R v Juric [2003] VSC 382

Supreme Court of Victoria|2003-08-13|Before: NETTLE J

Criminal law - Evidence - DNA evidence - Statistical evidence - Whether admissible -Whether prejudicial effect outweighs probative value - Whether jury must be satisfied of reliability of DNA evidence beyond reasonable doubt - Whether fair to admit results of DNA test at one locus when DNA tests at other loci rendered unreportable by contamination - Whether a DNA match probability as low as one in twenty is statistically significant.

1 The accused man, Miroslav Juric, stands charged with the murder of Orfeo Baldissera on 24 January 1996. He has twice before stood trial on that charge. He was convicted at the conclusion of the first trial in 1998 but the conviction was set aside on appeal. He was retried earlier this year but the jury were unable to agree upon a verdict. He is now to be tried a third time.

2 At the first trial the Crown adduced evidence concerning DNA testing carried out on a number of items found at the crime scene. It included evidence that crime scene examiners had found a latex glove tip in the deceased's motor car shortly after his death and that upon examination by the Victoria Police Forensic Services Centre it was ascertained that DNA on the glove tip may have come from the deceased and the accused but that it could not have come from the accused's two alleged accomplices, Evans and Tsagaris. Statistical calculations carried out on the basis of the test results were said to show that the chances of DNA having come from the deceased and the accused were more than 200 times greater than the chances of the DNA having come from the deceased and another person drawn from the male Caucasian population of Victoria. Two sets of tests were relied upon. The first, which was conducted in 1996, was directed to the HUMTHO1, FES, VWA, HLA-DQA1 and D1S80 sites and the second, which was conducted in 1998, used new technology called profiler plus and was centred on a number of additional sites.

3 At the first trial there was a conflict of expert evidence about the reliability of the DNA evidence. Dr Roberts and Ms Scott of the Forensic Science Centre expressed confidence that the results obtained at the HLA-DQA1 and D1S80 sites were reliable despite contamination that had been observed at the HUMTHO1, FES and VWA sites. Dr Atchison, who was called on behalf of the accused, was of the opinion that the contamination which had been observed at the HUMTHO1, FES and VWA sites rendered all of the results unreliable. Further criticism was directed at the profiler plus test results on the basis that they were so random as to be lacking in statistical reliability. The judge nevertheless admitted both the 1996 test results and the profiler plus test results as evidence for the Crown.

4 On appeal[1], the Court of Appeal ruled that the trial judge had been correct to admit the 1996 DNA evidence but that he should not have admitted the profiler plus evidence. In the Court's opinion it was well open to the jury to determine whether the contamination which was conceded to have infected the 1996 HUMTHO1, FES and VWA results rendered unreliable the 1996 HLA-DQA1 and D1S80 results, but that the jury could not on the basis of the material put before them properly and reasonably evaluate the differing and conflicting expert opinions expressed as to the profiler plus evidence.

5 Between the time of the appeal and the second trial, those who represented the accused obtained an opinion from Professor Boettcher that the contamination observed at the HUMTHO1, FES and VWA sites was likely to have come from the reagent which had been used to extract DNA from the glove tip for testing at all sites, including the HLA-DQA1 and D1S80 sites, rather than from the amplifying reagent or some other reagent used only at the HUMTHO1, FES and VWA sites. It followed in Professor Boettcher's opinion that none of the results could be relied upon or should be presented to the court. In light of that opinion, the Crown did not seek to tender any of the 1996 DNA test results at the second trial.

6 Since the second trial, however, the Crown has re-examined the 1996 DNA test results and it has reached the view that the 1996 D1S80 site result can still be relied upon, although the HLA-DQA1 site cannot be. The Crown wishes to contend at the third trial that, taken together with evidence which establishes Evans' and Tsagaris' involvement, the jury could and should be satisfied on the basis of the D1S80 site test result that there were three men involved in the murder of the deceased - Evans, Tsagaris and one other - and that one cannot exclude the possibility that the third man was the accused. The Crown wishes also to place before the jury a revised statistical calculation based only on the D1S80 site: that it is some twenty times more likely that the DNA came from the deceased and the accused than that it came from the deceased and another person drawn from the Victorian Caucasian population.

7 A ruling has been sought in advance of the trial as to whether the DNA evidence would be admissible. It is contended on behalf of the accused that the Crown's change of heart is unwarranted and that all of the results of the 1996 DNA testing are so clearly unreliable as to be inadmissible or alternatively that the probative value of the results is so much outweighed by their prejudicial effect that they should be excluded in the exercise of discretion.

8 A considerable amount of evidence has been given on voire dire by witnesses called by the Crown in support of the reliability of the results of the testing at the D1S80 site and by Professor Boettcher and Dr Sally Ann Harbison on behalf of the accused to establish its unreliability.

9 Robert Andrew, who is a forensic scientist employed by the Victoria Police Forensic Services Centre has practised as a forensic scientist for the last 13 years in the biology division of the Centre and for the last 5 years has led a team responsible for the preparation and analysis of samples submitted for DNA profiling. He deposed that he performed the DNA extraction on the samples originally submitted in connection with the deceased's death, including the latex glove tip said to have been found in the deceased's motor car shortly after his death (which was identified as item 28), and that he subsequently carried out testing at the D1S80, FES and HUMTHO1 sites on the DNA so extracted. He said that he used the same reagent for extracting DNA from item 28 for testing at each of those sites. He explained that he carried out the testing at the D1S80 site on 12 March 1996 and that at the same time he tested the extraction reagent separately using what is called a reagent blank. He observed alleles in the DNA extracted from item 28 at the 18 position and possibly also the 25 position, which he later amended to the 25c position, but he did not observe any alleles in the reagent blank. He concluded that the DNA extracted from item 28 contained DNA from two sources, one a man with DNA with an allele at the 18 position only (which it is said would correspond with the deceased) and another man with DNA with alleles at the 18 position and the 25c position (which would be consistent with the accused but exclude Tsagaris and Evans). He concluded on the basis of the tests of the reagent blanks that the reagent was not contaminated.

10 Mr Andrew deposed that on 15 March 1996 he loaded gels with DNA extracted from item 28 for testing at the HUMTHO1 and FES loci and for testing the extraction reagent at those loci. When he read those gels on 18 March 1996 he observed alleles in the reagent blanks at the HUMTHO1 locus at the 6, 7 and 9.3 positions and alleles at the FES locus at the 10, 12 positions. It was plain that both reagent blanks were contaminated with DNA.

