Reported Decision: 205 A Crim R 157 New South Wales Court of Criminal Appeal CITATION: Aytugrul v R [2010] NSWCCA 272 HEARING DATE(S): 27 May 2010 JUDGMENT DATE: 3 December 2010 JUDGMENT OF: McClellan CJatCL at 1; Simpson J at 124; Fullerton J at 238 DECISION: By majority: Appeal dismissed. CATCHWORDS: CRIMINAL LAW - appeal against conviction - whether a miscarriage of justice occurred because of the directions given with respect to DNA evidence - whether the verdict of the jury should be set aside on the ground that it is unreasonable or cannot be supported having regard to the evidence - consideration of the admissiblity of DNA evidence in light of s 135 and s 137 of the Evidence Act 1995 (NSW) - manner of expressing in non-scientific terms the conclusions to be drawn from DNA testing to the jury - jury verdict open and support by the evidence LEGISLATION CITED: Evidence Act 1995 Criminal Appeal Act 1912 CATEGORY: Principal judgment Burrell v R [2009] NSWCCA 193 Fleming v R (2009) NSWCCA 223 R v Carroll [2010] SASC 156 R v Doheny and Adams [1996] EWCA Crim 728; (1997) 1 Cr App R 369 R v Duke (1979) 1 A Crim R 39 CASES CITED : R v Gallagher [2001] NSWSC 462 R v Galli (2001) 127 A Crim R 493; [2001] NSWCCA 504 R v GK [2001] NSWCCA 413; (2001) 53 NSWLR 317 R v Karger (2002) 83 SASR 135 R v Noll [1999] VSCA 164; 3 VR 704 Weiss v R (2005) 224 CLR 300; (2005) HCA 81 A Ligertwood, "Avoiding Bayes in DNA Cases" (2003) 77 ALJ 317 JJ Koehler, "On Conveying the Probative Value of DNA Evidence: Frequencies, Likelihood Ratios, and Error Rates" (1996) 67 University of Colorado Law Review 859 "The Psychology of Numbers in the Courtroom: How to Make DNA-Match Statistics Seem Impressive or Insufficient" (2001) 74 Southern California Law Review 1275 JJ Koehler, A Chia & JS Lindsey, "The Random Match Probability (RMP) in DNA Evidence: Irrelevant and Prejudicial?" (1995) 35 Jurimetrics 201) TEXTS CITED: RB Korobkin & TS Ulen, "Law and Behavioral Science: Removing the Rationality Assumption from Law and Economics" (2000) 88 California Law Review 1051, 1088; A Tversky & D Kahneman, "Availability: A Heuristic for Judging Frequency and Probability" (1973) 5 Cognitive Psychology 207 G Hazlitt, "DNA Evidence: Unravelling the Strands" (2002) 14 Judicial Officers Bulletin 9, 66 M Findlay and J Grix, "Challenging Forensic Evidence? Observations on the Use of DNA in Certain Criminal Trials" (2003) 14 Current Issues in Criminal Justice 269 at 274 WC Thompson, "Are Juries Competent to Evaluate Statistical Evidence?" (1989) 52 Law and Contemporary Problems 4 at 13) R Jonakait, "When Blood is their Argument: Probabilities in Criminal Cases, Genetic Markers, and Once Again, Bayes' Theorem" (1983) University of Illinois Law Review 369 PARTIES: Yusef Aytugrul (Appellant) The Crown FILE NUMBER(S): CCA 2008/5115 COUNSEL: R J Button SC (Appellant) D Arnott SC (Crown) SOLICITORS: O'Brien & Hudson (Appellant) Director of Public Prosecutions (Crown) LOWER COURT JURISDICTION: Supreme Court LOWER COURT FILE NUMBER(S): 2008/5115 LOWER COURT JUDICIAL OFFICER: R A Hulme LOWER COURT DATE OF DECISION: 16 April 2009 IN THE COURT OF CRIMINAL APPEAL 2008/5115 McCLELLAN CJ at CL SIMPSON J FULLERTON J FRIDAY 3 DECEMBER 2010 AYTUGRUL, Yusef v R Judgment 1 McCLELLAN CJ at CL: The appellant was convicted after trial of the murder of Sevda Bayrak who died on 26 November 2005. He appeals his conviction on two grounds. The first ground relates to the use which was made of DNA evidence. In the second ground the appellant submitted that the verdict was unreasonable. The prosecution case 2 The Crown case was circumstantial. It alleged that the appellant's motive for the killing came from the failure of his relationship with the deceased. She had apparently refused his request to marry and the Crown submitted that he was upset when she formed a new relationship. 3 The appellant and the deceased met at a primary school attended by their children in 2002. They formed a relationship which ended in July 2003. The deceased had travelled to Turkey in April 2003 and returned to Australia in the latter part of June. 4 Between about September 2003 to April 2004 the deceased was in a relationship with Hasan Kilic. Between 22 May and 27 October 2004 the deceased again travelled to Turkey with her daughter. 5 The appellant had originally believed that the deceased would remain in Turkey indefinitely. However, he saw her in a street in Auburn in about February 2005 when they had a short conversation. When interviewed by the police the appellant said that by that time he had had no desire to rekindle a relationship of intimacy with the deceased. He said: "No I don't want to come together again, no, no." 6 The Crown said that this statement was not true and contrasted it with the poem which the appellant had published in the Turkish Weekly News on 29 June 2005 entitled "I Can Not Give Up." Parts of the poem read: "My heart burning for you like mad … Even if you don't want to remember my name, don't want to hear my voice, Even if you say give up, I can not give you up … Even if you cry all your hate, say give up, I cannot give up." 7 The Crown called evidence from which it was submitted that the defendant was harassing the deceased. 8 Mrs Bas, the deceased's half-sister, gave evidence that the deceased had told her that the appellant had arrived at the deceased's premises, a Housing Commission apartment in Merrylands, unannounced, between July and October 2005 with his son. She said that she understood that the appellant had claimed that his son wanted to play with the deceased's daughter and accordingly the deceased allowed the appellant to enter. The deceased told her that she had not given the appellant her address. The evidence established that he obtained the deceased's address at Merrylands (and her subsequent Parramatta address) from Mr Uluc, a friend who was convicted of illegally obtaining the information from Auburn Centrelink. Mr Uluc's computer carried the evidence which, combined with telephone records, proved beyond doubt that he was the source of the information. 9 Between July and October 2005 the deceased lived in Merrylands. During this time, her car was vandalised on several occasions. In July, a jelly like substance was found in the fuel tank. In August the tyres were deflated, a head light broken and black paint was applied to the car. On another occasion a fire was set in an adjoining garage. In September her vehicle's badges and number plates were stolen and paint was put on the car. In October, the car's brake line was cut. The Crown suggested that it was the appellant who had been responsible for this vandalism, in particular the stealing of the number plates. 10 There was also evidence that the deceased had told people that her front door buzzer and telephone landline would ring at night but no one would be there when she answered. On one occasion she allegedly saw a shadow on her balcony and later discovered a bin moved into a position which could have been used to assist someone climbing up to access the balcony. 11 Because of the problems at Merrylands the deceased asked to be transferred from her accommodation to another Housing Commission facility in North Parramatta. She moved on 25 October 2005. On 4 November 2005 the appellant obtained the deceased's new address from Mr Uluc. 12 On 4 November 2005, a week after the deceased had moved from Merrylands, two men came to the deceased's Merrylands address ostensibly wanting money for fixing the deceased's computer. The deceased did not own a computer. The evidence was that one of the men, who did not sound as if he was from an English speaking background, asked her next door neighbour where the deceased had moved to. The neighbour did not see the man's face. There was also evidence that a person had contacted St George Community Housing looking for the deceased and said that he wanted to drop off her computer. It was later that same day that the appellant obtained the deceased's address from Mr Uluc. 13 The deceased told others of her concern that the appellant was following her. She told Mrs Bas that at the beginning of 2005 her car was being followed by the appellant when she stopped and called to him. She told Mrs Bas that the appellant said that he wanted to continue their friendship stating that "if [you] don't accept it whatever will happen to you from now on you will be responsible for that." The deceased apparently interpreted this as a threat. On another occasion the deceased was shopping at IKEA at Homebush when she received a telephone call from the appellant. He apparently asked whether she "had five minutes" saying that he "just wanted to chat to her and he would never bug her again." The deceased apparently told the appellant that she was busy. He responded that she was not and that he knew that she was at Ikea shopping. About a month before her death the deceased told her stepmother that the appellant had been following her. Two weeks before her death she told a woman for whom she cleaned that earlier that day the appellant was behind her in the shopping centre. 14 On 11 November 2005 the appellant sent a text message to the deceased in which he said "Sevda you are sad and crying I feel it." That message was sent at 12.59 am. At 1.47 am the appellant sent a further message saying "I'm sure because I asked about your number plates you are blaming me for it." Followed by a further message at 1.57 am in which he said "Alright, understood, there's a god above. What can I say. I pray for your health." 15 The records also reveal that the appellant telephoned the deceased on that day at 1.11 am. The deceased sent the appellant six text messages between 1.10 and 1.44 am but the content of both the telephone call and the text messages is unknown. 16 As it happened, the deceased was crying at the time the appellant sent the text message at 12.59 am. Apparently she had been hit by her father who had become aware that she had formed a relationship with Erdal Tunc. The deceased's relationship with Mr Tunc was secret. He was married. The Crown submitted that because the appellant knew that the deceased was crying, he must have been in her vicinity at the time. 17 On 21 November 2005 the deceased apparently received a mobile telephone call from the appellant. At the time she was at Ms Yavuz' place. The appellant allegedly asked her why she had changed her number plates and suggested that it was so he could not follow her. 18 On 19 November 2005 a neighbour of the deceased observed her embracing a man who may have been Mr Tunc. 19 A few days later the neighbour heard a male voice moaning and what sounded like an argument coming from her apartment. Mr Tunc said that he could not remember having an argument with the deceased. 20 The deceased's daughter gave evidence at the trial in which she said that she had seen Mr Tunc in the deceased's bedroom on one occasion. The deceased had later promised her that Mr Tunc would not come again. Mr Tunc said this did not mean their relationship had ended but rather that he would not come when the daughter was present. The deceased's daughter also gave evidence that Mr Tunc had borrowed up to $8,000 of the deceased's savings to assist him with his gambling debts. Mr Tunc said this was merely a story which the deceased told her daughter in order to explain his presence at their home. 21 On 23 November 2005 the appellant called Centrelink on two occasions to allege that the deceased had defrauded Centrelink. Both these calls were made from the appellant's mother's house where he was living at the time. On 25 November the appellant rang the deceased's mobile phone at 9.06 am and again at 11.47 am. On this occasion he used a different mobile phone to the one he normally used to speak to the deceased. The events of 26 November 2005 22 On 26 November 2005 there were 7 telephone calls between Mr Tunc and the deceased. Mr Tunc visited the deceased at her home and remained there from approximately 6.30 pm to 8 pm. At 8.06:10 pm Mr Tunc's wife Nurgul called him on his mobile phone but he did not answer. He said that he was speaking with the deceased at this time. At 8.13:28 pm Mr Tunc called his wife and said that he was in his car. Thereafter he spent the evening with his wife at a friend's party, returning to the outside of the deceased's premises early the next morning. 23 From 8.04:45 pm to 8.06:13 pm, the deceased spoke by telephone with Mrs Bas. The deceased said that she was leaving home to pick up her daughter from Mrs Bas' home. They had a discussion about getting McDonalds for her daughter for dinner. This telephone call was the last known contact with the deceased. At about 9 pm Mrs Bas made several telephone calls to the deceased but they were not answered. 24 A neighbour of the deceased, Mr Jiang, said that at around 8 pm, he saw a man knocking on the deceased's door. He said that it continued for about 5 minutes. According to the Crown this event happened after Mr Tunc had left. It was suggested that the person knocking was the appellant. 25 Other neighbours of the deceased gave evidence at the trial. One neighbour heard a high pitched female scream between about 8.20 and 8.40 pm. Mr and Mrs Knowles who lived opposite the deceased reported hearing the same scream. Ms Paul-Hus said she heard three or four screams and "throughout these screams I heard other sounds like voices or yelling" coming from the direction of the deceased's building. She said that she first heard screams at about the time a Harry Potter movie started on television, which was at 8.30 pm. She said that "15 minutes after I heard the first screams, I heard a loud hysterical scream. It sounded like it came from the same place as the other screams came from." 26 At 12.09 am the following morning ie 27 November Mrs Bas called Mrs Tunc. The Tuncs were still at the friend's party. Mrs Bas went to the deceased's flat at 12.39 am. She said that she knocked on the door for between 5 and 10 minutes. She called the deceased's mobile phone and could hear it ringing on the other side of the door but there was no answer. Mrs Bas had no key to the premises and she returned to her car. 27 Mrs Bas called her husband Ogun at 1.07 am. At about the same time she saw Mr Tunc arriving at the flat. Mr Tunc gave evidence that he knocked on the door but could not get in. When Mr Tunc noticed that Mrs Bas was present he left and returned home. 28 In the morning of 27 November Mr Tunc tried telephoning the deceased at least 5 times. The first call was at 9.33 am and the last one at 12.44 pm. 29 In the meantime, at around 8 or 9 am, he went back to the deceased's unit and knocked on the door a few times. He also rang her mobile telephone. He could hear the ring tone but there was no answer. Being unable to find the deceased, Mr Tunc went back home. He again returned to the deceased's residence about midday. 30 The deceased's half-sister continued to try and contact her. Ultimately the police were alerted. They arrived at the deceased's premises at about midday. The deceased's body was located close to the door of the unit at 12.40 pm. She had been stabbed receiving multiple deep wounds, particularly to her abdomen and chest penetrating her lung. There was one stab wound near her anus and one to her buttocks. A blood stained white plastic shopping bag was found in her right hand. The bag had a number of cuts to it consistent with the stab wounds to her body. 31 The police identified two indentations between the front door and its doorframe. Some small wood filings were found on the floor in this location. The Crown inferred that the door had been jemmied open. 32 At 1.10 pm Mr Tunc telephoned Mrs Bas. He apparently told her that he thought the deceased may have committed suicide out of a sense of guilt for her affair with him. Mr Tunc said that on the evening of 27 November he had himself gone to his garage and connected a vacuum cleaner hose to the exhaust pipe of his car intending to kill himself but stopped when he thought about his children. 