COURT OF APPEAL FOR ONTARIO

Date: 2023-01-09 Docket: C65485

Before: Lauwers, Huscroft and Miller JJ.A.

BETWEEN

His Majesty the King Respondent

and

Loujack Café Appellant

Counsel: Myles Anevich and Peter Ketcheson, for the appellant Elise Nakelsky, for the respondent

Heard: December 12, 2022

On appeal from the conviction entered on November 29, 2016 and the sentence imposed on December 16, 2016 by Justice James A. Ramsay of the Superior Court of Justice, sitting with a jury.

Lauwers J.A.

[1] This appeal concerns violent attacks on five individuals at different times in downtown Hamilton that took place between March and December 2011.

[2] The appellant, Loujack Café, lived in Hamilton at the time of the underlying offences. He was convicted on four attempted murder charges and one aggravated assault charge, all related to random attacks on unsuspecting victims, all of whom sustained severe injuries. The appellant was also charged with two murders. The trial judge, then sitting as motion judge, severed the murder charges on February 18, 2014. One murder charge was withdrawn on March 20, 2015. The appellant was convicted on August 24, 2015 of the first degree murder of George Washington Burnett and a life sentence was imposed on him, which the appellant unsuccessfully appealed: R. v. Café, 2019 ONCA 775, 381 C.C.C. (3d) 98.

[3] The jury trial on the charges that were not severed ended with guilty verdicts. The appellant was sentenced to life imprisonment on the convictions for the four attempted murders and five years concurrent on the aggravated assault. No credit was given for any pre-sentence custody because the trial judge attributed that custody to the appellant’s murder conviction.

A. The Issues

[4] The appellant raises three issues:

1. The trial judge erred in refusing his application under s. 11(b) of the Canadian Charter of Rights and Freedoms to dismiss the charges based on undue delay under the principles of R. v. Jordan, 2016 SCC 27, [2016] 1 S.C.R. 631, particularly in his characterization and calculation of various delays and the “net delay”.
2. The trial judge erred in his treatment of cross-count similar fact evidence and in his jury instructions on its permissible use.
3. The sentence was unfit.

[5] I address each issue in turn. I would dismiss the conviction and sentence appeals for the reasons that follow.

B. Issue One: The Jordan application

[6] The appellant was arrested on January 27, 2012. It is common ground that the total delay in this case is 57 months, well beyond the 30-month standard prescribed in Jordan.

[7] This case proceeded under the principles set out by the Supreme Court of Canada in R. v. Morin, [1992] 1 S.C.R. 771. The trial on the charges underlying this appeal began on November 7, 2016, four months after the release of Jordan on July 8, 2016. The defence brought a motion at the outset of trial to dismiss the case under s. 11(b) of the Charter for undue delay. The appellant argues that delay was a live issue even before Jordan was released. In assignment court on October 17, 2014, the appellant addressed the court and asserted his s. 11 (b) rights, requesting that all his trial dates are set as soon as possible: “I do have an 11-B issue, because I have been in incarceration for more than 30 months. Now, I am being told I can set trial for the George Burnett case, the murder, … but I would like to set trials for the other cases, as well, and not hafta wait year after year …” The assignment court judge urged the appellant to take the matter up with his lawyer.

[8] The trial judge noted the appellant’s October 17, 2014 assignment court statement in his reasons on the Jordan application. At assignment court on December 19, 2014 the appellant repeated his complaint. However, in the absence of his counsel, the matter was put over to January 23, 2015. The endorsement on that date makes it plain that counsel had agreed that the murder trial should proceed on July 27, 2015, followed in due course by the trials on the remaining charges. On February 20, 2015, defence counsel appeared and agreed to February 8, 2016 for the three-week trial of these charges.

[9] The trial would have been completed well before the Supreme Court issued its decision in Jordan. However, defence counsel was not available for the February 8, 2016 trial date and it was vacated. At assignment court on November 20, 2015, the trial was re-scheduled for November 7, 2016 and the accused expressly waived his rights under s. 11(b) of the Charter on that occasion. The agent for defence counsel stated: “Yes, 11(b) is not an issue. Mr. Hicks and I were hoping that there would be something available sooner, rather than one year from now. Of course, again, 11(b) won’t be an issue, in any case.” It turned out that the rescheduled date was about 12 weeks after the Supreme Court released the Jordan decision.