11 Mr Andrew said that the fact that the reagent blanks were found then to be contaminated with DNA gave rise to a scientific possibility that the reagent itself had been so contaminated when used to extract DNA from item 28 for testing at the D1S80 site on 12 March 1996, but he discounted that possibility on the basis that the reagent blanks tested on 12 March 1996 showed no signs of contamination. He also considered that it was unlikely that the test run on 12 March 1996 had simply failed to detect any contamination then present in the reagent. As he observed, the same reagent had been used on that day to extract DNA from several samples apart from item 28 and no sign of contamination had been detected in those cases. In Mr Andrew's opinion, it followed that the most likely explanation of the contamination observed in the reagent blanks on 18 March 1996 was that it was introduced into the reagent blanks on a pipette used at or about the time of loading the gels on 15 March 1996 for testing at the HUMTHO1 and FES loci.

12 Ms Pamela Scott has been a forensic scientist since graduation from the Royal Melbourne Institute of Technology in 1979. She began working with DNA in 1995 and after qualification and authorisation she began to carry out DNA tests in 1996. She acted as the report officer for the DNA testing carried out on the samples collected in connection with the death of the deceased. She did not carry out any of the tests herself, but she checked some of them and she confirmed the conclusions which were reached. She expressed agreement with Mr Andrew as to the most likely cause of the contamination detected in the reagent on 15 March 1996 and expressed herself confident that one could rely upon the results of the testing of item 28 at the D1S80 locus.

13 Dr Henry Roberts is a senior forensic scientist employed by the Victoria Police at the Forensic Services Centre. He is highly qualified. He holds the degree of Bachelor of Arts in biology with chemistry from the University of York and the degree of Doctor of Philosophy in the area of biochemistry from the University of Oxford. He has been practising in the area of forensic biology continuously since 1981 except for a break between 1985 and 1988 when he practised in the field of forensic chemistry. Between 1988 and 2000 he was manager of the DNA laboratory at the Centre, and since 2000 he has been engaged upon more general biological examinations.

14 He expressed the opinion that there are several possible explanations for the fact that contamination was not observed in the reagent blank tested at the D1S80 locus on 12 March 1996 but was observed in the reagent blanks which were tested at the HUMTHO1 and FES loci on 15 March 1996. He said that the first possibility was that the D1S80 test run on 12 March 1996 was insufficiently sensitive to detect the contamination, in which event, he said, the contamination plainly did not affect the results of the test and they may be regarded as reliable.

15 The second possibility was that the D1S80 test failed in testing the reagent blanks on 12 March 1996. But Dr Roberts considered that the probability of that having occurred was so low that it should be ignored. In Dr Roberts' opinion it was unlikely because the positive control and several other samples which were amplified and typed with the same reagent produced positive results (which meant that the D1S80 test had not failed to detect alleles in those samples). He also noted that two reagent blanks, RB blood and RB hair run on 12 March 1996 were run in widely separated lanes on the gel and that neither had produced a positive result. He considered it to be most unlikely that the two RB tests would fail simultaneously in those circumstances.

16 The third possibility, and the one which Dr Roberts considered to be most likely, was that the contaminant DNA was introduced into the reagent blanks after the D1S80 test was carried out on 12 March 1996 and before the tests at the HUMTHO1 and FES loci were carried out on 15 March 1996 and read on 18 March 1996.

17 Dr Roberts said, however, that it made no difference so far as he was concerned how the contaminant got into the reagent blanks. If it did not occur until 15 or 18 March 1996 it was axiomatic that it could not have affected the efficacy of the tests conducted on 12 March 1996. If it did get in before that it was nevertheless improbable that it was in the samples tested on 12 March 1996, because no contamination was observed on that day either in the reagent blanks tested on that day or in twenty or so tubes tested in other matters on that day using the same reagent. And even if it were in the reagent on 12 March 1996, it was evident that it was present in such low concentrations as to be undetectable on that day, and hence it could not have affected the results of the tests of item 28 which were conducted on that day. In Dr Roberts' opinion, one could be confident in the results of the D1S80 test conducted on 12 March 1996. So far as he was concerned, there was no doubt at all about their reliability.

18 Professor Barry Boettcher was formerly professor of biochemistry at the University of Newcastle. He too is highly qualified in the area of DNA analysis and testing. He holds the degree of Bachelor of Science, majoring in biochemistry and genetics, from the University of Adelaide and a Doctorate of Philosophy also from the University of Adelaide. His doctoral thesis was on the subject of the genetical control of blood group substances in body secretions. Prior to his retirement, he taught genetics and genetic statistics in the University of Newcastle for many years.

19 Professor Boettcher was of the opinion that the contamination which had been discovered in the reagent blanks tested on 18 March 1996 cast such doubt on the reagent that without further testing of the reagent one could have no confidence that it was not contaminated when used to test at the D1S80 site on 12 March 1996. In his opinion the doubt thereby created was such as to render all results including the D1S80 results unreportable, meaning that it was not safe to base any conclusions upon them.

20 Professor Boettcher rejected as unacceptable the various possibilities postulated by Dr Roberts. He said that the fact that two reagent blanks tested on 18 March 1996 were the only contaminated tubes among the 24 or so tubes tested that day rendered it extremely unlikely that the contamination was independently introduced into the reagent blank tubes. He estimated that assuming that all tubes were loaded with the same pipette, the probability that the two reagent blank tubes but no others were contaminated from independent events was in the order of one 1/76,176 or 0.0013127%. He considered it to be a real possibility that the reason that the reagent blanks tested negative on 12 March 1996 was because the D1S80 test may have been too insensitive to detect the contamination. He said the fact that the FES and HUMTHO1 tests were sufficiently sensitive to pick up the contamination did not alter his view about that, because he considered that there was nothing to show that the D1S80 test was necessarily as sensitive as the FES and HUMTHO1 tests. He also considered it to be a possibility that the D1S80 test of the reagent blanks had failed, even though it was evident that it did not fail on the other samples tested on that day. He considered that one could not know what consequence may have resulted from contaminant present on 12 March 1996 in such small concentrations that it could not be detected. He appeared not to accept that it would make no difference to the test results.

21 Professor Boettcher criticised the scientific techniques which had been employed, including the fact that the positive and negative controls run on 12 March 1996 had been run on a different gel to the samples tested. He also observed that it was not until about 18 March 1996, at the earliest, that Mr Andrew apparently reached a state of confidence that the 25 allele was in fact a 25c allele, perhaps as a result of retesting item 28 on that day with the aid of a chemical comparator termed "a ladder". If that were so, he said, it followed that the determination that the DNA contained a 25c allele as opposed to a 25 allele did not occur until after contamination had entered the system, even on the hypothesis advanced by the Crown, and thus in Professor Boettcher's opinion the 25c result would be unquestionably unreportable. Moreover, he observed, when the sample was retested in September 1996 the results were at best a questionable 25c. It was plain, he said, that Mr Andrew was unable to achieve any certainty at that stage, one way or the other. In Professor Boettcher's opinion it was remarkable that the only occasion on which a 25c allele was said to have been observed with clarity was on or about 18 March 1996, for that was the very day on which contamination was first found in the system, and that that result was never able to be duplicated in an uncontaminated environment. It was in his opinion also unacceptable that although the laboratory paperwork implied that a second typer had observed a clear 25c allele and thereby confirmed Mr Andrew's observations, there was not anywhere to be found a documentary record in which the second typer had recorded his or her independent observation of the 25c. All one had was a form which purported to confirm the 25c result and some evidence that the protocol in the laboratory was that a result could not be signed off as confirmed unless it had been observed independently by a second typer.

