Opinion of the Court by Corrigan, J.

Here we resolve a conflict among the Courts of Appeal as to the proper standard of review when a defendant challenges a court’s decision to instruct on a concurrent intent, or “kill zone,” theory as applied to an allegation of attempted murder. (See People v. Canizales (2019) 7 Cal.5th 591 (Canizales).) We conclude that, although the Court of Appeal applied the proper standard, it erroneously concluded that sufficient evidence supported the giving of a concurrent intent instruction. We reverse the Court of Appeal judgment to the extent it affirmed the one attempted murder conviction that was based on that theory.

I. BACKGROUND

Early on April 16, 2015, defendant Ahmed Mumin robbed a San Diego convenience store and fatally shot customer Eric Schade. DNA testing of items recovered at the scene pointed to defendant as a suspect. The store clerk later identified him as the robber. Two days later, defendant was at an apartment complex where he asked a relative for a ride to “[a]nywhere,” saying the police were looking for him. Detectives, having discovered a connection between defendant and the complex, arrived and defendant ran when he saw them. Shortly thereafter, a burglary in progress was reported at the complex. A resident said defendant had a silver handgun, pushed on several windows, then hid a backpack and fled. Responding officers found a backpack hidden in nearby bushes. It contained defendant’s identification, a phone, and some nine-millimeter ammunition. An extensive search of the complex ensued. Five detectives and over a dozen uniformed officers wore tactical vests with a badge on the front and “Police” printed in large white letters. Investigators went to each apartment, loudly identified themselves as police, and directed residents to come outside. A police helicopter flew overhead to assist the search, which lasted approximately an hour. Detectives Jim Mackay and Luke Johnson approached a building with four adjacent doors leading to a community room and facing the area where defendant’s backpack was recovered. At trial, witnesses referred to the doors numerically, with Door 1 being farthest to the right. Believing the doors led to small rooms or storage units, Mackay went to Door 1 to see if it was locked. Detective Luke Johnson provided cover. He stood to the left of Door 1 generally in line with Door 2, and back some distance from the plane of the doorways. As a result Johnson was positioned about 25 feet away from Door 1 and to the left of it. Johnson testified he placed himself in line with Door 2 to avoid standing in a “fatal funnel” near Door 1. He explained: “If someone is going to shoot you, they’re going to shoot out the door that you opened. So you don’t want to be standing in that area.” Mackay stood by the right hinges of Door 1 and reached across to operate its handle. As he opened the door slightly, defendant fired once through the opening and twice through the closed Door 2. Neither officer was hit. All three bullets struck near dumpsters across a parking lot from where defendant’s backpack was found. In response, Mackay and Johnson took cover and returned fire. Johnson shot five times through Door 1 and Mackay fired three times at the same target. The shooting stopped. Defendant, who had been shot, complied with orders to come out of the community room. Inside, a Sig Sauer nine-millimeter pistol holding seven rounds was recovered. Two additional magazines, containing a total of 21 rounds, were also taken into evidence. Ballistics testing confirmed the gun was that used to kill Schade and shoot at the detectives. A criminalist testified about the various bullet holes and trajectories. All of the bullet holes found in Door 1 were made by shots fired from the outside. Two bullet holes in Door 2 were made by shots fired from inside the room. The third round recovered near the dumpsters likely went through the opening of Door 1 created when Mackay opened it. All three rounds defendant fired had hollow points, designed to mushroom on impact. At trial, in connection with the convenience store crimes, defendant was convicted of first degree murder, second degree robbery, and burglary. Various firearm allegations, along with special circumstance allegations of murder committed during a robbery and burglary, were found true.1 As to the apartment complex shooting, defendant was convicted of two counts of attempted, premeditated murder of a police officer, assault with a semiautomatic firearm, and assault on a police officer with a semiautomatic firearm. Attached enhancements were found

1 Penal Code sections 187, subdivision (a); 189, subdivision (a); 190.2, subdivision (a)(17)(A), (G); 211; 212.5, subdivision (c); 459; 460, subdivision (b); 12022.5, subdivision (a); 12022.53, subdivisions (b)–(d). true.2 Defendant was sentenced to life without the possibility of parole for the murder of Schade, along with consecutive terms totaling 55 years to life plus 41 years and four months. Sentence on the remaining counts and allegations was stayed. The Court of Appeal modified the judgment to vacate two counts of assault with a semiautomatic firearm because they were lesser included offenses of assault on the officers with that weapon. The judgment was otherwise affirmed. (People v. Mumin (2021) 68 Cal.App.5th 36, 62–63 (Mumin).)

