked them to check on Ms. Bradford, an elderly woman who lived alone. When Officer Scott arrived to investigate, he found the front door of the house locked. No one answered the bell or repeated knocks on the front door. Upon further investigation, Officer Scott found the back door standing open and the screen door shut but not latched. The eye and hook arrangement normally used to latch the screen door had been broken loose and was lying on the floor. Officer Scott stepped through the back door and into the kitchen. In the short, narrow hallway leading from the kitchen to the front bedroom, Officer Scott found Ms. Bradford's nude body. As a part of the ensuing homicide investigation, the police took photographs of the scene and searched for fingerprints and other evidence. Ms. Bradford's body was removed to the morgue for the purpose of conducting an autopsy. The police seized numerous items of evidence in an attempt to determine who had killed Ms. Bradford. The evidence was removed from the scene for further examination and testing. Many of the photographs were admitted into evidence for the purpose of showing the layout of the house, the scene of the crime, and the position of the body. The photos also revealed the amount and location of blood spattered and smeared on the walls and floors from which the jury could draw inferences as to the nature of the attack. The photos of the bedroom show bloodstains on the sheets and bedspread, a puddle of blood on the floor next to the bed, and a few bloody shoeprints. In addition to the shoeprints, the pictures show a bloody print made by a bare foot and many drops of blood on the floor in a trail away from the bed and toward the hall. Also visible in the pictures are a few pieces of newspaper scattered about the floor, some of which have spots of blood on them. The photos of the hallway show Ms. Bradford's body as it was found by the police—face down, head pointed towards *409 the bedroom and feet towards the kitchen, and laying at an angle so that it extended from one side of the hall to the other. Anyone traveling down the hall from the kitchen to the bedroom would be forced to step over the body. The pictures show streaks of blood on Ms. Bradford's legs and feet and a pool of blood under her face and chest. The pictures also reveal a number of drops of blood on the floor around Ms. Bradford's body. Other pictures show two small areas of blood on the walls above Ms. Bradford's feet on both sides of the hall, the blood in these areas is mostly in spots and drips, though there are some areas that look smeared. The pictures of the kitchen show a table with a number of items on it, including a wallet, which looks like it has blood stains on it. Pictures of the floor reveal more pieces of newspaper and a few bloody shoeprints, including one on the sill of the back door. The autopsy showed that Ms. Bradford died from a wound to the left side of her chest that severed her aorta and her trachea. The medical examiner concluded that the wounds were consistent with stab wounds from a sharp object, most probably a knife with a blade more than four inches long. All of the wounds were on the left side of the body, one appearing on the left arm. He characterized the wound to the arm as a "defensive wound;" i.e., one suffered in an attempt to ward off a blow. The medical examiner was of the opinion that Ms. Bradford died seven to twenty-four hours before her body was found. He explained that, while this range was his best approximation, the actual time of death could extend out of this seventeen hour range. Blood samples were taken from the body for the purpose of comparison with blood stains on evidence seized from the scene. The blood samples were delivered to the crime lab. Criminalist Margaret Owens tested the blood to "type" it with respect to five of the approximately one hundred enzymes present in human blood. Because these five enzymes were more resistant to the effects of temperature and drying, they could be detected and classified even after as long as a month outside of the body. Thus, the enzymes were most likely to be discernible in blood on the items taken from the house. Each of the five enzymes Ms. Owens tested had three possible types. For example, the enzyme abbreviated ESD could be type one, type two-one, or type two. Any particular person's blood would have only one of these types. Thus, the enzymes served as points of comparison between the blood sample taken directly from Ms. Bradford's body and any blood found on evidence seized from the house. Ms. Owens testified that fourteen percent of the black population in St. Louis has the same types for these five enzymes as Ms. Bradford did. In addition, Ms. Owens tested some of the pieces of evidence to determine whether they had traces of blood on them, whether the blood was human, and whether the enzymes matched Ms. Bradford's. Not all of the evidence seized had marks or stains to be tested. Each important piece of evidence will be discussed in turn. The police found a large kitchen knife in some weeds in the backyard of the house. The size of the knife was consistent with it being the murder weapon. The knife was examined for fingerprints, but no identifiable prints were present. Although there were no visible stains on the knife, Ms. Owens scraped the blade and handle with a piece of test paper and was able to obtain a very small sample. The sample proved to be blood but was too small to be tested to determine whether it was human blood or animal blood such as from cutting meat or poultry. The police seized some of the pieces of newspaper from the house. The three pieces of newsprint admitted into evidence each had at least a partial footprint. One piece of paper had a complete print made by a bare foot. The other two pieces of paper had shoeprints on them, which Criminalist Harold Messler testified came from two sets of shoes.[3] There was no evidence *410 as to how the prints compared to Mr. Grim's shoes. Ms. Owens tested the prints and determined that they were made in blood, that the blood was human blood, and that all five enzyme types matched the enzyme types in Ms. Bradford's blood. The most important piece of evidence was the wallet found on the kitchen table. Ms. Bradford's niece could not positively identify the wallet as her aunt's, but she identified a picture inside the wallet as being her aunt's husband, who had been dead for thirty-eight years at the time of the murder. The wallet could be held shut by a short strap fastened with a snap and, when opened, had a section for paper money; a place for coins, which also snapped shut; and a section for pictures that contained a plastic insert to hold the photos. The wallet had a few dark smears on the outside, and there were stains around each of the snaps that held the various parts of the wallet shut. Ms. Owens tested the stain on the coin flap of the wallet and found that it was blood, that it was human blood, and that four of the five enzymes matched those of Ms. Bradford. The test for the fifth enzyme was inconclusive; it did not reveal what type the enzyme was. According to Ms. Owens, seventeen percent of the black population in St. Louis shares these four enzyme types. Inside the wallet was a partial fingerprint left on the plastic insert directly opposite the picture of Ms. Bradford's late husband. Ms. Owens testified that the fingerprint appeared to be in blood that looked the same as the blood on the rest of the wallet. She further testified that she could definitely say, based on her visual examination, that there was blood on the insert. However, the fingerprint would have been destroyed if tests had been conducted to conclusively determine whether the print was in blood and whether any enzymes were present that matched Ms. Bradford's blood. Rather than destroy the fingerprint, Ms. Owens preserved the print using a "super glue" process. A St. Louis Police officer and a fingerprint expert from the FBI both testified that the fingerprint conclusively matched that of Mr. Grim's right thumb.[4] The FBI fingerprint specialist also testified that it is impossible to leave a fingerprint in dried blood. However, there was no evidence about the length of time it takes blood to dry, whether in small amounts, such as was on the wallet, or in larger amounts, such as were found in the bedroom and in the hallway, under the body. Officer Kroeck, the first evidence technician to arrive at the scene, did testify that the blood was dry when he got there, but that was at 11:34 p.m., about an hour after the body was discovered. Following the arguments based on this evidence, the jury deliberated for about six hours and returned a guilty verdict on the burglary count but reported that they had not reached verdicts on the other charges. After receiving a "hammer" instruction,[5] they went back to deliberate further on the other two counts. A little more than five hours later, they returned guilty verdicts on the murder and armed criminal action *411 charges.