Ward v The Queen [2018] VSCA 80

EVIDENCE - Admissibility - Opinion evidence - Where technical expert gave evidence of probability of mobile phone being in particular geographic area at particular time - Where probability expressed as percentage - Whether witness had 'specialised knowledge' to give evidence expressed as percentage - Evidence Act 2008 ss 76(1), 79(1).

CRIMINAL LAW - Appeal - Sentence - Armed robbery, attempted armed robbery and reckless conduct endangering life - Offender sentenced to 10 years and 6 months' imprisonment with non-parole period of 7 years and 6 months' imprisonment - Whether orders for cumulation excessive - Whether sentence manifestly excessive - Leave to appeal refused.

1 The appellant was charged with offences on three separate indictments. For ease of reference, we will refer to the indictments by their last two digits, being 6.1, 6.2 and 6.3.

2 On indictment 6.1, the appellant pleaded guilty to one charge of theft of a BMW motor vehicle on 18 February 2015.

3 On indictment 6.2, the appellant was charged with four offences which were also alleged to have been committed on 18 February 2015. He pleaded not guilty to those charges and stood trial in the County Court. On 14 December 2016, the jury returned a verdict of guilty on one charge of attempted armed robbery, one charge of armed robbery and one charge of reckless conduct endangering life. The jury acquitted the appellant on one charge of attempted armed robbery.

4 On indictment 6.3, the appellant pleaded guilty to one charge of theft of a registration plate, one charge of theft of a Ford motor vehicle, one charge of burglary and one charge of theft of cold and flu tablets from a pharmacy. The offences were committed on 5 March 2015 and 7 March 2015.

5 On 10 March 2017, the appellant was sentenced to ten years and six months' imprisonment. A non-parole period of seven years and six months was fixed.

6 On 29 August 2017, Whelan JA granted the appellant leave to appeal his conviction on a single ground, which is set out below. Whelan JA also refused the appellant leave to appeal his sentence.[1] The appellant has elected to renew his application for leave to appeal against sentence.

7 For the reasons that follow, the appeal against conviction will be dismissed and leave to appeal against sentence will be refused.

Circumstances of the offending - indictment 6.2

8 On 18 February 2015 at about 5:50 am, two men entered an Aldi supermarket in Prahran after smashing the glass front door.[2] The men were disguised. One of them was alleged to be the appellant. He had a shotgun. Staff members told the men that there was no money on the premises. The men left without taking anything. They drove off in a white BMW X1 which had been stolen. The incident was captured on CCTV, and the appellant was alleged to be the man wearing a backpack strapped to the front of his chest in that footage.

9 Later that morning, at 7:00 am, the two men carried out an armed robbery at a café in Malvern. The man alleged to be the appellant pointed a shotgun at one of the staff. Some $300 in cash was stolen.[3] The men drove off in the white BMW. This incident was also captured on CCTV. The appellant was alleged again to have been wearing a backpack strapped to his chest. Both offenders were carrying firearms.

10 Shortly after the armed robbery at the café, police pursued the white BMW along High Street in Malvern. One of the police officers in pursuit gave evidence that he saw the driver of the BMW through a partially open window and described him as a Caucasian with olive complexion and blonde spikey gelled hair. The BMW drove onto the wrong side of the road and mounted a curb before returning to the wrong side of the road near the intersection of Glenferrie Road and High Street.[4] The men were not apprehended.

11 On the afternoon of 18 February 2015, police located the white BMW in Kingsbury and placed the vehicle under surveillance. Shortly after 9:00 pm on the same day, the appellant was observed to enter the white BMW with a key and drive it around the streets of Kingsbury. He then met with an associate. Both men were observed removing a backpack from the boot of the white BMW. The appellant left his fingerprints on the vehicle. In an intercepted telephone call shortly before midnight on 18 February 2015, he described himself as driving around in a BMW X1.

13 The issue in relation to the charges on indictment 6.2 was identity - that is, whether the Crown could prove beyond reasonable doubt that the relevant offences were committed by the appellant. The Crown alleged that the appellant was the driver of the white BMW. There were no witnesses who identified the appellant as one of the offenders on 18 February 2015. Moreover, the identity of each offender could not be ascertained from the CCTV footage.

14 An aspect of the Crown's case on the issue of identity was circumstantial evidence with respect to the location of a mobile phone which, according to the Crown, the appellant had been using at the time of the offending. The Crown contended that, on the morning of 18 February 2015, the mobile phone had been communicating with cells on towers near the locations of the offences and at times proximate to the commission of those offences. The Crown adduced evidence from Oleg Prypoten, who is employed by Optus as a technical expert, about the statistical probability of a mobile phone being in a particular geographic area at a particular time. Mr Prypoten's evidence is the subject of the ground of appeal.

15 Initially, Mr Prypoten gave evidence in two Basha inquiries.[5] He eventually gave evidence in the presence of the jury on 7 and 8 December 2016. Before examining that evidence, it is convenient to revisit briefly the procedural background in this matter.

16 The trial commenced on 15 November 2016. On the application of the defence, the trial judge held a Basha inquiry in which several witnesses gave evidence. One of those witnesses was Mr Prypoten, who gave his evidence on 16 and 17 November 2015. His evidence is summarised below.

17 On 22 November 2016, a jury was empanelled. Soon thereafter, the trial judge discharged the jury after one of the jurors raised a concern about having to act as a trier of fact.

18 On 24 November 2016, a second jury was empanelled. On 25 November 2016, the trial judge discharged the jury after the jurors, and subsequently counsel, raised a concern about a member of the public using a mobile phone in court.

19 On 30 November 2016, a third jury was empanelled. Over the course of six days, a number of witnesses gave evidence, including David Finlay, who is another employee of Optus, and Mr Prypoten. Mr Prypoten gave his evidence again in a Basha inquiry on 7 December 2016. His evidence concerned the likely physical proximity of the relevant mobile phone, being the phone which the Crown alleged the appellant had been using at the time of the offending, to cells on certain towers in particular geographic locations at particular times.

20 Over an objection by the defence, the trial judge delivered a ruling in which he permitted the Crown to lead Mr Prypoten's evidence in the form of a statistical likelihood, or percentage probability, of a phone being within close proximity of a particular cell on a tower when communications between the phone and the cell were recorded. Mr Prypoten subsequently gave his evidence in the presence of the jury on 7 and 8 December 2016.

Summary of Mr Prypoten's evidence at the Basha inquiries

21 In order to understand the trial judge's ruling, and to place certain aspects of Mr Prypoten's evidence before the jury into context, it is necessary to summarise the evidence given by Mr Prypoten at the Basha inquiries.

22 As indicated above, Mr Prypoten gave evidence at the first Basha inquiry on 16 and 17 November 2016. His evidence was directed largely to explaining how a mobile phone communicates with cells located on a tower or, more specifically, a base station. He had produced a number of maps, which he described as 'plots', that represented the 'calculated areas for probable and possible cell coverage'. Each plot, which is a road map, is entitled 'Prediction of Probable (dark green) and Possible (light green) Cell Coverage area' for a certain geographical area. Certain zones are highlighted green and, within the highlighted zone, there may be a smaller zone that is highlighted in a darker shade of green. Two examples are extracted below.[6]

23 Mr Prypoten gave evidence that the plots were generated by a computer programme which takes into account many variables and calculates the probable and possible coverage areas for particular cells in the network. He described the dark green zones on the plots as areas in which coverage from the relevant cell was 'highly likely'. On the ability of a mobile phone to communicate with a particular cell from a particular place, Mr Prypoten gave evidence that the 'dark-green patch represents 95 per cent of probability for the call to be served by that cell'.

24 Mr Prypoten gave evidence at the second Basha inquiry on 7 and 8 December 2016. Before that inquiry, Mr Prypoten had produced 'prediction validation data' for some of the plots contained in his earlier statement.[7] The data took the form of road maps, each of which purported to show the areas where a phone determined a particular cell, identified by a five-digit number, 'to be the strongest cell'. The data was generated as a result of testing which was carried out using measuring equipment and an Optus van.

25 Defence counsel asked Mr Prypoten why an area of possible coverage, where a phone could detect a cell, was well outside the light green zone (a possible cell coverage area) shown on a corresponding plot.[8] Mr Prypoten responded: 'Yes. This is that 1 per cent of probability'.