22 Towards the end of his cross examination, Professor Boettcher also ventured the opinion that even if the results of the D1S80 test could be relied upon, he would not consider it to be clear that the results were as interpreted by Mr Andrew. Whereas Mr Andrew was of the clear view that the D1S80 test gel showed the presence of alleles at the 18 and 25c positions, Professor Boettcher said that he was only able to observe alleles at the 18 and 25 positions, with a thickening at the 25 position; possibly on the cathodal side. He agreed that the thickening which he observed on the cathodal side could be a 25c allele, but he said there were other possible explanations, including the possibility of a lack of uniformity in the electro-magnetic field imposed across the gel; thus resulting in distortion of the electro-phoresic projections.

23 Finally, Professor Boetthcer said that he regarded as meaningless the statistical calculation that it was twenty times more likely that the DNA came from the deceased and the accused than that it came from the deceased and someone else drawn from the Victorian Caucasian population. He pointed out that to say that it is twenty times more likely that the DNA came from the accused than from any other person in the Victorian Caucasian population is really to say no more than that the DNA could have come from one in every twenty people comprising the Victorian Caucasian population (or, roughly speaking, from any of 200,000 people in the State of Victoria). He added that, at least in scientific research, one should allow for a statistical deviation of plus or minus 5% when assessing the significance of any probability ratio and, further, that any probability ratio with a denominator of twenty or less is ordinarily regarded as statistically insignificant.

24 Dr Sally Ann Harbison gave evidence via video link from Auckland, New Zealand. Dr Harbison holds the degree of Bachelor of Science (Hons) in biochemistry, and a Doctorate of Philosophy, in molecular biology, from the University of Liverpool. She is the science leader of the Forensic Biology Group in the Institute of Environmental Science and Research Limited (ESR) at the Mount Albert Science Centre in Auckland, New Zealand. That laboratory is accredited by the Laboratory Accreditation Board of the American Society of Crime Laboratory Directors, which is commonly referred to as ASCLADLAB. Dr Harbison has worked in the forensic biology unit since she joined the laboratory in 1988 following three years of post doctoral research, and she has specialised in DNA profiling analysis since 1997.

25 Dr Harbison agreed with Mr Andrew and Ms Scott and Dr Roberts that the reagent blank that was amplified for D1S80 itself showed no evidence of contamination. But she was of the opinion that the testing had not excluded the possibility that the contamination which was observed at the FES and HUMTHO1 sites occurred in the system before or at the time that the D1S80 test was carried out.

26 In particular, Dr Harbison disagreed with Dr Roberts' observation that because the FES and HUMTHO1 tests had detected contamination, one should discount the possibility that the D1S80 test had been too insensitive to detect any contamination that was present. Dr Harbison pointed out that the size of each of the samples of DNA tested at the FES and HUMTHO1 sites was 30 ul, and thus twice the size of the 15 ul sample of DNA which was tested at the D1S80 site, and she said that it was possible, although she could not say so with any certainty, that if a larger sample had been used at D1S80, contamination would have been detected. In Dr Harbison's opinion, that, in itself, did not alter the result obtained at D1S80 but she thought that it suggested, perhaps, a level of unreliability.

27 Dr Harbison was also critical of Dr Roberts' method of calculation that it was 20 or so times more likely that the DNA in the sample came from the deceased and the accused than that it came from the deceased and another member of the Victorian Caucasian population. She deposed that if the ratio were calculated according to the method which is customarily used at ESR and widely used elsewhere, the D1S80 evidence rises no higher than that it is at least seven times more likely that the DNA on the glove tip came from the deceased and the accused than that it came from the deceased and someone selected at random from the Caucasian Victorian. Dr Harbison deposed that she considered that a likelihood ratio of seven provides such slight support in a case involving DNA evidence that it is extremely unlikely that ESR would report a statistic of that magnitude. It is more likely, she said, that ESR would say that the accused could not be excluded as being the source of the DNA, but that many other individuals also could not be excluded and that very little statistical weight should be given to such a result.

28 In her opinion, the DNA evidence provides such slight support when combined with the uncertainties about contamination and low levels and yields of DNA that the result should not have been reported or at the most should be reported as inconclusive. It should be taken as establishing no more, she said, than that the accused is not excluded as a contributor of DNA to the sample.

29 Mr Horgan SC who appears with Mr Gilligan for the Crown submitted that despite all of the evidence given on the voire dire, the position now remains essentially as it was when the matter was considered by the Court of Appeal in 2001. At the first trial there was a competition of expert opinions as to whether the contamination observed at the HUMTHO1 and FES loci rendered invalid or unreliable the results of the testing at the D1S80 locus. At that time Dr Roberts and Ms Scott were of the view that they did not whereas Dr Atchison was of the view that once contamination was observed at one site it rendered unreportable the results which had been observed at all other sites. The Court held, however, that the results of the testing at the D1S80 locus were admissible, because:

"Full explanations were given by the experts for their competing views and his Honour, in our view, was correct to conclude that those were matter `accessible' to the jury and capable of being resolved by them."[2]

30 Mr Horgan contended that is still the case, despite the change in the detail of the expert evidence on both sides. In Mr Horgan's submission, each of the experts who gave evidence on the voire dire provided a full explanation for their views, and related their views back to the primary technical evidence. Hence, if the same evidence were presented to a jury, the jury would have an adequate basis on the material put before them properly and reasonably to evaluate the differing expert opinions.

31 Mr Dane QC who appears with Mr Pearson for the accused submitted that the position now is very different to that which obtained when the matter was considered by the Court of Appeal. The competition of expert opinion is now between the views of Dr Roberts and Ms Scott and Mr Andrew on the Crown side, all of whom are of the opinion that the results at the D1S80 locus can be relied upon, and the views of Professor Boettcher and Dr Harbison on the accused's side, who are both of the view that the results cannot be relied upon: in part because of Professor Boettcher's thesis that once contamination is observed at one site it renders unreportable the results obtained at all other sites, but also because Professor Boettcher considers that the most likely point of contamination was before the D1S80 tests were undertaken (and hence that those results are directly compromised) and because Dr Harbison considers that one cannot exclude that possibility.