[6] We must now determine whether the evidence supports the verdicts. C. Sufficiency of the Evidence to Prove Second Degree Murder In considering whether the evidence is sufficient to support the jury's verdict, we must look to the elements of the crime and consider each in turn. Under the Dulany standard, we are required to take the evidence in the light most favorable to the State and to grant the State all reasonable inferences from the evidence. We disregard contrary inferences, unless they are such a natural and logical extension of the evidence that a reasonable juror would be unable to disregard them. Taking the evidence in this light, we consider whether a reasonable juror could find each of the elements beyond a reasonable doubt. In this case, we believe a reasonable juror could find beyond a reasonable doubt that Mr. Grim was guilty of second degree murder. The State chose to submit the charge of second degree murder in the disjunctive, permitting the jury to convict if it found that Mr. Grim aided and encouraged another person in committing the murder or if it found that Mr. Grim committed the murder himself.[7] For the sake of clarity, we will discuss the two disjunctive theories separately. First, the jury could have found that Mr. Grim was guilty as a principal. Essentially, the instruction required the jury to find 1) Mr. Grim stabbed Ms. Bradford to death and 2) it was his purpose to cause serious physical injury or death. The testimony of the medical examiner directly supported the conclusion that someone stabbed Ms. Bradford to death. The second element of the crime is supported because the evidence of the nature of the attack supports the conclusion that the attacker intended to cause Ms. Bradford serious physical injury; a person stabbing an elderly woman four times with a sharp object could scarcely intend anything less. The remaining issue is whether there is sufficient evidence to support a conclusion that it was Mr. Grim who stabbed Ms. Bradford. The evidence regarding the thumbprint on the wallet provides support for concluding that Mr. Grim was the one who stabbed Ms. Bradford. Because fingerprints are unique, it is possible for experts to compare a latent print found on an object with prints on file and establish the identity of the person who left the latent print. In this case, the experts were thorough in their explanation of how fingerprints worked and how the print found on the wallet matched the thumb of Mr. Grim. A fingerprint can also prove presence at a particular place because prints are left *412 through physical contact between people and things. If there is adequate evidence of where the object was when the person left the print, the print is evidence that the person was in the same place. The wallet was found inside Ms. Bradford's house following her murder and contained a picture of her long dead husband. From these facts the jury could conclude that the wallet was in the house before the murder, during the murder, and after the murder until the police took it to the lab. This conclusion is all the more reasonable because it is overwhelmingly improbable that anyone breaking into Ms. Bradford's home would bring along a wallet containing a picture of her late husband and then leave it behind for the police to find. This conclusion in turn supports the conclusion that Mr. Grim was inside the house when he left the print. A fingerprint, by itself, normally proves little more than identity and presence. However, in some cases the circumstances surrounding a print or the peculiar location or nature of the fingerprint provide additional evidence of how, when, or why the person touched the object. For example, in State v. Boyington, 831 S.W.2d 642 (Mo. App.1992), a bloody palmprint was found on the inside of a car trunk that contained a bullet-riddled corpse. This sort of location of a print tends to show that the person making the print was involved in the murder both because the print was in blood and because the print was inside the trunk with the body. In this case, the jury could conclude that the print was made in Ms. Bradford's blood. As a preliminary matter, the jury could conclude that the blood on the wallet was Ms. Bradford's. Ms. Bradford was stabbed to death and found face down in a pool of blood. There was blood on the floor in the bedroom and some blood on the walls near her body. Given the nature of the murder, the jury could conclude that all of the blood surrounding the body belonged to Ms. Bradford. Further, the laboratory tests indicated that the blood on the coin flap of the wallet matched Ms. Bradford's blood, at least as to the four enzymes that were testable on the wallet. Absent any other reasonable source of blood, the logical and sound inference from the evidence was that the blood on the coin flap of the wallet was Ms. Bradford's. The conclusion that the thumbprint was made in Ms. Bradford's blood follows from this conclusion because Ms. Owens' testimony that the fingerprint looked the same as the blood on the rest of the wallet supports the conclusion that it was the same; i.e., it was blood and it was Ms. Bradford's blood. In addition, Ms. Owens testified directly based upon her visual examination that there was blood on the insert. Though this testimony might not have the weight of the chemical tests performed on the other blood, on review we consider the evidence favorable to the State to be true. Thus, the print was made in Ms. Bradford's blood. Suppose that the police had caught Mr. Grim in the house minutes after the murder. The testimony of police officers that Mr. Grim had blood on his hands would be admissible for the purpose of showing that Mr. Grim committed the murder. The fact that there was blood on Mr. Grim's thumb would tend to prove that Mr. Grim committed the murder. Other inferences would be possible; having blood on his thumb would also be consistent with trying to assist Ms. Bradford or trying to determine whether she was dead. However, it is left to the jury to resolve which is true. In this case, there is no direct testimony that Mr. Grim had the victim's blood on his hands. However, there is evidence from which the jury could conclude that Mr. Grim was standing only a few feet from the dead body and that he left a print in the victim's blood. Of course, we are not deciding the admissibility of this evidence. Rather, we are faced with the question of whether this thumbprint in blood is enough to support the conclusion that Mr. Grim stabbed Ms. Bradford. That the print was in blood supports a number of inferences. One reasonable inference is that Mr. Grim got blood on his thumb as he held the knife and stabbed Ms. Bradford in the chest and arm. Another inference is that Mr. Grim came into the house after the murder, but before the blood was dry, touched some *413 wet blood, and then touched the inside of the wallet. However, we are bound to consider the inferences favorable to the State unless the contrary inference is such that it would necessarily give rise to a reasonable doubt in a reasonable juror's mind. See Dulany, 781 S.W.2d 52, 55. Nor should we consider this thumbprint in isolation from the other circumstances. To reach any of the pools of blood, a person coming through the back door would have to step over the body and walk into the bedroom, or would have to reach under Ms. Bradford's head or chest. Even the blood on the floor and walls is confined to areas immediately around the body and is in relatively small amounts. Further, the testimony that a fingerprint cannot be made in dried blood limits, to a certain extent, the time during which Mr. Grim might have wandered into the house and left his print. These circumstances reduce the persuasive force of the argument that Mr. Grim came in later and got blood on his thumb in some way other than by participating in the killing. Further, his print was left inside a wallet that was found closed. If he came along later, Mr. Grim chose to thumb through Ms. Bradford's wallet, close the wallet, lay it on the table, and leave without telling anyone about the crime. While this is a permissible inference, the question of which version of the facts actually happened is precisely the sort of issue that should be left to a jury to decide. The bloody thumbprint inside Ms. Bradford's wallet gives rise to a reasonable inference that Mr. Grim was involved in the murder. There are other potential inferences, but they are not so believable and inescapable as to convince us that they would give rise to a reasonable doubt in the mind of a reasonable juror. The surrounding circumstances further reduce the believability and the validity of the "innocent" inferences. If we voted as jurors rather than judges, we might well vote not guilty on this charge, but our role is fixed by the rule from Dulany. We decline to force ourselves into the posture of a thirteenth juror. We hold that the evidence in this case was sufficient to support the conclusion that Mr. Grim stabbed Ms. Bradford and, thus, was sufficient to support the conviction under the first theory of second degree murder. The second theory was based upon Mr. Grim's liability for aiding and encouraging or participating in the