26 During the second Basha inquiry, defence counsel submitted that the Crown should not be permitted to lead evidence from Mr Prypoten that uses percentages to describe statistical probabilities and possibilities. In other words, there was no scientific basis to say that a dark green zone in a plot, for example, represents a 95 per cent probability. Counsel took no issue with Mr Prypoten's use of expressions such as 'likely' or 'highly likely' when explaining the relevant probabilities. However, counsel argued that the statistics used by Mr Prypoten had not been verified by proper empirical testing. He said that expressing the probabilities in terms of percentages had no proper grounding in Mr Prypoten's evidence; they lent a false air of precision to that evidence.

27 In response to defence counsel's submissions, the trial judge said:

I want to make it clear that what my concern is, if someone is going to speak about probabilities and possibilities and high confidence, they are statements without empirical certainty but they are indicative of the position. If you are going to turn it into percentages, that changes the position. If in fact there is material upon which the percentages can be calculated, then I think I should allow the witness to give evidence as to that.

If it is really just a person saying 99 percent because that's what they mean with high confidence, then they should say 'high confidence'. If there is some uncertainty about this, I don't want a witness just plucking a figure out of the air because it would be inappropriate and I just wonder whether the better course is if the prosecution wants to lead evidence other than of high confidence, then the way to do it is to qualify the witness to do it.

28 The trial judge formed the view that the issue raised by defence counsel should be explored further with Mr Prypoten. In the event, Mr Prypoten was recalled to give evidence again in a Basha inquiry.

29 When asked for the basis of his evidence that the location of a mobile phone within a dark green zone on a plot could be determined with a probability of 95 per cent, Mr Prypoten said that there was no mathematical formula for that figure; 'it's just a representation of highly likely'. He said that the 95 per cent figure was based on 'experience' and that it had been used by expert witnesses from Optus over a number of years. As to the light green zones on the plots, Mr Prypoten said: 'Light green is unlikely so it's four per cent'.

30 Mr Prypoten was also asked what 'generates' the 95 per cent figure. He responded as follows:

Well, in the - in the computer, you put all the parameters in and one of them is the deviation of the model. So something we discussed before. So the tool, the computer software, is aware about the ranges within which signal was detected in every tile. And using that variation, it computes the value with 95 per cent probability of that value to be - to happen in that tile. So the 95 percent is the coverage probability or the value calculation probability which returned [sic] by the tool, by the computer software, and that's the number which was used for consistency.

31 Mr Prypoten accepted that, in order to determine whether the 95 per cent figure was an accurate prediction of probability in a geographic location, a statistically significant group of tests would need to be conducted. He also accepted that none of the plots which he produced purported to be the result of a statistically significant group of tests.

32 In the event, the trial judge permitted the Crown to lead evidence from Mr Prypoten that used percentages to describe statistical probabilities and possibilities. The trial judge delivered the following ruling:

On the basis of the evidence given by Mr Prypoten on the voir dire, it seems to me that the figure of 95 per cent is generated as a matter of input into an algorithm which is designed to produce a finding to that degree of certainty based on the algorithm. In that sense, it is an area in which expert evidence can be given. The question of the verification goes to the certainty in respect of the process, and at the end of the day, that is a question for the jury.

The jury might find that the algorithm is all right in the computer but not all right on the ground, but it is not a question of mere speculation and plucking a figure out of the air, it is a figure generated through a process which has a potential of a jury being satisfied, is the subject of some rigour, and is relied upon by the industry. So for those reasons, I am satisfied that the evidence should be given, but equally that it is open to be tested in terms of the way the jury should use it and the weight that the jury should give to it.

33 The appellant was granted leave to appeal against his conviction on the following ground:

The learned trial judge erred in permitting the prosecution to lead evidence of the statistical likelihood of a mobile phone being in particular geographic areas at particular times.[9]

34 As mentioned above, the issue at trial was whether the Crown could prove beyond reasonable doubt that it was the appellant who had committed the relevant offences. None of the witnesses had identified the appellant as one of the offenders, and the identity of each offender could not be ascertained from the CCTV footage. It was for these reasons that Mr Prypoten's evidence, which is examined further below, assumed some significance. It is necessary first to describe the burden of the other evidence which connected the appellant to the commission of the offences.

35 One of the witnesses who gave evidence before the jury was Mr Finlay, who is employed by Optus as a statement officer. His responsibilities include producing customer records and providing statements to explain those records. Mr Finlay gave evidence about a mobile phone number the last three digits of which are 480 ('the 480 number'). He said that it was an Optus number linked to an account in the name of Tegan Breheny. It was not in dispute that, at the time, Ms Breheny was the appellant's de facto partner. They lived together at 1/12 Mack Street, Reservoir.

36 Mr Finlay produced telephone records which showed that the mobile phone containing the 480 number had been communicating with cells on towers in suburbs such as Richmond, Toorak and Malvern at times proximate to the commission of the offences on the morning of 18 February 2015.[10] The defence did not challenge the admissibility of this evidence at trial.

37 Mr Finlay also produced a document known as a 'visitor location register'. He described it as a document which recorded communications, and the location of those communications, between a mobile phone and the Optus network. Mr Finlay said that the register has 'nothing to do with phone calls. It's just purely the communications between the handset and the mobile towers'. The contents of the visitor location register are elaborated below.

38 Several eyewitnesses who, at the time of the offending, were inside or near the Aldi supermarket in Prahran or the café in Malvern gave evidence that they saw the two offenders drive off in a white BMW immediately after the offending. One of the eyewitnesses, an Aldi employee, said that, after the incident, he recorded the registration plate number of the white BMW on the palm of his hand and on a piece of cardboard. The number was 1AA8ZF.

39 The CCTV footage of the incidents in Prahran and Malvern showed one of the offenders wearing a backpack which was strapped to the front of his chest. The CCTV footage of the incident in Malvern also showed the offenders driving to the rear of the café in a white BMW before leaving the vehicle to carry out the armed robbery.

40 On 18 February 2015 at 3:08 pm, several hours after the offending, police observed a white BMW, with the registration plate number 1AA8ZF, parked on Bradshaw Street in Kingsbury. This observation was recorded in a surveillance log which was tendered as an exhibit.

41 On 18 February 2015 at 8:55 pm, a taxi picked up a passenger from 1/12 Mack Street in Reservoir. The booking was made under the name of 'Paul', and it recorded a mobile phone number the last three digits of which are 248; the first seven digits are identical to those of the 480 number.[11] The taxi drove the passenger to the vicinity of Bradshaw Street in Kingsbury. This information was read from a statement, which was admitted into evidence by agreement, prepared by an employee of 13 CABS.

42 On 18 February 2015 at 9:03 pm, police observed a Caucasian male with 'spiky fair hair' and 'dark clothing' unlock the white BMW parked on Bradshaw Street in Kingsbury. He drove the BMW around the streets of Kingsbury for some time before meeting with an associate, Stephen Masson. It was not in issue at trial that the person observed to be driving the BMW at that time was the appellant and that the person with whom the appellant met was Masson.

43 On 18 February 2015 at 9:43 pm, police observed the appellant and Masson removing a backpack from the boot of the white BMW.

44 On 18 February 2015 at 11:52 pm, police intercepted a telephone call between the appellant and another associate, Mohammed Belhaj. During that call, the appellant said that he was driving around looking for a tyre and a rim for a BMW X1. A transcript of the telephone call was tendered as an exhibit and played to the jury.

45 Some fingerprints which were detected on the tailgate of the BMW matched the fingerprints of the appellant.

46 On 7 December 2016, Mr Prypoten commenced giving his evidence in chief before the jury. He gave evidence that he is a technical specialist who is employed by Optus. He holds a master's degree in radio engineering from the National Technical University in Ukraine. He has ten years' experience in the mobile telecommunication industry.

47 Mr Prypoten first described the operation of base stations and cells. He described a base station as a special piece of equipment which is located at a certain address and which provides coverage and service to Optus users. He described a cell as a logical division of the base station; one base station usually hosts many cells. Cells may provide specific coverage in particular directions. Each cell is connected to an antenna. Antennas point in different directions and sit at different heights. A cell 'will contribute to or define the frequency band and the power which is actually used to generate the coverage area'.

48 Mr Prypoten explained how a mobile phone communicates with a particular cell. He said that, when a call is made, the phone selects 'the most appropriate cell' - generally the cell which is nearest to the phone - and starts a communication with that cell. The phone may give priority to a cell on a certain network, such as the 3G network, if the phone itself is compatible with that network, such as a 3G phone. If the nearest cell is not working, the phone may communicate with a cell which is further away but operational. Next, the communication passes through a mobile switching centre before going to the recipient of the call. If the caller changes location, the communication will be handed over to a neighbouring cell, on a different base station.

