procedural. The Crown's proposed DNA evidence and associated statistical evidence were excluded. The court held the evidence was not sufficiently probative on the key contested issue (whether Coates, rather...
Key principles
Expert DNA evidence is admissible only if it is relevant to a fact in issue; where DNA evidence does not enable a jury to distinguish between competing persons who might have...
Where DNA testing shows multiple possible contributors and does not exclude another accused, the DNA evidence alone cannot support an inference that a particular accused left an...
Statistical conclusions expressed as likelihood ratios (for example '63 times more likely') are not relevant to an issue unless they address the actual proposition in contest; a...
Even where DNA evidence has some probative value, it may be excluded under the trial judge's discretion because of the risk of unfair prejudice from juror misestimation or misuse...
Issues before the court
Whether DNA evidence from a baseball cap found near the crime scene is relevant to proving that Coates, and not Gardner or an unknown person, left...
Plain English Summary
The Crown sought to admit DNA from a cap found near the getaway route, but the DNA could not definitively exclude another accused and suggested a possible third contributor. A likelihood-ratio comparing Coates plus an unknown against two unknowns did not resolve whether Coates, rather than Gardner, left the cap. Smith J held the DNA and statistical evidence either lacked necessary probative value on the contested issue or, if minimally relevant, posed unacceptable risk of juror misinterpretation and prejudice, and so excluded the evidence.
AI-generated legal information, not legal advice. Zoe can make mistakes. Check the cited source, and for advice about your situation consult a qualified Australian lawyer.
Deep Dive
1,742 words · generated 01/07/2026
What happened
This is a pre-trial admissibility ruling by Smith J in the Supreme Court of Victoria. The Crown sought to tender DNA evidence taken from biological material on a green baseball cap found in a laneway that the Crown alleges was used by armed robbers to move between getaway cars. The accused are Brian Lindsay Gardner and Michael Albert Coates.
Whether the Crown's statistical likelihood ratio evidence (that the DNA profile was 63 times more likely if from Coates + unknown than from two...
Whether expert statistical comparisons involving Gardner should be admitted when the expert expressed concerns about scientific validity.
Cited legislation
No linked legislation citations have been extracted yet.
Forensic testing by Ms Brown (a forensic scientist for the Crown) identified that the biological material on the cap had at least two contributors and perhaps three. The testing disclosed that Coates "could not be excluded as a contributor" whereas Gardner's status was equivocal: he "could have contributed" but the vWA locus data was inconclusive and did not reach reportable levels ([8], [15], [18]). The Crown sought to adduce two types of evidence from Ms Brown: (a) conventional evidence as to exclusions/non-exclusions (Coates not excluded, Gardner inconclusive), and (b) statistical evaluation, notably a likelihood ratio that the obtained typing was 63 times more likely if it came from Coates and one unknown person than if it had come from two unknown Victorian Caucasian persons ([8], [3(b)], [57]). Ms Brown expressed reservations about presenting certain hypothetical comparisons involving Gardner; she did not endorse some statistical comparisons as scientifically valid and the Crown therefore did not rely on a 10-times figure she had computed from three loci ([15], [51]).
Smith J was asked to rule whether the DNA evidence (and the statistical conclusions) were admissible. The judge analysed relevance and probative value, the problem of multiple contributors and timing of deposition, the limits of what the statistical comparisons actually addressed, and the risk of unfair prejudice due to juror misestimation of probabilistic evidence. The conclusion was that the DNA and statistical evidence either lacked probative value on the critical contested issue (whether Coates rather than Gardner left the cap) or, if of marginal relevance, should be excluded because of the risk of misuse and over-weighting by a jury. The Crown's evidence was excluded ([45], [57], [63], [69]).
Why the court decided this way
Central to Smith J's decision is the legal test of relevance: evidence must be capable, if accepted, of rationally affecting the probability of a fact in issue ([27]). The contested fact in this case was not simply that biological material existed on the cap, but whether Coates, as opposed to Gardner or an unknown third person, left the cap in the laneway at the relevant time linking him to the robbery ([36], [39]).