murder. Essentially, the jury had to find 1) someone else stabbed Ms. Bradford, 2) the person acted with the intent to cause serious physical injury, and 3) with the purpose of furthering the commission of the murder, Mr. Grim acted together with or aided and encouraged the other person. The first two elements are adequately supported by the evidence; Ms. Bradford was stabbed by someone, and whoever did it intended serious physical injury. The third element was supported by the evidence that the thumbprint was in Ms. Bradford's blood. One reasonable inference from this piece of evidence was that Mr. Grim actually stabbed Ms. Bradford. Another reasonable inference is that he helped the person that had the knife. The same "innocent" inferences described with respect to the first theory are relevant here but are, again, merely possible inferences and not necessary ones. A reasonable juror could have concluded beyond a reasonable doubt that Mr. Grim committed the crime of second degree murder, either through his own actions or by assisting another in the crime. The bottom line of the dissent's reasoning is that the inference to be drawn from the evidence that Mr. Grim was present before and at the time the murder was committed is of equal validity with the inference that sometime after the murder but before the blood dried, the defendant wandered into Ms. Bradford's backyard, saw the open door, entered and was confronted by the murder scene, in some manner acquired blood on his thumb, opened the billfold and made the thumbprint, then closed the billfold and left. We believe there are two fallacies in the dissent's rationale. First, these two inferences are not equally valid, the latter being so unique and unusual as to be implausible; and, second, we think that the dissent has ignored the requirement that "the Court accept[] as true all the evidence favorable to *414 the state, including all favorable inferences drawn from the evidence...." Dulany, 781 S.W.2d at 55. If an appellate court sets itself up to select between two or more acceptable inferences, it ceases to function as a court and functions rather as a juror, actually a "super juror" with veto powers. It is not the function of the court to decide the disputed facts; it is rather the court's function to assure that the jury, in finding the facts, does not do so based on sheer speculation. The bloody thumbprint of the defendant is adequate and sufficient evidence of the defendant's presence in Ms. Bradford's house at the time of her murder. Given that single fact and the physical evidence in this case, it is reasonable for a juror to conclude beyond a reasonable doubt that the defendant was guilty of second degree murder. While we have based our decision upon the law as stated in Dulany and the unique facts of this case, our decision is not entirely unprecedented; two cases decided under the circumstantial evidence rule are substantially similar. In State v. Maxie, 513 S.W.2d 338 (Mo.1974), a division of this Court faced a case with similar facts and upheld the sufficiency of the evidence to support second degree murder. In Maxie, the defendant's thumbprint was found on a cardboard box top in the victim's apartment amid various papers and debris scattered around the scene. While the defendant in Maxie did not leave his print in blood, he testified that he had never seen the outside of the apartment building, the inside of the apartment, the victim, the crime scene, or the box top. As explanation for the print on the box top, the defendant argued that he might have picked up the piece of cardboard and put it in the trash, from which the victim or her husband could have picked it up and brought it into their home. The jury was essentially left with two possible conclusions as to what happened, and the trial became a believability contest. In such a situation, this Court held that the evidence supported the jury's verdict. Even closer to the case at bar is State v. Gales, 507 S.W.2d 35 (Mo.App.1974). In Gales, an elderly woman who lived alone in the city of St. Louis was found murdered in her home. The defendant's fingerprint was found on a jewelry box in the house, and the evidence established that the defendant had some of the victim's property in his possession shortly before her body was discovered. The defendant's print was not in blood, so the print was less probative on the issue of involvement in the murder.[8] Although the police found the house locked up and had to break open a back door to enter the house and discover the body, the jury in Gales still had to find that the defendant had been present during the murder. In Gales, the defendant's possession of some of the victim's property was probative on the issue of whether the defendant had been inside the house but did not bear on the issue of when he was in the house. On evidence essentially equivalent to that in the present case, i.e., the fingerprint, the jury found, and the evidence supported the finding that, the death occurred in connection with the robbery. The evidence in this case was sufficient to support Mr. Grim's conviction for second degree murder. D. Sufficiency of the Evidence to Prove Armed Criminal Action In this case, the charge of armed criminal action was submitted by reference to the instruction for murder. As we have just explained, the evidence was adequate to support the conviction for second degree murder under either theory of liability. For the conviction for armed criminal action to be proper, the evidence had to support the additional finding that the murder was committed "by or with or through the use or assistance or aid of a dangerous instrument."[9] The evidence in this case *415 supported the conclusion that Ms. Bradford was stabbed with a sharp object, probably a knife. That evidence is sufficient to support the additional element contained in the crime of armed criminal action. E. Sufficiency of the Evidence to Prove First Degree Burglary As noted, a fingerprint usually proves little more than presence of a person in a particular place. Sometimes, however, presence in a particular place is a violation of the criminal code. There are a number of cases where a fingerprint supplemented with little else has been held sufficient to support a conviction for a crime focusing primarily on presence of the defendant in a forbidden place. See, e.g., State v. Schleicher, 442 S.W.2d 19 (Mo. 1969) (burglary); State v. Anderson, 671 S.W.2d 383 (Mo.App.1984) (burglary); State v. Sanders, 619 S.W.2d 344 (Mo.App.1981) (burglary); State v. Clemmons, 579 S.W.2d 682 (Mo.App.1979) (tampering); State v. Parker, 535 S.W.2d 126 (Mo.App.1976) (burglary). As to the charge of first degree burglary, the jury had to find beyond a reasonable doubt that 1) Mr. Grim knowingly entered unlawfully Ms. Bradford's house, 2) that he did so for the purpose of stealing, and 3) that while Mr. Grim was in the house, Ms. Bradford was there too and was still alive.[10] The first element, unlawful entry, is supported by the evidence provided by the thumbprint that Mr. Grim was in the house during the murder, and the evidence that the latch was broken from the screen door. The fact that Mr. Grim's thumbprint is on the wallet supports the conclusion that Mr. Grim had the intent to steal when he entered the house because it supports the inference that he looked for things of value while in the house. The third element of first degree burglary is also supported because, as explained with regards to the murder charge, the fact that the print is in Ms. Bradford's blood supports the conclusion that Mr. Grim participated in the murder. If Mr. Grim was in the house while Ms. Bradford was being stabbed, then Mr. Grim was in the house while Ms. Bradford was in the house. The conviction for first degree burglary is sufficiently supported that we will not disturb it on appellate review. Because Mr. Grim's convictions were adequately supported by the evidence, we deny his first point and turn to the second. II. ADMISSIBILITY OF THE KNIFE Mr. Grim's second point on appeal has to do with the admission, over his objection, of the knife found in Ms. Bradford's backyard. Mr. Grim claims there was no evidence to connect the knife to him or to the crime charged and, therefore, it was irrelevant and prejudicial to him. It is worth mentioning in passing that the knife was relevant to the crimes charged in that it formed a part of the body of the crime. Insofar as the knife was not tied to Mr. Grim in any way, it tends to weaken the *416 State's case against Mr. Grim rather than strengthen it. Thus, there would be little if any prejudice to Mr. Grim even if the knife were irrelevant. The point is denied. III. DEPRIVAL OF BATSON CLAIM Mr. Grim's third point of error relates to the way the trial court dealt with his Batson claim. Mr. Grim's attorney indicated that he believed that the prosecution had exercised its peremptory strikes in a discriminatory manner contrary to the teachings of Batson v. Kentucky, 476 U.S. 79, 106 S.Ct. 1712, 90 L.Ed.2d 69 (1986). The trial judge indicated that if the motion were made, he would quash