32 In Mr Dane's submission the degree of doubt which now attends the reliability of the D1S80 test results is of an altogether different order to that which was observed at the time of the first trial. It is now so great, he said, as to lead inexorably to the conclusion that the results are bereft of probative value. It follows accordingly that the results are inadmissible. Alternatively, Mr Dane submitted, the breadth and sophistication of the expert debate has now reached proportions that place beyond the capacity of a jury the means to discern which of the competing views is properly to be preferred[3]. That is true, in his submission, not only with respect to the question of the effects of the contamination, but also with respect to the differing views about the interpretation of the gel in which the D1S80 test results are embodied. Mr Dane pointed out that Mr Andrew was the only Crown witness qualified to read and interpret the gels, and he said that because his interpretation is different to Professor Boettcher's interpretation the issue could come down to a stark competition between two expert opinions about the meaning of the gel. Mr Dane contended that the only way in which the jury could resolve such a competition would be for the jury to read the gels themselves. But, Mr Dane said, the jury would be incapable of reading the gels themselves because they would not be qualified in the art. Hence there would not be a basis or at least an adequate basis for resolving the conflict of opinion.

33 Mr Dane further submitted that even if I should regard the evidence as having probative value and therefore as being admissible, the prejudicial effect of the evidence would so outweigh its probative value that I should exclude it. He contended that the aetiology of the contamination observed on 18 March 1996 is so uncertain that one must have the gravest doubt about the results of the tests conducted on 12 March 1996. Given, however, the cachet or mystique which attracts to the relatively new science of DNA typing, he submitted, the jury would almost certainly give to the results of the testing a weight which they will not bear. There is also a likelihood, he contended, that the Crown's expert witnesses would make a virtue of the inadequacies of their testing procedures, by emphasising to the jury that all results in which contamination had been found had been discarded, as if to suggest (contrary to the fact) that one could have complete confidence in the one result which had not been discarded, and further and in any event that it was unfair to allow the results of the D1S80 test to go to the jury when it is possible that the results of the other tests, ruined by contamination, may have excluded the possibility of the accused's involvement. To allow that, in Mr Dane's submission, would be in effect to allow the Crown to deprive the accused of a chance of defending the charge.

34 Additionally, Mr Dane contended, the admission of the evidence would create a risk of what he described as a "war" between the Forensic Services Centre, ESR and an expert from New South Wales, wherein the focus would become an attack upon and the defence of the laboratory standards of Forensic Services Centre, and that that would be a distraction for the jury which would in itself be prejudicial to the accused; especially when judged against the limited probative value of the evidence.

35 In any event Mr Dane said, the DNA evidence must be seen as a critical part of the Crown's case, and thus an indispensable link in any chain of reasoning that might lead to conviction or, alternatively, that it had the status of what has been referred to in some of the cases as an "additional fact". That being so, he submitted, the jury would have to be instructed that they could not act on the basis of the evidence unless they were satisfied beyond reasonable doubt as to its reliability. And since it was evident that one could not be satisfied beyond reasonable doubt about its reliability, it should be excluded at the outset. To do otherwise, Mr Dane said, would be to set up an Aunt Sally simply to enable the jury to knock her down again, but with all the attendant prejudice that the evidence might create in the mind of the jury.

36 Finally, Mr Dane submitted, even if I were disposed to allow any of the evidence to be put before the jury, I should exclude the statistical evidence. It was in Mr Dane's submission highly prejudicial and yet it had been shown on the evidence given on the voire dire to have virtually no reliability or statistical significance at all.

37 Mr Horgan replied that there has been no sea change in the degree of doubt which attends the results of the tests. He contended that it is rather a case where the appearance of doubt has been contrived by the taking of captious points about the detail and by avoiding the clarity of the substance. He pointed to the fact that when Professor Boettcher first expressed an opinion on the subject in his two reports of 17 and 19 February 2003, it was based on the false premise that the reagent blanks, RB Blood and RB Hair, had not been run on 12 March 1996 and he submitted that since the error in the Professor's assumption was exposed by Dr Roberts, Professor Boettcher had simply seized on feckless points of detail in an endeavour to support an opinion which is plainly no longer supportable. Mr Horgan submitted that Dr Harbison's evidence really did not take the matter any further. At its highest, he said, it is an opinion that one cannot exclude the possibility that contamination was present when the D1S80 test was run. Mr Horgan contended that there would be no difficulty in the jury making an informed decision as to which of the competing opinions is to be preferred. He said that all of the documentation will be accessible to the jury and hence they will be able to see for themselves the basis on which the Crown witnesses and Professor Boettcher and Dr Harbison have come to their respective conclusions. Mr Horgan also said that the jury should have no difficulty in deciding whether the 12 March 1996 test did yield a clear 25c allele. They will be able to look at the gel and with the aid of expert testimony be made to understand why the result is perceived as it is by each side. That is the sort of function, he contended, that juries perform every day.

38 Mr Horgan also rejected the idea that the jury would have to be instructed that they could not act on the basis of the DNA evidence unless satisfied beyond reasonable doubt of its reliability. He submitted that to conclude otherwise would be to confuse an indispensable link in a chain of reasoning with the evidence by which the indispensable link might be proved, and that any suggestion that there is in circumstantial evidence cases some category of "additional facts" of which the jury must be satisfied beyond reasonable doubt even though the "additional facts" are not indispensable links, is to make the sort of mistake about the meaning of Shepherd v The Queen[4] which was exposed in R v Kotzmamm[5].

39 Whatever doubt may now have been cast upon the results of the D1S80 test, I do not accept that it is enough to deprive the results of probative value. Despite the many criticisms which have been levelled against the results in the course of the voire dire, I consider that it would be open to a jury, on the basis of the evidence which I have heard, properly to come to the view that the results can be relied upon. That does not mean that I have come to that view myself or even that I think it probable that a jury would come to that view. I express no conclusion about that one way or the other. The point is simply that there are within the evidence which has now been directed to this exercise a number of rational bases for accepting the results of the D1S80 test, even if there are also rational bases to choose to reject them.

40 I also do not accept that the breadth or sophistication of the expert debate has yet reached proportions which would make it inaccessible to the jury. Certainly, Professor Boettcher has advanced a theory as to the point in time at which the reagent was first contaminated and, if that theory were accepted, it could result in the rejection as unreliable of all results obtained with the use of the reagent. To some extent, his opinion is backed up by the view of Dr Harbison that one cannot exclude the possibility that contamination was present at the time of the D1S80 test. But it could hardly be said that Dr Harbison was wholeheartedly dismissive of the results. When asked by Mr Dane as to what she said about the reliability of the results, Dr Harbison replied:

"Well, because different amounts of DNA have been used in the amplification, I don't agree that the results show that the contamination must have happened after the test. It may have happened before, or after, but it just - because a smaller amount of DNA was tested, it may not have been detected. That, in itself, does not alter the result obtained at D1S80."