Three interconnected problems defeated the Crown's case on admissibility:
Multiple potential contributors and inability to exclude Gardner. The DNA testing indicated at least two contributors and possibly three; if there were only two, the contributors could be Coates + unknown or Gardner + unknown. Gardner could not be conclusively excluded. Thus the DNA did not permit the jury to prefer Coates over Gardner as source of the sample without further evidence ([18], [42], [45]).
Timing of deposition. DNA cannot by itself provide timing: the Crown needed some evidence that the material was deposited shortly before the cap was left in the laneway. No such evidence was foreshadowed. Without temporal linkage, the presence of someone's DNA on an item found in a public or semi-public place cannot safely be taken as connecting that person to the offending act ([48]).
Misalignment between statistics and the contested question. The 63-times likelihood ratio compared two hypotheses, 'Coates + unknown' versus 'two unknowns', which, even if accurate, did not address the real dispute between Coates and Gardner. Thus that statistic did not bear on whether Coates, as opposed to Gardner, left the cap ([57]). Further, Ms Brown would not confidently give statistical evaluations involving Gardner on scientific grounds ([15], [51]).
Even if some marginal probative value existed, Smith J concluded that the complexity of probabilistic DNA evidence carried a real risk of juror misestimation or misuse. Juries could not be expected to apply ordinary commonsense to assess the impact of expressions like '63 times more likely' and the judge had not been directed to any simple analogy or method to assist jurors. In this context the judge exercised his discretion to exclude the evidence on prejudice grounds ([60], [63], [66], [69]).
Before and after state of the law
Before this ruling, general principles governing admissibility of expert and forensic evidence were well established: evidence must be relevant to a fact in issue, and the trial judge has a discretion to exclude even relevant evidence where its prejudicial effect outweighs probative value. Smith J applies these established doctrines to a modern forensic setting involving probabilistic DNA evidence ([27], [60], [69]). The judgment cites R v Stephenson ([27], f.n.1) for the relevance principle.
After the ruling, the practical import is that forensic DNA evidence must be carefully aligned to the precise factual contest in the case. Prosecution experts should not tender statistical comparisons that do not address the proposition the fact-finder must resolve; nor should experts provide statistical evaluations they lack scientific confidence in. Where DNA testing indicates multiple contributors or leaves open alternative sources, courts will scrutinise whether the DNA can, in combination with other evidence, rationally support the contested inference. Moreover, judges will be alert to the risk that juries may misinterpret complex likelihood ratios and will exclude such evidence where the prejudicial risk is significant ([45], [57], [63], [66], [69]).
The judgment emphasises judicial gatekeeping in the forensic evidence era: admissibility demands both logical relevance to contested facts and consideration of the jury's capacity to weigh complex probabilistic proof.
Key passages with plain-English translation
Passage: "Evidence may be said to be relevant where, if accepted, it could rationally affect, directly or indirectly, the assessment of the probability of the existence of a fact in issue." ([27])
Translation: Evidence is only helpful in the case if, when believed, it makes the fact the trial is trying to decide more or less likely.
Passage: "The evidence cannot have probative value in relation to the critical issue of Coates involvement unless there is evidence which, if accepted, could exclude Gardner." ([45])
Translation: The DNA can't prove Coates did it unless there is some reliable evidence that shows Gardner did not.
Passage: "I can see no basis for saying that [the likelihood ratio comparing Coates + unknown to two unknowns] is relevant to the actual issues in the case." ([57])
Translation: Saying the DNA is 63 times more likely if Coates was involved doesn't answer the real question, was it Coates or Gardner who left the cap.
Passage: "In my view there is considerable potential for the evidence to be misused or given too much weight." ([63])
Translation: There's a real danger the jury will put too much faith in the DNA statistics and get the wrong idea of how strong the evidence is.
Passage: "Commonsense and experience will not assist the jury in assessing and comparing the DNA evidence ... '63 times more likely' ... I have not been referred to any simple formula or analogies that might be used to understand the weight of this evidence." ([66])
Translation: Ordinary intuition won't help jurors understand complex probability numbers, and there's no easy way to teach them how to use those numbers correctly.
What fact patterns trigger this precedent
This ruling is directly engaged where:
DNA or other forensic evidence yields profiles compatible with more than one possible suspect, and the prosecution cannot exclude an alternative suspect by the forensic evidence alone (see [18], [42], [45]).