the jury and start from scratch. Apparently the trial court and Mr. Grim's attorney were operating under the assumption that the proper relief for a violation of Batson is to quash the jury. To the contrary, the proper remedy for discriminatory use of peremptory strikes is to quash the strikes and permit those members of the venire stricken for discriminatory reasons to sit on the jury if they otherwise would. Because all concerned misunderstood the proper relief under Batson and because Mr. Grim's attorney thought the jury was better for Mr. Grim than the jury that might have been chosen if a new panel were selected, he withdrew his motion. Mr. Grim now argues that this action of the trial court was improper. This point has not been properly preserved, however. Mr. Grim was entitled to challenge the strikes of the prosecution but did not do so. Had Mr. Grim's attorney asked for relief and then objected to the trial court's means of dealing with the problem, the point would be adequately preserved for our review. IV. IMPROPER PRESENTATION OF STATE'S EXHIBIT 60 Finally, as his fourth point of error, Mr. Grim claims that the trial court committed an error when it sent the jury State's Exhibit 60, a chart of the blood enzyme types found on different items in the house and showing Ms. Bradford's blood enzyme types. The chart had been admitted into evidence, but was inadvertently left out of the exhibits sent to the jury. The omission was not discovered until well into the jury's deliberations when the jury asked two questions. The jury's questions were "Was the thumbprint of Robert Grim found inside the wallet in Cora Bradford's blood?" and "Was the blood on the outside of the wallet Cora Bradford's?" The court responded that the jury must be guided by the evidence as it recalled it. However, it was at this time that the prosecutor discovered Exhibit 60 had not been sent to the jury. Faced with this situation, the court proposed to remedy it by sending the exhibit to the jury. Mr. Grim's counsel objected to the prejudicial effect of sending the exhibit immediately after the question, so the court delayed sending the exhibit another twenty minutes. Mr. Grim's attorney still objected to this irregular procedure, but was overruled. On appeal, Mr. Grim asserts this error as a reason for reversing his convictions and granting him a new trial. We hold that the irregularity, though to be guarded against and avoided, was not sufficiently prejudicial to Mr. Grim that it warrants reversal. Given the other evidence before the jury, including the testimony regarding the blood enzymes, and given the arguments made by counsel, any additional weight the jury may have given the exhibit because of the irregularity was not likely to have made a difference in their verdicts. The jury deliberated another one and one-half hours before coming back with the burglary conviction and then an additional five hours before returning the guilty verdicts for murder and armed criminal action. While it is impossible to tell exactly what weight the exhibit was given and exactly what effect the irregular presentation may have had, we remain unconvinced that the prejudice to Mr. Grim was sufficient to warrant reversal of the convictions. The judgment is affirmed. COVINGTON, HOLSTEIN, BENTON, PRICE and LIMBAUGH, JJ., concur; ROBERTSON, C.J., dissents in separate opinion filed. *417 ROBERTSON, Chief Justice, dissenting. The Due Process Clause of the Fourteenth Amendment to the United States Constitution prohibits the conviction of any defendant "except upon evidence that is sufficient fairly to support a conclusion that every element of the crime has been established beyond a reasonable doubt." Jackson v. Virginia, 443 U.S. 307, 314-15, 99 S.Ct. 2781, 2786-87, 61 L.Ed.2d 560 (1979). [Emphasis added.] With respect to the majority, the State's case against Robert Grim falls far short of this standard. By affirming these convictions, the majority rescues the State from the consequences of having charged crimes it could not prove. I submit that such is not the role of this Court and must, therefore, respectfully dissent. In holding that the reasonable doubt standard of proof is grounded in the constitution, the United States Supreme Court stated: [U]se of the reasonable doubt standard is indispensable to command the respect and confidence of the community in applications of the criminal law. It is critical that the moral force of the criminal law not be diluted by a standard of proof that leaves people in doubt whether innocent men are being condemned. In re Winship, 397 U.S. 358, 364, 90 S.Ct. 1068, 1072, 25 L.Ed.2d 368 (1970). The standard of proof beyond a reasonable doubt ... `plays a vital role in the American scheme of criminal procedure,' because it operates to give `concrete substance to the presumption of innocence to ensure against unjust convictions, and to reduce the risk of factual error in a criminal proceeding. Jackson, 443 U.S. at 315, 99 S.Ct. at 2787 (quoting Winship, 397 U.S. at 363, 90 S.Ct. at 1072). It is the duty of the trial court, on motion by the defendant, to determine that the threshold of reasonable doubt has been met before allowing the case to go to the jury. Where the trial court has failed in that duty and where the jury has returned a verdict of guilty, it is the unpleasant duty of the reviewing court to reverse the conviction. Respectfully, I believe the majority fails in that duty today. The constitutional protections embodied in the presumption of innocence and the requirement of proof beyond a reasonable doubt are not reserved simply for the morally blameless, nor are they reserved for use only in "easy" cases. Rather, "[u]nder our system of criminal justice, even a thief is entitled to complain that he has been unconstitutionally convicted and imprisoned as a burglar." Id. 443 U.S. at 323-24, 99 S.Ct. at 2791. The majority holds that the evidence presented in this case was sufficient to support convictions for second degree murder, armed criminal action and first degree burglary. In doing so, I contend that the majority indulges in probabilities and speculation. Because I do not believe that the evidence presented provides a sufficient basis for any reasonable juror to find every element of these crimes beyond a reasonable doubt, I would reverse each of the convictions entered in the trial court. I respectfully dissent. I. Standard of Review Before addressing the substance of Grim's appeal, the majority turns its attention to the standard of review. The State has requested, and the majority obliges, that this Court abandon the so-called "circumstantial evidence rule." Finding that "[n]o reason remains to perpetuate this different rule," the majority holds that "[w]e no longer need to hold circumstantial evidence cases to a higher standard than direct evidence cases." Maj.Op. at 406. Because I believe that the circumstantial evidence rule is not a higher standard of review, but is instead a means of subjecting circumstantial evidence cases to the same standard as other cases, I respectfully dissent from the majority's rejection of that analysis today. In discarding the circumstantial evidence rule, the majority asserts that the rule is a "higher standard" courts have applied to circumstantial evidence because of a "basic distrust," or a "societal distrust," of circumstantial evidence that is now "outdated." *418 Maj.Op. at 405, 406, 407. I disagree, unless due process is also "outdated." In one of the earliest cases employing this rule, this Court stated exactly the opposite of the majority's claim. "Although [a crime] may be established by circumstantial evidence alone, its certainty must be equal to that attained by direct evidence and it must be proved beyond a reasonable doubt." State v. Wheaton, 221 S.W. 26, 28-29 (Mo.1920). [Emphasis added.] I continue to believe that the circumstantial evidence rule is based on an understanding that, while circumstantial evidence is not inherently less probative than direct evidence, it is fundamentally different than direct evidence. Courts must account for this difference in judging the submissibility of criminal cases. See also State v. Lasley, 583 S.W.2d 511, 515-516 (Mo. banc 1979). The circumstantial evidence rule is a viable tool serving that purpose. Direct evidence is testimony, given under oath, as to the existence or nonexistence of an element of the crime concerning which the witness claims personal knowledge. Confronted with direct evidence, the jury's only function is to weigh the witness' credibility. If the jury believes the witness, the element has been established; if the jury does not believe the witness, the element has not been established. Weighing credibility and resolving competing versions of the facts are critical tasks that our society entrusts only to juries. Courts, especially appellate courts, recognize that these tasks are singularly within the competence of the jury and rarely, if ever, invade this province of the jury to hold a witness' testimony "unbelievable" as a matter of law. To be submissible, therefore, due process demands only that