33 At about 2 pm Mrs Bas telephoned her husband and told him that her sister was dead. About 10 minutes later she spoke to him again and told him that Mr Tunc had called her on her mobile telephone. After a further 10 minutes Mr Bas received a telephone call from Mr Tunc on the former's landline. Mr Bas said to Mr Tunc, "What were you doing at Sevda's place last night?" Mr Tunc denied being at the deceased's place and said, "I wasn't at Sevda's place last night. I was at my friend Selahattin's place. I was with him last night." Mr Bas said, "We saw you last night in front of the house. We are not stupid." When Mr Bas pressed Mr Tunc the latter said, "I was worried about her, that's why I came." Mr Bas said, "Why did you bolt when you saw my car?" and Mr Tunc answered, "It's late at night, I didn't want people to get the wrong idea, that's why I ran." 34 Mr Bas said that he did not tell Mr Tunc anything about how the deceased had died. He said that at that time he did not know the cause of her death and did not say anything about blood being present. 35 Mr Tunc said that he rang Mr Bas and Mr Bas starting accusing him of killing the deceased. He said that Mr Bas said to him, "What did you do to her?" and when Mr Tunc asked him what he meant Mr Bas said, "She's got blood all over her" and said something like "Were you screwing her? You killed her." 36 Mr Tunc attempted to tell Mr Bas that he could not do something like that but they stopped talking. Mr Tunc said that this was the first time he became aware that the deceased had been killed rather than committed suicide. 37 Later when Mr Tunc was arrested he asked the police whether the deceased was killed with a knife. He said that he knew that the deceased had blood all over her because Mr Bas told him this. At this time he had assumed that the deceased had slit her wrists. 38 The appellant was interviewed by the police. The second interview occurred on 9 March 2006. The Crown submitted that in that interview the appellant told three lies. He denied ever going to the deceased's Merrylands address with his son, claimed he did not know the deceased's address in Parramatta and said that he did not ask for or receive details of the deceased's address from Mr Uluc. The Crown also submitted that the appellant was not credible when he claimed that the deceased was a friend in the period before her death and when he denied any suggestion that he was following her. 39 Although Mr Tunc was originally arrested on suspicion of the murder he was later released. The appellant's explanations 40 The appellant said that on the evening of the deceased's death he was at home with his mother and eight year-old son. His mother gave evidence that this was the case. She said that they had dinner at the table although the appellant said that they had dinner whilst watching television. Although the Crown emphasised this discrepancy I do not believe it to be of any moment. It is quite likely that people who are endeavouring to accurately recall events could have a faulty recollection of that sort of matter. 41 There was evidence that at the date of the deceased's death the appellant's car had a faulty exhaust. Mr Okandan said that he fixed the exhaust on 28 November and although before this the car could have been driven, it would have been noisy and would be likely to have been heard by the deceased's neighbours if the appellant was in the vicinity. There was no evidence of any person hearing a noisy motor vehicle. 42 The appellant accepted that the letter and poem which were published were intended for the deceased. However he said that he did not intend to be aggressive or threatening but meant to rekindle the deceased's feelings towards him. There was evidence suggesting that the deceased loved poetry and that expressions of high emotion were not uncommon in Turkish culture. 43 At the time of the deceased's death the appellant was in a new relationship. The appellant's relationship with the deceased had ended 2½ years before her death and it was suggested that his new relationship was happy. He spent between 3 and 4 nights per week at the home of his new companion. On the evening of 26 November the appellant made a number of phone calls to his companion. 44 The defence emphasised that the evidence indicated there had been conflict between the deceased and her father. Given the complexity of her life, the defence suggested that just as no one knew of her relationship with Mr Tunc until after she had died, it was possible that there were other persons with whom she was involved. The defence emphasised the fact that there were two men that visited the Merrylands flat on 4 November who apparently had returned on 19 November looking for the deceased. Ground 1: A miscarriage of justice occurred because of the prejudicial way in which DNA evidence was expressed to the jury. Issues relating to DNA 45 The deceased's premises were carefully examined by forensic scientists and a hair was found on the thumbnail of the deceased. It was common ground that the hair became stuck to the nail as the blood on the nail was drying. The hair was accepted to be a hair which had been shed and not one which had been ripped out. The evidence was that it was possible that it found its way to the floor before the deceased was attacked and may have had no connection with her death. The hair was sent to the USA for mitochondrial testing. 46 The evidence indicated that the hair could have come from a male or female. All of the deceased's acquaintances were from the Turkish community. 47 DNA was also found on an abrasion under the chin of the deceased. That DNA profile was consistent with that of Mr Tunc although the DNA of at least two men was found under her chin. Apart from the DNA consistent with that of Mr Tunc, the other DNA was established to be not from the appellant. There was also fingerprints from unidentified persons found at the scene. 48 The appellant's motor vehicle was examined. He volunteered it for this purpose. He also volunteered to be interviewed. No blood was found in the car and there was evidence that it had not been cleaned for a long time. The appellant emphasised that when Mr Tunc's car was examined it looked like it had been cleaned. 49 The position of the Crown prosecutor was that proof of the hair being that of the appellant or connected to the murder was not an indispensable intermediate fact. He submitted that it was merely one of many strands in a cable. 50 When summing-up his Honour introduced the DNA evidence in the following manner: "…The expert evidence I am primarily concerned with at the moment is before you as part of all of the evidence to assist you with the question of whether that hair which was found on or under Ms Bayrak's thumbnail is consistent with having come from the accused. Clearly if it did that would very strongly link the accused with the murder because there is really no other explanation for how a hair that is consistent with being his could have got there." 51 After summarising the DNA evidence his Honour emphasised: "There is another important matter that you must consider in relation to this evidence and it does not involve evaluation of the evidence of the experts but it is a more fundamental matter. Did the hair come from the person who killed Sevda Bayrak? If it did not, it does not matter who it came from and all of that evidence of mitochondrial DNA takes the case nowhere." 52 His Honour reminded the jury that the hair was a shed body hair and that it may have been on or under the thumbnail and may have been picked up by the sticky blood from the carpet. 53 The Crown called Ms Gina Pineda (Technical Leader and Associate Director at Orchid Cellmark Inc, a private DNA company based in Dallas, Texas) and Professor David Balding (Professor in Statistical Genetics at Imperial College, London) to give expert evidence. The defence called Dr John Buckleton, a scientist employed by the New Zealand Government. All three gave evidence that the appellant could not be excluded as having been the donor of the hair, but their evidence differed as to the weight that could be attached to this conclusion. 54 The statistical evidence was presented to the jury in two forms: "random occurrence ratios" and "exclusion percentages". (In this appeal, the former have been referred to as a "likelihood ratios." Random occurrence ratios express the frequency with which a particular DNA profile is expected to occur in a population. The are expressed as "one in every X persons". An exclusion percentage expresses the conclusion as the proportion of people in that same population who would not be expected to have that DNA profile. 55 Ms Pineda's evidence was that the DNA profile taken from the hair was consistent with the appellant's DNA profile. Ms Pineda came to this conclusion having discounted a portion of the appellant's profile (which contained an extra "C" compared to the DNA profile found in the hair) because she accepted that there was a chance of variability in relation to that component of the profile. Professor Balding agreed that in the circumstances it was conventional to disregard such an inconsistency and to treat the two profiles as a perfect match. Dr Buckleton disagreed. In his opinion the inconsistency indicated a genuine difference between the profiles, which should not therefore be treated as a full match. 56 Ms Pineda gave evidence that Mr Tunc, whose DNA had been sampled, could be excluded as the source of the hair because his DNA profile did not match that found in the hair. Her evidence was that the DNA profile found in the hair and in the appellant's saliva occurred once in 4,839 persons on the SWGDAM database. Applying the standard 95% confidence interval she concluded that 1 in 1600 people in the general population (i.e. the world) would be expected to share the DNA profile found in the hair (random occurrence ratio). This figure was the "upper limit" which gave the most benefit to the appellant (the lower limit was 1 in 5000 individuals). Ms Pineda further stated that 99.9% of people would not be expected to have a DNA profile matching that of the hair (exclusion percentage). Ms Pineda appears to have rounded her result to one decimal point because the exclusion percentage corresponding to a random occurrence ratio of 1 in 1600 is 99.9375%. Unfortunately, this figure erroneously appears both as "99.5%" and "99.95" at different points in the trial transcript. 57 Professor Balding's evidence was that in the non-Turkish population, about 1 in 2000 people would be expected to share the DNA profile of the hair. He said that in the Turkish population, 1 in 50 (2%) or less of the population would be expected to share the DNA profile found in the hair (random occurrence ratio). These figures were also upper limits. 58 The evidence of Dr Buckleton placed emphasis on the ratio in which the relevant DNA profile might be found in the general population. Dr Buckleton's evidence was that 1 in 1000 people in the non-Turkish population, and between 1 in 50 and 1 in 100 people in the Turkish population, would be expected to have a DNA profile matching that extracted from the hair (random occurrence ratio). At the request of the trial judge, Dr Buckleton expressed "1 in 1000" as an exclusion percentage of 99.9%. "HIS HONOUR: Q: You indicated it should be reported more along the lines of 1 in 1000? A: Yes. Q: Ms Pineda as I recall put the 1 in 1600 in a different way but mathematically in effect saying the same thing, she referred to 99 point something per cent of the population being excluded – can counsel recall what it was, 99? DAVENPORT: 99.5 [99.9%] . HIS HONOUR: Q: So that's mathematically a different way of saying the same thing as 1 in 1600. Is it the case then for 1 in 1000 another way of saying that is what, 99%. A: Sure, if you want to do it that way. It's actually 99.99 so I don't know how the other was 99.5. Q: 99.5 I thought it was. A: Yeah, it must be 99.95 [99.9% or 99.9375%] because 1 in 1000 is clearly 99.9. Q: So the difference between the two of you, Ms Pineda and yourself, is she's saying 99.95 [99.9% or 99.9375%] and you're saying it should be more like 99.9, just on this mathematical exercise you're referring to - A: I consider the arguments regarding sampling and certainty to be marked with a lesser of the two evils. So in fact the difference, as you pointed out, is quite minimal." 59 There are numerous errors in the above passage. I have provided the relevant corrections in bold type. For accuracy's sake, it should be noted that the difference between the two experts' evidence lay between 0.0375% and 0.05%. 60 A few questions later, when Dr Buckleton was being asked about his last Power Point slide and what he meant by "upper limit", he gave this evidence: "A: Yep, I think 1 in 50 is probably conservative amongst Turks, the actual number is plausibly somewhere around 1 in 100 or 1 in 200 amongst Turks. The burning issue is how many of the pool of alternative perpetrators were Turkish? If they're all Turkish we should use that number, if only a fraction of them are we should weight ourselves between the two numbers David Balding gave." 61 In his Honour's summing-up, the trial judge made the following observations (emphasis added): "You have the opinion of Ms Pineda that, by reference to the SWGDAM database, she concluded this profile could be expected in 1 in 1600 people. Putting it in another way, 99.99% of the population could be excluded as possibly having this profile. … At one end, you have Ms Pineda's evidence that such a profile can be expected to be found in 1 in 1600 people or, looking at it from the reverse perspective, you would not expect it in 99.99% of people. At the other end you have the evidence of Dr Buckleton that among Turkish people you would expect to find the profile in something between 1 in 50 and 1 in 100 people or less, and the evidence of Professor Balding that you would expect it in 1 in 50 or less. Looking at their evidence in the reverse way, they are in effect saying you would not expect to find it in at least 98% of Turkish people. Of course, the less likely the expectation of finding the same profile in other people in the population, the more value the evidence has in establishing the probability that the hair came from the accused." 62 It is apparent that it was not correct to describe Ms Pineda's exclusion percentage as 99.99%. The figure used by Ms Pineda was 99.9%. Nothing turns on this issue. The defence's position is that the use of any percentages "close to 100" was impermissible. The question is whether the use of exclusion percentages between 98% and 99.9375% was erroneous. The appellant's submissions 63 The appellant submitted that the DNA evidence being expressed as exclusion percentages should have been rejected. It was submitted that, when expressed as percentages of close to 100, the evidence was unfairly prejudicial to the appellant and should have been rejected pursuant to s 135 or s 137 of the Evidence Act. 64 In R v GK [2001] NSWCCA 413; (2001) 53 NSWLR 317, GK was convicted of sexual intercourse with his stepdaughter who subsequently had a child. Expert evidence admitted at the trial indicated that the appellant "could be the biological father." The Crown sought to tender two Paternity Index results of 220,000:1 and 147,005:1 and their corresponding Relative Chance of Paternity results of 99.9995% and 99.9993% but the evidence was rejected by the trial judge. This Court concluded that although the Relative Chance of Paternity figures should not have been admitted, the Paternity Index figures should have been. The Court was concerned that the Relative Chance of Paternity figures were capable of misleading the jury. Sully J said at 341: "The bare percentage figures are so extraordinarily close to 100 per cent that there must be, at least, a real risk that jurors, even sensible jurors properly directed, would think that the difference between the percentage figures and 100 per cent was so miniscule that there could be no possibility that G.K. was not the father, when it is scientifically uncontroversial that the one thing that DNA testing cannot do is to exclude absolutely any possibility, however statistically trifling, to the contrary." 