49 Mr Prypoten also explained that a mobile phone will communicate with a particular cell as soon as the mobile phone is switched on. The mobile phone will communicate 'with the cell with the strongest signal and highest priority' - depending upon its compatibility with the network and proximity to the mobile phone. For a mobile phone to communicate with a cell, it is not necessary for a call to be made from the mobile phone.

50 Mr Prypoten also gave evidence about plots. He described plots as 'computer generated calculations of the cell coverage which is [sic] based on many assumptions'. The computer uses a large number of inputs, such as topographical area, digital terrain and the position of base stations. It takes into account how a signal is transmitted from an antenna and plots a 'tile' where signal strength is calculated for every cell. A plot, which resembles a road map, shows three areas:

One is the possible coverage; so called light-green. This is the area where the signal strength is sufficient to carry the call. And then there is a dark-green area where that cell is strongest. And then there is also a grey area where the call is still possible to take place.

51 At this stage, the Crown tendered a document which contained several plots titled 'Prediction of Probable (dark green) and Possible (light green) Cell Coverage area' for certain geographical areas. The plots were last updated in 2012. The final page of the document contained a table titled 'Table 1: Table showing Cell Name versus Cell ID numbers and the address of base station equipment where it is installed'. The table had four columns titled 'Cell Name', 'Cell ID', 'Site address' and 'Technology'.

52 The Crown also tendered two plots, each of which was an enlarged version of a plot contained in the document to which we have just referred. One plot showed a road map of Richmond South (figure 4), the other a road map of Toorak East (figure 6). The plots are as follows.

53 When asked what the dark green area on the Richmond South plot represented, Mr Prypoten said:

Dark green area represent the area where the - this cell, Richmond South OM2160U21A2 is the strongest cell and has enough signal level to cater for the call. This is area where it's highly likely for the call to be connected by this cell.

Q: Now you mentioned 'Highly likely'?

A: Yes.

Q: Sorry I cut you off, what did you mean by that?

A: It - it can be probably converted to 95 per cent probability of call to be connected to the cell in that area.

Q: 95 per cent probability of a call, mobile phone call?

A: Yeah, to happen in that area. If call was connected by that cell it is highly likely the phone was in the dark green area.

55 When asked how that 95 per cent probability figure was calculated, Mr Prypoten said that 'it is one of the parameters used by software program to calculate the signal level in every location'. The transcript reads as follows:

Q: So on a basis of 95 per cent probability, that dark green area is what you get?

A: Yes.

Q: On the plot?

A: Yes.

Q: That is a statistical probability?

A: Yes.

Q: Based on mathematical calculation?

A: Correct.

Q: Once you input the number of parameters that you've mentioned?

A: Yes, the tool should be properly calibrated and sort of enabled to do this calculations which was done for Optus.

56 As to the Richmond South plot, the following exchange took place:

Q: So the dark green patch, would you agree that it represents - does it represent 95 per cent probability for a particular call to be served by that cell, A2?

A: Yes.

Q: The one referring to Richmond South?

A: Yes.

Q: What about the light green area?

A: It will be around 4 per cent.

57 The prosecutor asked Mr Prypoten how he arrived at the figure of four per cent. Mr Prypoten responded that 'it's more from empirical and more from experience ... comparing between the life test data and calculation based on the tool'. By experience, he meant his experience and that of 'the Optus representatives which were doing the statements before in the court'.

58 Mr Prypoten confirmed that the 95 per cent probability figure was inputted as a parameter in the computer programme. He added: 'This is the so called - yeah, you can call it - what - accuracy of the calculation. Calculation - you know, trustability.' Mr Prypoten said that the figure of four per cent was not 'inputted'. When asked that how that figure was calculated, he said:

4 per cent is just - it's unlikely for the call to happen in the light green area, because it's only theoretical probability to go beyond the light green area; it's 1 per cent. So everything between 95 per cent and 1 per cent is, yeah, remaining 4 per cent.

Mr Prypoten said that the remaining one per cent, which covers the area outside the zones highlighted green, is also based on experience.

59 Mr Prypoten said that the plots represented a statistical model. He explained:

[I]n majority of cases, it will be accurate, but you will still have some anomalies or deviations where it will go beyond the calculated value, so too far beyond the calculated values. So in general, it is as accurate as it can be, and there was all measures applied to make it accurate. However, due to nature of radio signal and its deviation, you can't be certain, you can't be sure, that in this point in every instance of time you will have consistent and exactly the same result. No. You can only talk about probability. It's a high chance that you will be connected to this cell, but it's not warranted.

When asked what he meant by 'high chance', he replied: 'It's 95 per cent'. He confirmed that this figure applied to the dark green area. He also accepted that the signal strength from a particular cell is affected by meteorological conditions and that these factors are taken into account by the computer programme.

60 Mr Prypoten was asked about another plot which was in evidence. He summarised what each of the shades of green and the colour grey indicated: 'dark green area means highly likely for the device to be in that area. The light green area is unlikely and grey area is just theoretical possibility.'

61 The prosecutor drew Mr Prypoten's attention to the visitor location register which had been produced earlier.[12] Mr Prypoten explained its purpose as follows:

So when user is registered in the network, there is a database which storing [sic] details of every active subscriber. And the database is periodically updated for each user who is the location information. So they realise the visitor location register is the database which is holding last known location of the phone.

62 The visitor location register comprises two pages, each of which contains a table with seven columns: 'Report Time', which lists dates and times; 'Radio Access Info', which lists the type of network to which the user is connected - either the number 1 (2G network) or 2 (3G network); 'MSISDN Number' which is the mobile phone number; 'IMEI', which is akin to the serial number of the phone itself; 'Lac', being a number which represents a 'local area code'; 'Sac', the abbreviated form of 'service area code'; and 'Cell'. A particular cell can be identified through the combination of the 'Lac' and 'Sac' numbers or, in the case of 2G calls, the combination of the 'Lac' number and the 'Cell number'. The data contained in the visitor location register correspond with the plots.

63 The visitor location register shows that, on 18 February 2015 at 5:21 am, some 30 minutes before the offending in Prahran, the mobile phone alleged to be in the possession of Mr Ward, and containing the 480 number, communicated with a cell on a tower on Plenty Road in Preston. At 5:28 am, the mobile phone communicated with a cell on a tower on High Street in Kew East. At 5:31 am, it communicated with a cell on a tower on Cotham Road in Kew. The next entry, at 5:54 am,[13] showed that the mobile phone communicated with a cell on a tower on Church Street in Cremorne (Richmond South), some 1.5 kilometres from the Aldi supermarket in Prahran. At 6:01 am, it communicated with a cell on a tower on Lansell Road in Toorak. Another communication was detected at 6:05 am with a tower on Tooronga Road in Malvern. At 6:17 am, the mobile phone communicated with a cell on a tower on Glenferrie Road in Malvern. Finally, at 6:24 am, it communicated with a cell on a tower on Toorak Road in Toorak.

64 Mr Prypoten also explained that Optus tested its network regularly. The testing involved an Optus employee travelling in a van with a mobile phone or scanner and taking field measurements.

65 Mr Prypoten was cross-examined at length about the various plots and data which had been produced. In particular, defence counsel asked Mr Prypoten about the field measurement maps which are extracted above (figures 4 and 6).[14] In relation to one of the maps, the following exchange took place:

Q: You have got no data underlying this material, there's just this map?

A: Yeah, I don't have the table explaining each and every point. Yes.

Q: Nor do we know what would have happened if the van had driven all the way up Glenferrie Road - - -?

A: Oh, yeah.

Q: - - - because it just didn't?

A: Yeah, it didn't. Yeah.

Q: Nor do we know what would have happened if it drove all the way down Toorak Road - - -?

A: Yes.

Q: - - - because it didn't?

A: It didn't, yeah.

66 In one particular field measurement map (figure 4), defence counsel observed that a testing van had detected a connection between a mobile phone and a cell in an area which was highlighted light green in the corresponding plot. The transcript reads as follows:

Q: If you look just at figure 4, based on your earlier evidence you might have concluded that there was only a less than 4 per cent change of a mobile phone in that location connecting with the Richmond South tower, because it's in the light green zone?

A: Yes.

Q: Now that the van drove out there and actually measured it, we can say that, at least on that day, it was a 95 per cent chance of connecting?