direct evidence be such that any reasonable juror could believe it. With circumstantial evidence, however, the State asks more from the jury. There, the jury is first asked to believe the State's witness, just as with direct evidence. Then, if the jury is persuaded that the witness is truthful, the jury is asked to infer, from the evidentiary facts, the existence of some essential element of the crime charged. This second step, the process of deducing the existence of an element from an evidentiary fact, is unique to circumstantial evidence. It is this step that is subjected to the test of reasonable doubt by the circumstantial evidence rule. Few, if any, cases can be proved without some reliance on deduction. For instance, it is often said that a defendant's culpable mental state must be proved (absent an admission) by circumstantial evidence. Thus, courts have determined that certain acts give rise to an inference of certain mental states, as a matter of law. The process of deduction in cases based entirely on circumstantial evidence, however, is much more involved. There, the State is asking the jurors not only to infer the defendant's culpable mental state, but also to infer the defendant's criminal acts that underlie the inference of mens rea. This multi-leveled inference demands scrutiny to ensure that a reasonable jury could conclude the defendant's guilt, based on the circumstances shown, beyond a reasonable doubt. Where the State's proof is entirely circumstantial, the circumstantial evidence rules requires only that submissibility be determined, not just in terms of whether any reasonable juror could believe the State's witnesses, but also whether any reasonable juror could draw the inferences the State requires in order to convict. Not only must a reasonable juror be able to infer the essential elements from the circumstances shown, but the strength of this inference must be such that a reasonable juror could be convinced of those elements beyond a reasonable doubt. Finally, and most important, where guilt depends on the piling of inference upon inference—as it does in this case—the sum of the inferences must be strong enough to convince a reasonable juror of the defendant's guilt beyond a reasonable doubt. In short, I read the circumstantial evidence rule simply as a means of determining when a reasonable doubt exists as a matter of law in cases based entirely on circumstantial evidence. Despite the assertions of the majority, I believe that this has *419 always been the interpretation, and the role, of the circumstantial evidence rule in this state. Because the rule continues to have value in ensuring that the due process standard of Jackson is met, and because the majority has not shown that the rule fails in that task—either in the present case or in decades of actual practice[1]—I see no reason for abandoning it.[2] While the majority will carry the day on this point, it should be understood by my colleagues on the trial and appellate benches that it is the rule that is abandoned today and not the duty that underlies it. II. Sufficiency of the Evidence In determining on appeal whether a case was constitutionally submissible, it is (or should be) irrelevant that the jury returned a verdict of guilty. Because we cannot know how or why a jury reaches its verdict, our review is limited to whether the State has introduced sufficient evidence for any reasonable juror to have been convinced of guilt beyond a reasonable doubt. Jackson, 443 U.S. at 319, n. 13, 99 S.Ct. at 2789, n. 13. To ensure that reviewing courts do not engage in futile attempts to weigh evidence or judge credibility, "the factfinder's role as weigher of the evidence is preserved through a legal conclusion that upon judicial review all of the evidence is to be considered in the light most favorable to the prosecution." Id. at 319, 99 S.Ct. at 2789. In light of these standards, the only evidence in this case linking Robert Grim to any of the charged crimes is a partial thumbprint, matching Grim's right thumb, lifted from inside a wallet found in the victim's house. That's it. From this single piece of evidence, the majority holds that a juror could reasonably infer the following: First, that Grim unlawfully entered the home of Ms. Bradford with the purpose of committing a crime therein and was inside the home sometime prior to her death (first degree burglary); Second, that, with the purpose of killing, or seriously wounding, Ms. Bradford, Robert Grim stabbed her or aided someone else in stabbing her to death (second degree murder); and Third, that he did so with a dangerous or deadly weapon (armed criminal action). The majority holds not only that a reasonable juror could infer the above based solely on the thumbprint, but also that a reasonable juror could be convinced of these facts beyond a reasonable doubt. The majority's reasoning is as follows: 1. Grim's thumbprint was in the victim's blood. Maj.Op. at 412. 2. It follows, then, that Grim stabbed the victim to death, Maj.Op. at 412, or aided whoever did the stabbing. Maj. Op. at 413. —Therefore, he is guilty of murder in the second degree and armed criminal action. 3. If follows, then, that since Grim killed, or helped kill, the victim, he must have been in the house prior to her death. Maj.Op. at 415. —Therefore, he is guilty of burglary in the first degree. With all due respect in this horrendously difficult case, I believe the majority, in its detailed and lengthy analysis, has been distracted by the question of whether it is possible to conclude Grim is guilty in this case. We already know that it is possible—the jury's verdict tells us that. The inquiry at issue—one we are bound by the constitution and our oaths to make—is *420 whether a juror could reasonably conclude from the evidence that Grim was guilty and whether such a juror could conclude this beyond a reasonable doubt. This is a question of law to be determined, in the first instance, by the trial court and then on review by an appellate court. In this case, I believe that the majority's first conclusion is permissible, barely. As discussed below, the second and third conclusions are the products of mere speculation and are, therefore, constitutionally insufficient to support a conviction. With respect to my colleagues, I would reverse all three convictions. A. The "bloody" thumbprint There was evidence from which the jury could conclude that the substance on the outside of the wallet was Ms. Bradford's blood. Expert testimony established that substance as blood, as human blood and as containing chemical properties found only in approximately seventeen percent of the Black population in St. Louis. Expert testimony also established that Ms. Bradford's blood contained these same chemical properties. The fact that the human whose blood was on the outside of the wallet and Ms. Bradford were both members of a relatively small demographic group, coupled with the nature and extent of Ms. Bradford's wounds, provided a sufficient basis from which a reasonable juror could have inferred that the blood on the outside of the wallet came from Ms. Bradford. As to the print-bearing substance on the inside of the wallet, however, there is much less to go on. There was no objective evidence that the substance was blood, or that it was human blood, or that it was of the same "type" as Ms. Bradford's blood. The only evidence offered by the State as to the nature of this substance was criminologist Margaret Owens' testimony that the substance "looked like" the blood on the outside of the wallet. She did not testify, nor could she have, that it "looked like" human blood or that it "looked like" blood of any specific "type." Rather, the jury—and now this Court—is left to determine as to the nature of this substance by deduction. Given that the jury could reasonably infer that the blood on the outside of the wallet belonged to Ms. Bradford, should we allow the jury to use that inference as a basis from which to infer that the substance inside the wallet was also her blood? "An inference is a logical and reasonable conclusion of a fact not presented by direct evidence but which by the process of logic and reason, a trier of fact may conclude exists from the established facts." State v. Hyde, 682 S.W.2d 103, 106 (Mo.App. 1984), cert. denied, 471 U.S. 1056, 105 S.Ct. 2120, 85 L.Ed.2d 484 (1985). As this Court has said: An inference is a conclusion drawn by reason from facts established by proof; a deduction or conclusion from facts or propositions known to be true.... A supposition is a conjecture based on the possibility that a thing could have happened. It is an idea or a notion founded on the probability that a thing may have occurred, but without proof that it did occur. [Emphasis added and quotation marks omitted.] Draper v. Louisville & N.R. Co., 348 Mo. 886, 156 S.W.2d 626, 630 (1941). This Court has long been skeptical of convictions that depend upon inferences drawn from other inferences, i.e., "inference stacking." See State v. Rector, 328 Mo. 669, 40 S.W.2d 639, 643 (1931); State v. Knight, 296 S.W. 367, 369 (Mo.1927). However, when a court reverses such a conviction, it is important to realize that the ground of decision is not the so-called rule against "inference stacking" but rather that the evidence is not sufficient to support a finding of guilt beyond a reasonable doubt. The [inference stacking] rule is essentially based on the fact that an inference cannot be based on insufficient evidence. It has been said to be not a rule of general application but a rule of reason governing only when the proved facts and their reasonable implications furnish no basis for agreement or disagreement *421 by persons of average intelligence as to whether the factum probandum has been established, or where there is some doubt as to the first inference. It has been said to be at most a convenient way of guarding against what is regarded as attenuated reasoning on the part of the jury or evidence thought to be too remote or uncertain or lacking in probative force. However, there is in fact no absolute rule of law that forbids the resting of one inference on facts whose determination is the result of other inferences. On the contrary, inferences may be based on facts whose determination is the result of other inferences, so long as the first inference is based on such evidence as to be regarded as a proved fact and the conclusion reached is not too remote or conjectural. [Emphasis added.] 