65 His Honour continued (at 341): "…it would have led me, in the present case, to have excluded the raw percentage figures, whether upon the basis of s 137 or s 135, not because of a fundamental dissatisfaction with the statistical sufficiency of the relevant data bases, but because of the combined effect of the following considerations: [1] that the residual risk of unfairness deriving from the subliminal impact of the raw percentage figures would have been, to me, both real and unacceptable; and [2] that the admission of the Paternity Index figures would have given the Crown every fair opportunity of persuading the jury to accept the Crown contention that there was no real possibility that G.K. was not the father." 66 Mason P who agreed with Sully J said (at 331): "…part of the difficulty would be removed if 99.9993% was transposed to a statement that there was a 0.0007% probability or chance of C's father being anyone other than GK. Even such adjustment would leave me of the view that the evidence is unduly prejudicial in its impact." 67 In R v Galli (2001) 127 A Crim R 493; [2001] NSWCCA 504 the appellant was convicted of having sexual intercourse with a 26 year-old who had a serious intellectual disability. There were five men who had contact with the victim at her group home where the sexual intercourse took place: three carers and two other patients. The two patients were eliminated as suspects, one being wheelchair bound and the other being blind. Two carers were eliminated by DNA testing. Expert evidence of DNA testing provided in relation to the appellant a Combined Paternity Index of 2,400,000:1, a Relative Chance of Paternity of greater than 99.9999% and a Relative Chance of Non-Paternity of less than 0.00001% in relation to the appellant. 68 The 99.9999% figure was not put before the jury, the Crown accepting that the evidence could be misleading or confusing as contemplated by s 135(b) of the Evidence Act. However, 99.9999% was mentioned in the course of a record of interview, the transcript of which was before the jury, and by the trial judge in his summing-up. Although the Crown argued that this was a fleeting reference, this Court applied GK and intervened. In relation to a Paternity Index of 50:1 that featured in submissions, Spigelman CJ said in obiter at [72]: "…if a figure of 98% was put to a jury, it is likely that many jurors would regard that as very significant evidence pointing to the accused, even though the Paternity Index ratio was very low, so that numerous persons in the general community could share the DNA profile." 69 In the result, the Court applied the proviso on the basis that in the circumstances of the case, the unchallenged Paternity Index being 172 to 1 was of such force that a conviction was inevitable (at [118]). 70 It was submitted in the present case that the DNA evidence was a very significant part of the prosecution case and would have been in the forefront of the jury's mind. So much was said to be plain from the portions of the summing-up to which I have previously referred at [50], [51] and [61] above. The Crown submissions 71 The Crown drew attention to the exchange between defence counsel and the trial judge with respect to the admissibility of Ms Pineda's evidence. The objection to the evidence was initially framed in terms that it was inappropriate to use the SWGDAM database, which Ms Pineda had used, because it was a North American database. Subsequently, counsel phrased the real problem as follows: "DAVENPORT: Your Honour, with respect the real problem that arises in my submission is that if this evidence goes before the jury in these terms, then it has, or may, have the effect of appearing to them to be somewhat like the 'prosecutor's fallacy', that is it's 99 percent sure that it is the accused. HIS HONOUR: It wouldn't be expressed in those terms. None of us would express it in those terms. DAVENPORT: It's interesting the Crown served on me this morning a memo that was sent to their expert in England this morning, and one of the questions that was asked by the Crown was whether the result was expressible in any other way: 'If this result is not expressible in that kind of way, that is the 99.9, we will need to be able to justify the result … distinction.' CROWN PROSECUTOR: They are not my words, they are the words of my instructing solicitor. HIS HONOUR: We all know the authorities on this point, and what is the correct way of expressing these things, and what is an incorrect and misleading way of expressing these things, I am sure the jury is not going to be misled by any of it." 72 The Crown submitted that when understood in its context, defence counsel did not take issue with the "subliminal impact of the raw percentage figures". 73 I do not think this argument has merit. Although perhaps inelegantly expressed, counsel clearly raised a concern as to the use of a percentage in relation to the DNA evidence. In any event if it was necessary I would grant leave pursuant to rule 4 to raise the issue. 74 The substantive response of the Crown to the appellant's submission commenced with a consideration of Ms Pineda and Dr Buckleton's evidence. The Crown emphasised that in GK, the results of the statistical analysis gave a relative chance of paternity, i.e. the likelihood that the complainant had GK's child. The evidence was that "for [the complainant] to produce a child with this genetic pattern, it is about 500 million times more likely that she would do that with [GK] as the father than she would another random person that was the father". That was not what occurred in the present case. It was submitted that because in the present case Ms Pineda gave evidence of a percentage of the population who could not have contributed to the DNA profile, not who was likely to have contributed to it, the present case could be distinguished. The remaining 0.1% of the population comprised a large population who could have been the source of the hair. 75 With respect to the decision in Galli, the Crown submitted that the difficulty in that case arose from the fact that evidence which could have been given by the relevant expert and was given on the voir dire was not given during the trial. That expert indicated that he was saying that the relative chance of paternity was 99.999% but he was not saying that the accused was 99.999% likely to be the father. 76 In the present case the trial judge instructed the jury in the following terms: "When you are considering the evidence of these experts you should bear in mind that none of them are saying that the hair found on Ms Bayrak' thumb is the accused's. None of them are saying that the mitochondrial DNA profile found in the hair establishes that it definitely came from him. What you have, at its highest, is evidence that you may or may not accept establishes that the profiles are the same and that then goes on to say that because of this the accused cannot be excluded as a source of the hair. It does not say he is necessarily, but it does say that it is possible that it came from him." 77 The Crown emphasised that in the passages in the summing-up where his Honour referred either to 99.99% or 98% his Honour did not say that this was the percentage chance that the hair belonged to the appellant. Rather his Honour said that 99.99% or 98% of the population could be excluded from having the same DNA profile. An examination of the transcript of the summing-up confirms that this was the approach taken by the trial judge. The Crown emphasised, correctly, that no redirection was sought from his Honour warning the jury about reasoning that there was a 99.99% chance that the hair came from the appellant. It was submitted that this was explained because the jury were clearly directed that the DNA evidence did not establish that the hair was the appellant's, but simply that he could not be excluded as a source of it. Further, that it was not suggested (as it was in GK and Galli) that the percentage figures connoted the likelihood that it was the appellant's hair. Rather, it was made clear that they simply represented those who could not have contributed to the DNA profile. The Prosecutor's Fallacy 78 The Prosecutor's Fallacy is a failure in statistical reasoning that invites the jury to assume that a DNA statistic provides a statistical likelihood that the incriminating DNA belongs to the suspect and therefore that the suspect is guilty. The mathematical basis for the fallacy lies in incorrectly disregarding the fact that the suspect's DNA sample was selected for analysis because of other (non-DNA) evidence said to link him or her to the crime. In the English case of Doheny and Adams (1997) 1 Cr App R 369; [1996] EWCA Crim 728 at 372–373 Phillips LJ (as he then was) explained the fallacy as follows (at 372 - 373): "It is easy, if one eschews rigorous analysis, to draw the following conclusion: (1) Only one person in a million will have a DNA profile which matches that of the crime stain. (2) The defendant has a DNA profile which matches the crime stain. (3) Ergo there is a million to one probability that the defendant left the crime stain and is guilty of the crime. … Taking our example, the prosecutor's fallacy can be simply demonstrated. If one person in a million has a DNA profile which matches that obtained from the crime stain, then the suspect will be 1 of perhaps 26 men in the United Kingdom who share that characteristic. If no fact is known about the defendant, other than that he was in the United Kingdom at the time of the crime the DNA evidence tells us no more than that there is a statistical probability that he was the criminal of 1 in 26. The significance of the DNA evidence will depend critically upon what else is known about the suspect. If he has a convincing alibi at the other end of England at the time of the crime, it will appear highly improbable that he can have been responsible for the crime, despite his matching DNA profile. If, however, he was near the scene of the crime when it was committed, or has been identified as a suspect because of other evidence which suggests that he may have been responsible for the crime, the DNA evidence becomes more significant. The possibility that two of the only 26 men in the United Kingdom with the matching DNA should have been in the vicinity of the crime will seem almost incredible and a comparatively slight nexus between the defendant and the crime, independent of the DNA, is likely to suffice to present an overall picture to the jury that satisfies them of the Defendant's guilt." 79 This passage was cited with approval by Mason P in GK at [48]. 80 In Australia, a jury may use DNA profiling evidence to compare the likelihood of the DNA sample having come from the defendant with the likelihood of it having come from a randomly selected member of the community. Evidence presented in that form is called a "likelihood ratio" or "relative likelihood". It was recently explained by Sulan J in R v Carroll [2010] SASC 156 at [28]: "A DNA profile taken from an evidence sample is compared to a sample provided by an individual. If the DNA profile taken from an evidence sample does not match the DNA profile of a person, then that individual can be conclusively excluded as being the source of the DNA from the evidence. If the profiles of the evidence sample and an individual do match, then there are two competing possibilities to explain the matching DNA profiles. The first possibility is that the DNA profile match has occurred because the DNA has originated from the person in question. The second possibility is that the DNA match has occurred by chance. That is, that there is someone else in the population who just happens to have the same DNA profile as the person in question. The probability of the evidence (ie probability of the sample matching the known or unknown person) given each of these scenarios is calculated using statistical analysis. A population database is used to provide an indication of the relevant prevalence of each of the alleles that were observed in the population. The given ratio of the two probabilities is called the likelihood ratio." 81 In R v Karger (2002) 83 SASR 135 the following likelihood ratio was approved by the South Australian Court of Criminal Appeal at [147]: "the crime stain is greater than 90 billion times more likely to match the profile of the accused if he left it, rather than if an unknown, unrelated person left it." 82 Clearly, as Doyle CJ observed at [15], evidence in this form must be treated with caution. " For the purposes of expressing the likelihood ratio, two factual hypotheses are compared. One is the prosecution hypothesis that the appellant was the source of the bloodstain and of the incriminating DNA. The other hypothesis is the defence hypothesis that an unrelated person is the source of the bloodstain and of the incriminating DNA. The statistical evidence calculates the probability of the occurrence, in either case, of the match found between the incriminating DNA and that of the appellant, and expresses the competing probabilities as a ratio." 83 While DNA evidence has been permitted to take the form of a likelihood ratio, it would be an example of the fallacy for the jury to transpose that ratio into the odds of a defendant's guilt. The fallacy is well illustrated by the following example: "If the experts say there is a one in 200,000 chance of DNA coming from a randomly selected member of the community this [does] not mean there is [a] 199,999 to one chance that it came from the accused. If one in 200,000 people match that DNA and there are 20 million people in Australia then one might reasonably opine that there are 100 people in Australia that could have left that DNA. There is then one chance in 100 that the DNA came from the accused." (A Ligertwood, "Avoiding Bayes in DNA Cases" (2003) 77 ALJ 317 at 319). 84 Although it was not submitted that the Prosecutor's Fallacy was committed in the present case, if the jury accepted that the likelihood of obtaining a match between the DNA profile taken from the hair and the DNA profile from another randomly selected member of the Turkish community was 1 in 50, it would have been wrong for them to conclude that the appellant was 50 times more likely than not to have been the source of the hair, and therefore 50 times more likely than not to have been guilty. I am satisfied that his Honour adequately instructed the jury to avoid this approach in his summing-up. 85 The question raised in ground 1 of this appeal is whether the presentation of the DNA evidence in the form of exclusion percentages misled the jury which would not have been the case if the analysis was confined to the corresponding random occurrence ratios. Research on the persuasive power of probabilistic formulations 86 There are various equally mathematically valid ways of expressing the same DNA statistic. Using a random occurrence ratio of 1 in 1000, and taking the population of Australia as 21 million, these include: 1. 1 in 1000 people would be expected to have the DNA profile found in the hair specimen. 2. 999 out of 1000 people would not be expected to have the DNA profile found in the hair specimen 3. 0.1% of people would be expected to have the DNA profile found in the hair specimen. 4. 99.9% of people in Australia would not be expected to have the DNA profile found in the hair specimen. 5. 21,000 people in Australia would be expected to have the DNA profile found in the hair specimen. 6. 20,979,000 people in Australia would not be expected to have the DNA profile found in the hair specimen. 87 Formulations 1 and 2 express random occurrence rates as frequencies. Formulation 3 expresses what might be termed an "inclusion percentage". Formulation 4 is an exclusion percentage. Formulations 5 and 6 simply transpose the random occurrence rate onto an actual population, in this case the population of Australia. Formulations 1, 3 and 4 were adopted in the course of the trial. Formulation 5 was also used, although the exemplar population was not that of Australia but rather of a football stadium. The appellant complains only about the use of formulation 4. 