[10] The trial judge was fully familiar with all of the information. In his view:

Looking at the case from a bird's eye view, the Crown pushed for an early trial from day one . They mounted a vigorous defence to the motion to sever the counts. Severance of the counts necessitated a certain amount of delay.

The court was diligent to keep the matter from being forgotten. Monthly assignment court appearances were required. Counsel were constantly required to inform the court about availability. …

The accused, on the other hand, was in no hurry. He made the understandable choice to prefer his right to counsel of choice over his right to a trial within a reasonable time. In his personal protests to the contrary in the fall of 2014, he was going against his counsel's advice. Neither the court nor the Crown could very well intervene between the accused and his chosen counsel. He had to sort that out himself. When he retained Mr. Hicks, he agreed to the chosen dates, and expressly waived his right under 11(b) on two occasions.

Essentially , then, this trial was delayed because the accused kept changing lawyers. The counsel chosen were all successful lawyers with busy practices. The delay beyond prima facie acceptable limits was solely caused, or waived by the defence, as the case may be.

[11] The trial judge explained his reasoning for that conclusion. He did not reach the ground of “transitional exceptional circumstance” established in Jordan although counsel argued it.

[12] Taking into account defence delay, the trial judge computed the net delay to be 26 months, which is within the Jordan standard. However, the trial judge’s decision contains an obvious mathematical error because he said: “ The total delay is 57 months. The defence delay is 49 months. The net delay is 26 months.” I would ascribe no importance to the mathematical error. The trial judge’s reasons are more than adequate to explain his finding that the Jordan standard was met.

[13] The appellant invites this court to undertake a detailed Jordan analysis followed by a detailed Morin analysis, arguing that this case was beyond the acceptable time limits on either standard. I would not accept this invitation.

[14] Instead, I would follow this court’s approach in R. v. Earle, 2021 ONCA 34, 399 C.C.C. (3d) 302. In Earle, this court set out the governing principles concerning the application of s. 11(b) of the Charter, which I need not repeat, and then made several observations concerning Jordan’s “transitional exceptional circumstance”. Speaking for the court, I noted, at paras. 14-15:

In Jordan, the Supreme Court held that the transitional exception is to be applied contextually “when the Crown satisfies the court that the time the case has taken is justified based on the parties’ reasonable reliance on the law as it previously existed” even where, as here, “the delay exceeds the [Jordan] ceiling” (emphasis in original): at para. 96, referring to Morin. The court continued: “This requires a contextual assessment, sensitive to the manner in which the previous framework was applied, and the fact that the parties' behaviour cannot be judged strictly, against a standard of which they had no notice.” The concluding sentence is especially telling: “Of course, if the parties have had time following the release of this decision to correct their behaviour, and the system has had some time to adapt, the trial judge should take this into account.”

The changes wrought by Jordan were intended to be sweeping, but not abrupt: “Ultimately, for most cases that are already in the system, the release of this decision should not automatically transform what would previously have been considered a reasonable delay into an unreasonable one”: at para. 102. The court added, at para. 98: “We rely on the good sense of trial judges to determine the reasonableness of the delay in the circumstances of each case.”

[15] I then observed, at para. 33:

This is a case to which the transitional exceptional circumstance identified in Jordan plainly applies. Because the trial started so soon after the release of Jordan, within 10 weeks, this case did not require the intensive analysis under Jordan and then under Morin undertaken by the parties and the trial judge.

[16] This court agreed with the trial judge’s observation in Earle about the impact of Jordan: ten weeks after Jordan’s release, the parties had no “realistic opportunity” to adapt to the new framework. This case, at 12 weeks, is the same. To paraphrase para. 36 of Earle, the release of Jordan did not automatically transform what would previously have been considered a reasonable delay into an unreasonable one.

[17] I would take no issue with the trial judge’s determination under Jordan, to which this court must defer. The appellant has not pointed to any palpable or overriding factual errors, or any errors of law that would warrant this court's interference. But I would add that any such error would, in any event, have brought into play Jordan’s “transitional exceptional circumstance”, as I have explained, especially given the defence’s express s. 11 (b) waiver on October 30, 2015 that was repeated on November 20, 2015.