Asked then by Mr Dane: "Does it put it into any question?", Dr Harbison confined herself to:

"I think it tends to suggest, perhaps, a level of unreliability."

Contrastingly, Mr Andrew and Ms Scott and Dr Roberts hold firmly to the opinion that the D1S80 results can be relied upon and they are able, as they have demonstrated on the voire dire, to source their opinions to identified objective facts.

41 Certainly, both views of the matter and the facts on which each is based are complex and not easy to master. They involve biochemical techniques and mechanisms which are well beyond the ordinary experience of most men and women. But although they are complex and perhaps difficult, I do not consider that they would be incomprehensible to any attentive lay person once explained as they have been to me in the course of the evidence. Moreover, as was said recently by Gummow and Callinan JJ in Velevski v The Queen[6]:

"[182] Juries are frequently called upon to resolve conflicts between experts. They have done so from the inception of jury trials. Expert evidence does not, as a matter of law, fall into two categories: difficult and sophisticated expert evidence giving rise to conflicts which a jury may not and should not be allowed to resolve; and simple and unsophisticated expert evidence which they can. Nor is it the law, that simply because there is a conflict in respect of difficult and sophisticated expert evidence, even with respect to an important, indeed critical matter, its resolution should for that reason alone be regarded by an appellate court as having been beyond the capacity of the jury to resolve."

42 If the evidence were adduced it might be necessary and it certainly would be desirable for each member of the jury to be provided with copies of the dozen or so pages of laboratory records which are said to support each side's view of the issues, for without those it is not easy to follow what each expert witness says about them. And if there were to be a dispute about the reading of the gel, the jury may need to be provided with the gel itself and to have it explained to them with the aid of some sort of paper or other diagram or guide as to the locations of the 18 and 25 and 25c positions. That would give them a head start in knowing where to look on the gel when assessing the experts' oral testimony. But there is nothing particularly sophisticated or out of the ordinary in any of that. Those sorts of aids are common enough in practice and they are the sorts of things which ss. 19 and 20 of the Crimes (Criminal Trials) Act 1999 expressly contemplate may be given to a jury[7].

43 The question of whether the prejudicial effect of the evidence outweighs its probative value is more complex. For as the Court of Appeal remarked when dealing with the matter in 2001[8], there are cases in which the simplicity of the opinion expressed cannot be allowed to obscure the difficulties which have been encountered in the testing process. While at one end of the spectrum there will be cases where the sample is so pure and unadulterated that clear typings can be obtained at a large number of DNA sites giving rise to statistical improbabilities running into the millions or even billions, there are cases at the other end of the spectrum cases like R v Tran[9] and R v Lewis[10] where the results of testing are so attended by uncertainty that they are to be excluded from the jury's consideration either on the basis that there is no reliable foundation accessible to the jury upon which they could properly assess the reliability of the opinions expressed by the experts or alternatively on the basis that the prejudicial effect of the evidence far outweighs its probative value.

44 In R v Tran the Crown sought to introduce evidence of DNA profiling to establish a connection between the accused and the murder victim. There was evidence that the victim had been raped before she had been killed. Body samples were sent to a DNA testing laboratory, including a bloodstain and swabs for the victim's body, bloodstains from the victim's boyfriend, bloodstains from the accused and bloodstains from the victim's parents. The Crown's experts conducted DNA testing on all samples, but most of them proved inconclusive. In the end the Crown relied on only one of the tests. In that test two identifying bands were very faint and the subject of considerable dispute between expert witnesses and even the Crown's expert conceded that they were not capable of statistical weight. McInerney J held that it was not open to the jury to prefer one expert opinion against another because the evidence was so unreliable. It was uncertain whether the source of the DNA was male or female cells and despite the fact that the accused was of Vietnamese origin, there was no database which included persons of Vietnamese or South-East Asian ethnicity. His Honour also considered to be relevant the facts that there was only one result in four, that the result could not be duplicated, that the accused was deprived of the opportunity of determining whether the test was accurate, and that there was no other material to corroborate the findings. In his Honour's opinion the competing views which had been expressed about the existence of an upper faint band and the criticisms of the scientific testing which had been carried out would have had such a tendency to produce a misleading and confusing impression for the jury as to render the evidence inadmissible. His Honour considered that the jury would not be in a position to determine the issues, even given the wide discretion given to juries. They could do no more than speculate about competing possibilities. Alternatively, if the evidence were admissible, his Honour considered that its prejudicial effect might far outweigh its probative value and thus that it should be excluded from the jury in the exercise of discretion. In that connection the judge referred with approval to the observation of Maurice J in R v Lewis [11] that :

"Forensic evidence, especially if it goes to a vital issue implicating an accused person in the commission of an offence, may often have a prejudicial effect on the minds of a jury which far outweighs its probative value. The jury, being people without scientific training, may often be impressed by an expert's qualifications, appointments and experience and the confident manner in which he expresses his opinions. And yet it ought not be left to such matters alone to provide a foundation for the jury making an assessment of the probative value of forensic evidence, particularly where there are conflicts in expert testimony, or where it is acknowledged that other experts of more or less equal distinction are unlikely to agree.".

45 Finally, his Honour observed that in any direction to the jury:

"...it could only be put if the jury accepted the validity of this testing as part of the chain of circumstances that could be taken into account in determining the guilt or otherwise of the accused, and a jury would have to be warned that this, in the circumstances, is by no means conclusive, that is the implication or association with him." (Emphasis added).

46 In Lewis v The Queen[12] the Court of Appeal of the Northern Territory upheld an appeal on the basis that forensic evidence identifying a suspect from a bite bruise left on a victim's flesh had not been established to be scientifically reliable and that a conviction based on such evidence was unsafe and ought to be quashed. The essential facts of the decision are conveniently summarised in the judgment of Maurice J as follows:

"This is a quite extraordinary case on its facts; extraordinary because, of the three eye witnesses who were called by the Crown, one positively excluded the appellant as a suspect. No satisfactory explanation for his evidence consistent with the accused's guilt was suggested by the Crown. Other things being equal, his evidence must have given rise to a reasonable doubt on the vital issue of identification. Even without this remarkable contradiction in the Crown's case, it would have been extremely dangerous for the jury to have relied upon the evidence of the other two eye witnesses. No doubt this was recognised by the Prosecutions Section of the Department of Law and goes a long way to explaining why the trial was put off for eighteen months after appellant was committed. The Crown needed forensic evidence; all it had was a photograph of bruise marks left by five teeth on the prosecutrix's boyfriend's chest and the decision of the Queensland Court of Criminal Appeal in R v Carroll (1985) 19 A Crim R 410 unanimously denouncing expert evidence purporting to connect marks of his kind to a suspect as unsafe. The Crown appears to have been waiting to see if special leave to appeal from the Carroll decision was granted by the High Court. It was not. In spite of this, relying upon the leading forensic odontologist, Mr Bernard Grant Sims whose evidence in the Queensland case had been declared unsound for the purposes of obtaining a conviction the Crown decided to press on."[13]