A: You see, you're referring to one spot.

Q: Yes?

A: I'm referring to the whole area.

Q: Yes, I understand that. I am referring to, at the moment, that one spot?

A: Yeah. Yeah, you're looking at one exceptional case. Yes.

TJ: What counsel is asking you, I think, is that at the time that this recording was taken in that spot ... the prediction would have been that it only had a 4 per cent chance of being the strongest cell, but it turned out it was, in that spot, on that day, 95 per cent chance. That is the position that counsel is putting to you?

A: Yes, you can say that. Yes.

Q: Not only 'can say it', it's simply correct?

A: No, because the 5 per cent statement - sorry, the 4 per cent statement probability applies to the whole area, and it's - so it just happened to be that all those 4 per cent, or higher chance of probability, they just are combined in there. Okay.

Q: Whatever reason, the probabilities on that day went from less than 4 per cent to 95 per cent?

A: Yes, there is an anomaly in there. Yes, correct.

67 At this point, the trial was adjourned until the following morning. Defence counsel had indicated that he had further questions to ask of Mr Prypoten.

68 On 8 December 2016, the cross-examination of Mr Prypoten resumed. Mr Prypoten was taken through evidence which he had given earlier in a Basha inquiry.[15] It will be recalled that, in his evidence in the second Basha inquiry, Mr Prypoten accepted that, in order to determine whether the probability of 95 per cent was an accurate prediction in a geographic location, a statistically significant group of tests would need to be conducted. He also accepted that none of the plots which he produced purported to be the result of a statistically significant group of tests.

69 Mr Prypoten corrected the evidence which he had previously given in relation to the 95 per cent figure and how that figure was calculated. His revised percentage figure for the dark green area was now 91 per cent rather than 95 per cent. Defence counsel reminded him of the answers which he had given the previous day in response to questions about where he had obtained the figure of 95 per cent, including answers which he had given in the absence of the jury. Mr Prypoten said:

There was a change of circumstances since those answers and clarification which I just provided. I got clarification where those 95 per cent are coming from.

He added that the evidence which he had previously given was his 'understanding yesterday' based on his 'yesterday knowledge'.

70 Mr Prypoten was not re-examined. Defence counsel did not ask the trial judge to reconsider his earlier ruling. Nor did he ask the trial judge to give the jury any specific direction about the evidence concerning the 95 per cent figure, or the revised 91 per cent figure.

Closing addresses and trial judge's charge to jury

71 In his closing address, the prosecutor told the jury that it should infer that the appellant's mobile phone was communicating with mobile phone cell towers 'in the relevant areas at the relevant times, whether he's making calls or not'. The prosecutor summarised aspects of Mr Prypoten's evidence as follows:

But, members of the jury, the first thing the prosecution wants to put at this stage is that when a call is made, based on Mr Prypoten's evidence, the mobile phone connects with the strongest cell for that phone, usually the cell which is closest to the device, the mobile phone, with certain exceptions. And he explained that. And then said as a rule of thumb, Mr Prypoten said, the stronger the cell means it's closer to the phone. And even when a call or an SMS is not being made but the phone is on, it is communicating with the cell with the strongest signal and highest priority. You heard Mr Prypoten explain all that.

72 The prosecutor then referred to the plots which were in evidence. He said:

Now, a lot of time was spent on the statistical plots ... with the two blown up plots, which are used to predict coverage in a particular area. The light and dark green areas, with percentage probabilities. You've heard a lot about that ... You've also heard evidence of field measurement data, which was done in order to validate these predictive plots ...

73 The prosecutor also reminded the jury of the visitor location register and took them through that document at length. He also drew their attention to call records which were produced by Mr Finlay and asked them to infer, based on these documents, that the appellant was using the 480 number in particular places at particular times.

74 In his closing address, defence counsel noted that the Crown had no evidence that the appellant was at the locations at which the offending took place at the times at which it was alleged that the offending took place; 'the best the prosecution can do is say that a phone which was registered to his girlfriend was communicating with phone towers broadly in those areas, broadly at those times'.

75 As to Mr Prypoten's evidence, defence counsel said:

Now, it was irresistible when Mr Prypoten gave evidence, to point to the real limitations in that material. The exceptions and inaccuracies in the predictive plots that he used. The areas which were well outside what was designated as the possible area of mobile phone coverage, which actually did have coverage, and in some instances was the strongest cell, the fact that the presentation of this evidence in terms of mathematical probabilities eventually completely collapsed, and was exposed as not being empirically validated at all. But you'll be pleased to hear, I'm sure, I'm going to resist the temptation to take you back to that material, and back to the exhibits which were tendered through Mr Prypoten. You've heard the evidence recently, and you'll be familiar with the limitations in that evidence.

76 In his charge, the trial judge did not specifically address the issue with respect to the percentage probabilities cited by Mr Prypoten in his evidence. Instead, the trial judge gave the following general direction about expert witnesses:

When someone comes into the court and gives evidence from the witness box, I explained this to you in the course of the trial but I think it is worth repeating. The starting point is that they are competent to give evidence and that they can give relevant evidence, so someone like, for example, who has been the subject of an armed robbery can come to court and give evidence about what happened. 'I saw this, I felt this, this happened'.

This is within the competence of a normal person to speak about what they saw, what they felt and what they heard. People who come to court and express opinions though, have to be experts, and here you have had expert evidence from a person who takes fingerprint evidence, a person who examines fingerprint evidence. A person who examines DNA evidence. Telecom people who know about phones.

Now, all of those people cannot say a word to this court unless they establish their expertise. Once they have established their expertise, they can give evidence. But because they are expert, does not make their evidence necessarily persuasive. All it does is, and the expertise of a person qualifies them to express an opinion. But you do not throw out your common sense when someone comes into the court as an expert and expresses an opinion.

If someone came into the court as an expert and said, 'Well, I can tell you is my expertise. It never rains in Melbourne. Okay. I am a meteorologist. I have studied this for a year. We do not get any rain in Melbourne ever', you would not find that - the person may well be - have all the qualifications in the world, but you would not find that too persuasive when you walk out and you get hit by the rain.

So, it is for you to assess the persuasiveness, the reliability of any evidence given in the court, and do not be overborne by someone's qualifications. The qualifications just allow them to speak on the topic. Thereafter, they have to be persuasive and you are the judges of that. And so, you must keep that in mind.

77 The appellant contended that Mr Prypoten's evidence was of critical importance in the trial. He argued that Mr Prypoten should not have been permitted to give evidence by reference to the 95 per cent figure. He said that the figure had not been verified by proper empirical testing and that the fact that Mr Prypoten had sought clarification, in the course of giving evidence, as to what the mathematical figures represented illustrated the infirmities in this aspect of his evidence. According to the appellant, the evidence of statistical likelihoods being led resulted in a substantial miscarriage of justice.

78 The appellant further contended that, in the absence of proper verification, the precise statistical probabilities upon which the Crown relied could neither be confirmed nor convincingly disproven. The appellant argued that the jury should not have been invited to rely upon the percentage probabilities as accurate and that there was an unacceptable danger of the jury being irrationally impressed with that evidence.

79 The appellant also contended that the other evidence linking the appellant to the commission of the offences was far from compelling. He said that there was cogent evidence pointing to the possible involvement of others in the offending.

80 The respondent contended that the trial judge's ruling was correct. It observed that the statistical probability of a mobile phone being in a particular geographic area at a particular time was but one aspect of Mr Prypoten's evidence. The respondent argued that, in the context of the evidence as a whole, it did not assume the importance now being suggested. The respondent also pointed out that the statistics assumed no importance at all during closing addresses and the trial judge's charge. Even if the trial judge fell into error by permitting the evidence to be adduced, argued the respondent, no substantial miscarriage of justice had occurred.

81 A curious feature of the present case is that the challenge to the admissibility of Mr Prypoten's evidence was not framed in terms of pt 3.3 of the Evidence Act 2008 ('the Act') and, in particular, s 79(1), which renders admissible certain evidence of an opinion. Nor was the admissibility of that evidence challenged by reference, in terms, to the exclusionary rule in s 137.

82 Part 3.3 of the Act is concerned with opinion evidence. Section 76(1) states a general exclusionary rule for evidence of an opinion. It provides:

Evidence of an opinion is not admissible to prove the existence of a fact about the existence of which the opinion was expressed.