II. DISCUSSION

Defendant argues the trial court improperly instructed on the concurrent intent theory of liability for attempted murder of both detectives. Before we turn to the proper standard for reviewing those claims, we summarize the relevant substantive law. A. Concurrent Intent and the Law of Attempted Murder “In every crime or public offense there must exist a union, or joint operation of act and intent, or criminal negligence.” (Pen. Code, § 20.)3 The mental state, or mens rea, that must accompany each crime is an element of the offense. The mental state required for the crime of murder is the existence of malice, which may be either express or implied. (See Pen. Code, §§ 187,

2 Penal Code sections 187, subdivision (a); 189, subdivision (a); 245, subdivisions (b), (d)(2); 664, subdivision (e); 12022.5, subdivision (a); 12022.53, subdivisions (b)–(d). Defendant was additionally convicted of being a felon in possession of a firearm and ammunition (Pen. Code, §§ 29800, subd. (a)(1); 30305, subd. (a)(1)). 3 Our discussion here does not touch on strict liability offenses or crimes of negligence. subd. (a); 188.) Malice is express when “there is manifested a deliberate intention to unlawfully take away the life of a fellow creature.” (Pen. Code, § 188, subd. (a)(1).) It is implied when “a defendant act[s] with conscious disregard of the danger to human life.” (People v. Knoller (2007) 41 Cal.4th 139, 156; see People v. Smith (2018) 4 Cal.5th 1134, 1165; Pen. Code, § 188, subd. (a)(2).) Because malice may be implied, second degree murder does not require a specific intent to kill. To elevate that offense to murder in the first degree on a malice theory, the defendant must act with a specific intent to kill that is formed willfully, deliberately, and with premeditation. (Pen. Code, § 189, subd. (a); see CALCRIM No. 521.) Special rules apply when a person tries but fails to complete an offense. An attempt “consists of two elements: a specific intent to commit the crime, and a direct but ineffectual act done toward its commission.” (Pen. Code, § 21a.) In the context of attempted murder, People v. Bland (2002) 28 Cal.4th 313 (Bland) observed: “The mental state required for attempted murder has long differed from that required for murder itself. Murder does not require the intent to kill. Implied malice — a conscious disregard for life — suffices.” (Id. at p. 327.) However, attempted murder requires a specific intent to kill. Bland first examined the contours of the doctrine of transferred intent as it applied to a completed murder. Under the doctrine, if a defendant intended to kill A but inadvertently killed B, the intent to kill A is deemed to transfer to the killing of B, so that the defendant is guilty of B’s murder. (See Bland, supra, 28 Cal.4th at pp. 320–321, citing People v. Scott (1996) 14 Cal.4th 544, 546.) Bland then built on that principle to hold that, even when a defendant succeeds in killing his intended target, his intent to kill extends to any others he actually kills. (See Bland, at pp 321–326.) Bland imposed a limit on its holding, however. While the intent to kill one target transfers to others actually killed, the doctrine will not extend that lethal intent to others who may be assaulted or injured but do not die. To be guilty of the attempted murder of a person who survives, the defendant must intend to kill that survivor. (See id. at pp. 326–328.) It is essential to keep clear the distinction between the sufficiency of implied malice to support a murder conviction when the victim dies and the requirement of a specific intent to kill in order to support a charge of attempted murder when the victim survives. As Bland explained: “Someone who in truth does not intend to kill a person is not guilty of that person’s attempted murder even if the crime would have been murder — due to transferred intent — if the person were killed. To be guilty of attempted murder, the defendant must intend to kill the alleged victim, not someone else. The defendant’s mental state must be examined as to each alleged attempted murder victim. Someone who intends to kill only one person and attempts unsuccessfully to do so, is guilty of the attempted murder of the intended victim, but not of others.” (Bland, supra, 28 Cal.4th at p. 328.) Bland clarified that, although an intent to kill cannot transfer to surviving victims, a defendant may nevertheless be liable for attempted murder under a concurrent intent theory: “[T]he fact the person desires to kill a particular target does not preclude finding that the person also, concurrently, intended to kill others within what [Ford v. State (Md. 1993) 625 A.2d 984] termed the ‘kill zone.’ ‘The intent is concurrent . . . when the nature and scope of the attack, while directed at a primary victim, are such that we can conclude the perpetrator intended to ensure harm to the primary victim by harming everyone in that victim’s vicinity. For example, an assailant who places a bomb on a commercial airplane intending to harm a primary target on board ensures by this method of attack that all passengers will be killed. Similarly, consider a defendant who intends to kill A and, in order to ensure A’s death, drives by a group consisting of A, B, and C, and attacks the group with automatic weapon fire or an explosive device devastating enough to kill everyone in the group. The defendant has intentionally created a “kill zone” to ensure the death of his primary victim, and the trier of fact may reasonably infer from the method employed an intent to kill others concurrent with the intent to kill the primary victim. When the defendant escalated his mode of attack from a single bullet aimed at A’s head to a hail of bullets or an explosive device, the factfinder can infer that, whether or not the defendant succeeded in killing A, the defendant concurrently intended to kill everyone in A’s immediate vicinity to ensure A’s death. The defendant’s intent need not be transferred from A to B, because although the defendant’s goal was to kill A, his intent to kill B was also direct; it was concurrent with his intent to kill A. Where the means employed to commit the crime against a primary victim create a zone of harm around that victim, the factfinder can reasonably infer that the defendant intended that harm to all who are in the anticipated zone. This situation is distinct from the “depraved heart” [i.e., implied malice] situation because the trier of fact may infer the actual intent to kill which is lacking in a “depraved heart” [implied malice] scenario.’ ” (Bland, supra, 28 Cal.4th at pp. 329–330, quoting Ford, at pp. 1000– 1001, fn. omitted.) In Bland the defendant fired into a car with three occupants. The driver, who appeared to be the primary target, was fatally shot. The two passengers survived. Bland concluded that, though the jury found the defendant primarily wanted to kill the driver rather than the passengers, “it could reasonably also have found a concurrent intent to kill those passengers when defendant and his cohort fired a flurry of bullets at the fleeing car and thereby created a kill zone. Such a finding fully supports attempted murder convictions as to the passengers.” (Bland, supra, 28 Cal.4th at pp. 330–331.) People v. Canizales, supra, 7 Cal.5th 591 reaffirmed Bland’s concurrent intent theory but articulated its contours and limitations. It also explained how the theory relates to proving the specific intent to kill required for attempted murder. “Bland’s adoption of the kill zone theory meant that a prosecutor charging attempted murder in a multivictim case had an additional, alternative ground by which to prove the requisite intent to kill. Under appropriate facts, the prosecutor could attempt to show either that the defendant’s intent to kill one or more alleged victims arose independently of his actions toward any other victim, or that the defendant’s intent to kill an untargeted victim arose concurrently with his intent to kill a primary target.” (Canizales, at p. 603). And it noted that “there are evidentiary bases, other than the kill zone theory, on which a fact finder can infer an intent to kill for purposes of attempted murder liability that do not depend on a showing that the defendant had a primary target . . . .” (Id. at p. 608, citing People v. Stone (2009) 46 Cal.4th 131, 140 (Stone), and People v. Smith (2005) 37 Cal.4th 733, 743.) The Canizales case involved a gang-related shooting at a neighborhood block party. Earlier that day, Canizales had encountered two rival gang members, Pride and Bolden. Canizales and fellow gang members, including codefendant Windfield, went to the party, where Windfield fired five shots. He killed one partygoer but missed Pride and Bolden. Canizales was convicted as an aider and abettor of Windfield’s attempted murder of Pride and Bolden. (Canizales, supra, 7 Cal.5th at pp. 598–601.) There was substantial evidence that Pride was the primary target. Canizales explained the concurrent intent theory relies on circumstantial evidence to establish that the defendant acted with the specific intent to kill not only the primary target but also everyone in the kill zone. “[W]hen the prosecution’s theory substantially relies on circumstantial evidence, a jury must be instructed that it cannot find guilt based on circumstantial evidence when that evidence supports a reasonable conclusion that the defendant is not guilty.” (Canizales, supra, 7 Cal.5th at p. 606; see CALCRIM No. 225.) Moreover, Canizales took care to point out that the concurrent intent theory is a separate and particularly narrow one. It “may properly be applied only when a jury concludes: (1) the circumstances of the defendant’s attack on a primary target, including the type and extent of force the defendant used, are such that the only reasonable inference is that the defendant intended to create a zone of fatal harm — that is, an area in which the defendant intended to kill everyone present to ensure the primary target’s death . . . and (2) the alleged attempted murder victim [who was a secondary target] was located within that zone of harm. Taken together, such evidence will support a finding that the defendant harbored the requisite specific intent to kill both the primary target and everyone within the zone of fatal harm.” (Canizales, supra, 7 Cal.5th at p. 607; see id. at p. 597.) By definition then, a kill zone is an area which a defendant intentionally creates in order to kill all those within it to ensure the primary target’s death. Canizales emphasized that when the theory is relied upon, there must be evidence the defendant intended to kill a primary target. “ ‘[W]ithout a primary target, there cannot be concurrent intent because there is no primary intent to kill as to which the intent to kill others could be concurrent.’ ” (Canizales, supra, 7 Cal.5th at p. 609, quoting People v. Medina (2019) 33 Cal.App.5th 146, 155.) Canizales noted that multiple factors bear on whether the circumstances of the attack sufficiently show a defendant intended to create a kill zone around a primary target. These include “the type of weapon used, the number of shots fired (where a firearm is used), the distance between the defendant and the alleged victims, and the proximity of the alleged victims to the primary target.” (Canizales, at p. 607.) The far more commonplace act of firing one or a few shots at a group may supply the actus reus for a number of crimes. But, standing alone, it does not support a conclusion that the shooter intended to create a kill zone around the primary target in order to ensure that primary target will die. (See People v. Perez (2010) 50 Cal.4th 222, 232; Stone, supra, 46 Cal.4th at p. 138.) Without substantial evidence showing the defendant acted with intent to kill a primary target, the essential basis for a concurrent intent analysis is not satisfied. Canizales went on to hold that a concurrent intent instruction would only have been warranted if there was substantial evidence that, “if believed by the jury, would support a reasonable inference that defendants intended to kill everyone within the “ ‘kill zone.’ ” (Canizales, supra, 7 Cal.5th at pp. 609– 610.) The opinion then turned to the evidentiary record. Windfield was 100 feet away from Pride when he fired a handgun at him. He hit neither Pride nor Bolden. That fact was not alone dispositive. But considered along with the limited number of shots fired, Winfield’s distance from Pride, and the open area into which he fired, there was insufficient evidence to support a conclusion that the defendants intended to create a zone of fatal harm around Pride. (See id. at pp. 611–612.) In light of the potential for misapplication of the concurrent intent theory, Canizales emphasized “that going forward trial courts must exercise caution when determining whether to permit the jury to rely upon the kill zone theory. Indeed, we anticipate there will be relatively few cases in which the theory will be applicable and an instruction appropriate. Trial courts should tread carefully when the prosecution proposes to rely on such a theory, and should provide an instruction to the jury only in those cases where the court concludes there is sufficient evidence to support a jury determination that the only reasonable inference from the circumstances of the offense is that a defendant intended to kill everyone in the zone of fatal harm. The use or attempted use of force that merely endangered everyone in the area is insufficient to support a kill zone instruction.” (Canizales, supra, 7 Cal.5th at p. 608; see also id. at p. 597.) B. Conflict in the Courts of Appeal Regarding Application of Canizales A conflict has arisen in the Courts of Appeal as to the proper approach to be applied in reviewing a trial court’s decision to give a concurrent intent instruction. In re Rayford (2020) 50 Cal.App.5th 754 (Rayford) involved a habeas petition seeking relief following Canizales. Rayford and codefendant Glass were at a party attended by Shadonna and Donisha Williams, among others. Glass argued with one Perry. Shadonna and Donisha left the party, gave Perry a ride to another location, then returned to their own home, where they lived with their mother Sheila. Rayford, Glass, and a group of 10 to 30 young men later came to the Williams residence. There were 13 people in or around the structure, including a number of children. Several people came outside and Glass said he wanted to fight with Perry, whom he erroneously believed was in the house. Sheila said Perry was not present and there would be no fight. As she tried to get her family and friends back inside, multiple shots were fired by Glass and Rayford. The house was hit a number of times and two people were struck. No one was killed. Sheila asked Glass why he shot at the house and he said: “ ‘That’s what you bitches get.’ ” (Id. at p. 763.) Glass admitted he went to the house to fight Perry, and Rayford admitted he came to watch. Each denied bringing a gun or firing any shots. (Id. at pp. 761–764.) Each was charged with 11 counts of attempted murder with gang and firearm allegations. The court gave an instruction on concurrent intent (CALJIC No. 8.66.1) and the prosecution relied on that theory. (Rayford, at pp. 764–765.) The prosecutor admitted Perry was not present at the home and he was not a victim. He urged that Sheila was one of the “ ‘primary victims,’ ” along with the two victims who were hit. (Id. at p. 765.) Following convictions on all allegations, Rayford and Glass were sentenced to 11 consecutive life sentences for the attempted murders plus 220 years for the enhancements. (Ibid.) Rayford’s direct appeal was resolved before Canizales was decided. The court affirmed both attempted murder convictions, holding that there was sufficient evidence of a primary target and creation of a kill zone. (Rayford, supra, 50 Cal.App.5th at p. 766.)4 It reasoned that Sheila and a number of others could have been the defendants’ primary targets because Sheila had treated Glass with disrespect and that others who tried to protect Perry disliked the defendants. (Ibid.; see People v. Rayford (July 18, 2006, B179017) [nonpub. opn.], petn. for review den. (Nov. 1, 2006, S146142).) On habeas, the Rayford court considered two questions: 1. whether the trial court should have given a concurrent intent instruction; and 2. whether the given instruction was legally sound. It correctly held that, in light of Canizales, the instruction itself was deficient. As a result, it properly granted habeas relief. (See Rayford, supra, 50 Cal.App.5th at pp. 782– 784.) It is the manner in which the court reviewed the trial court’s decision to instruct on concurrent intent at all that missed the mark and is part of the conflict among the courts of review that we must resolve here. Rayford pointed to the following language from Canizales: “[I]n determining whether ‘there is sufficient evidence from which the jury could find that the only reasonable inference is that the defendant intended to kill (not merely to endanger or harm) everyone in the zone of fatal harm,’ we consider the circumstances surrounding the shooting, including ‘the type of weapon used, the number of shots fired (where a firearm is used), the distance between the defendant and the alleged victims, and the proximity of the alleged victims to the primary target.’ ” (Rayford, supra, 50 Cal.App.5th at p. 779, quoting

4 The court, however, reversed the gang enhancements for insufficient evidence, as well as firearm enhancements dependent on that allegation. (See Rayford, supra, 50 Cal.App.5th at p. 766.) Canizales, supra, 7 Cal.5th at pp. 597, 607.) The Court of Appeal observed it had previously affirmed the defendants’ convictions because “the circumstances of the shooting here support a reasonable inference the shooters intended to kill everyone in the zone of fatal harm around Sheila.” (Rayford, at p. 779.) On habeas review, however, the court held the new authority of Canizales dictated a different outcome. It cited the “only reasonable inference” language from Canizales then reasoned: “[O]ther circumstances support a reasonable alternative inference more favorable to Rayford and Glass, that the shooters acted not with the specific intent to kill everyone in and in front of the house, but with conscious disregard of the risk Sheila and her family and neighbors might be seriously injured or killed.” (Ibid.) Because there was another possible inference that the Court of Appeal itself considered reasonable, it held the concurrent intent instruction should not have been given. (Id. at p. 781.) Rayford’s analysis of the issue suggested that, under Canizales, a trial court should not instruct regarding concurrent intent, even if a reasonable supporting inference could be drawn from the evidence, if the court itself concludes another reasonable inference could be drawn that does not support a concurrent intent theory. As we explain below, this approach misconstrues the differing roles of the court and jury. The Court of Appeal in this case disagreed with Rayford’s analysis.5 The court reasoned: “Canizales does not depart from,

5 The opinion below was authored by Chief Justice Guerrero while she served as an associate justice for the Fourth District Court of Appeal, Division One. She has recused herself from this case. and instead reaffirms, established principles governing a trial court’s decision to instruct on a theory of liability and an appellate court’s review of such a decision. The trial court must determine whether the evidence would support a jury determination that the only reasonable inference was that the defendant held the requisite intent. If a trial court’s decision to instruct is challenged on appeal, we must make the same determination on de novo review. But, in so doing, the issue is not whether we believe the only reasonable inference from the evidence is that the defendant had the requisite intent — just as, in other substantial evidence contexts, the issue is not whether we believe the defendant to be guilty beyond a reasonable doubt. The issue is whether the evidence would support such a determination by the jury. Under these circumstances, it is well established that the evidence supports a jury determination that an inference is the only reasonable inference if we conclude it is at least a reasonable inference. We disagree with Rayford to the extent it holds otherwise.” (Mumin, supra, 68 Cal.App.5th at p. 47.) In this case, the Court of Appeal properly differentiated between the roles of court and jury. “The distinction between the jury, on one hand, and the appellate court, on the other, reflects the fundamental rule that the jury, not the appellate court, must be convinced of the defendant’s guilt beyond a reasonable doubt. A jury must acquit a defendant if a reasonable alternative interpretation of the evidence suggests innocence, because it necessarily creates reasonable doubt . . . . But if an appellate court identifies a reasonable alternative interpretation based on its own review of the evidence, it does not necessarily compel reversal, because an appellate court need not be convinced of a defendant’s guilt beyond a reasonable doubt. Instead, as noted, the appellate court asks whether the jury could have found the defendant guilty beyond a reasonable doubt. It is the jury, of course, that sees and hears the evidence. The appellate court has only the cold record before it. An appellate court’s ability to identify a reasonable alternative inference suggesting innocence does not mean that the jury, viewing the evidence live at trial, could not have rejected that inference as unreasonable.” (Mumin, supra, 68 Cal.App.5th at p. 50.) The court observed that Canizales, in applying the principles it set forth, “did not require that the appellate court itself determine whether the inference supporting the instruction was the only reasonable inference. Instead, it explained, ‘an instruction on the kill zone theory would have been warranted in this case only if there was substantial evidence in the record that, if believed by the jury, would support a reasonable inference that defendants intended to kill everyone within the “kill zone.” ’ [Citation.] In other words, ‘ “ ‘evidence must appear in the record which, if believed by the jury, will support the suggested inference.’ ” ’ [Citation.] An appellate court need not determine that such an inference is the only reasonable inference.” (Id. at p. 51.) The Court of Appeal stated Canizales’s “only reasonable inference” formulation “is simply another way of saying that the evidence must support a defendant’s guilt beyond a reasonable doubt. It does not imply any change to the established standard of review where the prosecution relies on circumstantial evidence. If the evidence supports a reasonable inference of the requisite intent, it necessarily follows that the jury could find it was the only reasonable inference.” (Id. at p. 52.) Thus, “to the extent [the] Rayford [court] believed that one reasonable inference from the evidence would support a kill zone instruction under Canizales, but a reasonable alternative inference would not, the correct result would have been to uphold the instruction.” (Id. at pp. 53–54.) This is so because we presume the jury followed the instruction that it could rely on circumstantial evidence to convict only if it finds the inference pointing to guilt is the only reasonable one based on its evaluation of the evidence. C. The Substantial Evidence Standard Applies to Giving or Reviewing a Concurrent Intent Instruction In light of this conflict, the question becomes who must determine whether the inference pointing to guilt is the only reasonable inference: the judge or the jury? Established precedent dictates that the final determination is for the jury alone. Yet, if the jury has the last word on that point, what is the role of the trial court in deciding how the jury should be instructed as to available theories, and how should that trial court ruling be considered on appeal? Defendant contends a standard more stringent than substantial evidence must be satisfied before a court may instruct on a concurrent intent theory under Canizales. He urges that both trial and appellate courts themselves must conclude from the evidence “that the only reasonable inference is that the defendant intended to create a zone of fatal harm . . . .” (Canizales, supra, 7 Cal.5th at p. 597.) The argument fails. Canizales did not depart from the traditional substantial evidence inquiry. The analysis at both the trial and appellate level looks to whether substantial evidence supports giving the challenged instruction. In the context of a criminal case, the substantial evidence standard stems from the requirement that a criminal conviction necessitates “proof beyond a reasonable doubt of every fact necessary to constitute the crime . . . .” (In re Winship (1970) 397 U.S. 358, 364.) That safeguard “dates at least from our early years as a Nation.” (Id. at p. 361.) In holding that due process requires such proof, the high court in Winship explained that “[t]he reasonable-doubt standard plays a vital role in the American scheme of criminal procedure. It is a prime instrument for reducing the risk of convictions resting on factual error. The standard provides concrete substance for the presumption of innocence — that bedrock ‘axiomatic and elementary’ principle whose ‘enforcement lies at the foundation of the administration of our criminal law.’ ” (Id. at p. 363; see also Apprendi v. New Jersey (2000) 530 U.S. 466, 476–478.) “Winship presupposes as an essential of the due process guaranteed by the Fourteenth Amendment that no person shall be made to suffer the onus of a criminal conviction except upon sufficient proof — defined as evidence necessary to convince a trier of fact beyond a reasonable doubt of the existence of every element of the offense.” (Jackson v. Virginia (1979) 443 U.S. 307, 316 (Jackson).) Jackson clarified that, in order to preserve that due process right, the critical inquiry in a criminal appeal “must be not simply to determine whether the jury was properly instructed, but to determine whether the record evidence could reasonably support a finding of guilt beyond a reasonable doubt. But this inquiry does not require a court to ‘ask itself whether it believes that the evidence at the trial established guilt beyond a reasonable doubt.’ [Citation.] Instead, the relevant question is whether, after viewing the evidence in the light most favorable to the prosecution, any rational trier of fact could have found the essential elements of the crime beyond a reasonable doubt. [Citation.] This familiar standard gives full play to the responsibility of the trier of fact fairly to resolve conflicts in the testimony, to weigh the evidence, and to draw reasonable inferences from basic facts to ultimate facts. Once a defendant has been found guilty of the crime charged, the factfinder’s role as weigher of the evidence is preserved through a legal conclusion that upon judicial review all of the evidence is to be considered in the light most favorable to the prosecution.” (Id. at pp. 318–319, fn. omitted; see People v. Cuevas (1995) 12 Cal.4th 252, 260–261; People v. Johnson (1980) 26 Cal.3d 557, 578.) Thus, an appellate court retrospectively inquires whether a rational trier of fact could have found the defendant guilty beyond a reasonable doubt, based on all the evidence when viewed in the light most favorable to the prosecution. “Sufficiency review essentially addresses whether ‘the government’s case was so lacking that it should not have even been submitted to the jury.’ ” (Musacchio v. United States (2016) 577 U.S. 237, 243; see Pen. Code, § 1118.1.) In determining whether a case should be submitted to the jury in the first instance, the trial court’s inquiry mirrors an appellate court’s analysis regarding the sufficiency of the evidence, except the trial court must necessarily assess prospectively whether substantial evidence exists from which a rational jury could convict the defendant beyond a reasonable doubt. A corollary of that rule is a court may instruct the jury regarding a particular theory of conviction only if substantial evidence would allow a rational jury to find that theory supported by the facts. For example, we have repeatedly applied the substantial evidence standard when evaluating whether a court must instruct on lesser included offenses. “Instruction on a lesser included offense is required only when the record contains substantial evidence of the lesser offense, that is, evidence from which the jury could reasonably doubt whether one or more of the charged offense’s elements was proven, but could find all the elements of the included offense proven beyond a reasonable doubt.” (People v. Moore (2011) 51 Cal.4th 386, 408–409; see, e.g., People v. Steskal (2021) 11 Cal.5th 332, 345; People v. Vargas (2020) 9 Cal.5th 793, 827; People v. Lopez (2020) 9 Cal.5th 254, 269; People v. Westerfield (2019) 6 Cal.5th 632, 718.) Similarly, courts have observed that “[t]he test for determining whether instructions on a particular theory of guilt are appropriate is whether there is substantial evidence which would support conviction on that theory.” (People v. Nguyen (1993) 21 Cal.App.4th 518, 528; see also People v. Campbell (1994) 25 Cal.App.4th 402, 408.) In making these determinations, the trial court does not inquire whether it would convict of a lesser crime than that charged or whether it would find a given theory has been proven. Instead the court considers whether a rational jury could so decide. The question here is whether Canizales departed from the traditional substantial evidence test and articulated a new