88 All formulations express the same DNA statistic, but plainly enough, they are not all equally compelling. To say, "99.9% of people in Australia would not be expected to have the DNA profile found in the hair" appears more damning than to say, "1 in 1000 people would be expected to have the DNA profile found in the hair". Counsel for the appellant was alive to this issue during the voir dire: "HIS HONOUR: If I understand correctly, I thought you said at the outset of this that you didn't object to the opinion being expressed that the accused was not excluded. You didn't object to her saying the profile has been observed in one in 4839 individuals. DAVENPORT: That's correct. HIS HONOUR: I take it from that you would not object to her expressing it in more conservative terms of one in 1600. DAVENPORT: No, Your Honour HIS HONOUR: But you object to 99.99 [sic] per cent. DAVENPORT: Because it has a connotation that is very different to the reality. HIS HONOUR: It means the same thing. DAVENPORT: It may very well, your Honour, but it has a very different impact upon a jury's consideration. That is 99.9 per cent. Now, I must say I had to get reassured by Dr Buckleton that is one in 16000 [sic] because it sounds as though you are including such a huge proportion of the population when you put in figures of one in 1600. It is a very different proposition. That is in the ear of the listener, you can see, for example, 16000 people and you have ten of them who may have the same mitochondrial profile. HIS HONOUR: If the jury received evidence only in terms of that, of one in 1600, they could do this exercise for themselves. They could say if it was one in a hundred, that is one per cent. If one in thousand, that is .1 of per cent. One in 1600 is less than .1 of a per cent." 89 JJ Koehler has written widely on the differing persuasive power of probabilistic formulations (see, eg, JJ Koehler, "On Conveying the Probative Value of DNA Evidence: Frequencies, Likelihood Ratios, and Error Rates" (1996) 67 University of Colorado Law Review 859; "The Psychology of Numbers in the Courtroom: How to Make DNA-Match Statistics Seem Impressive or Insufficient" (2001) 74 Southern California Law Review 1275; JJ Koehler, A Chia & JS Lindsey, "The Random Match Probability (RMP) in DNA Evidence: Irrelevant and Prejudicial?" (1995) 35 Jurimetrics 201). In a 2001 article entitled "The Psychology of Numbers in the Courtroom: How to Make DNA-Match Statistics Seem Impressive or Insufficient" he observes: "The bottom line appears to be this: When the statistic is framed in the language of probability (e.g., 0.1%) in a way that highlights a particular suspect's chance of matching by coincidence, it tends to be persuasive. But when the statistic is framed in the language of frequencies (e.g., one in one thousand) in a way that highlights the chance that others will match by coincidence, it is much less persuasive. Similarly, DNA match statistics that target an individual suspect are more persuasive than the equivalent statistic that targets a broader population." 90 In the same article, Koehler discusses his "exemplar cuing theory" which holds that the more easily a juror can envisage other coincidental matches with the crime scene DNA, the less impressed he or she may be by the DNA profiling evidence. "[T]he perceived probative value of a statistical DNA match (and, by extension, other forensic match evidence) depends on the ease with which triers of fact can imagine examples of others who would also match the DNA profile. When triers of fact find it hard to imagine examples of others who might match by chance, the evidence will be treated as compelling proof that the matching suspect is the source of the recovered DNA evidence. But when such matches are easier to imagine, the evidence will seem less compelling" (at 1280). 91 Koehler's theory derives from behavioural science literature suggesting that individuals are more likely to regard an outcome as probable when they can easily call to mind instances or examples of that outcome (see, eg, RB Korobkin & TS Ulen, "Law and Behavioral Science: Removing the Rationality Assumption from Law and Economics" (2000) 88 California Law Review 1051, 1088; A Tversky & D Kahneman, "Availability: A Heuristic for Judging Frequency and Probability" (1973) 5 Cognitive Psychology 207). 92 Koehler usefully identifies two factors which may affect the persuasive potential of DNA statistics. The first is the way the statistic is targeted. A DNA statistic targeted towards the suspect is more compelling than one targeted towards a large reference population. This is not a matter arising in the present case, however it may be an issue where evidence in the form of a likelihood ratio is sought to be adduced at a trial. Koehler writes: "To illustrate, suppose a juror hears that there is a one in one hundred thousand chance that a defendant who is not the source of the genetic evidence would match by coincidence. The juror will probably not give much thought to the possibility that the match was a coincidence because one in one hundred thousand is very close to zero. This might constitute overwhelming proof of identity in the juror's mind. Now consider a juror who hears that one in every one hundred thousand people in, say, Houston who are not the source will match coincidentally. This juror may reason as follows: 'If one in every one hundred thousand people in Houston match, then dozens of people from Houston would match, as would thousands of others throughout the U.S. For this reason, I do not find the evidence convincing as proof of identity" (at 1283). 93 The second factor is the manner in which the statistic is framed. Koehler suggests that formulations which contain frequencies (e.g. 1 in X people) invite jurors to adopt a broad, "outside" view in which other coincidental matches are more easily called to mind (at 1284). On the other hand, formulations containing probabilities (e.g. 99.9% of the population) encourage a narrower outlook and discourage consideration of coincidental matches. 94 An important qualification to Koehler's theory applies where the random occurrence ratio is exceptionally small: "Where DNA incidence rates are so small that exemplar generation is difficult under any target and frame formulation, the way in which the statistics are presented is less significant" (at 1297). 95 Koehler concludes (at 1299-1300): "Some may find it alarming that the wording of a statistic can affect the chance of a defendant's guilt in the eyes of a juror. Yet this observation is consistent with a large body of research on the psychology of numbers. Much of this research indicates that people think heuristically rather than probabilistically. That is, when presented with quantitative information, we do not perform algebraic computations and arrive at solutions by using tenets of logic and probability theory. Instead, we evaluate quantitative evidence via mental shortcuts and other rules of thumb. In the case of DNA evidence, the ease with which we can imagine scenarios or examples of a match other than the suspect may be the heuristic of choice." The present case 96 The present case highlights the caution with which ostensibly objective evidence must be approached by the Crown. 97 Because certain forms of expressing the statistics carry greater persuasive potential than others, it is easy to see why the Crown in any trial may wish to present DNA statistics in a manner that targets the offender and expresses a percentage rather than a frequency. That is what occurred in the present case. Once the exclusion percentage had been adduced in the course of Ms Pineda's evidence, it was left to defence counsel to undo the damage in cross-examination. That took the form, unsurprisingly, of an example framed within the terms of formulation 5 above: "Q. So, on one day, with 16,000 people in the football stadium, you might have 100 people with the same mitochondrial DNA as Mr Aytugrul? A. And on other days you might have none." 98 DNA profiling is an empirical scientific method. In an accusatorial system of justice, the duties of the Crown demand that all evidence and, in particular, complicated expert evidence be presented fairly to the accused. Where it strongly implicates the accused, it will speak for itself. Where it does not, the Crown should not have the advantage of the "subliminal impact" of statistics to enhance the probative value of the evidence. To do so may come at the cost of a fair trial. 99 To my mind and for the reasons identified in GK and Galli, his Honour should have excluded the exclusion percentages from the evidence, all of which invited a subconscious "rounding-up" to 100. It was not sufficient for his Honour to warn the jury against the potential misuse of the percentages. The exclusion percentage figures were too compelling. To my mind his Honour's directions would not have eliminated the risk of unfair prejudice to the appellant (GK at 341; see also Mason P at 331). That prejudice substantially outweighed the probative value of the evidence. 100 There is no doubt that DNA profiling evidence will continue to figure with increasing prominence in police investigations and criminal trials across Australia. 101 Contemporary difficulties with DNA evidence tend to relate to its intelligibility rather than its integrity (See, eg, G Hazlitt, "DNA Evidence: Unravelling the Strands" (2002) 14 Judicial Officers Bulletin 9, 66; JJ Koehler, "On Conveying the Probative Value of DNA Evidence: Frequencies, Likelihood Ratios, and Error Rates" (1996) 67 University of Colorado Law Review 859, 884; R v Gallagher [2001] NSWSC 462). As Findlay and Grix write: M Findlay and J Grix, "Challenging Forensic Evidence? Observations on the Use of DNA in Certain Criminal Trials" (2003) 14 Current Issues in Criminal Justice 269 at 274: "the presentation of evidence concerning Deoxyribo Nucleic Acid (DNA) requires demystification of a language of science unfamiliar to lawyers and to juries in particular. In addition, the language of analysis expects the lawyer and the juror to venture into the specialist and foreign realm of genetics and statistics, which in its clearest representation can be confusing. A central focus of confusion is the relevance of the probability ratio that DNA analysis provides, and its translation into evidence of identification (see R v Gali [2001] NSWCCA 504)." 102 Issues related to juror comprehension of statistical evidence are not confined to DNA evidence. Indeed, the basis of many forensic sciences, including DNA profiling, the comparison of hair, paint, glass, fibres, particles and bite marks, lies in statistical theory (WC Thompson, "Are Juries Competent to Evaluate Statistical Evidence?" (1989) 52 Law and Contemporary Problems 4 at 13). In 1983 one commenter observed (referring to the United States), "our criminal justice system is now at the threshold of an explosion in the presentation of mathematical testimony" (R Jonakait, "When Blood is their Argument: Probabilities in Criminal Cases, Genetic Markers, and Once Again, Bayes' Theorem" (1983) University of Illinois Law Review 369 cited in WC Thompson, "Are Juries Competent to Evaluate Statistical Evidence?" (1989) 52 Law and Contemporary Problems 4 at 10). The response to the difficulty with the intelligibility of DNA evidence is not to banish all statistical evidence from the courtroom but to rationally determine the probabilistic formulations which are appropriate for use in a criminal trial. 103 It is not necessary for me to determine the admissibility of likelihood ratios, as these were not in issue in this appeal. However, in my view, it will continue to be a matter for the trial judge to assess whether the probative value of a likelihood ratio outweighs its prejudicial effect. Where exceptionally low odds are presented to the jury (such as 1 in 10 billion), a similar problem may occur to that which occurs in the use of high exclusion percentages. Jurors may incorrectly assimilate a low likelihood ratio with a 0.0% chance that the crime scene DNA came from anyone but the accused. That is not an example of the Prosecutor's Fallacy, because it involves an assimilation of two statistics rather than the incorrect use of one, but it comes close. Ground 2: The verdict of the jury should be set aside on the ground that it is unreasonable, or cannot be supported, having regard to the evidence. 104 The appellant submitted that the appellant's conviction was not supported by the evidence. On the other hand the Crown submitted that even if there were problems arising from the DNA evidence which meant that it should have been excluded the circumstantial case was so strong that this Court should exercise the proviso and dismiss the appeal. 105 The test to be applied by this Court when it is submitted that a verdict is unreasonable or cannot be supported having regard to the evidence is settled. In Fleming v R (2009) NSWCCA 223 I said at [61]: "The approach which this Court must adopt when this ground of appeal is raised has been considered on many occasions. I discussed it briefly in Dodds v R (2009) NSWCCA 78 where I said: 'The appellant submitted that the verdict of the jury was unreasonable or cannot be supported by the evidence: Criminal Appeal Act 1912 s 6(1). The correct approach to this question by an appeal court has been considered by the High Court on a number of occasions: M v R (1994) 181 CLR 487; MFA v R (2002) 213 CLR 606 at 614 at 615. I summarised the relevant principles in R v Habib [2005] NSWCCA 223 and discussed them in Kaliyanda v R [2007] NSWCCA 300. The court's task is to consider whether upon the whole of the evidence it was open to the jury to be satisfied beyond reasonable doubt that the accused was guilty. The joint judgment in MFA offered the following by way of guidance: 'The majority in M pointed out that '[i]n most cases a doubt experienced by an appellate court will be a doubt which a jury ought also to have experienced'. In such a case of doubt, it is only where the jury's advantage of seeing and hearing the evidence can explain the difference in conclusion about the accused's guilt that the appellate court may decide that no miscarriage of justice has occurred: If the evidence, upon the record itself, contains discrepancies, displays inadequacies, is tainted or otherwise lacks probative force in such a way as to lead the court of criminal appeal to conclude that, even making full allowance for the advantages enjoyed by the jury, there is a significant possibility that an innocent person has been convicted, then the court is bound to act and to set aside a verdict based upon that evidence' [56]. In their conclusion their Honours acknowledge that it is not uncommon for some aspects of the evidence in a trial to be less than wholly satisfactory although juries are usually well able to evaluate conflicts and imperfections of evidence. "In the end, the appellate court must ask itself whether it considers that a miscarriage of justice has occurred authorising and requiring its intervention" [96]." 106 I have previously summarised the Crown case. It is important when considering a circumstantial case to have regard to the totality of the relevant circumstances (Burrell v R [2009] NSWCCA 193 at [53]-[64]). The appellant's submissions tended to deal with each issue separately rather than consider the evidence as a whole, and raise some matters which in my view were of no importance. 107 The relevant circumstances include the fact that the appellant had very strong feelings for the deceased throughout 2005. He was prepared to express them through a poem published in a newspaper. I am also persuaded that the appellant was effectively "stalking" the deceased. He sought out her address and watched and followed her, making it plain to her on occasions that this was what he was doing. Although he had formed a new relationship, he was plainly upset by the termination of his relationship with the deceased and, although seeking to renew that relationship, was also driven to both annoy and do harm to her. The appellant was not truthful with the police with respect to his contact with the deceased. Of significance are the lies he told in an attempt to hide the fact that he obtained details of her address from Mr Uluc. 108 The depth of the appellant's confused emotions in relation to the deceased is revealed by the complaints he made to Centrelink about her three days before her death. There were undoubtedly attempts at harming her. It is known that on the day before her death he telephoned the deceased twice from a telephone number that was not known to the deceased, although the content of these telephone calls is unknown. 