"Given the recency of the Carroll case and the novelty of forensic evidence positively identifying a suspect from bruise marks left on a victim's flesh, it behove the Crown in this case to carefully lay the ground for the reception of the opinions expressed by Mr Plummer and Mr Sims. It could only do this by proving the scientific reliability of the exercise they carried out. From my reading of the evidence, the Crown did not even attempt to do this; instead, it chose to rely on the witnesses' qualifications and experience in the field of forensic dentistry generally, and, in particular, upon the impressive curriculum vitae of Mr Sims and the confident way in which he expressed his conclusion."[14]

48 To similar effect, Muirhead J, with whom Asche J agreed, was of the opinion that:

"It may be that this type of forensic evidence is becoming important in the forensic field but as I understand the evidence there is no established universal view as to its reliability in identifying, as opposed to excluding, a suspect. Bruising or its degree must be variable when teeth and the body are in motion and the admissibility of expert evidence in this situation may be a matter of some doubt. Speaking for myself I do not consider it was of itself a sufficient foundation for conviction. There may be cases where it will be a link in a chain of circumstantial evidence, but here upon my view there was no chain. There was no other reliable circumstantial evidence, save for the appellant's presence at Squires earlier that night, linking him with the rape."[15]

49 In view of the approval which the Court of Appeal accorded to each of these decisions when it dealt with this matter in 2001, I begin my approach to the discretion to exclude the DNA evidence by reference to the sorts of considerations identified in each decision.

50 Starting first with Tran, it is apparent that the results of the D1S80 test are to some extent affected by the sorts of considerations to which McInerney J referred. Just as in Tran, only one result is available - the others having been shown to be unreportable - the result cannot be duplicated (because there is no longer left any of the sample or extraction reagent which was used), there is therefore no opportunity for the accused to determine whether the test was accurate, and there is no other forensic material to corroborate the findings. But the problems here are not as great as they were in Tran. On that part of the evidence given on the voire dire about which there is no dispute, two bands were observed at the D1S80 locus, albeit that there seems now to be some question in Professor Boettcher's view (although not in the opinion of Dr Harbison) as to whether the second band was a 25c or a 25 with thickening on the cathodal side. There is no question about whether the 18 and 25c or 25 alleles came from male cells. On the Crown experts' view of the matter, the results yield a high probability that the glove tip from which the DNA was extracted contained DNA from two sources, one of whom could have been the deceased (because the deceased had DNA with an allele at the 18 position at the D1S80 locus), and the other of whom could not have been either of the two men in whose company the deceased is alleged to have murdered the deceased (because, while Evans has an allele at the 25 position at the D1S80 locus, neither he nor Tsagaris has an allele at the 25c position at the D1S80 locus). Professor Boettcher and Dr Harbison do not argue with the logic of that analysis. They differ only in the extent to which the results of the D1S80 test can be treated as unaffected by contamination and hence reliable. There is little question about the relevance of the database. The calculations have been developed on the basis of the general Victorian Caucasian database, and it is not suggested and it does not appear that the accused is other than Caucasian. It is, however, less certain that the results enable one to conclude with any confidence that the DNA is twenty times more likely to have come from the deceased and the accused than from the deceased and someone else. While the accused has DNA with alleles at the 18 and 25c positions at the D1S80 locus, on Dr Roberts' calculations so do one in every 22 or so males in Victoria and on Dr Harbison's calculations the ratio could be as low as one in seven.

51 Do those considerations lead to the view that the prejudicial effect of the evidence is outweighed by its probative value? Apart from the statistical evidence, which I shall put to one side for the time being, I do not think so.

52 If one were to take at face value some of the things which were said in Tran and Lewis, it could be that all of the DNA evidence would be excluded in the exercise of discretion. The evidence goes to an important issue implicating the accused in the death of the deceased. The jury, being people without scientific training, may well be impressed by the experts' qualifications, appointments and experience and the confident manner in which they are likely to express their opinions. There are significant conflicts in expert testimony. And it does have to be acknowledged that they are experts of arguably more or less equal distinction who are unable to agree. These are all things which were said in Lewis, and thus by adoption in Tran, to point in the direction of exclusion.

53 But what was said in Tran and Lewis has to be read in context. As has been seen, Lewis was an extreme case. There was really only one expert asserting an opinion and it was not sourced to the facts. He had already been the subject of considerable criticism at the hands of the Queensland Court of Criminal Appeal. Neither the weaknesses in the expert's opinion nor the fact that he had been criticised were disclosed to the jury. The jury were left without basis or indeed reason to question the evidence and to decide for themselves whether it should be accepted. And apart from the forensic evidence there was really nothing else to link the accused to the offence. In the result there was a high probability that the jury would accord to the evidence far more importance than it deserved.

54 The position in Tran was also a good deal different to that which I find here. There are similarities of the kind to which I have already referred, but they relate to problems in nature no different to those which the Court of Appeal has already held were not sufficient to exclude the 1996 DNA evidence from the jury. Whereas in Tran there was not a basis for the jury to determine the reliability and weight to be accorded to the DNA evidence, here for the reasons which I have already given I consider that there is. The two sets of experts are able to demonstrate by reference to the facts how they have reached their conclusions.

55 There are indications in Tran and perhaps also in Evans of a line of authority that DNA evidence should be excluded from the jury unless there is a basis upon which the jury could be satisfied of its reliability beyond reasonable doubt. There are also a number of cases at first instance and on appeal in which it has been said that a jury should be instructed that they are not to act on the basis of DNA evidence unless they are satisfied of its reliability beyond reasonable doubt. If that were correct, the evidence would have to be kept from the jury unless it could be seen that there is a basis upon which they can be so satisfied. But in this case, I do not consider that it is correct.

56 In Tran McInerney J referred with approval to the observation of Hunt J in the unreported decision of Elliot, that:

"If scientific testing in the particular case is unreliable or if it has a tendency to produce a misleading or confusing impression for the jury, or if the weight to be afforded in the results is so minimal as to preclude the jury being satisfied beyond reasonable doubt that the Crown has established the fact which it seeks to prove then clearly I have a duty to exclude it from; the jury - whether it is a result of ruling that the evidence is inadmissible or whether it is excluded in the exercise of my discretion." (Emphasis added.)