31A C.J.S. Evidence § 116 (1964). See also Wills v. Berberich's Delivery Co., 345 Mo. 616, 134 S.W.2d 125, 129 (1939) (same). Applying this reasoning, it appears that the jury could reasonably conclude that the substance inside the wallet was Ms. Bradford's blood. The inference that it was Ms. Bradford's blood on the outside of the wallet is sufficiently free from doubt that the secondary inference, i.e., that it was her blood inside the wallet, is not too remote or conjectural. B. Grim murdered the victim The majority explains this "inference" as follows: That the print was in blood supports a number of inferences. One reasonable inference is that Mr. Grim got blood on his thumb as he held the knife and stabbed Ms. Bradford in the chest and arm. Another inference is that Mr. Grim came into the house after the murder, but before the blood was dry, touched some wet blood, and then touched the inside of the wallet. However, we are bound to consider the inferences favorable to the State unless the contrary inference is such that it would necessarily give rise to a reasonable doubt in a reasonable juror's mind. See [State v.] Dulany, 781 S.W.2d 52, 55 [(Mo. banc 1989) ]. Maj.Op. at 412-413. I believe that the majority has, in this brief statement, misstated the law[3] and confused the word "inference" with the word "possibility." Respectfully, I do not believe "possibility" has any place in a legal construct founded on reasonable doubt. An inference, as previously discussed, is a "conclusion drawn by reason from facts established by proof." Draper, 156 S.W.2d at 630. The two "inferences" acknowledged by the majority above are not drawn from "facts established by proof." Instead, they are drawn from an underlying inference—"that the print was in blood"—which was, as just discussed, the product not of proven facts but of inferences drawn from inferences drawn from proven facts. At this point, the inference "pile," barely strong enough to support the conclusion just discussed, collapses under the sheer weight of its own speculation. Thus it seems apparent to me that what the majority refers to above as "inferences" are not really inferences at all. They are simply two contradictory hypotheses—two possibilities—both of which are consistent with the evidence and either of which adequately explains the "bloody" thumbprint. They are what this Court condemned in Draper as "supposition;" that is, "conjecture based on the possibility that a thing could have happened.... [or] founded on the probability that a thing may have occurred but without proof that it did occur." Draper, 156 S.W.2d at 630. *422 [Emphasis added.] I thought that the question of whether convictions may be supported by supposition was no longer open for decision. See State v. Carter, 36 S.W.2d 917, 919 (Mo.1931) (court cannot "resort to conjecture, suspicion and surmise" to affirm conviction); State v. Hardy, 326 Mo. 969, 34 S.W.2d 102, 103 (1930) ("It is not the province of a jury to choose between suspicions, and a verdict based upon suspicion will not be permitted to stand."); State v. Huff, 317 Mo. 299, 296 S.W. 121, 122 (1927) (jury must not be allowed simply to "choose between suspicions"). The majority, in an attempt to turn its reasonable theory of guilt into a reasonable inference of guilt, relies upon the totality of the circumstances in which the print was found. When the police arrived, the wallet in which Grim's thumbprint was discovered was lying in plain view on a table just inside a wide-open back door. The door had stood unguarded for not less than seven, and perhaps as many as twenty-four, hours. The outside of the wallet was stained with what we can reasonably infer to be the victim's blood. Supra at 408. Inside of the wallet was a thumbprint formed in the smallest trace of what we can reasonably infer was also the victim's blood. Supra at 408. There were two sets of shoeprints and one footprint on the floor, none of which was linked in any way to Grim. The majority explains that, to reach any of the "pools" of blood, someone coming along after the murder would have had to "step over the body and walk into the bedroom, or would have to reach under Ms. Bradford's head or chest." Maj.Op. at 413. Why anyone would be concerned with "reach[ing] any of the pools of blood" is not explained by the majority. The only amount of blood that we "know" was on Grim's hand was the amount left behind as a thumbprint; an amount so small that its source need hardly have amounted to a "pool." That Grim may have picked up this trace of blood from some other source, such as the outside of the wallet, is studiously ignored by the majority. Next, the majority makes passing reference to the fact that, since "a fingerprint cannot be made in dried blood," this "limits, to a certain extent, the time during which Mr. Grim might have wandered into the house and left his print." Maj.Op. at 413. The State failed to introduce any evidence of how long blood takes to dry to the point it cannot form a print. The only testimony that the blood was ever dry comes from an investigator who did not arrive on the scene until more than an hour after the crime was discovered. For all we know or may reasonably infer, the blood was still wet when the first officer got there; a time which, we are told, was seven to twenty-four hours after Ms. Bradford was killed. I can hardly see how this limits, to any extent, the time during which Grim might have left the thumbprint but not been guilty of the crimes charged. Instead, it indicates a great deal of time. The majority also remarks on the fact that the wallet was found closed. Maj.Op. at 413. Evidently, the majority finds it incredible, assuming that Grim had entered the house after the murder to steal the contents of a wallet lying in plain view from outside, that he would "thumb through Ms. Bradford's wallet, close the wallet, lay it on the table, and leave without telling anyone about the crime." Id. Indeed, the majority asserts that this explanation is "implausible." Maj.Op. at 413. Apparently, the majority believes it implausible that a burglar would close a wallet behind him but plausible that a killer would. What is implausible is the majority's explanation. In its resort to the totality of the surrounding circumstances, the majority inexplicably ignores the fact that, given the two sets of shoeprints, we know at least one person other than Grim was in Ms. Bradford's house after her death, and maybe two. The possibility that Grim came onto the scene after the murder but before the blood was dry is strengthened immeasurably by the State's evidence affirmatively showing that others were there. Unfortunately, the majority ignores that part of the State's case which it cannot reconcile with guilt. *423 Ultimately, the majority reveals the true basis of its decision. After acknowledging that there are two competing explanations for the "bloody" thumbprint, and after ignoring the fact that the evidence provides no basis for electing between the two, the majority dismisses Grim's innocence as "so unique and unusual as to be implausible." Maj.Op. at 413. This statement alone proves the fallacy of the majority's position; innocence is presumed, it is guilt which must be proven. The majority has stood the presumption of innocence on its head. So long as the State can articulate a reasonable hypothesis of guilt, one which is not inconsistent with the proven circumstances, the majority seems to indicate that the case is submissible unless, and until, the defendant's innocence appears beyond a reasonable doubt. I am of the opinion that since innocence is ordinarily presumed, the State labors under a burden of producing evidence of guilt beyond a reasonable doubt. Thus, in purely circumstantial