109 I accept the Crown's submission that the appellant's contact with the deceased in the weeks leading up to her death involved an element of intimidation. The text messages sent on 11 November, one of which concluded with the words, "What can I say, I pray for your health" can be interpreted as threats. 110 For my part I would discount the evidence of vandalism of the deceased's motor vehicle. After the deceased left Merrylands the evidence indicated that there were no further problems in the units. There was nothing to connect those problems with the appellant. 111 To my mind, Mr Tunc can be excluded as the killer. The deceased spoke on her mobile phone with Mrs Bas between 8.04.45 at 8.06.13 pm. During this time Mr Tunc's wife rang on his mobile phone but he did not answer. The deceased told her half-sister that she would be shortly on her way to collect her daughter and gave no indication that there was any problem. Mr Tunc telephoned his wife at 8.13 pm, only seven minutes after the deceased terminated the call to her half-sister. I find it inconceivable given the severity of the wounds inflicted on the deceased that a difficulty between her and Mr Tunc could have emerged after 8.06, he then stabbing her many times, telephoning his wife at 8.13 without giving any indication of a problem, following which he went with his wife to a friends' party. There was no suggestion of any blood being found on Mr Tunc, his clothes or his motor vehicle. 112 Furthermore, at the relevant time, the neighbour Mr Jiang saw a male knocking at the deceased's door but there was no answer. There was evidence that the door was "jemmied." The fact that the "jemmy marks" were not noticed that night by Mrs Bas or Mr Tunc is not surprising. The deceased was killed at about 9 pm and her half-sister was first there around midnight. She said that as she moved in the stairwell she saw Mr Tunc approaching the units and not leaving the flat. She rang her husband at 1.07 am. This sequence of events is consistent with Mr Tunc having left the party he was at at 12.30 am and then going out to see if the deceased was alright. There is nothing to suggest that Mr Tunc would have made the jemmy marks on the door late at night. 113 The appellant emphasised that Mr Tunc had made a statement which suggested that he was aware that the deceased had been killed with a knife at a time when the police had not released this detail. The Crown's response was to emphasise that Mr Tunc said that he first thought that the deceased had cut her wrists, gaining this impression from the sequence of events which he observed when he went to her premises around midday. He said he saw the police and a locksmith and rang Mrs Bas who was crying. He then spoke to Mr Bas who on Mr Tunc's version said to him, "she's got blood all over her" and something like "Were you screwing her? You killed her." The jury must have accepted this explanation, which seems to me to be reasonable. The jury also received evidence from Mr Tunc that he considered suicide. Although it was possible that he contemplated this course because he was the killer, it was equally explicable as an act of despair at her death and an expectation, as has happened, that their affair would be revealed. 114 The appellant emphasised that apart from the hair which, when analysed, could have been the appellant's, DNA consistent with that of Mr Tunc was found on the deceased's chin. I do not believe this to be of any significance given that they were having an affair and he admitted that he had been with the deceased until shortly after 8 pm. 115 Furthermore, Mr Tunc made no effort to disguise the fact that he had been with the deceased on the evening of 26 November. This was surprising. It is plain that he was concerned that his relationship with the deceased would be discovered which explains some of his movements when he became concerned about the deceased's welfare. However, if he was responsible for her death it would be expected that he would deny that he had been at the deceased's premises at about the time she was killed. He had no reason to believe that anyone would have been able to establish that he had been there. Furthermore if he was responsible for her death it is most unlikely that he would have returned late at night out of concern for her welfare. If he knew he was the killer it would be more likely that he would seek to distance himself from the deceased's premises. 116 Eliminating Mr Tunc as the killer does not mean that the appellant must have committed the crime. It could have been another person although there is no evidence to suggest who that might have been. But the question remains whether the admissible evidence proved the Crown case against the appellant. 117 The appellant submitted that there were five matters which were profound problems in the prosecution case, making the appellant's conviction unreasonable: First, it was submitted that each of the three lies relied upon by the Crown as showing consciousness of guilt were explicable on another basis; Secondly, it was submitted that the campaign of damage at the flat of the deceased could not safely be linked to the appellant; Thirdly, it was submitted that the DNA evidence did not strongly implicate the appellant, yet it was the only evidence that directly connected the appellant to the crime scene; Fourthly, it was submitted that the alibi of the appellant cannot be discounted beyond reasonable doubt; and Fifthly, it was submitted that the involvement of Mr Tunc could not be ruled out. DNA consistent with his profile was found on an abrasion under the deceased's chin. Mr Tunc's seeming knowledge about the crime scene cannot be effectively explained on the Crown case. There were also some indications that his relationship with the deceased was not as happy as he claimed. His suicide attempt was consistent with consciousness of guilt. 118 I accept that the lies may not have been told out of a consciousness of guilt. It is possible that the appellant's lies in relation to Mr Uluc may have been told out of a desire to protect his friend. I also accept that the earlier damage to the deceased's motor vehicle cannot with certainty be linked to the appellant. 119 In relation to the DNA evidence although the hair could have come from other persons, particularly Turkish persons, it confirms that the appellant could have been present at the time that the deceased was killed. There was no evidence that the appellant had previously been to the deceased's premises at Parramatta and if the hair was from the appellant it could only have been deposited at about the time of her death. 120 I accept that the discrepancy between the appellant's evidence and that of his mother is of little consequence, however the jury had to assess the strength of the alibi evidence. It is plain that they did not accept it. Conclusion 121 I have carefully considered all of the evidence. Although the surrounding circumstances raise significant suspicion that the appellant may have committed the offence his conviction is dependent on the DNA evidence. I accept that when that evidence is considered this was a significant Crown case. However, because, in my view, use of the exclusion percentages had the potential to overwhelm the jury I would order a new trial. In Weiss v R (2005) 224 CLR 300; (2005) HCA 81 at [45] the High Court said: "Likewise, no single universally applicable criterion can be formulated which identifies cases in which it would be proper for an appellate court not to dismiss the appeal, even though persuaded that the evidence properly admitted at trial proved, beyond reasonable doubt, the accused's guilt. What can be said, however, is that there may be cases where it would be proper to allow the appeal and order a new trial, even though the appellate court was persuaded to the requisite degree of the appellant's guilt. Cases where there has been a significant denial of procedural fairness at trial may provide examples of cases of that kind." 122 To my mind that approach is apposite to the present case. 123 I have already indicated that in my view ground 1 should be upheld. Orders 1. Appeal upheld. 2. The appellant's conviction quashed. 3. Order a new trial. 124 SIMPSON J: I have read in draft the judgment of McClellan CJ at CL. For the reasons that follow, I am of the view that Ground 1 of the appeal (concerning the manner in which the DNA evidence was expressed) ought to fail. I agree that Ground 2 (that the verdict was unreasonable and cannot be supported on the evidence) ought to fail. I set out below my reasons for coming to that view. Ground 1: the DNA evidence 125 The circumstances have been set out at length in the judgment of McClellan CJ at CL; I propose to refer to some additional matters. It is unnecessary to refer to the Prosecutor's Fallacy, because it is not here asserted that that infected the approach to the DNA evidence. 126 The DNA evidence in this case was unusual in this respect. Commonly, the sample from which DNA is drawn for testing against that of a suspect is taken from the victim or from the crime scene. There is a high degree of probability, sometimes tantamount to certainty, that the source of the DNA sample (sometimes called "the crime stain") is the perpetrator of the crime: that is, that once the source of the crime stain is identified, so is the offender. 127 Illustrations of this are to be found in two decisions of this Court to which reference will later be made (R v GK [2001] NSWCCA 413; 53 NSWLR 317 and R v Galli [2001] NSWCCA 504; 127 A Crim R 493). There are many others. DNA taken from victims of sexual assault, from blood left at the scene of a robbery, from clothing known to have been worn by an offender, and from a cigarette smoked by an offender, are further examples. The circumstances provide varying, but usually very high, degrees of probability that the source of the crime stain DNA is the person who committed the crime. But it is not always so, and it is not so in this case. 128 It could not safely be concluded, from the DNA evidence alone, that the person from whom the hair originated was Ms Bayrak's murderer. The most the DNA evidence could do, in the circumstances of this case (assuming that the jury were satisfied that the expert evidence linked the appellant to the hair), was to constitute one relevant circumstance in a circumstantial case. 129 The relevant ground of appeal is: "Ground 1: a miscarriage of justice occurred because of the prejudicial way in which DNA evidence was expressed to the jury." It is apparent from the appellant's submissions, however, that the ground was presented as an admissibility ground: the appellant challenges the admission of one aspect of the DNA evidence. 130 The police investigation involved extensive DNA testing. DNA evidence was given by four witnesses, three of them called by the Crown. This does not imply that there was unnecessary repetition, nor any serious conflict. The three DNA witnesses called by the Crown, on the whole, dealt with different aspects of testing. Dr John Buckleton, the DNA expert called on behalf of the appellant, essentially commented upon the evidence of the Crown witnesses. 131 I propose to outline the evidence given by the witnesses, particularly so far as it is relevant to this ground of appeal. I will not do so in the same order in which the witnesses were called. 132 Ms Michelle Franco is a forensic biologist employed in the Division of Analytical Laboratories in NSW. 133 Over a period of time Ms Franco received from investigating police a variety of items for scientific testing. These included many items, or swabs, from inside or outside of Ms Bayrak's home unit; clothing from Ms Bayrak, from Erdal Tunc and from the appellant; and entomological and body samples from Ms Bayrak. One of the items was a cigarette butt found in front of the entrance landing to the unit. There were also two buccal swabs provided by the appellant, one from Erdal Tunc (with whom Ms Bayrak was conducting an illicit extra-marital relationship), and others from other individuals associated with Ms Bayrak. Ms Franco performed nuclear DNA testing on many of these items. 134 The clothing provided by Mr Tunc and by the appellant was subjected to preliminary testing for blood, with no result – that is, on no item of clothing provided by either man was the presence of blood identified. 135 DNA that convincingly linked Mr Tunc to the cigarette butt was identified. Mr Tunc also could not be excluded as the male source of DNA found on a fingernail clipping from Ms Bayrak, nor from swabs taken from her neck. (Since it is known that Mr Tunc had visited Ms Bayrak earlier on the evening of her death, these last findings are hardly surprising and of no real significance.) 136 The appellant was excluded as the male contributor to DNA identified in swabs from Ms Bayrak's neck and on a fingernail clipping. 137 In sum, so far as the appellant is concerned, on no item from Ms Bayrak's body or unit was there initially identified any DNA consistent with having originated from him. 138 In the course of the examinations at the laboratory, a colleague of Ms Franco's observed, with one of the fingernail clippings, a black hair that had been found at post mortem. It was stored in a plastic jar in a sealed plastic bag. 139 Dr Susan Bennett was a scientific officer with the NSW Police Force. She had a particular professional interest in the examination of textile fibres, hair and textile damage. On 19 December 2005 the hair was transmitted to her for examination. Dr Bennett identified the hair as a human body hair, possibly an arm hair, and unlikely to have originated from Ms Bayrak. The hair had no root (indicating that it was probably a hair that had been shed as part of the normal bodily process). The fact that it had no root was significant because it made the hair unsuitable for nuclear DNA analysis of the kind undertaken by Ms Franco. Dr Bennett therefore recommended that it be submitted for mitochondrial DNA analysis, a process that is not available in Australia or New Zealand. 140 It was, accordingly, submitted to a private DNA laboratory in Texas, USA. There it was examined by Ms Gina Pineda. Ms Pineda has relevant expertise in the analysis of mitochondrial DNA. Also submitted to her, for comparison purposes, was a reference sample of saliva taken from the appellant, and one from Ms Bayrak. 141 Ms Pineda gave to the jury a fairly lengthy exposition of the process of mitochondrial DNA examination. In conducting her examination, Ms Pineda followed guidelines issued by a "Scientific Working Group on DNA Analysis Methods" ("SWGDAM"). These guidelines, inter alia, deal with reporting the results of DNA testing. 142 The first conclusion Ms Pineda reached (or the first she expressed) was that the DNA on the hair was not consistent with the DNA from Ms Bayrak. Ms Bayrak was, accordingly, excluded as the donor of the hair. 143 The second conclusion expressed by Ms Pineda was that the profile of the DNA on the hair was consistent with the profile of the DNA sample provided by the appellant. The appellant (and all his maternal relatives) could therefore not be excluded as the source of the hair. In reaching this conclusion, Ms Pineda acknowledged that the DNA profile from the hair was not identical with that from the appellant's saliva example. There was one point of divergence. She regarded the discrepancy as immaterial, and of no significance. She rejected the possibility that the discrepancy might suggest that the hair and the saliva came from two different individuals. (This was a conclusion with which Dr Buckleton disagreed.) The discrepancy between the two DNA samples, and the conflict between Ms Pineda and Dr Buckleton, play no part in this appeal. Accordingly, while recognising that the two samples were not truly identical, it will be convenient to refer to the DNA profiles under consideration (that on the hair, and that from the appellant's saliva sample) as a single profile. 