[18] I would dismiss this ground of appeal.

C. Issue Two: Cross-count Similar Fact Evidence

[19] I begin with a brief contextual synopsis of the evidence concerning the five individuals who were attacked, then set out the trial judge’s view of the similarities on which he based his two rulings on the use of cross-count similar fact evidence, and next address the appellant’s more detailed arguments.

[20] The individuals who were attacked:

1. Kyle Warden approached the appellant in the early morning hours of March 19, 2011 and asked to purchase marijuana. On Mr. Warden’s account, he followed the appellant through an alleyway, where the appellant stabbed him three times in the neck, and when Mr. Warden turned to run, the appellant stabbed him twice in the back. Mr. Warden suffered serious injuries including an internal jugular injury.
2. David Arsenault was walking home from a Salvation Army soup truck on November 7, 2011 at 10:30 p.m., when he was attacked from behind and stabbed. There were wounds to his shoulder, neck, and fingers. Mr. Arsenault underwent emergency surgery, including having an artery cut from his leg and put into his shoulder, and suffered nerve damage.
3. Jennifer McGrath was walking home from the YMCA, on November 10, 2011, just after 7:30 p.m., when she was attacked from behind. She suffered two stab wounds to her upper back, a wound behind her right ear and down her neck, a soft tissue injury around her hairline, a neck fracture, and a skull fracture.
4. Christian Savory was walking to a scheduled methadone clinic appointment early in the morning on December 12, 2011. After meeting a friend at Tim Hortons on his way to the appointment, Mr. Savory was attacked by an individual who drove by on a bike and inflicted a single wound to Mr. Savory’s lower back and then drove off. Mr. Savory was stabbed between his spleen, lung, and main artery, near his kidney.
5. Jin Xu was walking with her six-year-old daughter on December 28, 2011, at 5:15 p.m., when an individual came up from behind her and with a single push knocked her to the pavement. Her face hit the pavement and she suffered serious injuries to her face, jaw and teeth. She lost consciousness and sustained a fracture of her maxilla bone (above her front teeth). She lost one tooth and two other teeth were broken.

(1) The trial judge’s view of the similarities of the acts

[21] The Crown’s factum sets out a summary of the trial judge’s reasoning about the similarities in the attacks. The appellant accepts this list as an accurate recapitulation of the trial judge’s reasoning:

[The trial judge] … decided that the evidence on counts 2, 3, and 4 was admissible in relation to the others as similar fact evidence after evaluating the objective unlikelihood of coincidence based on numerous factors, including:

• proximity in time and place – downtown Hamilton, over the fall months of 2011
• 3 vulnerable victims (only 4 random attacks in the area during that time frame)
• the victims were walking alone at night
• random attacks by a stranger
• similarity in detail and circumstance – no robbery, motiveless crimes
• description of perpetrator – male, black, hoodie, dark clothing
• high degree of similarity in the manner in which the crimes were committed
  • use of weapon (a knife)
  • victims were stabbed from behind
  • he inflicted life threatening injuries, ie. stab wounds to the upper back, neck, head
• perpetrator fled and left no evidence at the scene
• no collusion – the victims did not know each other.

In both rulings, the judge noted the difference that a bicycle was used during one offence which he concluded was “incidental” and did not “necessarily detract from the similarity in the legal sense.”

[22] The trial judge ultimately ruled that the attacks on Arsenault, McGrath, and Savory could be put to the jury as cross-count similar fact evidence given the “high degree of similarity” between them.

(2) The appellant’s arguments on the cross-count similar fact evidence

[23] The appellant makes four arguments in challenging the cross-count similar fact evidence. First, the trial judge did not adequately carry out or explain his reasoning in the mandatory probative/prejudicial calculus. Second, the evidence should not have been admitted as cross-count similar fact evidence because the acts were not “strikingly similar”: R. v. Arp, [1998] 3 S.C.R. 339. This complaint is also said to taint the jury charge. Third, the trial judge misapprehended the similar fact evidence. Fourth, the appellant argues that the trial judge erred in admitting two types of propensity evidence: (1) statistical evidence based on criminal data, and (2) the appellant’s statement to police on the Xu count. After setting out the governing principles, I address each of these arguments in turn and explain why I would reject them all.