83 Section 79(1) of the Act establishes the following exception to the opinion rule:

If a person has specialised knowledge based on the person's training, study or experience, the opinion rule does not apply to evidence of an opinion of that person that is wholly or substantially based on that knowledge.

84 In Honeysett v The Queen,[16] the High Court stated that the first condition of admissibility in s 79(1) of the Act directs attention to the existence of an area of 'specialised knowledge', which is to be distinguished from matters of 'common knowledge'.[17] 'Knowledge', in this context, refers to an 'acquaintance with facts, truths, or principles, as from study or investigation', and it connotes more than subjective belief or unsupported speculation.[18] The second condition requires that the opinion be wholly or substantially based on that knowledge. This condition recognises that it will sometimes be difficult to separate from the body of specialised knowledge on which the expert's opinion depends 'observations and knowledge of everyday affairs and events'.[19] As the Court explained, it is sufficient that the opinion is substantially based on specialised knowledge which is based on training, study or experience.[20] The opinion must be presented in a way that makes it possible for a court to determine that it is so based.[21]

85 As a starting point, it is to be observed that much of Mr Prypoten's evidence, including his evidence with respect to the probability that a mobile phone was in a particular geographic area at a particular time, was evidence of an opinion. Such evidence is prima facie inadmissible by the operation of s 76(1) of the Act. The onus of establishing the conditions of admissibility rested on the Crown, as the party seeking to adduce that evidence.[22] One exception to the opinion rule upon which the Crown could have relied is found in s 79(1), which concerns opinions based on specialised knowledge. It was not suggested that any of the other exceptions in pt 3.3 were relevant in the present case. In order to lead such evidence, the Crown had to satisfy the two conditions of admissibility in s 79(1); it had to establish, first, that Mr Prypoten had specialised knowledge, based on his training, study or experience, about the relevant probabilities and, secondly, that his opinion was wholly or substantially based on his specialised knowledge. In the present case, the Crown did neither of these things.

86 During the hearing of the appeal, counsel for the appellant accepted that defence counsel had not, at any stage during the trial, challenged Mr Prypoten's expertise. It seemed to have been assumed that Mr Prypoten had specialised knowledge, within the meaning of s 79(1), to give evidence with respect to certain, albeit unspecified, matters. Defence counsel had not objected to the use of the computer programme which generated the dark green and light green zones on the plots in respect of which Mr Prypoten gave evidence. Furthermore, defence counsel had not objected to Mr Prypoten's use of expressions such as 'likely' or 'highly likely' to describe the probability that a mobile phone was in a particular geographic area at a particular time. Instead, defence counsel objected to Mr Prypoten's expression of those probabilities as a percentage; the challenge, it seems, was based on the lack of empirical basis for expressing the evidence of probabilities as a percentage.

87 In his oral submissions, counsel for the appellant contended that Mr Prypoten's expression of the relevant probabilities as a percentage did not fall within the exception in s 79(1) since it had not been established that Mr Prypoten had specialised knowledge such as to enable him to express the relevant probabilities as a percentage. The percentage figures, said counsel, were unfounded in so far as they lacked an empirical basis and were likely to be misleading or to cause the jury to draw impermissible inferences against the appellant. Alternatively, counsel argued that Mr Prypoten's evidence of the relevant probabilities was not wholly or substantially based on any specialised knowledge which he had on matters about which he had given evidence earlier. Counsel also contended that, if the expression of the relevant probabilities as a percentage fell within the exception in s 79(1), the evidence should nevertheless be excluded under s 137 on the basis that the probative value of the evidence is outweighed by the danger of unfair prejudice to the appellant.

88 All that was adduced about Mr Prypoten's training, study and experience was that he was a technical specialist who is employed by Optus, holds a master's degree in radio engineering and has ten years' experience in the mobile telecommunication industry. In a statement filed by the Crown before the first Basha inquiry, Mr Prypoten deposed that he was responsible for the planning and design of the Optus 2G, 3G and 4G networks. He deposed also that he had made statements as an expert witness 'with more than 20 court appearances' to date and that he has been attending 'skill up' training sessions, provided by mobile phone equipment vendors, every 3-6 months. Mr Prypoten was not cross-examined with respect to any of his qualifications or experience.

89 There was no dispute at trial that, broadly speaking, Mr Prypoten had some form of specialised knowledge within the meaning of s 79(1). However, from his evidence at both Basha inquiries and before the jury, it is difficult to discern the precise matters to which that specialised knowledge related. There was no attempt to explain the connection between that specialised knowledge and Mr Prypoten's training, study and experience, much less that that knowledge was based on his training, study and experience. This made it all the more difficult to determine whether, if at all, Mr Prypoten's opinions, and in particular the impugned evidence, were based on that specialised knowledge.

90 It seems to us that any 'knowledge' which Mr Prypoten purported to have, in so far as that knowledge formed the basis of an opinion of the probability, expressed as a percentage, that a mobile phone was in a particular geographic area at a particular time, amounted to nothing more than unsupported speculation. The basis of the 95 per cent figure, which was said to represent the probability that a mobile phone in the dark green zone on a plot was in a particular geographic location at a particular time, is a matter which was unclear at the time when the trial judge delivered his ruling and which remains even less clear in the light of Mr Prypoten's revision of the 95 per cent figure.

91 True it is that, when asked during the second Basha inquiry what 'generates' the 95 per cent figure, Mr Prypoten referred to the fact that Optus employees inputted certain variables, including 'the deviation of the model', into a computer programme which 'computes the value with 95 per cent probability'. Further, Mr Prypoten had said that 'the 95 percent is the coverage probability or the value calculation probability which returned [sic] by the tool, by the computer software'. However, Mr Prypoten also gave evidence that there was no mathematical formula for the 95 per cent figure and that it was 'just a representation of highly likely'. He added that the 95 per cent figure was based on 'experience' and that it had been used by expert witnesses from Optus over a number of years. He also accepted that none of the plots in evidence purported to be the result of a statistically significant group of tests, which had to be conducted in order to determine whether the 95 per cent figure is an accurate prediction.

92 It seems to us that these concessions are plainly inconsistent with the trial judge's conclusion that 'the figure of 95 per cent is generated as a matter of input into an algorithm which is designed to produce a finding to that degree of certainty based on the algorithm'. In our opinion, the trial judge erred in admitting Mr Prypoten's evidence with respect to the probability, expressed as a percentage, that a mobile phone was in a particular geographic area at a particular time.

93 We would add, by way of observation, that the Crown's approach to adducing Mr Prypoten's evidence, and the defence's approach to challenging the admissibility of that evidence, were unusual and are not to be encouraged. They are approaches riddled with unexplored assumptions. They have given rise to unnecessary difficulties in applying the statutory tests for the exclusion and admissibility of opinion evidence. Whether a witness has specialised knowledge based on his or her training, study or experience, the matters to which that knowledge relates and the nexus between evidence of a particular opinion and the witness's specialised knowledge are matters which must be addressed at trial by reference to the language of the Act. They are not matters to be ascertained for the first time by an appellate court by reference to the record of the trial.

94 As is plain, Mr Prypoten's evidence was evidence of an opinion, and contained several opinions at that. Part 3.3 of the Act governs the admissibility of opinions. Parliament has expressed the opinion rule in s 76(1) such as to direct attention to why the party tendering the evidence says it is relevant.[23] As indicated above,[24] as the party seeking to adduce evidence of an opinion, the Crown bore the onus of establishing the conditions of admissibility under the Act, by reference to the language which Parliament has adopted in the Act. Any challenges to the admissibility of that evidence should likewise have been mounted by reference to the statutory language.[25] In the present case, it appears that there was no challenge to the fact that Mr Prypoten had specialised knowledge based on his training, study or experience. Further, the defence appears to have conceded that he was qualified to give the opinion that the probability that a mobile phone was in a particular location at a particular time was 'likely' or 'highly likely'. However, the defence did not concede that he had been qualified to give the opinion expressed as a percentage.

95 Section 276(1) of the Criminal Procedure Act 2009 provides that this Court must allow an appeal against conviction if the appellant satisfies the Court that:

(a) the verdict of the jury is unreasonable or cannot be supported having regard to the evidence; or

(b) as the result of an error or an irregularity in, or in relation to, the trial there has been a substantial miscarriage of justice; or

(c) for any other reason there has been a substantial miscarriage of justice.