144 Ms Pineda's second conclusion called for a further, statistical, analysis; in this case Ms Pineda performed a "counting method". This involved searching an existing database in order to ascertain the frequency with which the DNA profile occurs in the database. The database she used consists of DNA profiles of 4839 individuals of different ethnic backgrounds. All individuals were North American. No Australians are represented and, to Ms Pineda's knowledge, no Turks. It is known as the SWGDAM database. 145 Initially, Ms Pineda obtained a nil result – that is, she concluded that that profile did not occur at all in the DNA of any of the 4839 individuals whose DNA constitutes the SWGDAM database. However, after considering a report provided by Dr Buckleton, Ms Pineda re-searched the database and found that there was indeed one match – that is, that DNA profile occurred once in the 4839 individuals there represented. 146 The recognised process then involves the application of what is called a "95% confidence interval" (the terminology used varies, but 95 percent is constant). It is unnecessary to explain this (and Ms Pineda did so only cursorily). No issue was taken with the correctness of this methodology. 147 On Ms Pineda's analysis, the result of the application of that test was that the DNA profile could be expected to be found in one in approximately 1600 individuals (in the world). Ms Pineda then expressed the same conclusion in another way. By inverting the frequency of the profile, she was able to state how many people (in the world) could be expected not to have DNA of that profile. This she called an "exclusion probability". Her conclusion was that 99.9 percent of the (world's) population could not have the DNA profile involved in this case. It is important to understand that this does no more than express the earlier conclusion in a different way. 148 Although I have paraphrased Ms Pineda's evidence, because of its importance to the first ground of appeal, I consider it appropriate to quote her (relevant) evidence precisely. She said: "Well, our final conclusion, after analysis of this data, the first thing we can see is that this hair is not consistent with the victim. So, the victim is excluded as being a donor of this hair found in her right hand thumbnail. When we make a comparison to the accused, Aytugrul, we can see that the profiles are consistent with each other. Therefore our conclusion is that Yusef Aytugrul and all his maternal relatives cannot be excluded as a donor to this hair sample." (T 277) 149 It was because of that conclusion that the statistical analysis was performed. As to that, Ms Pineda said: "When we performed this additional statistical analysis, we saw that applying the 95 confidence interval we can expect to see this profile most commonly; one in approximately 1600 individuals. So, what this means is if we were to take another database, we would expect to see this profile not more than one in 1600 individuals. If you take that frequency of the profile and inverted it, another way of looking at it is to say we expect to see it most commonly in one in 1600 people ; how many people could not have contributed to this profile. That is what we call an exclusion probability and we give this a percentage, a percentage of the population who could not have contributed to this profile, based on the database we are using and that percentage came up as a 99.9% of the population could not have this profile that we have encountered in this case." (T 278-279, bold added) 150 In further explaining the "95% confidence interval", the transcript records Ms Pineda as saying: "Well because we see it once in 4839 individuals in this particular database, we want to get an idea that to account for the size of this database. If we had a different database how often would we see this profile again? And to account for the size of the database we apply this confidence interval that tells us basically 95% of the time you're going to see this profile not more often than one in 1600 individuals. And that is the upper limit of that range. So that is the most conservative value that we can give you which gives the most benefit to the accused. It's the most frequently that we can expect to see this profile in the separate database. So the number does go down considerably, and again that is done to be on the conservative side, and what we give you is the upper limit of the interval, which means that that's the most common that you can expect to see this type. The 99.9% figure is just the inverse of that. One in 1600 is the frequency of occurrence – the expected frequency. The 99.9% is the percentage of the proposition (sic – population) who could not have contributed to that sequence. So it's the same information, just looking at it in two ways ." (T 285, bold added) 151 She said that the upper limit to the range of the frequency of occurrence of the profile was one in 1600; the lower limit was one in approximately 5000 individuals. 152 Ms Pineda was later asked to undertake a similar exercise by comparing the DNA on the hair with DNA provided in a sample by Mr Tunc. The result was that Mr Tunc was excluded as the source of the hair. 153 A further refinement on Ms Pineda's evidence was provided by Professor David Balding, whose specialty is in statistical genetics, and who gave his evidence by video link from London, where he is employed. 154 Using the SWGDAM database, and another (European) database ("EMPOP") (and drawing on an exercise undertaken by Dr Buckleton) Professor Balding considered two further possibilities. One was the probability that some other person, of European (but not Turkish) origin could have a DNA profile that matched the DNA on the hair; the other was the probability that a person of Turkish origin could have the same profile. Like Ms Pineda, he applied the 95 percent confidence interval. He concluded that, among non-Turkish Europeans, the frequency of the profile was one in 2000 or less. This was a "cautious upper limit", and he was "confident" that the true frequency would be less than that. 155 Turning then to the probability that a Turkish individual other than the appellant would have that DNA profile, Professor Balding said that he could be less confident about the frequency of the profile because there was less data available from Turkey, but that, knowing that the suspect was Turkish, it was reasonable to conclude that others of Turkish origin would have that profile. That is because all of his relatives should have the same maternal ancestor (and, presumably, would share genetic characteristics). Professor Balding did not explain the significance of the female line but did emphasise that the ancestral line would extend by five or ten generations such that the individuals who shared the profile would be unlikely to know they are related. 156 Again relying in part on Dr Buckleton's researches, Professor Balding was of the view that the DNA profile in question would be present in less than two percent (or one in 50) Turkish individuals. This, also, was a cautious assessment because of the size of the available database. Dr Balding thought the true match probability was likely to be less than two percent. (On those figures, the exclusion percentage would be 98 percent – that is, 98 percent Turkish people would be expected not to have that profile.) 157 Dr Buckleton is a New Zealand scientist qualified on behalf of the appellant. He works in forensic interpretation. 158 Dr Buckleton's evidence was the least easy to follow. He was somewhat critical of Ms Pineda's methodology, particularly her use of the SWGDAM database (because it is North American based). As I have mentioned above, he also disagreed with Ms Pineda's rejection of the possibility that the two samples of DNA emanated from different individuals. He did not, however, come to the view that the two samples did come from different individuals; he merely accepted the possibility that they did. He accepted that the appellant could not be excluded as the source of the hair. He did not quantify the possibility that the two samples may have come from different sources. 159 As against Ms Pineda's assessment that the DNA profile in question could be expected to appear in one in 1600 of the (world's) population, and also applying a 95 percent confidence level limit, Dr Buckleton came to a figure of one in 1000. That is, his conclusion was that that profile could be expected to be found in one in 1000 individuals in the world. In his evidence in chief, he did not give a corresponding exclusion figure. He did that, at the conclusion of his cross-examination, in response to questions asked by the trial judge. The exclusion figure which he assessed was 99.5 percent: that is 99.5 percent of the (world's) population would be expected not to have that DNA profile. As to the probability of a person of Turkish descent other than the appellant having the DNA profile in the hair, he agreed with Professor Balding that a conservative estimate of two per cent was reasonable. 160 Quite apart from the different ways of formulating or expressing the results of the DNA testing, given that there was only one match (or one possible match according to Dr Balding) on the SWGDAM data and none on the EMPOP database, the agreed position of the three experts was that mitochondrial DNA of the appellant was rare. 161 The above is a reasonable summary of the evidence. I have set it out at some length because it will be necessary to consider it in relation to both grounds of appeal. 162 It is now convenient to outline a few basic propositions. The DNA evidence was properly admitted. DNA evidence is virtually useless, or incomprehensible, to a jury unless interpreted: see R v Noll [1999] VSCA 164; 3 VR 704, per Callaway JA. The role of the expert witnesses, having presented the factual (scientific) results of DNA testing, is to explain to the jury what those results mean – that is, to outline any conclusions that can be drawn from them, and ultimately, to show how the results impact upon the likelihood that the person accused or suspected is the source of the sample said to link him or her with the crime, or the crime scene. (It is not the role of the expert witnesses to express their own conclusions with respect to that question: R v Doheny and Adams [1996] EWCA Crim 728; (1997) 1 Cr App R 369). The task is to give this explanation in a way that non-scientific members of a jury can comprehend. 163 Scientists have different ways of explaining the results of their findings to members of the non-scientific community. This is no more and no less than an educative function often performed by an expert witness in a criminal trial. 164 As McClellan CJ at CL points out at [86], there are various ways of expressing, in non-scientific terms, the conclusions to be drawn from DNA testing. These include: ● frequency ratio; ● inclusion percentage; ● exclusion percentage. Some formulations are likely to have greater impact than others. That merely means that some formulations have a greater educative force or persuasive appeal than others; or that some are more colourful, or more easily comprehended, than others. Provided that what is contained in the formulations is accurate, I see no reason to prefer one method of expression over another. By referring to accuracy I do not mean to suggest that evidence is inadmissible if it is incorrect: I mean that, provided the various means of expressing the conclusions correspond accurately with one another, there is no reason to prefer one over another. 165 It is to be remembered that evidence of this kind is often to be given by way of explanation to jurors, the vast majority of whom, it may be assumed, are unversed in scientific matters. It often has (as in this case) a considerable level of complexity. 166 It is, in my view, therefore beneficial (provided that, in translating one formulation to another, accuracy is not sacrificed) for the conclusions drawn to be expressed in ways that can readily be comprehended by juries. It is also not to be overlooked that this evidence is usually given orally (although increasingly frequently with the assistance of PowerPoint or other electronic or technological aids) to juries who are attempting to absorb it as it is given. That is a great deal to ask of a lay jury. If they can be assisted by easily understood explanations, such explanations ought to be available to them – always provided that the quest for comprehensibility does not lead to oversimplification or confusion, or to misleading, in the explanation. 167 In giving her "1 in 1600" evidence, Ms Pineda was expressing her conclusions in terms of the frequency ratio. In giving her "99.99%" evidence she was expressing exactly the same conclusions in terms of the exclusion percentage. Although Dr Buckleton arrived at different figures, he, similarly, expressed himself in terms of the frequency ratio and (prompted by the trial judge) the exclusion percentage. 168 Dr Buckleton disagreed with Ms Pineda's calculation that the DNA profile could be expected to be found in one in 1600 profile in the world. He did not disagree with the accuracy of the translation of that statistic to a 99.99 percent exclusion ratio. Indeed, in performing his own calculations, be implicitly adopted it. 169 As I understand the submission of the appellant in the present case: it was acceptable for one formulation (the frequency ratio) of the result of the testing to be put before the jury, but unacceptable for another formulation (the exclusion ratio) to be put before the jury. The appellant does not challenge the admissibility of the evidence framed in terms of the frequency ratio; what he challenges is the admissibility of the (same) evidence framed in terms of the exclusion percentage. 170 Again on the important proviso that the exclusion percentage formulation is an accurate translation of the scientific data (when expressed in terms of the frequency ratio), I am at a loss to understand why a jury ought not to be assisted by having the evidence couched in the language most likely to be meaningful to lay recipients. 171 For present purposes, the debate centres around the admission of the exclusion percentage formulation, which is not suggested to be mathematically or statistically inaccurate (if the starting point of one in 1600 is accepted). So far as I can glean, it was not contended that the evidence was, strictly, inadmissible. Rather, it was contended that it ought to have been excluded under the provisions of either s 135 or s 137 of the Evidence Act 1995. Exclusion of evidence under either of those sections depends upon a conclusion that the evidence in question might be unfairly prejudicial, might be misleading or confusing, or might cause or result in undue waste of time (s 135) or that its probative value is outweighed by the danger of unfair prejudice to the accused (s 137). Section 135 is in the following terms: " 135 General discretion to exclude evidence The court may refuse to admit evidence if its probative value is substantially outweighed by the danger that the evidence might: (a) be unfairly prejudicial to a party, or (b) be misleading or confusing, or (c) cause or result in undue waste of time." Section 137 is in the following terms: " 137 Exclusion of prejudicial evidence in criminal proceedings In a criminal proceeding, the court must refuse to admit evidence adduced by the prosecutor if its probative value is outweighed by the danger of unfair prejudice to the defendant." 172 I have earlier categorised the ground of appeal as "an admissibility ground". That is not strictly accurate. As I have indicated above, it seems to me that the real issue is the contention that the evidence ought to have been excluded under one of these provisions. 173 However, it seems to me appropriate to make the following observations. What is admissible or inadmissible is evidence. The evidence is the DNA evidence. It is also the expert evidence interpreting the DNA evidence. 