The forensic evidence includes likelihood ratios or other probabilistic statements that compare hypotheses that do not align precisely with the contested proposition before the jury (see [57]).
There is no independent evidence linking the time of deposition of the forensic material to the time of the alleged offending act (see [48]).
The expert expresses reservations about the scientific validity of certain hypothetical comparisons offered by the prosecution (see [15], [51]).
In such fact patterns, a judge should examine relevance narrowly and consider exclusion if juror misestimation is a realistic risk.
How later courts have treated it
The judgment itself does not record subsequent appellate consideration or later treatment by other courts. The ruling is an example of judicial caution in admitting complex probabilistic forensic evidence. Practitioners should identify any appellate or higher-court pronouncements for authoritative extension or constraint of Smith J's approach. Absent later authorities cited in this ruling, its immediate application is to pre-trial admissibility assessments in analogous circumstances where DNA evidence leaves alternative contributors unexcluded and statistical comparisons do not address the contested question.
Still-open questions
Will appellate courts endorse this approach consistently? This is unresolved in the text. The ruling is at first instance and applies established relevance and prejudice principles; whether appellate courts will refine standards for admissibility of complex likelihood-ratio evidence is not addressed here.
How should courts instruct juries on likelihood ratios when such evidence is both relevant and necessary? Smith J concluded there was no simple formula to help jurors and excluded the evidence for that reason ([66]). The judgment leaves open the question of whether appropriate judicial directions or expert assistance (for example a statistical expert called to explain likelihood-ratio meaning carefully) could render such evidence admissible in a future case.
What level of scientific confidence or validation is required before an expert may present particular hypothetical comparisons? The judgment relies upon the expert's own reluctance as persuasive ([15], [51]) but does not articulate precise thresholds for admissibility of particular statistical computations. That remains an area for further judicial or evidentiary guidance.
To what extent can corroborating circumstantial evidence (for example observations of the accused's behaviour, CCTV, timing, or witnesses) cure the probative defects identified here? Smith J emphasised that absent evidence excluding Gardner or establishing timing, the DNA could not prove Coates left the cap ([45], [48]). But the ruling leaves open the precise quantum and nature of additional evidence required to render similar DNA statistics admissible.
How should courts handle likelihood ratios of differing scope (e.g. full-locus tests versus partial-locus comparisons)? The judgment mentions a 10-times figure based on three loci that the Crown did not rely upon because of scientific doubts ([51]). The ruling does not set a rigid rule for partial profiles; it suggests scientific validity and relevance to the contested proposition are decisive.
A ruling is sought as to the admissibility of DNA evidence sought to be led by the Crown.
[3]
The Crown wishes to lead evidence from Ms Brown, a forensic scientist, as to the results of DNA testing of biological material found on a green baseball cap which in turn was found in the laneway used by the alleged armed robbers to go from one getaway car to another.
[4]
The Crown wishes to lead the following evidence:
[5]
(a) (i) the DNA testing was inconclusive as to whether Gardner could be excluded as a source of the biological material but indicated that he could have contributed to it;
[6]
(ii) Coates could not be excluded as a contributor to the material;
[7]
(b) (i) that the DNA typing obtained would be 63 times more likely if it had come from Coates and one unknown individuals than if it had come from 2 unknown individuals chosen at random from the Victorian Caucasian population; (ii) that that finding considered in isolation provided moderate support for the proposition that the biological material originated from Coates and one unknown randomly selected individual from the Victorian Caucasian population.
[8]
I note that this particular finding involved effectively consideration of the results of four of the nine examinable areas.
[9]
Both the Crown and Coates asked Ms Brown to do a number of statistical analyses of different hypothetical combinations which included Gardner. I was informed by counsel for the Crown that Ms Brown had concerns about the scientific validity of the hypothetical comparisons and wished to do no more, in any evidence in chief, than provide the above statistical comparison between the hypothesis of Coates and one unknown individual as contributors on the one hand and 2 unknown individuals on the other as the sources of the biological materials. Counsel informed the court that Ms Brown was reluctant to advance statistical evaluations of hypothetical combinations including Gardner because the information available left the issue of exclusion inconclusive -- that is unlike Coates, there was apparently insufficient evidence to be able to say that he could not be excluded. The difficulty apparently arises from the area known as vWA where there were indications consistent with Gardner being a contributor but insufficient to cross the threshold required to permit them to be used. Apparently a possible explanation for this was an absence of sufficient material. In that area there were indications of a possible 16 which was consistent with Gardner but also of a 14 which was not. Neither, however, was reportable and Ms Brown treated that area examination as inconclusive.