96 In the light of our conclusion that the trial judge erred in admitting Mr Prypoten's evidence with respect to the probability, expressed as a percentage, that a mobile phone was in a particular geographic area at a particular time,[26] the inquiry shifts to whether that error resulted in a substantial miscarriage of justice.[27]

97 The principles governing this question are not in dispute. They were set out by the High Court in Baini v The Queen[28] and summarised by this Court in Andelman v The Queen,[29] where Maxwell P, Weinberg and Priest JJA said:

• Section 276 is not to be interpreted solely by reference to the interpretation given to the common form criminal appeal proviso in Weiss v The Queen.[30] Comparing 'a statute with its legislative predecessor ... is only a useful exercise if doing so illuminates the actual text of the new provision'.[31]
• There is no 'single universally applicable' definition of a 'substantial miscarriage of justice'. That is because the possible kinds of miscarriage of justice dealt with by s 276(1) are 'too numerous and too different to permit prescription of a singular test'.[32]
• With respect to ss 276(1)(b) and (c) (with which this appeal is concerned), the types of substantial miscarriage of justice include cases where 'there has been an error or an irregularity in, or in relation to, the trial and the Court of Appeal cannot be satisfied that the error or irregularity did not make a difference to the outcome of the trial'.[33]
• Those paragraphs also cover cases where 'there has been a serious departure from the prescribed processes for trial'.[34]
• A 'substantial miscarriage of justice' may occur where there has been a 'departure from process' even if the 'verdict was open or it is not possible to conclude whether the verdict was open'.[35]
• The question whether there has been a 'substantial miscarriage of justice' 'may be affected by the strength of the prosecution case'. In such cases, however, the Court of Appeal must be aware of the 'natural limitations that attend the appellate task'.[36]
• A finding that the conviction was 'inevitable' is merely 'relevant' to the Court's determination of whether there has been a substantial miscarriage of justice. It does not conclude the issue.[37]
• If it is submitted that the verdict was inevitable, the appellant need show 'no more than that, had there been no error, the jury may have entertained a doubt'.[38]
• In assessing 'inevitability', the Court of Appeal must 'decide that question on the written record of the trial'.[39]
• In cases such as Baini where evidence has been wrongly admitted or excluded, the Court cannot determine that there has been no substantial miscarriage of justice 'unless it determines that it was not open to the jury to entertain a doubt as to guilt. Otherwise, there has been a substantial miscarriage of justice because the result of the trial may have been different (because the state of the evidence before the jury would have been different) had the error not been made'.[40]

98 In our opinion, the admission of the impugned evidence did not result in a substantial miscarriage of justice. Three points should be made.

99 First, the impugned evidence assumed little, if any, importance to the outcome of the trial. At any rate, it assumed much less importance than was suggested by the appellant. The controversy surrounding the impugned evidence boiled down to the expression of the probability that a mobile phone was in a particular geographic area at a particular time in quantitative terms rather than qualitative terms. The defence had not objected to terms such as 'highly likely' to describe that probability. We are not persuaded that the conversion by Mr Prypoten of the term 'highly likely' into the 95 per cent figure (which was subsequently reduced to 91 per cent on the day after Mr Prypoten gave that evidence, after he reconsidered the matter) was at all significant in the jury's determination of whether the appellant committed the offences of which he was convicted. Moreover, the unexplained basis of either the 95 per cent figure, or the revised 91 per cent figure, would have led the jury to reduce whatever little weight that might have been attached to that particular evidence almost to a nullity.

100 Secondly, the expression of the probability, in percentage terms, that a mobile phone was in a particular geographic area at a particular time received no attention during closing addresses and the trial judge's charge to the jury. In his closing address, defence counsel seized on the opportunity to cast doubt over that evidence, tellingly observing that it 'completely collapsed, and was exposed as not being empirically validated at all'. The trial judge's general direction about expert witnesses and warning of the danger of the misuse of expert evidence in his charge to the jury was a proportionate response to the prosecutor's and defence counsel's treatment of that evidence in their closing addresses.

101 Thirdly, there was ample evidence from which the jury could draw an inference of the appellant's guilt. That evidence, which is set out in detail above, may be summarised as follows:

(a) The 480 number was registered in the name of Ms Breheny. On 18 February 2015, Ms Breheny was the appellant's de facto partner, and they were living together at an address in Reservoir.

(b) The offending took place on 18 February 2015 in Prahran (at about 5:50 am) and Malvern (at about 7:00 am). Between 5:21 am and 6:24 am on 18 February 2015, a mobile phone containing the 480 number, which was alleged to be in the possession of the appellant, communicated with cells on towers in Preston (which, of course, is close to Reservoir, where the appellant resided) (5:21 am), Kew East (5:28 am), Kew (5:31 am) (which, of course, indicates a transition from the northern suburbs to the eastern suburbs, either at Burke Road or the Chandler Highway), Cremorne (Richmond South) (which, of course, is close to Prahran, where the Aldi supermarket was located) (5:54 am), Toorak (6:01 am), Malvern (6:17 am) and then Toorak (which is adjacent to Prahran) (6:24 am). The movement of the telephone communications to towers at times and locations proximate to the locations in which the offending took place was highly significant. So too was the fact that the communications before 5:54 am indicated that the appellant had been travelling in the direction of the Aldi supermarket in Prahran shortly before the offending there took place. After that time, it appears that the appellant had been travelling in the general direction of the café in Malvern at which the armed robbery took place.

(c) During the offending, one of the offenders had a backpack strapped to his chest, and the two offenders fled both scenes in a white BMW with the registration plate number 1AA8ZF. Police located the white BMW in Bradshaw Street, Kingsbury, several hours after the offending.

(d) On 18 February 2015, shortly before 9:00 pm, a booking for a taxi was made under the name of 'Paul' and recorded a number which, save for the last three digits,[41] was identical to the 480 number. The taxi picked up a passenger from the appellant's residential address in Reservoir, where he lived with his de facto partner (the owner of the 480 number), and drove the passenger to the vicinity of Bradshaw Street, Kingsbury.

(e) On 18 February 2015, shortly after 9:00 pm, police observed the appellant entering the white BMW with a key and driving it around the streets of Kingsbury. Police later observed the appellant and an associate removing a backpack from the boot of the white BMW.

(f) A police officer, Senior Constable Vince Naidu, who attended the scene of the armed robbery at the café in Malvern at about 7:30 am on 18 February 2015, and witnessed the pursuit of the white BMW, gave a description of the driver as a Caucasian with olive complexion and blonde spikey gelled hair. A police log, which recorded the movements of the BMW at about 9.00 pm that same evening, described the driver as having 'spikey fair hair'.

(g) Some fingerprints which were detected on the tailgate of the white BMW matched the fingerprints of the appellant.

(h) The appellant was recorded shortly before midnight on that same day, speaking on his partner's phone, and telling the recipient of the call that he was in a BMW X1, and needed a rim and a tyre. That would accord with any damage that might have been sustained to the vehicle earlier that day when it mounted a curb during the police pursuit.[42]

102 This was a circumstantial case, but an extremely powerful one. It would be a remarkable coincidence for the appellant's partner's phone to have communicated sequentially with the towers along the route described, corresponding with the times of the offences, were there to be no connection between the movement of that phone and the offences themselves.[43] As the High Court made clear in R v Hillier,[44] in dealing with a circumstantial case, individual pieces of evidence should not be looked at in isolation. Had this been an application for leave to appeal against conviction on the basis that the verdict was unsafe or unsatisfactory, the question for this Court would have been whether it was open to the jury, on consideration of the whole of the evidence, to be satisfied beyond reasonable doubt that the appellant was guilty. In determining whether there was a substantial miscarriage of justice, and as part of that process, evaluating the strength of the prosecution case, the whole of the evidence must be considered.

103 As indicated above, Mr Prypoten should not have been permitted to express his opinion as to the high likelihood of the mobile phone being within the dark green areas of the relevant plots in percentage terms. We note that, in a trial conducted in a different matter in the Supreme Court in October 2017, Coghlan JA, having first permitted Mr Prypoten to use such percentage terms, considered the matter and instructed the jury to disregard his evidence in so far as it was expressed in that way. Coghlan JA recognised the deficiencies in that witness's capacity to convert a qualitative assessment into one which misleadingly signified a level of confidence in statistical terms which could not be supported by his specialised knowledge.[45]

104 Having regard to the entirety of the evidence, we consider that, notwithstanding the fact that Mr Prypoten's evidence as to percentages should not have been given in that form, the convictions on the charges involving the attempted armed robbery, the armed robbery and the reckless conduct endangering life on indictment 6.2 were inevitable. The difference between the expression of the probability that the mobile phone containing the 480 number was in a particular geographic area at a particular time as a percentage, and the expression of that probability in qualitative terms (such as 'highly likely') was, at best, marginal. Whatever significance the jury may have attached to the percentage figures must have been dissipated during cross-examination, particularly as Mr Prypoten showed himself unable, coherently, to explain how he had arrived at the initial figure of 95 per cent and his subsequent downward revision of that figure. The matter of percentage did not receive any attention during closing addresses or the trial judge's charge. All things considered, the jury could not have entertained a reasonable doubt as to the appellant's guilt.