174 What may be rejected under either s 135 or s 137 is evidence that is otherwise admissible, but which is assessed, by the trial judge, to have one or more of the characteristics mentioned in those sections – that is, under s 135, that its probative value is substantially outweighed by the danger that it might be unfairly prejudicial, be misleading or confusing, or cause or result in undue waste of time; and, under s 137, that its probative value is outweighed by the danger of unfair prejudice to the accused. (The only differences between s 137 and s 135(a) are that s 135 applies to both civil and criminal proceedings, and confers a discretion on the trial judge to reject the evidence in question where the danger of unfairness substantially outweighs probative value; s 137 applies to criminal proceedings only, and, where the condition is met – ie the trial judge assesses that the probative value of the evidence is outweighed by the danger of unfair prejudice to the accused – exclusion of the evidence is obligatory. 175 Here, the expert interpretation of the DNA evidence, whether put as a frequency ratio, or as an exclusion percentage, is exactly the same evidence. The content of the evidence does not vary. 176 It was not suggested that the evidence of exclusion percentages was misleading or confusing, or such as to cause or result in undue waste of time. That being the case, s 135(b) and (c) can be put to one side. Apart from the differences I have mentioned, s 137 and s 135(a), essentially raise the same issue. Both sections are concerned with the content of the evidence, and not the manner in which it is expressed. (Although I accept it is probably the case that a trial judge retains a general discretion to protect against unfairness, and that may extend to the manner in which evidence is presented.) 177 How can evidence expressed in one way be such as not to attract the operation of s 135 or s 137 (ie be such that its probative value is not outweighed by the danger of unfair prejudice) but, when expressed in another way, become unfairly prejudicial? 178 The debate appears to arise from the decision of this Court in R v GK [2001] NSWCCA 413; 53 NSWLR 317. 179 GK came before the Court by way of a Crown appeal under s 5A(2) of the Criminal Appeal Act 1912. 180 The issue in that case was the identification of the biological father of a child born to the stepdaughter of the accused. The Crown proposed to call DNA evidence as to the statistical likelihood that the accused was the father of the child. After a voir dire hearing, part of the evidence was excluded. The ruling was that evidence that testing did not exclude the possibility of paternity would be admitted; but evidence of "likelihood" and of "percentage" would be excluded. This was because: "… there is a very considerable risk that the jury notwithstanding any blandishments on my part may take the practical position of a punter and say this is not just a mere certainty but an absolute certainty." The trial judge ruled against the admission of "the arithmetical figures". The accused was acquitted. 181 This Court was therefore considering the correctness of the ruling to exclude the evidence. 182 The DNA evidence in that case was presented (on the voir dire) in two different ways. One was called "the paternity index". In that case, the evidence was that it was 220,000 times more likely that the accused was the biological father of the child than anybody else taken at random in the population. Contrary to the ruling at trial, this Court held that the evidence in that form was admissible, and ought to have been admitted. 183 The second way in which the evidence was presented was called "relative chance of paternity". Based on exactly the same information, the proposition above (the paternity index) was translated to "the relative chance of paternity [of the accused] was 99.99993%" (sic). 184 There was no suggestion that that figure was in any way inaccurate, or failed to reflect the paternity index figure. 185 Sully J declared himself "more ambivalent" about whether the evidence in that form ought to have been admitted. (I frame it in that way, because the focus of the decision in that case, as in the present, was not on the admissibility of the evidence, but on whether, notwithstanding its apparent admissibility, it ought to have been excluded under either s 135 of s 137 of the Evidence Act 1995.) 186 Sully J concluded: "99 I am more ambivalent about the percentage quantifications of the Relative Chance of Paternity. Mr. Goetz [the scientific expert] himself volunteered that he thought that the Paternity Index was the easier of the two statistical measures to understand, and I myself agree with that view. The bare percentage figures [the Relative Chance of Paternity] are so extraordinarily close to 100 per cent that there must be, at least, a real risk that jurors, even sensible jurors properly directed, would think that the difference between the percentage figures and 100 per cent was so miniscule that there could be no possibility that [the accused] was not the father; when it is scientifically uncontroversial that the one thing that DNA testing cannot do is to exclude absolutely any possibility, however statistically trifling, to the contrary." 187 In the following paragraph his Honour concluded that, pursuant to either s 137 or s 135, (and expressly not because of a fundamental dissatisfaction with the statistical sufficiency of the relevant databases), he would have excluded the raw percentage figures. He said he would have done this: "… because of the combined effect of the considerations: [1] that the residual risk of unfairness deriving from the subliminal impact of the raw percentage figures would have been, to me, both real and unacceptable; and [2] that the admission of the Paternity Index figures would have given the Crown every fair opportunity of persuading the jury to accept the Crown contention that there was no real possibility that [the accused] was not the father." 188 Mason P (at [60]) agreed, adding: "Mr Goetz himself recognised that the relative chance of parentage percentage was ' more complicated to explain '. I shall not endeavour to explain why, lest I reveal my own misunderstanding of relative chance of paternity expressed as a percentage, beyond observing that part of the difficulty would be removed if 99.9993% (sic) was transposed to a statement that there was a 0.0007% probability or chance of [the child's] father being anyone other than [the accused]. Even such adjustment would leave me of the view that the evidence is unduly prejudicial in its impact." (italics in original) Dowd J also agreed. 189 Senior counsel for the appellant also relied, in support of the decision in GK, on the decision of this Court in R v Galli [2001] NSWCCA 504; 127 A Crim R 493 (per Spigelman CJ, with whom Sully and Adams JJ agreed). Galli was another case in which paternity was the issue. By reason of profound intellectual disability, the victim was unable to give any evidence as to paternity of the child she conceived. Evidence of a kind very similar to that the subject of the decision in GK was available (and through the same witnesses). However, in Galli, the Crown did not seek to adduce the evidence of relative chance of paternity. 190 The Crown accepted, on the authority of GK, that the relative chance of paternity figures were not admissible. The Court was therefore not called upon to consider the correctness of GK (see [68]), nor its applicability. Although the Crown had accepted that the relative chance of paternity figures ought not to be adduced, and did not directly seek to have that evidence adduced, the accused had been asked about the figures in the course of an interview with police, the transcript of which was before the jury. In that way the evidence that the Crown had agreed not to tender was put before the jury. At [72], Spigelman CJ appears to have found, for the purposes of that case, that the available relative chance of paternity figures in fact contained a potential to mislead or cause confusion, as envisaged by s 135. At [50] his Honour said: "The danger in evidence of this character is that its very precision and concreteness suggests an exactness which a statistical distribution does not have. There will be circumstances in which this may mislead the jury and lead it to give the evidence greater weight than it ought be given ." (bold added) 191 But this followed a reference to Noll, and the observation that: "[48] Unless evidence can be given in a numerical form, a jury is unlikely to appreciate the probative value of the DNA tests … [49] Where, as here, there is a well-established and widely recognised scientific basis for conducting the relevant computations, it is appropriate that the jury be assisted in this manner." 192 At [51] the Chief Justice quoted King CJ in R v Duke (1979) 1 A Crim R 39 at 41: "There may be unusual cases in which the judge has reason to fear that the jury will be overawed by the scientific garb in which the evidence is presented and will attach greater weight to it than it is capable of bearing ." (bold added) 193 No "circumstances" of the kind to which the Chief Justice referred (at [50]) were here identified. Nor was there here any reason to think the case was unusual, or that the jury would give the exclusion percentage evidence greater weight than it was capable of bearing. The fact is, given the statistical analysis of the DNA evidence, the evidence was capable of bearing very great weight. It was not conclusive that the hair came from the appellant, but it was very powerful evidence to that effect. 194 During the course of argument on the appeal, the presiding judge pressed senior counsel to identify the unfairness (ie the s 135 or s 137 factor that ought to have operated to lead to the exclusion of the evidence) which, on his case, attended the admission of the evidence of the exclusion percentage. He was unable to identify any. He relied upon the decisions in GK and Galli, but could not specify any unfairness (T 22). 195 Indeed, what underlies the appellant's submission is that the evidence, framed as a percentage unlikelihood that the DNA was not that of the appellant, removes any possibility of misleading or confusion, and put starkly before the jury the results of the DNA testing in language they could readily comprehend. 196 GK is not binding authority that evidence of the kind there under consideration is never admissible; it is a conclusion that, on the facts of that case, either s 135 or s 137 could be invoked in order to exclude the relative chance of paternity evidence. Nor is Galli authority for the proposition. Neither is authority for the proposition that "exclusion percentage" evidence is never admissible, nor that such evidence, though admissible, must inevitably be excluded under either s 135 or s 137. 197 In GK, Mason P said: "37 It is not the judicial function to give the accused a fighting chance of gaining an acquittal, as if the trial were a horse race and the judge were a handicapper deciding how much weight to place in the Crown's saddlebags. The admissibility of evidence is not in an inverse ratio to its probative effect. 38 … 39 Accordingly, if relevant DNA statistical evidence is tendered through a witness of due expertise then its probative weight cannot itself be a ground for withholding it from the jury. Indeed its very significant probative weight is a factor in favour of admission notwithstanding the capacity of extremely high odds to carry a prejudicial overlay." 198 The evidence, put as it was, was prejudicial: all Crown evidence is intended to be prejudicial. That is why it is tendered. I have been able to discern nothing that suggests that the evidence before the jury, framed as it was, was unduly or unfairly prejudicial, or confusing or misleading such as to raise for consideration either s 135 or s 137. 199 Nor am I of the view that there was any deficiency in the way in which the jury was directed in relation to the DNA evidence. 200 I would reject Ground 1 of the appeal. Ground 2: unreasonable verdict 201 The relevant facts and circumstances are set out in the judgment of the Chief Judge, and, other than as appears necessary, I do not propose to restate them. 202 In my opinion, that the ground should fail is best illustrated by a chronological examination of the facts implicating the appellant. In dealing with this evidence, I have not overlooked that a good deal of it was necessarily given in hearsay form. Nor have I overlooked that there is a degree of uncertainty about the timing of some of the events described. Certain of the facts have been derived from two interviews of the appellant by police, which were in evidence. With those limitations in mind, it seems to me that the following are the relevant facts and circumstances, in as near to chronological sequence as I can ascertain. 203 The relationship between the appellant and Ms Bayrak began sometime in 2002. Her daughter Tutku and his son Mert attended the same primary school. Ms Bayrak was then living in Rose Hill. The relationship became quite intense on the part of the appellant and he proposed marriage. Ms Bayrak rejected the proposal. She terminated the relationship in July 2003. There is little evidence of the impact of the termination upon the appellant at that stage; there was no evidence that he was unduly distressed or otherwise affected by the end of the relationship. In early November Ms Bayrak entered into a short term relationship with another man, Hasan Kilic. That relationship ended, apparently amicably, in April 2003. There is no evidence that the appellant was aware of this relationship, or, if he were, that it affected him in any way. 204 In 2003 and again in 2004 Ms Bayrak travelled to Turkey, where, on the second occasion, she remained for some weeks. She returned in June 2004. There was evidence from which it could be inferred that the appellant was aware of her having travelled to Turkey, but not, initially, of her return. He appears to have learned of her return by a chance encounter with her in the Auburn shopping centre in early 2005. (Evidence of this is contained in the appellant's first interview with police.) According to him, he then saw Ms Bayrak four or five times in the Auburn area. He said that he did not ask, and Ms Bayrak did not tell him, where she was then living. 205 Ms Feliz Yavuz, who employed Ms Bayrak as a casual cleaner, gave evidence of an occasion, in May or June 2005, when Ms Bayrak showed her a fruit platter that the appellant had given her as a gift. Ms Bayrak told Ms Yavuz that she did not wish to accept it, but also did not wish to hurt the appellant's feelings. 206 On 17 May 2005 Ms Bayrak and her daughter Tutku took up public accommodation in Crown Street, Merrylands. They remained in that accommodation until 25 October 2005. While they were living there, there were a number of incidents of which evidence was given. Her car was damaged on more than one occasion. On one such occasion the brake linings were cut, on another a jelly like substance was placed in the petrol tank. On another the number plates were stolen. (There is nothing, other than suspicion, in the evidence to link the appellant with any of these incidents.) 207 On 5 June 2005 the appellant telephoned Ms Bayrak. The content of this call is not known. The following day, 6 June, Mr Mehmet Uluc conducted an unauthorised search of Centrelink records and gained access to Ms Bayrak's address. Mr Uluc was an acquaintance of the appellant who was employed by Centrelink. He was charged with and pleaded guilty to a number of offences arising from this and subsequent unauthorised searches. Mr Uluc's evidence was quite unsatisfactory: initially, he denied having carried out the searches at the request of the appellant; subsequently, he admitted that he had done so, at least on 4 November. (Telephone calls from the appellant's telephone to Mr Uluc's telephone left open a clear inference that on each occasion that Mr Uluc accessed Ms Bayrak's personal details, it was at the instigation of the appellant. Mr Uluc denied knowing Ms Bayrak, and there was no evidence to contradict that.) 208 In any event, the evidence showed the following: on 6 June 2005, at 11.08am, a call was made from the appellant's mobile telephone to Mr Uluc's mobile telephone. At 11.20am Mr Uluc made an unauthorised search of Ms Bayrak's personal details. 