(3) The governing principles on the use of cross-count similar fact evidence

[24] In R. v. MacCormack, 2009 ONCA 72, 241 C.C.C. (3d) 516, at para. 46, Watt J.A. made the trenchant observation that “[t]he general principles that control the exceptional admission of evidence of similar acts have been thoroughly chewed over in a long line of previous authorities and are not in desperate need of rumination here.” That observation has not discouraged further rumination. The appellant identifies R. v. ZWC, 2021 ONCA 116, 155 O.R. (3d) 129 as particularly illuminating in the use of the concepts of “moral prejudice” and “reasoning prejudice”.

[25] As will become evident, notwithstanding the express appeal to ZWC, the appellant’s actual arguments bear on two related problems that Watt J.A. addressed in MacCormack: how the degree of similarity, on the one hand, and linkages between the accused and each alleged similar act, on the other hand, are to be assessed by the trier of fact. The appellant’s argument in MacCormack, that they must be kept separate to avoid bootstrapping, was not accepted. Watt J.A. said, at paras. 57-58:

The “similarity” issue that lies at the threshold of the decision about the admissibility of evidence of similar acts is to be based on an examination of the acts themselves. After all, it is the high degree of similarity between or among the acts that overcomes the improbability of coincidence: Perrier at para. 21; Arp at para. 48. In general, the “similarity” issue is to be decided without reference to evidence linking the accused to each alleged similar act (“linkage evidence”): Perrier at para. 21; Arp at para. 49.

The general rule that linkage evidence is not to be taken into account in determining the “similarity” issue is not unyielding: R. v. Woodcock (2003), 177 C.C.C. (3d) 346, at paras. 80-81 (Ont. C.A.). There may be cases where examination of the linkage evidence should form part of the “similarity” decision: Woodcock at para. 81. Further, in many cases, the fact that a trial judge first considers whether there was any evidence linking the accused to the alleged similar acts will not amount to reversible error: Woodcock at para. 82. It is not often easy to draw a bright line that distinguishes between evidence that demonstrates similarity among the acts and evidence that shows an accused's involvement in them.

[26] Watt J.A. added, at para. 59, that the link must be demonstrated as a precondition to admissibility. He noted, drawing on para. 53 of Arp, that “[i]n a trial on a multi-count indictment, the link between an accused and an individual count will be relevant to the issue of identity on the other counts that disclose the required degree of similarity in the manner in which those offences were committed”.

(4) The application of the governing principles on the use of cross-count similar fact evidence

[27] I now turn to address the appellant’s four arguments.

(a) Were the trial judge’s reasons adequate?

[28] First, the appellant argues that this court should impose on trial judges a requirement that they effectively “show their homework” when serious charges are in play and similar fact evidence is proposed to be used. Trial judges should be compelled to set out the details of their reasoning process, particularly how the probative value of the evidence relates to its prejudicial effect, how moral prejudice and reasoning prejudice relate on the facts, and how probative value and prejudicial effect are to be balanced. The appellant argues that the imposition of this requirement would obviate the kind of error alleged in this case: the trial judge made an “error in form because his analysis was not expressed” and made an “error in substance” because he failed to balance probative value against prejudicial effect.

[29] There is no error in this case and no need to impose such an obligation on trial judges. In this particular case, the trial judge was manifestly conversant with the legal principles, as his two rulings on cross-count similar fact evidence show, giving assurance to the presumption that he knows the law.

(i) The first cross-count similar fact evidence ruling

[30] The trial judge’s first cross-count similar fact evidence ruling was in the context of the defence’s pre-trial motion to sever the counts. After severing the murder counts the trial judge attended to the Crown’s cross-count similar fact evidence application. He cited several authorities and stated that he was applying the test in Arp to the similar fact evidence that was possibly relevant to the identity of the accused.

[31] The trial judge considered together the Warden, Arsenault, McGrath, Savory attacks and took the view that “ [t]he fou r stabbings have a high degree of similarity”, noting that they involved “ peo pl e walking on the street in Hamilton who were at tac ked by a stranger for no apparent reason.” He said: “The s tranger came upon them suddenly, on one occasi o n on a bicycle, and struck a blow or two with a knife, but in the case of count seven [Xu], with a fist.” He added: “Such random attacks are rare in Hamilton.” The trial judge took the view that: “ it would be open to a jury to infer that o ne person committed them all.”