[10]
Before turning to the submissions, I note that it is common ground that the DNA evidence establishes that there must have been two contributors to the biological material and that three contributors cannot be excluded. If, however, there were only two contributors, they could not comprise the two accused because there were findings made at two areas which matched neither accused. Thus if only two people supplied the biological material, it could be only Coates and one unknown person or Gardner and one unknown person.
[11]
Counsel for the Crown submitted that the baseball cap was an item found at a secondary crime scene, close to the red utility that was set on fire and on the path taken by the alleged armed robbers between the two alleged getaway cars. The Crown submitted that the evidence would be relevant if a connection can be established between the baseball cap and one of the robbers. Counsel submitted that the fact that the DNA material found linked the baseball cap to the accused Coates was relevant probative and admissible. Counsel submitted that the inference would be open that the person fleeing the scene left the baseball cap behind, that the person who had the cap was connected to the crime and that person was Coates.
[12]
Thus the Crown relies on the DNA evidence to connect Coates to the armed robbery. But it can only do that if the DNA evidence alone, or in combination with other evidence, is capable of giving rise to the inference that Coates left the cap behind in the laneway during the robber's flight.
[13]
Evidence may be said to be relevant where, if accepted, it could rationally affect, directly or indirectly, the assessment of the probability of the existence of a fact in issue. Determining relevance, however, is not simply a matter of logic. The common law requires that evidence be of sufficient relevance[1] before it is admissible. Here the issue arises in a particular context. The evidence of the baseball cap, and its location, and the DNA evidence, is said by the Crown to be relevant because, in combination, it can give rise to the inference that Coates left the baseball cap in the laneway while escaping after the robbery.
[14]
Counsel for Coates submitted that the evidence has no probative value and is therefore irrelevant and inadmissible. In the alternative, it is put that if there is any probative value it is very slight and the prejudicial effect of the evidence will outweigh any probative value.
[15]
Plainly the points in issue in the case need to be identified and the probative value, if any, to those issues that the DNA evidence may have needs to be determined.
[16]
There is no issue as to the involvement of Gardner in the armed robbery or his presence in the laneway at the relevant time. There is, however, an issue as to the involvement of Coates in the armed robbery and his presence in the laneway at the relevant time. The baseball cap has the potential to link Coates to the armed robbery if there is evidence from which it would be open to find, in conjunction with other evidence if necessary, that he left the baseball cap in the laneway at the relevant time.
[17]
If the situation was that the DNA found on the baseball cap could be connected only to Coates and no one else, an inference would be open that he had left his baseball cap in the laneway some time not long before the police inspection and that he was, therefore, one of the alleged robbers. But that is not this case.
[18]
The DNA evidence set out in paragraph 3 (a) above, if accepted, establishes that biological material was found on the internal head band of the baseball cap and that Coates, as a source, could not be excluded. But, as counsel for Coates submitted, the DNA evidence, if accepted, also establishes that Gardner could have contributed to it and his exclusion was equivocal -- in other words it is unclear whether he cannot be excluded as a source of the DNA. It is put that, therefore, the jury cannot assume, as counsel submitted the Crown had in fact done, that the baseball cap was not connected to Gardner at the relevant time.
[19]
As to this aspect of the DNA evidence, it does not itself provide a basis upon which a conclusion, preliminary or otherwise, could be drawn connecting the baseball cap to Coates to the exclusion of Gardner. There is no other evidence foreshadowed by the Crown that would enable a jury to make a choice excluding Gardner as the wearer or carrier of the baseball cap. The evidence cannot have probative value in relation to the critical issue of Coates involvement unless there is evidence which, if accepted, could exclude Gardner.