105 The appeal against conviction should be dismissed.

106 We now turn to consider the application for leave to appeal against sentence. After a plea hearing on 14 December 2016, the trial judge sentenced the appellant on 10 March 2017 as follows:

Indictment 6.3: Aggregate term of 7 years' imprisonment

107 Having been refused leave to appeal against his sentence by Whelan JA, the appellant has elected to renew his application for leave on the following ground:

The orders for cumulation made by the learned sentencing judge for the offences on indictments [6.1] and [6.3]:

(a) breached the principle of totality; and

(b) resulted in a manifestly excessive total effective sentence.

108 At the conclusion of oral argument before this Court, counsel for the appellant informed the Court that the appellant sought to amend his application for leave to appeal to add a proposed ground of manifest excess across the total effective sentence. Senior counsel for the respondent did not object to this course. In the event, the Court directed the parties to file and serve written submissions in relation to the proposed ground.

109 The additional proposed ground is in the following terms:

The individual sentences imposed on each of the charges on each of the Indictments, the orders for cumulation made, the resulting total effective sentence and the non-parole period fixed are manifestly excessive.

110 The respondent did not object to the proposed amendment. Accordingly, we will grant the appellant leave to add the second proposed ground in the terms set out above.

Circumstances of the offending - indictments 6.1 and 6.3

111 As mentioned above, the appellant pleaded guilty to one charge of theft on indictment 6.1. The Crown alleged that the appellant had stolen a white BMW X1, with the registration 1AA8ZF, sometime between 8:30 pm on 12 February 2015 and 8:00 am on 13 February 2018. The vehicle had been stolen from Cyril Street in Elwood. The Crown did not allege that the appellant himself stole the vehicle from Cyril Street.

112 The circumstances of the offending the subject of the charges on indictment 6.3 may be summarised as follows.

113 On 1 March 2015, Trevor Rowe parked his grey Ford XR 6, registration UNA797, outside Victoria Park in Collingwood. He left his bag and car keys in a change room. Later that day, he discovered that his car keys had been stolen. He reported the theft to police.

114 On 5 March 2015, police observed the appellant and an associate stealing a registration plate (1AT2FW) from a 2013 Ford Falcon.[46] Later that day, police observed the appellant driving the grey Ford XR 6 which belonged to Mr Rowe. The stolen registration plate was attached to the rear of the car.[47]

115 On 7 March 2015 at about 2:50 am, the appellant and one unknown person attended a pharmacy in Sunshine. The appellant had driven them in a vehicle that matched the description of the vehicle stolen in Collingwood on 1 March 2015 (UNA797). At the time of arrival, the store was closed for trading and the front entrance was locked. The appellant remained in the vehicle.

116 The appellant reversed the vehicle up the front ramp of the pharmacy and struck the front entrance glass doors with the rear of the vehicle, causing the glass to smash. The unknown person then entered the pharmacy with a wheelie bin. The appellant remained in the vehicle.[48] Once inside the pharmacy, the unknown male filled the wheelie bin with approximately 30 boxes of cold and flu tablets. He fled the store via the smashed front entrance and returned to the vehicle driven by the appellant. The appellant and the unknown person drove away from the scene.[49]

117 The trial judge said that the appellant has a significant criminal history which included offences involving dishonesty and violence.[50] The offences included (a) convictions in May 2013 with respect to the theft of a motor vehicle, going equipped to steal, possession of an unregistered firearm and ammunition and dealing with the proceeds of crime; (b) convictions in February 2012 of making a false document and using the false document to prejudice another; (c) a conviction in April 2010 of the attempted theft of a motor vehicle, burglary and theft; (d) convictions in May 2009 of theft of a motor vehicle and reckless conduct endangering life; and (e) multiple offences in 2007 of dishonesty and violence including an aggravated burglary and intentionally causing serious injury for which the appellant was sentenced to 14 months' imprisonment in a youth justice centre.

118 The trial judge noted that, on 28 April 2009, the appellant was sentenced in the County Court to four years and nine months' imprisonment (six months of which was served concurrently) on one charge of armed robbery and on one charge of possessing an unregistered firearm as a prohibited person. A non-parole period of two years and three months was fixed.[51]

119 The trial judge said that the armed robbery involved the appellant entering a bakery armed with a sawn-off shotgun which he pointed at a woman who was managing the bakery.[52] The appellant demanded that she provide him with the contents of the till. As the trial judge observed, the offending bears a striking similarity to the charge of armed robbery on indictment 6.2.[53]

120 The trial judge turned to the appellant's personal history.[54] The appellant was raised by his mother. His parents separated when he was young. At the age of 15, the appellant left the family home to live in a friend's house.

121 As to the appellant's educational and work history, the trial judge said that the appellant dropped out of school in Year 9 and started apprenticeships in carpentry and electronics. He did not complete either of these apprenticeships. However, he holds certificates in construction, carpentry, sand blasting and powder coating. He worked as a meatworker, starting in 2007, and continued in that role for some years before he sustained an injury to his right hand.

122 The trial judge said that the appellant has a long history of abusing alcohol and other drugs.[55] He started drinking alcohol at the age of 13 and eventually started using cannabis and methamphetamine. At the time of the present offending, the appellant had been using two or more grams of methamphetamine daily, combined with the daily use of cannabis. During the plea, counsel for the appellant said that the appellant has tried to control his drug use since his incarceration and even sought a transfer to another prison for the reason that the current prison was 'full of drugs and I wanted to get away from it'.[56]

123 A report dated 17 February 2017 prepared by Dr David Ball, a forensic psychologist, was tendered on the plea on the appellant's behalf. The trial judge summarised the contents of this report.[57] Dr Ball said that the appellant presented with no symptoms of depression or anxiety and showed no obvious signs of mental illness. Dr Ball described the appellant as having an impaired capacity to exercise good judgment and a lengthening criminal history in association with a history of substance abuse and living on the fringes of the community. He described the appellant as no longer satisfying any DSM-V diagnostic criteria for substance use disorders but as presenting with the DSM-V diagnostic criteria for Antisocial Personality Disorder.

124 In his report, Dr Ball considered the appellant to have the potential of being employable immediately. However, he said that the appellant would require monitoring, supervision and support when he is released into the community. He also said that the appellant would need structured and intensive drug and alcohol relapse prevention treatment.

125 The trial judge noted that, in respect of the three charges on indictment 6.2, 'it is clear from your statements made to Dr Ball that you maintain your innocence and present with no true remorse'.[58]

126 The trial judge said that he gave due weight to the fact that there was some delay in the matter coming to trial, through no fault of the appellant.[59]

127 The trial judge accepted that the charges of armed robbery and attempted armed robbery lacked sophistication and were not well planned.[60] As to the charge of attempted armed robbery, the trial judge did not accept that the offending was based on an impulse as it 'involved a co-offender, the use of a stolen vehicle, the use of a sledgehammer, balaclavas and weapons'.[61] He also said that it is unlikely that the armed robbery of the café in Malvern occurred by chance, having regard to the location of the back entrance of the café which is accessed via a narrow laneway.[62]

128 The trial judge described the armed robbery as 'an instance of serious offending which falls at the more significant end of the range of activity involved in offences of that nature'.[63] Despite there being no victim impact statements, the trial judge was satisfied, having regard to the CCTV footage of the offending, that the armed robbery must have exposed the victims to considerable trauma.[64]

129 As to the charge of reckless conduct endangering life, the trial judge said that the appellant's driving exposed passengers in other vehicles and pedestrians to possible death.[65] He considered the offence to fall within the upper range of the activity involved in this type of offending.[66]

130 The trial judge also considered the charges of theft and burglary on indictment 6.3 as involving serious offending: the offending was planned; it involved damage to property; and a significant quantity of medication was stolen.[67]