evidence cases, where the State's evidence does not preclude a plausible, innocent explanation for the circumstances, I would hold that a reasonable doubt exists as a matter of law and, thus, the case is not submissible. In the present case, the State's evidence gives rise to two plausible explanations for the circumstances shown; one guilty and the other innocent. There is simply no basis in the evidence for choosing one over the other. Not even the Dulany rule, oft-cited by the majority as urging deference to the State's position, contemplates that a reviewing court is bound to provide the evidence the State lacks by resorting to speculation. The majority, faced with an almost complete absence of evidence, seeks support in prior cases. With apparent relief, the majority remarks that its result is "not entirely unprecedented." Maj.Op. at 414. The majority first cites State v. Maxie, 513 S.W.2d 338 (Mo.1974), cert. denied, 420 U.S. 930, 95 S.Ct. 1132, 43 L.Ed.2d 402 (1975), in which this Court affirmed a second degree murder conviction arising out of a gruesome double-homicide. See also State v. Lindsey, 507 S.W.2d 1, 2 (Mo. banc 1974) (first degree murder conviction affirmed for Maxie's accomplice). In Maxie, the Court held that the State's case, based largely on fingerprint evidence, was sufficient when coupled with the defendant's testimony. The Court explained its holding, stating that, by offering testimony, the "defendant has ruled out any reasonable possibility that his fingerprint was ... [made] under circumstances other than at the time of the homicide." Id. at 343. [Emphasis added.] Unlike the defendant in Maxie, however, Grim has not eliminated the "innocent" explanation that the print was left after Ms. Bradford was killed. The majority also cites State v. Gales, 507 S.W.2d 35 (Mo.App.1974), in which the court of appeals affirmed a single conviction for felony murder. Maj.Op. at 414. Compared to the present case, the evidence in Gales was overwhelming. There, the defendant was seen with a television and fan, which had been taken from the victim's house, and was arrested with several article's of the victim's jewelry on his person. Additionally, the defendant's fingerprints were found on the victim's jewelry box. Id. at 36-37. As noted by the majority, there was no possibility, in Gales, that the defendant came along after the killing and left his print. Maj.Op. at 414. The victim's house was found locked on the morning following the burglary and murder. Gales, 507 S.W.2d at 36. Thus, the court held that the circumstances showed, beyond a reasonable doubt, that the fingerprints could only have been left at the time the crime was committed. Id. at 37. Prior to today, the rule in cases relying primarily on fingerprint evidence was that fingerprints found in the place where a crime is committed, under such circumstances that they could only have been impressed at the time the crime was committed —corresponding to those of the accused—may be sufficient proof of identity to sustain a conviction. State v. Thomas, 452 S.W.2d 160, 163 (Mo. 1970). [Emphasis added.] The holding in Maxie is consistent with this rule; there, the defendant eliminated any "innocent" *424 explanation for his prints. Similarly, Gales is consistent with Thomas in that the short period of time before the crime was discovered, and the fact that the perpetrator left the victim's house locked, provided some basis for concluding that the defendant's print must have been left during the crime. It is unclear why the majority fails to apply Thomas to this case, especially after citing and relying on Maxie and Gales—both of which expressly rely on Thomas. The State's proof fails to link Grim's print with evidence of other circumstances tending to exclude beyond a reasonable doubt Grim's theory of innocence, that he came along after the Ms. Bradford was dead. Therefore, I would hold that Thomas controls and, on that basis, would reverse. C. Grim was in the house while the victim was alive To prove first degree burglary in this case, the State must present evidence sufficient to find, beyond a reasonable doubt, that Grim was inside Ms. Bradford's house while she was there and still alive. Maj. Op. at 414. The majority affirms the submissibility of the State's case on the following reasoning: [T]he fact that the print is in Ms. Bradford's blood supports the conclusion that Mr. Grim participated in the murder. If Mr. Grim was in the house while Ms. Bradford was being stabbed to death, then Mr. Grim was in the house while Ms. Bradford was in the house. Maj.Op. at 415. Thus, once the majority's argument for submitting the murder charge is accepted, the burglary charge is submissible a fortiori. In my view, the burglary charge is not submissible because the State's evidence shows no more than that Grim was in the victim's house after her death.[4] As noted, the only piece of evidence linking Grim to any of the crimes charged is the "bloody" thumbprint. And, as noted, a fingerprint cannot be left in dried blood. Therefore, given the evidence, and drawing all reasonable inferences in favor of the State, one can conclude that Grim was in the victim's house at some point in time after she was killed and before the blood dried, a period which may have been as long as twenty-four hours. The evidence does not show, however, when Grim arrived or when he left. Nor does it provide any basis for inferring that, because he was there after she was killed he must have been there while she was being killed. We just don't know. Worse, we have no basis from which to reasonably deduce the answer. Again, the majority confuses an "inference," that which is logically drawn from the evidence, with a "possibility" that is not inconsistent with the evidence. Again, given that we may conclude only that Grim was in the house after Ms. Bradford's murder, two contradictory explanations arise: first, that Grim was there when she was killed, and second, that he came along after she was killed. This is not the first time this Court has faced such a situation and, prior to today, we have recognized that when the evidence is no more consistent with guilt than it is with innocence, reasonable doubt exists as a matter of law and the case may not go to the jury. State v. Roberts, 709 S.W.2d 857, 862 (Mo. banc), cert. denied, 479 U.S. 946, 107 S.Ct. 427, 93 L.Ed.2d 378 (1986); State v. Prier, 634 S.W.2d 197, 200 (Mo. banc 1982); State v. Black, 611 S.W.2d 236, 240 (Mo.App.1980). In fact, the court of appeals has expressly *425 stated that this rule is a necessary incident to the constitutional standard in Jackson. State v. Gardner, 737 S.W.2d 519, 521 (Mo. App.1987) ("federal constitutional requirement of establishing guilt beyond a reasonable doubt dictates the rule in the Black case take precedence over the usual [Dulany] standard of review in appeals alleging insufficient evidence"). I agree. Therefore, I would hold that the State's evidence in this case provides no basis for finding that Grim was in Ms. Bradford's house while she was alive, beyond a reasonable doubt. It is possible, given the evidence, that he was. It is just as possible, on the basis of the evidence, that he was not. Such a case cannot, constitutionally, go to the jury if the due process guaranties of the constitution are to be given their historic meaning. III. No one likes an "unsolved" crime; no one likes to see a defendant acquitted when he may have "done it." Common sense may point to Grim; evidence beyond a reasonable doubt does not. This Court's duty is to ensure that criminal convictions are based on constitutionally sufficient evidence—evidence from which guilt beyond a reasonable doubt may be found. When this duty rests too heavily, there is comfort in the knowledge that we are not the first to bear it. We have carefully analyzed all the [evidence] and, while it may be said that it is calculated to arouse a suspicion of guilt of this defendant, yet suspicions or even strong probabilities of guilt do not authorize a conviction. Giving every circumstance its full force, ... it falls far short of ... [that] upon which the citizen should be deprived of his liberty. The law should be universal in its application; it should be applied to the humble and the exalted alike. This defendant may be guilty, but the facts as disclosed by the record in this cause fail to show it. If there was any substantial evidence upon which to base this verdict, it would not be disturbed; but, in view of the insufficiency of the testimony to authorize this conviction, we must decline to sanction it. State v. Scott, 177 Mo. 665, 76 S.W. 950, 952 (1903). The majority frets that this dissent weighs the evidence and would have this Court sit as a "super jury." I regret that the majority so misunderstands what I say. Our job is not to act as a "super jury," but neither is it our job to abandon to the jury our responsibilities as guardians of the law. Every appellate court addressing an attack on the submissibility of a criminal case bears the responsibility of deciding whether the State has presented sufficient evidence from