[32] The trial judge wrestled with whether the Xu count should be included in the cross-count similar fact evidence:

[O]n an indictment charging four random stabbings, I do not see the addition of a less serious and somewhat similar attack as … sufficiently prejudicial to require its severance. Any potential prejudice, in my view, can be cured by a proper instruction.

[33] Despite this observation, the trial judge excluded evidence on the Xu count from consideration by the jury as cross-count similar fact evidence. He commented:

The use of a fist, rather than a knife, however, strikes me as essential to the nature of the offence. It’s possible that an accused could descend to a lower level of aggravated crime in his offence pattern and, in fact, he said that that is what happened. I do not, however, consider the crime to be sufficiently similar to meet the test in Arp.

[34] There can be no doubt from this discussion that the trial judge was deeply conversant with the facts in the case and with the law, and was also focussed on the possible prejudice to the appellant.

(ii) The second cross-count similar fact evidence ruling

[35] The trial judge revisited the first ruling in the pre-charge conference in light of the evidence at trial. He heard detailed legal and factual submissions – extending to about 50 transcript pages – on the elements of the jury charge, with special focus on cross-count similar fact evidence. The trial judge then excluded the evidence of the Warden assault from the jury’s consideration as cross-count similar fact evidence.

[36] The trial judge found that two significant features of the Warden attack were different – Warden knew the appellant was the attacker, and it was not a surprise attack from behind. He found that the Warden attack “ cannot be said to be similar to the attacks on Arsenault, McGrath, or Savory in the sense of Arp ” because the “high degree of similarity” is not there.

[37] The trial judge also considered the issue of linkages between the accused and the similar acts:

In the case of Arsenault, we have a photograph taken at Aikman Avenue, right nearby in time, and in place, to the attack on Mr. Arsenault, in which the jury could infer that this is the accused and they could, although they don’t have to, also make the inference that he is the person who did the attack.

In the case of McGrath, we have a text message of November 21 st , 2011, which could be taken as an admission of involvement.

In the case of Savory, we have the posting, or sharing of a video of the attack on Savory described as, “my video”, and the access of it, ‘48 times’, again, by the accused, at least it’s open to them on the evidence that it could’ve been the accused who did that, and, to me, it makes the linkage.

[38] The trial judge went further in his explanation in the second ruling:

I have to expand on my similar fact application, because the Arsenault, McGrath and Savory counts do have a high degree of similarity, or at least the jury could find that they have high degree of similarity, high enough to meet that threshold, and while I wrestled with this, I think there is a linkage in the evidence on each count to the accused. There is a linkage sufficient to leave the possibility of similar fact across count two, then, that would not simply be bootstrapping.

[39] In light of that explanation, the trial judge declined to reconsider his severance ruling, noting that the three remaining counts are “of the same seriousness, and I think with the proper instructions about disposition evidence, and so on, that the jury will be able to handle these without prejudice to Mr. Café.”

[40] The appellant submits that pulling together the similar fact evidence and the links is impermissible bootstrapping. But this objection does not survive scrutiny under MacCormack, discussed above.

[41] There is no doubt that the trial judge was familiar with the law to which counsel drew his attention. It is obvious from the argument and the colloquy that the trial judge’s attention was focused on the evidence’s probative value and its prejudicial effect, and how they were to be balanced. This is plain especially in the last line quoted in the preceding paragraph: “I think with the proper instructions about disposition evidence, and so on, that the jury will be able to handle these without prejudice to Mr. Café.” A high degree of deference is owed to the trial judge’s decision on this probative/prejudicial calculus: Arp, at para. 42; R. v. B. (C.R.), [1990] 1 S.C.R. 717. This ground of appeal fails.

(b) Did the trial judge err in his treatment of cross-count similar fact evidence and in his jury instructions on its permissible use?

[42] The appellant’s second argument is that the evidence should not have been admitted as cross-count similar fact evidence because the attacks were not “strikingly similar”: Arp.