57 In R v Green[16] the New South Wales Court of Criminal Appeal stated that:

"First the jury was required to be told that it had to be satisfied (beyond reasonable doubt) that there was a match between the two profiles. If they concluded that there was a match, then it would then have to be made clear to the jury that the only inference they could draw from such matching was that the appellant could not be excluded, and that therefore it was possible that he was responsible for the semen stains. Further, it would also need to have been made clear to the jury that the matching results could not in the absence of other evidence, prove beyond reasonable doubt that the appellant was the person responsible for the stains." (Emphasis added.)

58 In R v Pantoja[17], Abadee J observed to similar effect that:

"First, the jury was required to be told that it had to be satisfied (beyond reasonable doubt) that there was a match between the two profiles..."

"...it is important to emphasise that a match obtained by any blood tests - DNA or otherwise - between the suspect and the offender does not establish that the two are the one and the same person. It establishes no more than that the accused could be the offender. However, any blood test which positively excludes the suspect as the offender, if there is a reasonable possibility that the test is correct, must necessarily exclude the suspect completely, notwithstanding that a match has been obtained by other blood tests... (7) I have assumed that, in the particular case, it is necessary for the Crown to establish beyond reasonable doubt that the appellant was not excluded as the suspect, and that the probability of the match having occurred by chance or coincidence is very small. That is certainly what was required in the present case in the reasoning process towards an inference of guilt: Shepherd [1990] HCA 56; (1990) 170 CLR 573 at 579 - 580; [1990] HCA 56; 51 A Crim R 181 at 184-5. It will not always be the case that these facts will have to be established beyond reasonable doubt, in which case the words in italics in the test will require modification." (Emphasis added.)

59 The matter was dealt with again in the second R v Pantoja appeal[18] by BJ James J as follows:

"...(Hunt CJ at CL in the first appeal) expressly recognised that it would not always be the case that these facts would have to be proved beyond reasonable doubt and when they did not have to be proved beyond reasonable doubt the words in italics in the text (`if there is reasonable possibility that the test is correct') would require modification. His Honour held that in the present case these facts would have to be proved beyond reasonable doubt, in the context of concluding that `apart from the results of the forensic tests, the evidence against the appellant that he had had sexual intercourse with his sister-in-law was not strong'....I do not consider that at the second trial the evidence against the appellant that he had had sexual intercourse with his sister-in-law, apart from the evidence of the results of the forensic tests on the vaginal swabs, was `not strong' or that, if the Crown did not succeed in establishing beyond reasonable doubt that the appellant was not excluded as a suspect by the DNA testing of the vaginal swabs, the appellant should be necessarily excluded as being the person who had sexual intercourse with Ms Trujillo on 15 October 1992. In my opinion it would have been quite possible for the jury at the second trial properly to be satisfied beyond reasonable doubt that the appellant had had sexual intercourse with Ms Trujillo on 15 October 1992, without being satisfied beyond reasonable doubt that the appellant was not excluded by the evidence of the testing of the vaginal swabs. It would have been open to the jury at the second trial to reach the conclusion that the appellant had had sexual intercourse ...by taking into account (other evidence which was identified)."

60 Nevertheless, in R v Fletcher[19] Lee J referred with approval to what had been said in Green and also to what had been said by Cripps JA in Pantoja and observed that the need to instruct the jury according to what was said in those cases had been set out in Australian Criminal Trial Directions[20]. His Honour noted the observations of the Queensland Court of Appeal in R v Jones [21], to the effect that it may not always be necessary to instruct a jury that they should only act on the basis of DNA evidence if satisfied beyond reasonable doubt of its reliability, but he was persuaded that in the case which was before him it was appropriate so to instruct the jury.

61 Similarly, the South Australian Full Court in R v Karger[22] recently approved a trial judge's direction to the jury that strongly implied that the jury should be satisfied of the reliability of the DNA evidence beyond reasonable doubt before acting upon it[23].

62 In most of these cases it has been assumed or it has been clear that the fact sought to be established by the use of DNA evidence in question was an indispensable link in the chain of logic leading to conviction, in the strict sense essayed in Shepherd v The Queen[24]. If DNA evidence is the only evidence of such a fact and perhaps also where any other evidence of the fact is "weak" in the sense in which that word was used by Hunt CJ at CL in Pantoja, it is to be expected that the jury will be instructed that they must be satisfied beyond reasonable doubt of the reliability of DNA evidence before they could be satisfied of the fact sought to be proved. But as Hunt CJ at CL pointed out, it is otherwise where there is other evidence of the fact sought to be proved and it is open to the jury on the basis of the totality of the evidence to be satisfied of the fact beyond reasonable doubt, even if none of the individual pieces of evidence relied upon has that effect. BJ James J makes the same point in the second Pantoja appeal and his Honour's analysis was expressly approved by Batt JA, with whom Phillips CJ and Callaway JA agreed, in R v Kotzman[25].

63 What is the position here? Putting aside the statistical evidence, the fact which the Crown seeks to establish with the aid of the DNA evidence is that three men were involved in the death, not two as the accused alleged during the course of the last trial. That fact is an indispensable link in the chain of reasoning in the Crown case. Given that it is known that Evans and Tsagaris were involved, one could not logically be satisfied that the accused was guilty of murdering the deceased unless satisfied that there were at least three men involved in the death of the deceased. It follows that before the jury could convict the accused they would have to be satisfied of that fact beyond reasonable doubt.

64 That does not mean, however, that the jury would have to be satisfied of the reliability of the DNA evidence beyond reasonable doubt. While the DNA evidence may be regarded as an important part of the Crown case, in the sense of an important piece of evidence, and while there is authority that a crucial piece of evidence may be regarded as an "additional fact" which, even though it may not be an essential link in the chain of logic, has to be proved beyond reasonable doubt[26], the Court of Appeal has said that that is not the law. There are no "additional facts... that have to be proved beyond reasonable doubt even though they are not, in the strictly logical sense, indispensable links in a chain of sequential reasoning"[27]. The correct analysis is as explained by BJ James J in the second Pantoja appeal. Therefore, unless one is dealing with a case like Lewis or Tran, where there is no evidence apart from the DNA evidence or, perhaps, where the Crown case apart from the DNA evidence is "weak", it is not necessary that the reliability of DNA evidence be established beyond reasonable doubt. In this sort of case the DNA evidence may be regarded as one among a number of facts, none of which is necessarily established beyond reasonable doubt, but on the basis of the totality of which a jury may be satisfied beyond reasonable doubt. The point is explained by Callaway JA in R v Noll[28].