[20]
The Crown faces other problems. First, the DNA testing points to a third person having contact with the baseball cap. Secondly, the DNA evidence cannot itself provide evidence as to the time when the biological material was deposited on the baseball cap. There being more than one potential source of DNA, there needs to be some evidence on which it would be open to find that the DNA material that can be connected to Coates was placed by him on the baseball cap shortly prior to it being left in the laneway before the necessary inferences could be drawn. There does not, however, appear to be any evidence on which it would be open to reach such a conclusion . In addition, other apparently undisputed evidence will show the existence of a friendship and association between the accused, including the use by Gardner of Coates' home and material from it. Thus, undisputed evidence exists which will support the possibility that Gardner had borrowed the baseball cap from Coates - assuming that it belonged to Coates and not the third person or Gardner.
[21]
One of the hypotheticals tested by Ms Brown compared the likelihood that the material was deposited by Coates and one unknown person rather than Gardner and one unknown person. Comparing information obtained from three areas only, Ms Brown has calculated that the former is ten times more likely. Ms Brown, however, is concerned about the statistical validity of comparisons of hypotheses involving Gardner and the Crown does not rely upon it.
[22]
I have considered whether the proposed evidence might be relevant, when considered with the other evidence, on the basis that it could give rise to the inference that Coates was an active participant in the robbery. In my view, again, the foreshadowed evidence could not do so.
[23]
I turn to the evidence (above 3 (b)) as to the likelihood of the DNA having been contributed by Coates and an unknown person as compared to two unknown persons. I can see no basis for saying that this evidence is relevant to the actual issues in the case. If it was to be assumed that that evidence proved beyond reasonable doubt that Coates supplied the material as opposed to unknown persons, it does not address, directly or indirectly the real issue in the case which is whether Coates, as opposed to Gardner, left the baseball cap behind in the laneway.
[24]
Counsel for Coates also submitted that if there be some probative potential, and therefore relevance, the evidence should be excluded because of its prejudicial effect which counsel submitted would outweigh any probative value. Counsel submitted that it would be impossible to explain in any clear way to the jury how to deal with the DNA evidence itself and the conclusions to be drawn from it. Counsel for the Crown, however, submitted that while evidence of DNA and the related statistics is difficult it would be possible to present it to a jury in an understandable way so that there would be no risk of misuse of the evidence or misestimation of its weight by the jury to the unfair prejudice of the accused Coates.
[25]
In my view there is considerable potential for the evidence to be misused or given too much weight. It is difficult, if not impossible, without proper grounding in the statistical analysis of DNA, to form a view as to the significance of the probative difference between the statements that the findings had the result that Coates could not be excluded as a contributor to the biological material whereas Gardner could have been a contributor but it was unclear whether he could not be excluded as a contributor. [2]
[26]
In the typical circumstantial case, the jury can call on its collective commonsense and experience to assess the combined impact of the circumstances it considers. Commonsense and experience will not assist the jury in assessing and comparing the DNA evidence - both the general opinions and the specific conclusions such as
[27]
"63 times more likely", or "10 times more likely". Counsel have not been able to suggest to me how the jury might be assisted in the task of determining how much weight to attach to such conclusions in a way relevant to resolving the question whether the Crown has proved its case beyond reasonable doubt. I have not been referred to any simple formula or analogies that might be used to understand the weight of this evidence.
[28]
For the foregoing reasons, it appears to me that, if either aspect of the evidence has relevance, it should be excluded in the exercise of the discretion because of the risk of prejudice flowing from mis-estimation of the weight of this evidence.
[2] For recent discussion see Andrew Ligertwood "Avoiding Bayes in DNA Cases" (2003) 77 ALJ. The statistical information is not presented in the form of relative frequency but rather in the form of a likelihood ratio.
Parties
Applicant/Plaintiff:
R
Respondent/Defendant:
Gardner & Coates [2003] VSC 153
AI Analysis
Outcomeprocedural
Disposition:
The Crown's proposed DNA evidence and associated statistical evidence were excluded. The court held the evidence was not sufficiently probative on the key contested issue (whether Coates, rather than Gardner, left the baseball cap) and, in any event, any marginal probative value was outweighed by the risk of prejudice and juror misestimation (see [45], [57], [63], [69]).