[43] At trial, the appellant argued that the jury charge was inadequate because it did not use the words “strikingly similar” in instructing the jury on the permissible use of cross-count similar fact evidence related to the Arsenault, McGrath and Savory counts. After hearing submissions, the trial judge gave this re-instruction:

More perhaps on a legal point, I have told you that similar act evidence is the sort of evidence that offences are so similar that it defies coincidence that different persons committed the offence and that is a high degree of similarity . If I did not get that across, then I meant to.

And I should also say that in the case of McGrath and Arsenault, they did not actually talk about knives. So, if you find that they were stabbed with a knife you would be drawing an inference from the nature of their wounds. That is circumstantial evidence of a knife being used, as opposed to direct evidence, okay? [Emphasis added.]

[44] The words “strikingly similar” are not magic words that must be incanted by a trial judge instructing the jury on the permissible use of cross-count similar fact evidence. The underlined words used by the trial judge: “so similar that it defies coincidence that different persons committed the offence” and “a high degree of similarity” are adequate. Although the expression, “strikingly similar”, is often used to describe the high degree of similarity required for admission of similar act evidence, it is not the only appropriate descriptor: see e.g., R. v. Durant, 2019 ONCA 74, 144 O.R. (3d) 465, paras. 89 and 98.

[45] That said, in their submissions to the jury both the Crown and the defence used the expression, “strikingly similar”. I have no doubt that the jury got the message from all the sources. The follow-on instruction on McGrath being stabbed with a knife was meant to account for McGrath’s testimony that she felt that she had been struck by a “blunt instrument”, as the defence pointed out in his objection to the original charge. Whatever McGrath’s subjective feeling was, the forensic evidence was that she had been stabbed and so properly was part of the similar fact group. This shows that the trial judge was alive to the issues and knew how to frame them for the jury. The trial judge’s decision to leave the cross-count similar fact evidence to the jury is entitled to deference.

[46] Relatedly, the appellant argues that the lurking danger of impermissible propensity reasoning was not adequately addressed in the jury charge. I disagree and note that the trial judge gave a limiting instruction:

If you find that the similarities in the offences are such that it is likely that the stabbings were committed by the same person, you may consider the evidence pertaining to counts 2, 3 and 4 when deciding your verdict on the other counts. If not, you must decide each count separately and base your decision only on the evidence that pertains to the count you are deciding. Either way, you must not convict Mr. Café on any count unless you are satisfied that the Crown has proven his guilt on that count beyond a reasonable doubt.

[47] There is no reason to conclude that the jury did not understand and give effect to this instruction.

(c) Did the trial judge misapprehend the similar fact evidence?

[48] The appellant’s third argument is that the trial judge misapprehended the similar fact evidence. This argument invites the court to consider the minutiae of the similarities and differences. The similarities found by the trial judge were set out at length above. The appellant also concedes that the trial judge did make reference to the dissimilarities in his charge.

[49] The trial judge instructed the jury to “[c]onsider the similarities and dissimilarities in the manner in which the offences were allegedly committed.” He repeated this caution in slightly different words: “Consider the similarities and dissimilarities amongst the alleged attacks and the importance of the feature that is similar or dissimilar.” After reviewing the similarities, he turned to the dissimilarities or differences:

Mr. Arsenault’s and Mr. McGrath’s attacker was on foot. Mr. Savory’s attacker rode a bicycle. Again, Mr. Savory’s attacker was dressed in a purple coat.

And there are details of the description that differ from person to person. There is also a difference in the sex of the victims and the number of stab wounds on Mr. Savory.

[50] He noted that if the jury was not persuaded by the similarities that “it is likely that the stabbings were committed by the same person” then the jury must fall back to consider each count separately “and base your decision only on the evidence that pertains to the count you are deciding”, always keeping in mind the Crown’s burden to prove guilt beyond a reasonable doubt. The trial judge ended his review by repeating the caution:

I have summarized what I think are the main similarities and dissimilarities. It is for you to say what features of the attacks are similar or dissimilar and how important any similarity or dissimilarity is. It is your view that counts, not mine.