209 On a date that cannot be precisely ascertained, the appellant and his son made an unannounced call on Ms Bayrak and her daughter as their Crown Street unit. The reason given was that the appellant's son missed his friend Tutku and wished to see her. The timing of the visit, in conjunction with the evidence of Mr Uluc's search of the Centrelink records, left the clear inference that the appellant obtained Ms Bayrak's address from Mr Uluc. 210 In his interviews, the appellant denied ever having visited Ms Bayrak at Crown Street. The evidence that he did in fact do so was very powerful, and the Crown relied upon his denial as a material lie evidencing consciousness of guilt. The jury was fully directed about this and no complaint is made about the directions. 211 On 29 June 2005 the poem submitted by the appellant to the Turkish newspaper by the appellant was published. This clearly indicated that he retained, or had re-developed, strong feelings for Ms Bayrak. 212 As a result of the various incidents that had taken place at Crown Street, Ms Bayrak and her daughter were allocated alternative housing at Parramatta. On 25 October they moved to the new premises at Betts Street. 213 In October or November 2005 Ms Bayrak was shopping at a shopping centre at Homebush. The appellant called her on her mobile telephone. She told him that she was busy, but he replied that he knew where she was, that she was shopping at Ikea. He said that he wanted to chat to her, that it would only take five minutes, and he would never "bug" her again. (Telephone records of calls from the appellant to Ms Bayrak's phone suggest that this may have been on 29 October 2005.) An inference was available that the appellant was keeping Ms Bayrak under surveillance. 214 On 1 and 2 November the appellant telephoned Mr Uluc five times. The content of these calls is not known. However, on 4 November Mr Uluc twice searched Ms Bayrak's records. The second of these searches was at 2.32pm; at 2.35pm Mr Uluc telephoned the appellant. It is a reasonable inference that he conveyed to the appellant Ms Bayrak's new address. 215 Three days later, on 7 November, Ms Bayrak telephoned the appellant, and later that evening (at 10.48pm) sent him a text message. The content of these communications is not known. 216 On 9 November 2005 Ms Bayrak told one of her employers, Edrun Gundus, that she had been at the Auburn shops and that the appellant had been following her. He asked "don't I deserve a hello?". 217 On (an unspecified) occasion, while she was living at Crown Street, Merrylands, Ms Bayrak told Ms Gundus that she had heard a noise and seen a shadow at her balcony. The following morning she found that a rubbish bin had been pulled to underneath the balcony, apparently to provide access or a viewing platform. (There was nothing to link the appellant with this.) 218 Just two weeks before Ms Bayrak's death (11 November) she and Mr Tunc were at her unit in Betts Street, Parramatta. Ms Bayrak was telling Mr Tunc about an argument she had had with her father. She was upset and crying. The appellant sent her a text message in the following terms: "Sevda you are sad and crying I feel it." 219 This was just before 1am. In the next 45 minutes the appellant sent six text messages to her telephone. It could readily be inferred that the appellant was in the vicinity of her unit, sufficiently nearby to be able to see her state of distress. 220 On 21 November 2005 Ms Bayrak was at Ms Yavuz's home. She took a mobile telephone call. She told Ms Yavuz that the appellant had rung and asked her why she had changed her number plates, and if it was so that he could not follow her. She said that she told him that she had changed the number plates because hers had been stolen. She told Ms Yavuz that she did not know how the appellant knew that the number plates had been changed. (That the appellant made this enquiry would tend to exculpate him from involvement in at least one of the instances of vandalism. If he had removed the number plates, why would he ask her that question?) 221 On 23 November 2005 the appellant telephoned Centrelink to report that Ms Bayrak was working and in receipt of income that she did not disclose for social security purposes. 222 On 25 November the appellant twice telephoned Ms Bayrak but she did not answer the phone. That day his current girlfriend went to the Gold Coast for a holiday. 223 It was late on the following day that Ms Bayrak was murdered. 224 Those circumstances together show, in my opinion, a considerable degree of interest in Ms Bayrak on the part of the appellant. It may well be that it appeared to the jury to have escalated in intensity as time progressed. 225 When those facts are put together with the DNA evidence, the overall picture presented is one upon which it was open to the jury to conclude that the appellant was the perpetrator of the murder. 226 It is true that there are countervailing circumstances. 227 As part of its case, the Crown relied upon three lies which it claimed had been told by the appellant, evidencing a consciousness of guilt. These concerned his denial that he had visited the Crown Street address; a claim that he did not know Ms Bayrak's address at Betts Street; and a claim that he did not ask for or receive details of her address from Mr Uluc. 228 The second and third of these may well have been seen as equivocal; the manner in which the appellant obtained Ms Bayrak's address involved Mr Uluc in criminality. No doubt the appellant was well aware that obtaining the address in the way he did, at the least, involved an impropriety on the part of Mr Uluc. It is more difficult to explain the clear lie that he did not ever go to the Merrylands unit. The evidence that he did so was very strong indeed. 229 There was also evidence, never explained, that Mr Uluc had searched Ms Bayrak's Centrelink records as early as 2001, before the appellant met her. 230 By way of final submission, senior counsel for the appellant put five distinct propositions. These were: firstly, that the three lies relied upon by the Crown were explicable on another basis. As to the second and third of the lies, I would agree; I have already made my comments in relation to the first. 231 Secondly, it was submitted on behalf of the appellant, the campaign of damage at Ms Bayrak's Merrylands address could not safely be attributed to the appellant. I would accept this. 232 Thirdly, it was submitted, the DNA evidence does not strongly implicate the appellant, and yet, it was said, that is the only evidence that directly connects the appellant with the crime scene. I find this proposition difficult to accept. The appellant's case was that he had little or no physical contact with Ms Bayrak for some time. 233 One possible explanation put forward to the jury about the hair was that it may have been picked up on Ms Bayrak's thumbnail after she had been stabbed, and while her blood was still liquid. This depended upon the hair having been present at the scene (see the judgment of McClellan CJ at CL at [40]). That is hardly a convincing explanation, given that the appellant claimed never to have been at the Betts Street unit. The submission thus depended further upon the evidence that there was a 1 in 100 possibility that its source was somebody other than the appellant. 234 The fourth submission made on behalf of the appellant was that his alibi could not be discounted beyond reasonable doubt; that the discrepancies were minor and natural; and that it was unlikely – even incredible – that the appellant would falsely nominate his neighbour as an alibi witness soon after the offence. These are legitimate points to make, but are points that were properly put to the jury and no doubt considered by them in the context of the whole of the evidence. I do agree that the discrepancies could be attributed little weight. 235 The final submission pointed to the possible involvement of Mr Tunc. It is true that considerable suspicion was thrown upon Mr Tunc and it was a central plank of the appellant's defence that Mr Tunc was, or might have been, the perpetrator. Again, however, this is a matter that was properly put before the jury and considered by them. 236 Having regard to the chronology of events outlined above, I am satisfied that the verdict of guilty was one which was open to the jury upon the whole of the evidence. 237 Accordingly, I agree with the Chief Judge that this ground of appeal should be rejected. I would dismiss the appeal against conviction. 238 FULLERTON J: I have read the judgments of McClellan CJ at CL and Simpson J. I agree with Simpson J that Ground 1 of the appeal should fail and with her Honour's reasons for coming to that view. I also agree with Simpson J that the appeal against conviction should be dismissed. 239 I also agree with McClellan CJ at CL and Simpson J that Ground 2 should fail and with their Honours reasons for coming to that view. I wish, however, to add my own reasons for sharing their view that on a review of all the evidence it was open to the jury to find the appellant guilty of murder. 240 The chronology of objective facts and circumstances to which their Honours have each referred in their respective judgments is, in my view, illustrative of the appellant's sustained intimidation and harassment of the deceased in the months leading up to her death. In addition, his account to police when interviewed first in December 2005 and again in March 2006, and in particular the way he described his relationship with the deceased over the course of the questioning, is inconsistent with what the evidence otherwise reveals about the intensity of his obsessional interest in her. Quite apart from the specific lies he told to police as evidencing a consciousness of guilt, the jury were entitled to consider the way he represented his relationship with the deceased in the context of other evidence bearing upon that issue. 241 I propose to refer to some of that evidence and, for ease of reference, to the paragraphs in Simpson J's judgment where some of the events and incidents he was questioned about are summarised. 242 The interviews were conducted in English. As is clear from the transcribed recordings there were occasions when the police needed to rephrase some questions and where the appellant's syntax was imperfect. That said, there was no suggestion that he needed the assistance of an interpreter or that his answers reflected any lack of understanding of what he was being asked. 243 In paragraph [203] Simpson J noted that the relationship between the appellant and the deceased terminated in July 2003 at which time there was little evidence of its impact on him and no evidence that he was unduly distressed or otherwise affected by the end of the relationship. He did however inform police in the December 2005 interview that he was shocked and disappointed when the relationship was terminated and when the proposal that they marry did not eventuate. He said he believed that was because the deceased wanted to live in Turkey and he did not. He also told police that despite his disappointment he was not interested in resuming the relationship with the deceased when he learnt that she had returned from Turkey in June 2004 as he had been hurt twice by her leaving him to go to Turkey and he did not want to rekindle his affections and have her leave him again. He went on to tell police that he had no romantic feelings for her on the occasions that he saw her in the Auburn area thereafter and, further, that while on some occasions where they had chance contact with one another in the street they would exchange pleasantries, on other occasions they would not speak at all. 244 On any view of the evidence in the case, what the appellant describes as a casual acquaintance with a former lover is inconsistent with the evidence of his determined efforts, from at least as early as June 2005, to locate her through the Centrelink data base by illicit means and in stark contrast with the tone and content of the poem he submitted for publication in June 2005. 245 In so far as the incident in October or November 2005 at the Ikea shopping complex at Homebush (to which Simpson J referred at [213]) is concerned, the appellant told police that although he made the telephone call it was the deceased who suggested that they have a cup of coffee together to which he agreed. He told police that they spoke amicably about work but did not discuss their private lives. In regards to the deceased's account of the appellant following her on another occasion (referred to by Simpson J at [216]), he denied that he had done so or that she had ever tried to avoid him. To the contrary. He said that whenever they had telephone contact or met on the street they would talk as friends. 246 He said that they had spoken on the telephone a couple of times after having a coffee at the Ikea complex but that as at the date of the interview on 7 December 2005 he had no contact with her for two months. He said he sent her a message on her birthday on 10 November and that this was the last contact with her before her death on 26 November. There is no record of a message sent or received on the deceased's birthday. There is however evidence of calls the appellant made to the deceased on 25 November, the content of which is unknown. 247 In paragraphs [218] - [219] of Simpson J's judgment her Honour referred to the text message sent by the accused at about 1am on 11 November which said "Sevda you are sad and crying I feel it" (at a time when she was in fact inside her unit and tearful). Her Honour also referred to six text messages sent thereafter, the content of which is not known but from which it was open to conclude that the appellant had the deceased under surveillance. I note that in a seventh message sent by the appellant at 1.47 am he said "I'm sure because I asked about your number plates you are blaming me for it". In the intervening calls between 1am and 1.47am, in my view it is open to infer that the appellant sought to or succeeded in engaging the deceased in dialogue about the fact that she had changed the number plates on her car after they were stolen. The appellant was unable to offer any explanation to police as to why he thought that the deceased might have suspected him of stealing the number plates. On the other hand, since it appears the deceased thought that the appellant was responsible, even if that was not borne out by the evidence, it indicates something of her attitude towards him and her concerns that he was behaving in a threatening and provocative manner entirely inconsistent with the casual friendship that he was intent on conveying when interviewed by police. 248 In so far as the call at 1am on 11 November is concerned the appellant told police that the deceased had told him that she was not happy with her job and was looking for other work and he was referring to that when he claimed to empathise with her sadness. He said he had the same motivations when he sent a message at 1.57 that same morning which said "Alright. Understood, there's a God above. What can I say. I pray for your health". Of that text he said that he had particular concerns for the deceased's mental health because of her work situation. The jury would have been entitled to reject this account as untrue given that within two weeks of 11 November he reported the deceased to Centrelink because she was working and in receipt of income that she did not disclose for social security purposes even if they were not prepared to read the text sent at 1.57am as a threat as contended for the Crown. ********** DISCLAIMER - Every effort has been made to comply with suppression orders or statutory provisions prohibiting publication that may apply to this judgment or decision. The onus remains on any person using material in the judgment or decision to ensure that the intended use of that material does not breach any such order or provision. Further enquiries may be directed to the Registry of the Court or Tribunal in which it was generated.