131 The trial judge was satisfied that he should 'significantly' discount the sentence that he would otherwise have imposed on the charges of theft and burglary by reason of the appellant's plea of guilty in circumstances where the Crown had conceded that the case against the appellant was not all that strong.[68] The trial judge said that the appellant was also entitled to a discount, by reason of his plea of guilty, on the sentence that the trial judge would have otherwise imposed on the charge of theft on indictment 6.1.[69]

132 The trial judge accepted that the appellant is well supported by his family, particularly his mother, and that he has a number of skills which provide him with reasonable prospects for employment.[70] He also accepted that, while in prison awaiting his trial, the appellant had taken steps to control his drug dependency and that this, together with the support from his family, 'potentially speaks positively on the issue of your rehabilitation prospects'.[71] However, the trial judge said that the appellant's long history of offending and the nature of the present offending led him to be guarded as to whether the appellant has positive prospects of being rehabilitated.[72]

133 The trial judge said that, given the nature of the offending, 'issues of punishment deterrence and community protection [are] at the forefront' in sentencing the appellant.[73] He also referred to his obligation to ensure that the totality of the appellant's sentence remains just and appropriate in respect of the whole of the offending.[74]

134 In his written submissions, the appellant contended that the six months' cumulation ordered on the charge of theft of a motor vehicle on indictment 6.1 was excessive and breached the principle of totality. The appellant argued that it would have been appropriate for his sentence on that charge to be wholly concurrent. He said that, in sentencing him on the charges of armed robbery and attempted armed robbery, the trial judge had taken into account the appellant's use of a stolen vehicle. Accordingly, so it was submitted, the need for care in considering the appropriate level of cumulation on this charge was particularly acute.

135 As to the proposed ground of manifest excess, the appellant contended that the sentences imposed were manifestly excessive in view of the appellant's family support; his employment history; his pleas of guilty to the charges on indictments 6.1 and 6.3; his completion of a 12-hour drug and alcohol course while on remand, as well as the work which he had undertaken in custody (both of which were brought to the attention of the trial judge on the plea); the principles of totality and parsimony; and the effect of delay in the appellant's case.

136 There are three reasons why the appellant's contentions in support of the first proposed ground of appeal must be rejected.

137 First, the charges on indictment 6.3 involved conduct which was separate and distinct from the conduct the subject of the charges on the other indictments. The offending itself could only be described as serious. An order for concurrency on any of charges 2 or 3 on indictment 6.3, in all of the circumstances, would have been lenient and would not have adequately reflected the appellant's overall criminality. We do not consider that the orders for cumulation offended the principle of totality or resulted in a total effective sentence which was manifestly excessive

138 Secondly, as to the order for cumulation on the charge of theft on indictment 6.1, it appears that the trial judge took into account the matter of the stolen vehicle for a limited purpose when sentencing the appellant on the charge of attempted armed robbery on indictment 6.2.[75] More precisely, the matter was taken into account in so far as it showed that that particular offending was not committed as a matter of impulse.[76] Again, we are not persuaded that the order for cumulation offended the principle of totality or resulted in a total effective sentence which was manifestly excessive.

139 Thirdly, the appellant has prior convictions for theft of a motor vehicle committed between 2005 and 2013. Thus, specific deterrence assumed an important role in assessing the appropriate sentence and, in particular, the orders for cumulation on the charge of theft on indictment 6.1 and charge 2 on indictment 6.3.

140 It remains to consider the second proposed ground of appeal. In our opinion, this ground is without substance. It was reasonably open to the trial judge to impose the sentence which he did. In the light of the gravity of the offending, the appellant's moral culpability and his plea of not guilty on indictment 6.2, we consider that the sentence of 10 years and 6 months' imprisonment with a non-parole period of 7 years and 6 months was within the permissible range of sentences available to the trial judge.

141 In particular, we consider that each of the individual sentences imposed was within range. Both the attempted armed robbery and the armed robbery were serious examples of these offences. Each involved the use of a firearm and carried maximum penalties of 20 and 25 years' imprisonment, respectively. The trial judge assessed the gravity of the offending as being at the serious end of the range. That conclusion was plainly correct. The motor vehicle pursuit was also serious, as it involved a chase through a number of intersections at high speed. The appellant has a poor record and had previously been sentenced for armed robbery.

142 The appellant's election to renew his application for leave to appeal against sentence should be refused.

[1] Ward v The Queen [2017] VSCA 222.

[2] These circumstances gave rise to charge 1 on indictment 6.2.

[3] These circumstances gave rise to charge 3 on indictment 6.2.

[4] These circumstances gave rise to charge 4 on indictment 6.2.

[5] See R v Basha (1989) 39 A Crim R 337.

[9] Ward v The Queen [2017] VSCA 222.

[10] At the time of his arrest on 7 March 2015, the appellant was in possession of two other telephones, neither of which had used the 480 number. Police did not recover a phone which had used the 480 number.

[11] In his closing address to the jury, the prosecutor contended that the last three digits of the mobile phone number were erroneously recorded.

[12] See [37] above.

[13] This offending at the Aldi supermarket in Prahran took place at 5.50 am.

[14] See [52] above.

[15] See [31] above.

[16] (2014) 253 CLR 122 ('Honeysett').

[18] Ibid 790-1 [23] (emphasis in original) (citations omitted). See also Dasreef Pty Ltd v Hawchar [2011] HCA 21; (2011) 243 CLR 588, 602-3 [32] (French CJ, Gummow, Hayne, Crennan, Kiefel and Bell JJ) ('Dasreef').

[19] Honeysett (2014) 253 CLR 122, 132 [24], quoting Velevski v The Queen [2002] HCA 4; (2002) 187 ALR 233, 268 [158] (Gummow and Callinan JJ).

[20] Honeysett (2014) 253 CLR 122, 132 [24].

[21] Ibid, citing HG v The Queen [1999] HCA 2; (1999) 197 CLR 414, 427 [39] (Gleeson CJ).

[22] Dasreef [2011] HCA 21; (2011) 243 CLR 588, 602-3 [31].

[23] Dasreef [2011] HCA 21; (2011) 243 CLR 588, 602 [31].

[24] See [85] above.

[25] See, eg, Tuite v The Queen [2015] VSCA 148; (2015) 49 VR 196.

[26] See [92] above.

[27] See Criminal Procedure Act 2009 s 276(1).

[28] [2012] HCA 59; (2012) 246 CLR 469 ('Baini').

[29] [2013] VSCA 25; (2013) 38 VR 659.

[30] [2005] HCA 81; (2005) 224 CLR 300.

[31] Baini [2012] HCA 59; (2012) 246 CLR 469, 478 [20] (French CJ, Hayne, Crennan, Kiefel and Bell JJ).

[40] Ibid (emphasis in original) (citations omitted). For a recent decision by the High Court on the application of the common form of the proviso (as distinct from that now set out in s 276(1) of the Criminal Procedure Act 2009), see Kalbasi v Western Australia [2018] HCA 7, which concerned a misdirection by a judge that was held not to occasion a substantial miscarriage of justice.

[41] See fn 10 above.

[42] See [10] above.

[43] The appellant did not give evidence in the trial. The notion that some other person, the actual offender, might have had possession of the appellant's partner's phone earlier in the day, while committing the offences, but the phone had somehow been restored to the appellant later that same day, is almost fatuous. In any event, the defence put forward was that the evidence of the movements of the phone was unreliable, not that the appellant had loaned the phone to some other person.

[44] [2007] HCA 13; (2007) 228 CLR 618.

[45] Coghlan JA instructed the jury to disregard that part of Mr Prypoten's evidence in which he sought to convert an assessment of 'high probability' into a percentage figure, noting that those percentages do not reflect the percentage of the likelihood of the calls made from those areas. Rather, the dark green areas were the product of a computer generated algorithm designed for a wholly different purpose: the planning of coverage which could be achieved by the particular configuration of the towers and cells throughout Melbourne. Transcript of Proceedings, DPP v Tamvakis (Supreme Court of Victoria, Justice Coughlan, 9 October 2017).

[46] These circumstances gave rise to charge 1 on indictment 6.3.

[47] These circumstances gave rise to charge 2 on indictment 6.3.

[48] These circumstances gave rise to charge 3 on indictment 6.3.

[49] These circumstances gave rise to charge 4 on indictment 6.3.

[50] DPP v Ward [2017] VCC 226 [15] (Judge Saccardo) ('Sentencing remarks').

[75] See [127] above, citing Sentencing remarks [35].