which the jury can find all of the elements of the crime on the basis of the facts presented, not speculation or emotion. This appellate responsibility is not one of weighing the evidence anew as a jury, but of making sure that the jury's deliberations comport with due process. Every case does not go to a jury; sometimes the State fails to make its case. Respectfully, submissibility review is an appellate court's only opportunity to safeguard this requirement of the Constitution. To abandon this role to the jury is not only to make the jury the final arbiter of the facts, which it is, but also of the law, which it is not. Our society long ago rested its faith in a system that declares it far better to let a guilty person go free than to imprison an innocent one. It is this commitment to the rule of law that separates us from the barbarians. I respectfully dissent. NOTES [1] As pointed out by the State in its brief, other examples abound. See, e.g., State v. Francis, 199 Mo. 671, 98 S.W. 11 (1906). [2] The full instruction is as follows: Circumstantial evidence is the proof of facts or circumstances that give rise to a reasonable inference of other facts that tend to show the guilt or innocence of the defendant. Circumstantial evidence should be considered by you together with all the other evidence in the case in arriving at your verdict. You should not find the defendant guilty unless the facts and circumstances proved are consistent with each other and the guilt of the defendant, and inconsistent with any reasonable theory of his innocence. It is the second paragraph about which the State complains. [3] Some of the shoeprints were of a honeycomb pattern from a Reebok shoe. Both of the pieces of paper had at least part of a honeycomb print on it. The piece of paper found in the kitchen also had a shoeprint with a waffle pattern. [4] Both experts testified that no two fingerprints are alike and that when fingerprints have more than eight points of similarity an expert may properly conclude that the prints are from the same person. At the request of the prosecutor, the St. Louis police officer counted as many points of similarity as he could and found 16. The FBI specialist found 25 points of similarity. [5] The instruction, based on the then current version MAI-CR3d 312.10, read as follows: You should make every reasonable effort to reach a verdict, as it is desirable that there be a verdict in every case. Each of you should respect the opinions of your fellow jurors as you would have them respect yours, and in a spirit of tolerance and understanding endeavor to bring the deliberations of the whole jury to an agreement upon a verdict. Do not be afraid to change your opinion if the discussion persuades you that you should. But a juror should not agree to a verdict that violates the instructions of the Court, nor find as a fact that which under the evidence and his conscience he does not believe beyond a reasonable doubt to be true. [6] Actually, this was the second jury to consider these charges against Mr. Grim. The first jury deliberated for about eight hours spread over two days before concluding that they could not reach a verdict. After a mistrial was declared because of the hung jury, the trial now under review was conducted with a newly chosen jury. [7] Actually, the State submitted and argued first degree murder as well, but the jury found Mr. Grim not guilty of that crime. The State elected not to proceed on a felony-murder theory and submitted a second degree murder instruction that, in relevant part, read as follows: A person is responsible for his own conduct and he is also responsible for the conduct of another person in committing an offense if he acts with him with the common purpose of committing that offense, or if, for the purpose of committing that offense, he aids or encourages the other person in committing it. . . . . . . As to Count I, if you find and believe from the evidence beyond a reasonable doubt: First, that on or about July 28, 1988, in the City of St. Louis, State of Missouri, the defendant or another person caused the death of Cora Bradford by stabbing her, and Second, that it was the defendant's or another person's purpose to cause serious physical injury to or to cause the death of Cora Bradford, then you are instructed that the offense of murder in the second degree has occurred, and if you further find and believe from the evidence beyond a reasonable doubt Third, that with the purpose of promoting or furthering the commission of that murder in the second degree, the defendant acted together with or aided or encouraged another person in committing that offense, then you will find the defendant guilty under Count I of murder in the second degree. [8] The victim was not stabbed to death; thus, there was not blood in which to make a print. [9] The relevant part of the actual instruction was as follows: As to Count II, if you find and believe from the evidence beyond a reasonable doubt: First, that defendant committed the offense of murder in the second degree, as submitted in Instruction No. 6, and Second, that defendant or another person committed that offense by or with or through the use or assistance or aid of a dangerous instrument, then you are instructed that the offense of armed criminal action in connection with the offense of murder in the second degree has occurred, and if you further find and believe from the evidence beyond a reasonable doubt: Third, that with the purpose of promoting or furthering the commission of that armed criminal action in connection with the offense of murder in the second degree, the defendant acted together with or aided or encouraged another person in committing that offense, then you will find the defendant guilty under Count II of armed criminal action in connection with the offense of murder in the second degree. [10] The relevant part of the actual instruction was as follows: As to Count III, if you find and believe from the evidence beyond a reasonable doubt: First, that on or about July 28, 1988, in the City of St. Louis, State of Missouri, the defendant knowingly entered unlawfully in an inhabitable structure located at 3133 Rolla Place and possessed by Cora Bradford, and Second, that defendant did so for the purpose of committing the crime of stealing therein, and Third, that while the defendant was in the inhabitable structure Cora Bradford was present in the structure and Cora Bradford was not a participant in the crime, then you will find the defendant guilty under Count II of burglary in the first degree. [1] The majority makes cryptic reference to the "inconsistencies," "confusion" and "unfairness" that have been "caused" by the use of the circumstantial evidence rule by appellate courts. Maj.Op. at 406, 407, 407. A review of the cases in which this Court and the court of appeals have applied the rule, however, reveals neither inconsistency nor confusion and, in fact, neither the State nor the majority has cited to any specific examples of either. [2] I agree with the majority that, in light of our jury instruction defining the reasonable doubt standard, there is no need to continue giving the second paragraph of MAI-CR3d 310.02 concerning circumstantial evidence. Maj.Op. at 408. But to conclude, as the majority has, that the circumstantial evidence rule is of no use in judging the submissibility of a case on appellate review is to throw the proverbial baby out with the bath. [3] The majority states that it is "bound" by those inferences favorable to the State. This is not, and never has been, the standard of review. In judging submissibility, the Court is only "bound" by those inferences that are reasonable and that can be drawn from the evidence. See, e.g., State v. Davis, 814 S.W.2d 593, 594 (Mo. banc 1991), cert. denied, ___ U.S. ___, 112 S.Ct. 911, 116 L.Ed.2d 812 (1992); State v. Sumowski, 794 S.W.2d 643, 645 (Mo. banc 1990). See also Jackson, 443 U.S. at 319 n. 12, 99 S.Ct. at 2789 n. 12; United States v. Henson, 939 F.2d 584 (8th Cir.1991). Respectfully, I suggest that the majority's conclusions, while in some sense reasonable, they are in no sense drawn from the evidence. [4] It should be noted that the evidence presented by the State was sufficient to prove that Robert Grim had committed burglary in the second degree, punishable by up to seven years in prison. The State's circumstantial evidence would support, beyond a reasonable doubt, a reasonable juror's belief that Grim "knowingly enter[ed] unlawfully or knowingly remain[ed]" unlawfully in Ms. Bradford's house "for the purpose of committing the crime therein." Section 569.170, RSMo 1986. The State did not charge this crime, however, electing instead to charge Grim with burglary in the first degree. Additionally, the State charged Grim with murder in the first degree and armed criminal action. Only after two trials, nineteen hours of jury deliberation and a "hammer" instruction, was the State successful in obtaining guilty verdicts on the armed criminal action and burglary charges and a guilty verdict on the lesser-included offense of second degree murder.