[51] The appellant highlights differences not picked out by the trial judge that he argues amount to material misapprehensions. Regarding the Savory attack, the attacker used a bicycle but the others involved a person on foot. The trial judge drew this difference to the jury’s attention. Savory got the colour of the bicycle wrong. Of course, as counsel conceded in argument, the fact that the appellant owned a bicycle was not unimportant. The Savory attack took place one month after the others, but it is not clear why this should matter. The Savory link, as the trial judge noted, was the video of the attack, which the appellant posted and called “my video”.

[52] McGrath was certain that she had just passed the man who stabbed her, but she did not pick the appellant out of a photo array. The evidence showed this method for identifying a person works best within 12 to 24 hours of the encounter with the witness, not two and a half months later, as was the case here. Although the trial judge did not refer to this in his charge, the trial Crown did in his jury address. The jury was aware of this detail. The McGrath link noted by the trial judge was a text message, quite apart from another witness.

[53] The appellant took particular issue with the Arsenault attack, pointing out that the jury address suggested the attacker was a black male wearing a hoodie. But Arsenault was unable to give a description of the attacker and drew the inference that someone he had seen 45 minutes earlier was the attacker. The Arsenault link identified by the trial judge was a photograph taken at Aikman Avenue, right nearby in time, and in place, to the attack on Mr. Arsenault, from which the jury could infer the identity of the accused.

[54] Apart from the trial judge’s analysis, these matters were exhaustively canvassed by trial counsel for the Crown and the defence. None of them amounts to a misapprehension of the evidence by the trial judge. This ground of appeal fails.

(d) Did the trial judge improperly admit propensity evidence?

[55] The appellant’s fourth argument is that the trial judge erred in admitting two types of propensity evidence: (1) statistical evidence based on criminal data, and (2) the appellant’s statement to police on the Xu count.

(i) The admissibility and use of statistical evidence

[56] Several police officers who testified said that, in their experience, random stabbings are rare. That rarity was proffered in support of the proposition that the form of attacks attributed to the appellant was as a distinctive hallmark of the attacker. To buttress this point, the Crown called the evidence of Kristi Hawley, who worked with the Hamilton Police Service as a crime analysis coordinator. Her evidence was that in 2011, there was a total of 279 police reports mentioning sharp-edged weapons. Of these, 79 were non-lethal stabbings, 26 were classified as “stranger stabbings” (where the perpetrator is a stranger to the victim), with four being “random” and the rest having other motives ascribed to them. The appellant was charged with all four “random” stabbings.

[57] The appellant argues that the trial judge erred in admitting Ms. Hawley’s evidence. The appellant argues that this was opinion evidence and presumptively inadmissible. In my view, Ms. Hawley’s evidence was not opinion evidence. She did not offer any analysis or an opinion. Ms. Hawley’s evidence was purely factual; all it did was describe data confirming the evidence of the police officers that random stabbings are rare, randomness being a valid distinguishing feature of the stabbings with which the appellant was charged. The evidence was properly admitted.

[58] When defence counsel asked Ms. Hawley to offer an opinion on whether the statistics pointed to one or to several perpetrators, the trial judge, in the presence of the jury, clarified that Ms. Hawley could not offer such opinion: “I don’t think she can say one way, or another. She can just say what’s recorded.” This effectively cautioned the jury that Ms. Hawley was not offering an opinion on whether the attacks were committed by the same perpetrator. This caution safeguarded against the prejudicial reasoning that because the accused was the only person charged with the four “random” attacks, he must therefore be guilty of committing those attacks.

[59] The trial judge further instructed the jury on the use of statistical evidence in the jury charge. He cautioned that Ms. Hawley’s “ evidence only goes to support the idea that this sort of attack is not common. You can use this when deciding whether the random aspect of the crime is distinctive”. This instruction makes it clear that the statistical evidence was a contextual factor to be considered by the jury in determining whether it was likely that one person committed all three attacks.

(ii) A statement the appellant made to the police in relation to the Xu investigation

[60] The Crown introduced part of a statement the appellant made to the police in relation to the Xu investigation:

Question: … am I correct in understanding what, what you’re saying is that when, when you decide that you’re going to do something you are intending to kill her?

Answer: no, I am just attacking them. It’s, um, am I intending to kill them? Yes, I might be. I’m not gonna run from nothin’. When these voices piss me off, I tell them that I’m gonna do something about it. I don’t, I don’t have the intent of killing, but if it goes to killing I go there.