standard for giving an instruction or reviewing that decision on appeal. In support of his argument, defendant attempts to rely on language in Canizales that a concurrent intent instruction should only be given “where the court concludes there is sufficient evidence to support a jury determination that the only reasonable inference from the circumstances of the offense is that a defendant intended to kill everyone in the zone of fatal harm.” (Canizales, supra, 7 Cal.5th at p. 608; see also id. at p. 597.) Canizales does not support the conclusion defendant urges. He overlooks the fact that Canizales expressly referred to evidence supporting the jury’s findings. In deciding whether to give the instruction, the court considers whether substantial evidence exists from which the jury could draw the required inference. If so, the instruction is justified. Ultimately, it remains for the jury to determine whether that inference is the only reasonable one. The trial court may not preemptively substitute its view of the evidence for that of the jury. The standard is the same for appellate review. Rather than viewing any passage from Canizales in isolation, it is important to examine the context in which the case referred to “the only reasonable inference.” As the Court of Appeal below noted (Mumin, supra, 68 Cal.App.5th at p. 47), the “only reasonable inference” language derives from Canizales’s observation that “[t]he kill zone theory looks to circumstantial evidence to support a permissive inference regarding a defendant’s intent. This is not unusual. As we have described on many occasions, intent to kill often must be inferred from circumstantial evidence surrounding the crime. [Citation.] And when the prosecution’s theory substantially relies on circumstantial evidence, a jury must be instructed that it cannot find guilt based on circumstantial evidence when that evidence supports a reasonable conclusion that the defendant is not guilty.” (Canizales, supra, 7 Cal.5th at p. 606.) For this proposition, Canizales cited People v. Bender (1945) 27 Cal.2d 164, which reasoned that a jury cannot rely on circumstantial evidence to convict if a reasonable inference from such evidence points to innocence. Bender went on to observe that its analysis “enunciates a most important rule governing the use of circumstantial evidence,” and “it should be declared to the jury in every criminal case wherein circumstantial evidence is received.” (Id. at p. 175.) Canizales also pointed to CALCRIM No. 225, the standard jury instruction on circumstantial evidence, which explains in part that “[i]f you can draw two or more reasonable conclusions from the circumstantial evidence, and one of those reasonable conclusions supports a finding that the defendant did have the required (intent/ [and/or] mental state) and another reasonable conclusion supports a finding that the defendant did not, you must conclude that the required (intent/ [and/or] mental state) was not proved by the circumstantial evidence.” In other words, if a juror draws an inference that is not consistent with guilt and concludes that inference is reasonable in light of persuasive evidence, he or she must conclude there is reasonable doubt. Canizales consistently kept its focus on what a jury could determine. It did not mention or suggest a deviation from settled law or hint at a different analytical framework for deciding whether to instruct on concurrent intent or for reviewing that decision on appeal. Immediately after this discussion regarding the jury’s role in evaluating circumstantial evidence, Canizales concluded “that the kill zone theory for establishing the specific intent to kill required for conviction of attempted murder may properly be applied only when a jury concludes” the defendant intended to kill his primary target, the nature of his attack demonstrated he also intended to assure the target’s death by creating a zone of fatal harm around him, and that an alleged attempted murder victim was in that zone. (Canizales, supra, 7 Cal.5th at p. 607, italics added.) “Taken together, such evidence will support a finding that the defendant harbored the requisite specific intent to kill both the primary target and everyone within the zone of fatal harm.” (Ibid.) Canizales then listed several factors “the jury should consider” when “determining the defendant’s intent to create a zone of fatal harm and the scope of any such zone . . . .” (Ibid.) Canizales included the “only reasonable inference” caveat to explain how the jury should evaluate circumstantial evidence relating to the defendant’s intent. This discussion of the jury’s role does not suggest a departure from the traditional substantial evidence standard in evaluating whether the court has properly instructed on a particular theory of conviction. Further, as the Court of Appeal here reasoned, Canizales’s application of its stated standard confirms it was applying ordinary substantial evidence principles. (See Mumin, supra, 68 Cal.App.5th at pp. 48–52.) Canizales explained at the outset that “an instruction on the kill zone theory would have been warranted in this case only if there was substantial evidence in the record that, if believed by the jury, would support a reasonable inference that defendants intended to kill everyone within the ‘kill zone.’ ” (Canizales, supra, 7 Cal.5th at pp. 609– 610.) To justify such an instruction, “the record would need to include (1) evidence regarding the circumstances of defendants’ attack on Pride that would support a reasonable inference that defendants intentionally created a zone of fatal harm around him, and (2) evidence that Bolden was located within that zone of fatal harm.” (Id. at p. 610.) We ultimately concluded “the evidence concerning the circumstances of the attack (including the type and extent of force used by Windfield) was not sufficient to support a reasonable inference that defendants intended to create a zone of fatal harm around a primary target.” (Id. at p. 610.) Canizales also noted, “Because we conclude that the evidence here is insufficient to support a finding that defendants intended to create a zone of fatal harm, we have no occasion to determine the scope of any such zone given these facts.” (Id. at p. 611, italics added.) Canizales’s repeated references to a lack of substantial evidence to support “a reasonable inference” or “a finding” that the defendants intended to create a kill zone constituted nothing more than an application of the traditional substantial evidence standard, which asks whether the evidence is such that a rational trier of fact could have found guilt beyond a reasonable doubt. (See Jackson, supra, 443 U.S. at pp. 318– 319.) The substantial evidence standard “gives full play to the responsibility of the trier of fact fairly to resolve conflicts in the testimony, to weigh the evidence, and to draw reasonable inferences from basic facts to ultimate facts.” (Jackson, supra, 443 U.S. at p. 319.) As we have stated in the context of postconviction review for evidentiary sufficiency: “Conflicts and even testimony which is subject to justifiable suspicion do not justify the reversal of a judgment, for it is the exclusive province of the trial judge [in a court trial] or jury to determine the credibility of a witness and the truth or falsity of the facts upon which a determination depends.” (People v. Maury (2003) 30 Cal.4th 342, 403.) Ultimately, it is within the jury’s