65 Here the DNA evidence is one fact among a number of facts on which the Crown seeks to rely in order to establish that there were three men and that the accused was one of them[29]. Other facts include the evidence of Evans and Tsagaris (albeit that they are accomplices), objective and subjective circumstantial evidence as to the accused's movements and behaviour shortly before and after the time of the alleged offence, and tape recordings of conversations in which the accused is said to have made admissions. In those circumstances, I do not regard the Crown case as "weak". I accept the possibility that a jury may be not be satisfied beyond reasonable doubt about the reliability of the DNA evidence, or indeed beyond reasonable doubt of any of the facts upon which the Crown intends to rely in order to establish that there were three men and that the accused was one of them. But even so, the jury might still properly be satisfied beyond reasonable doubt by the totality of the evidence that there were three men and that the accused was one of them.

66 Plainly, the jury could not take the DNA evidence into account unless they were satisfied of its reliability. It is important evidence and therefore they would have to reach a state of "reasonable satisfaction" about its reliability in the light of the nature and consequences of the fact to be proved[30]. But in the circumstances of this case, I do not accept that that warrants the exclusion of the DNA evidence from the jury. Despite the apparent conflict of expert opinion, I consider that it would be open to the jury to reach a state of reasonable satisfaction that the DNA evidence can be relied upon.

67 I turn next to the question of fairness. Is it fair to allow the DNA evidence to go to the jury when it consists of the results at only one site and where, to adopt Mr Dane's formulation, the contamination at other sites may have robbed the accused of a chance of exclusion by the results at those other sites? Is there a risk that the Crown experts will so much turn the shortcomings of their laboratory procedures into the virtue of excluding the results of the corrupted sites as to make the results of the D1S80 test appear more reliable than they are? I do not think so. While I suppose that each of those things is a theoretical possibility, I do not regard them as realistic possibilities. I start from the position that the problems which Mr Dane identifies were inherent in the DNA evidence when the Court of Appeal upheld its admissibility in 2001. I add that each of the problems has been canvassed to a greater or lesser extent in the course of extensive cross examination of the Crown experts on the voire dire. The response has been clear. That is not to say that I am convinced by what they have said, or to say that I am not convinced. For present purposes I do not conceive that to be my function. It is to say, however, that having now heard all the evidence and assuming it were repeated at trial, I see no chance of the jury being left in any doubt about the proportions of the problems or unable properly to reach a view about the conclusions which they should draw.

68 I am also not persuaded that the possibility of any "war" of the kind foreseen by Mr Dane constitutes sufficient risk of prejudice to the accused to outweigh the probative value of the DNA evidence. I do not see that relevant cross examination of the Crown experts would be likely to distract the jury from the task in hand and I exclude as unlikely the possibility that those who represent the accused would engage in cross examination of any other kind.

69 Finally, on this point, it appears to me that the argument about depriving the accused of a chance is somewhat misdirected. It is in essence like saying that if an eye witness had seen three men going into the deceased's house on the night in question, but did not see any of their faces, his evidence ought be excluded because, if he had seen their faces, he might have been able to say that the accused was not one of them. I regard that as an untenable proposition.

70 There remains the question of whether the statistical evidence should be excluded on the basis that a probability of one in twenty or one in seven is statistically insignificant, but of considerable prejudicial effect. Despite what was said by Hunt CJ at CL in Evans, I am hesitant to conclude that evidence should be withheld from a jury simply on the basis that it has the potential to mislead or confuse. As the decision of the Court of Appeal in this case implies, the difficulties which attach to potentially misleading or confusing evidence should ordinarily be dealt with by the use of appropriate directions. Nevertheless, there must and does come a point at which some technical evidence has the capacity to be so misleading or confusing, whatever directions may be given about it, that its probative value is exceeded by its prejudicial effect. At that point it is to be excluded from the jury [31].

71 In my opinion, one reaches that point in this case with the statistical calculation that it is twenty times more likely that the DNA came from the deceased and the accused than that it came from the deceased and some one else drawn from the Caucasian population of Victoria. That seems to me to be evidence that is so inherently capable of misleading or confusing the jury, whatever directions may be given, that it should be excluded.

72 As the authorities make plain, and as the evidence given by Dr Roberts bears out, the sorts of probability ratios that are ordinarily encountered in DNA evidence are in the order of one in several thousand if not in several million. Indeed it is because of the facility with which those sorts of results have been and continue to be produced on a regular basis that DNA evidence has acquired the wide spread reputation for reliability if not infallibility that it enjoys in popular culture. Given the popular perception of the importance of a DNA match, a jury might well treat as being of real significance a calculation that the chance of the DNA having come from the accused is twenty times more likely than the chance of it having come from someone else. In point of fact, however, the probative value of a match ratio of one in twenty is not much greater than nil.

73 As Professor Boettcher explained, to say that it is twenty times more likely that the DNA came from the accused than from any other person in the Victorian Caucasian population is really to say no more than that the DNA could have come from one in every twenty people comprising the Victorian Caucasian population (or, roughly speaking, from any of 200,000 people in the State). Moreover, if Dr Harbison is correct, the ratio when properly calculated is only one in seven. Assuming that one should allow for a statistical deviation of plus or minus 3% or 5%, as was suggested, it means it could be anyone.

74 In the result, I rule that so much of the DNA evidence as consists of the probability ratio will be excluded from the jury, but otherwise that the DNA evidence may be admitted. I shall hear counsel on the form in which the evidence should be adduced so as to achieve that objective.

[1] Rv Juric (2002) 4 VR 411

[3] cf Chamberlain v The Queen [No 2] [1984] HCA 7; (1984) 153 CLR 521 at 558, per Gibbs CJ and Mason J

[5] [1999] VSCA 27; [1999] 2 VR 123 at 139 [51] - [54], per Batt JA, with whom Phillips CJ and Callaway JA agreed.

[6] [2002] HCA 4; (2002) 76 ALJR 402 at [179] - [182]; see also Hoy v R [2002] WSCA at 275 at [226] - [228]

[7] cf R v BAH [2002] VSCA 164 esp. at [17]

[16] (Court of Criminal Appeal (NSW), 26 March 1993,unreported, at pp.41-2)

[20] Butterworths - Glissan and Tilmouth at para. [3-800-15-75]

[23] ibid at [47]; see also Andrew Ligertwood, Avoiding Bayes in DNA cases (2003) 77 ALJ 317 at 318

[26] R v Murphy (1985) 4 NSLWR 42; Penney v The Queen [1998] HCA 51; (1998) 72 ALJR 1316; Gipp v The Queen (1998) 194 CLR 193 at [76], [79] and [80], per McHugh and Hayne JJ in diss.

[27] [1999] VSCA 27; [1999] 2 VR 123 at 140 at [54], per Batt JA

[29] cf R v Gardner and Coates - Ruling 2 [2003] VSCA 153 at [14] and [15]

[30] R v Kotzmann supra at p.141 [56], per Batt JA

[31] R v Gardner and Coates supra at [19] and [20]

At a glance

Court

Decision date

Before

Source

Judgment (115 paragraphs)

Parties

Cases Cited (7)