The Voice keeps challenging me to see if I’m a strong person, strong enough to do something very, very, very wrong, but then I did it, and it continues to bother, me and it continues to bother me. So, I’m like, so, I stopped doing stuff that was so aggressive. I stopped doing it. My crimes became less aggressive ”. [Emphasis added.]

[61] The appellant does not take issue with the admissibility of the first part of this passage, but argues that the underlined text should have been redacted. It was unduly prejudicial, he argues, because it implicitly encouraged the jury to use the statement beyond the Xu count and apply it to the other counts. The trial judge noted that the passage related to Xu, who was punched, not stabbed, and explained why the appellant “did something different” to Xu. The trial judge stated that the passage was “ not unfairly prejudicial and I don’t see how it would be taken out of context and being taken to refer to other things that are not on the indictment”, which in the context of the argument leading to the ruling referred to the murder charges, “or that are not relevant”, again in the context of the argument on the ruling referred to the Arsenault, McGrath and Savory counts.

[62] In relation to Xu, the statement serves as a form of confession, as counsel for the appellant conceded in argument, but he submitted that its prejudicial effect outweighed its probative value in the jury’s possible cross-count use of this evidence in considering the Arsenault, McGrath and Savory counts. It was plainly highly probative respecting the Xu count. In my view, the trial judge’s limiting instruction quoted earlier was sufficient to prevent prejudicial misuse of the passage by the jury. This ground of appeal fails.

D. The Sentence Appeal

[63] An appellate court can intervene to vary a sentence only if: (1) the sentence is demonstrably unfit; or (2) the sentencing judge made an error in principle that impacted the sentence: R. v. Friesen, 2020 SCC 9, 391 C.C.C. (3d) 309, at paras. 25-26.

[64] At trial, the Crown sought life sentences for the attempted murders and the 14-year maximum sentence for the aggravated assault. The defence submitted that 12-14 years was appropriate for each attempted murder, with one year consecutive for the aggravated assault. Further, the defence submitted that the appellant was entitled to just under five years credit for time spent in pre-trial custody.

[65] The trial judge sentenced the appellant to life imprisonment on the attempted murders and five years concurrent on the aggravated assault. He gave no credit for any pre-sentence custody because this time was attributable to the conviction and sentence for life imprisonment on the appellant’s severed first-degree murder charge.

[66] The appellant asserts that the trial judge erred in principle by not adhering to principles of restraint, totality, and proportionality. I disagree. The trial judge acknowledged and considered these sentencing principles. General and specific deterrence were primary principles of sentencing in this case. The trial judge found there to be no mitigating factors. As to aggravating factors, he placed particular weight on the vulnerability of the victims and the appellant’s lengthy criminal record. On vulnerability, the trial judge noted that these were “cowardly attacks” and “sneak attacks”. The trial judge notes that the appellant chose “[v]ulnerable victims” and his motivation for attacking them was not clear. The sentencing judge rejected the appellant’s evidence that he was “hearing voices” and stated that this was made up “in the hopes of getting a lighter treatment from the court.” Further, the offences had serious consequences and “[a]ll the victims’ lives took a downturn.”

[67] The trial judge concluded that the severity of the offences warranted the imposition of a life sentence. He held that “the offender’s crimes are so serious and his moral blameworthiness is so great that it [was] proportionate to impose a life sentence.” Further, “the protection of the public” required imposing sentences of life imprisonment and a delay of parole for ten years under s. 743.6 of the Criminal Code. The severity of the aggravated assault and the lasting injuries warranted the imposition of a sentence of five years imprisonment concurrently.

[68] On totality, the trial judge agreed with defence counsel that a consecutive sentence would not be appropriate. On proportionality, he considered an array of factors and held that the circumstances of the case and the need to protect the public required the imposition of a life sentence.

[69] The sentences imposed were fit and are entitled to appellate deference. This court’s intervention is not warranted.

E. Disposition

[70] I would dismiss the conviction appeal. I would grant leave to appeal sentence but would dismiss the sentence appeal.

Released: January 9, 2023 “P.D.L.”

“P. Lauwers J.A.”

“I agree. Grant Huscroft J.A.”

“I agree. B.W. Miller J.A.”