exclusive province to determine whether an inference that may be drawn from the evidence is, in fact, the only reasonable one, a determination that depends on its resolution of conflicting evidence and weighing the credibility of witnesses. As we recognized long ago, “[i]t may be confidently declared that, founded upon the evidence, the jury not only is authorized to make any logical and reasonable deduction, but also that the jury is the exclusive judge of the weight and value of the inference that may be drawn by it . . . .” (Hamilton v. Pacific Elec. Ry. Co. (1939) 12 Cal.2d 598, 602.) The standard for which defendant advocates would allow the trial court to usurp the jury’s province. Under the defense theory, the court could draw its own inferences and compare them to other possible inferences. Such an approach would have the court resolve conflicts in the evidence and weigh the testimony of witnesses. Yet “[t]o do so would be a clear usurpation of the jury’s exclusive function.” (Hicks v. Ocean Shore Railroad, Inc. (1941) 18 Cal.2d 773, 781.) Defendant suggests a heightened instructional standard is necessary to address Canizales’s concern that a jury may misapply the concurrent intent theory. Canizales cautioned that “trial courts must be extremely careful in determining when to permit the jury to rely upon the kill zone theory,” observing: “As past cases reveal, there is a substantial potential that the kill zone theory may be improperly applied, for instance, where a defendant acts with the intent to kill a primary target but with only conscious disregard of the risk that others may be seriously injured or killed. Accordingly, in future cases trial courts should reserve the kill zone theory for instances in which there is sufficient evidence from which the jury could find that the only reasonable inference is that the defendant intended to kill (not merely to endanger or harm) everyone in the zone of fatal harm.” (Canizales, supra, 7 Cal.5th at p. 597.) The best protection against a jury’s improper application of a concurrent intent theory is to make sure that it receives proper instruction on that theory and only in the narrow and particularized circumstances to which it may legitimately be applied. But, in emphasizing the duty to protect against juror confusion or improper application, Canizales did not articulate a different or higher standard of review in concurrent intent cases. Rather, by explaining how the jury should apply conventional circumstantial evidence rules, and by focusing on the physical factors informing whether a defendant has exhibited an intent to create a kill zone (see discussion post), Canizales emphasized that only a rare set of circumstances would warrant a concurrent intent instruction and “there will be relatively few cases in which the theory will be applicable and an instruction appropriate.” (Canizales, supra, 7 Cal.5th at p. 608.) Courts should reserve instruction to those fairly unusual cases where the facts fit the theory. They should also take care to “describe the contours and limits of the kill zone theory . . . .” (Id. at p. 609.) In urging that a trial court can pre-weigh, or a reviewing court can reweigh, the evidence, defendant misconstrues an important part of the Canizales analysis. On review the test remains whether substantial evidence had been presented to support a reasonable inference by the jury “that defendant[] intended to create a zone of fatal harm around a primary target.” (Canizales, supra, 7 Cal.5th at p. 610.) We disapprove In re Rayford, supra, 50 Cal.App.5th 754 to the extent it suggested a different standard of review applies. D. A Concurrent Intent Instruction Was Not Supported by Substantial Evidence Although we decline defendant’s invitation to change the standard of review, we agree that a concurrent intent instruction should not have been given based on the evidence here. Justification for instructing on concurrent intent requires substantial evidence that: 1. the defendant intended to kill a primary target; 2. he concurrently intended to achieve that goal by killing all others in the fatal zone he creates; and 3. the alleged attempted murder victim was in that zone. These requirements protect against an improper attempted murder conviction based only on a conscious disregard for life. “Evidence that a defendant who intends to kill a primary target acted with only conscious disregard of the risk of serious injury or death for those around a primary target does not satisfy the kill zone theory. . . . [T]he kill zone theory does not apply where ‘the defendant merely subjected persons near the primary target to lethal risk. Rather, in a kill zone case, the defendant has a primary target and reasons [that] he cannot miss that intended target if he kills everyone in the area in which the target is located. In the absence of such evidence, the kill zone instruction should not be given.’ ” (Canizales, supra, 7 Cal.5th at p. 607.) Canizales found significant that the attack there “occurred at a block party on a wide city street, not in an alleyway, cul-de-sac, or some other area or structure from which victims would have limited means of escape.” (Id. at p. 611.) Clearly, defendant’s conduct endangered Johnson’s life. When defendant fired, Johnson stood to the left and away from Door 1 and in front of Door 2. Two of defendant’s three shots went through Door 2 at a point about four feet off the ground. Although Johnson was not hit, he easily could have been. However, though Johnson was placed in harm’s way, a concurrent intent instruction would only have been warranted if defendant acted not with a mere conscious disregard for life but with a specific intent to kill. To prove that intent, the prosecution invoked the concurrent intent theory. The Canizales factors demonstrate that reliance was misplaced. First, defendant fired three shots from a handgun, fewer than the five shots found insufficient in Canizales. (See Canizales, supra, 7 Cal.5th at pp. 610–611.) Though we do not suggest categorically that a kill zone can never be created with a relatively small number of shots, that number must be considered together with the area into which they are fired. For example, People v. Dominguez (2021) 66 Cal.App.5th 163 involved the firing of “21 shots into a small space enclosed on three sides.” (Id. at p. 187.) Concurrent intent cases involving vehicle shootings have been described as employing a “flurry” (Bland, supra, 28 Cal.4th at p. 331) or “hail” (People v. Tran (2018) 20 Cal.App.5th 561, 567) of bullets. Canizales contrasted its five-shot case with the facts of two other cases. One involved a residential shooting with “[a]t least 50 bullet holes . . . .” (People v. Vang (2001) 87 Cal.App.4th 554, 558.) In another, the defendant “fired as many as ten shots at four people standing in close proximity to one another.” (Washington v. U.S. (D.C.Cir. 2015) 111 A.3d 16, 24; see Canizales, at p. 610.) The fact a defendant chose to shoot into a confined space or at a defined group in close proximity to each other strengthens the inference that the creation of a kill zone was intended. In that sense, the size of the confined space or the close grouping of the targets helps define the intended zone of fatal harm.6 Although three shots might be sufficient to create a zone of fatal harm around a primary target in a confined space or in the midst of a tight group, the situation here is the converse.

6 The Attorney General asserts Mumin did not shoot more than three rounds because he was wounded by the officers’ return fire. But the inference of intent to create a kill zone is drawn from what the defendant did and the amount of force actually used. While there may be a plausible explanation for why he did not fire more shots, it is not sufficient to argue he might have kept shooting and from that speculation infer that he would have intentionally created a kill zone. Canizales teaches we may only evaluate the intent to create a kill zone based upon the actual circumstances. (See Canizales, supra, 7 Cal.5th at pp. 609–611.) The defendant fired from the confines of the community room out into an open area. As in Canizales, the openness of the space created opportunities to escape or to take protective cover. On the facts presented here, it is difficult to identify the scope of an alleged zone of fatal harm other than to say it was outside the room where defendant was hiding. Second, neither detective was hit. Canizales cautioned that whether a defendant intentionally created a lethal zone does not turn on “the effectiveness or ineffectiveness of the defendant’s chosen method of attack.” (Canizales, supra, 7 Cal.5th at p. 611.) However, the absence of injury to a secondary target is one factor to consider in determining whether the defendant intended to kill secondary targets. It is the overall nature of the zone created and the means used to ensure lethality that often provides the most telling evidence of the actor’s intent. As the Bland court pointed out in embracing the concurrent intent doctrine, the theory rests on an inference drawn from “ ‘the nature and scope of the attack,’ ” which was “devastating enough to kill everyone in the group.” (Bland, supra, 28 Cal.4th at pp. 329–330.) Third, it is true that defendant here was much closer to Mackay than the defendants in Canizales. (See Canizales, supra, 7 Cal.5th at p. 611.) If defendant were near Door 1 when he fired, he would also have been close to Mackay, on the other side of the door. However, Johnson was at least 25 feet away and off to the side, in line with Door 2. This fact is relevant because, generally, the farther away a secondary victim is from a primary target, the greater the force necessary to demonstrate an intent to create a lethal zone encompassing both the primary target and the others the defendant is alleged to have concurrently intended to kill. Here it is undisputed that defendant fired three shots from a handgun into an open area, where the alleged primary and secondary targets were positioned at least 25 feet apart and neither was hit. Given the open area, neither the number of shots fired nor the weapon used are sufficient to suggest defendant acted to create a kill zone around Mackay in order to ensure his death. Had Johnson died, the jury could have convicted defendant of his murder based either on a theory of transferred intent or implied malice. What makes this case different is the particular showing of specific intent to kill required to sustain an attempted murder charge. As we clarified in Canizales: “[T]he kill zone theory does not apply where ‘the defendant merely subjected persons near the primary target to lethal risk.’ ” (Canizales, supra, 7 Cal.5th at p. 607.) The Court of Appeal upheld the attempted murder conviction as to Johnson, reasoning as follows. “Based on the evidence, the jury could reasonably have found the following: Mumin armed himself with a semiautomatic firearm and hollow-point bullets. Hollow-point bullets are particularly damaging based on their design. Mumin had recently fatally shot a nonthreatening individual who would not comply with his demands. After trying and failing to escape from his apartment complex, Mumin hid in the community room with his loaded firearm. He heard numerous police officers calling around the apartment complex, as well as a police helicopter overhead. When Mackay began to open Door 1, Mumin believed the police had found his hiding place. In rapid succession, Mumin fired through the opening at Door 1 and swept over to Door 2, firing two more shots that penetrated through the closed door and struck objects some distance away. The jury could reasonably conclude, based on this evidence, that Mumin was unsure exactly where the police officer opening the door was located and intended to create a zone of fatal harm in front of both double doors, killing anyone in that zone in order to ensure that the police officer (Mackay) would be killed as well. It was the last stand of a desperate killer who had endured more than an hour in the community room listening and waiting for police to find him.” (Mumin, supra, 68 Cal.App.5th at pp. 57–58.) The analysis is overbroad. Many of the cited circumstances might support an independent intent to kill or explain why Mumin would be motivated to shoot and kill Mackay to avoid arrest. The circumstances also support an inference that Mumin had the capacity to create a zone of fatal harm using a semiautomatic handgun with hollow point rounds. Yet the essential fact in question remains defendant’s intent. This inquiry lies at the heart of the concurrent intent analysis. To convict of attempted murder on a concurrent intent theory, it is not enough to conclude that the defendant intended to kill the target and simply ran the risk that others might be killed. The Attorney General argues that “Mumin rapidly shot three bullets at Doors 1 and 2, the entrances closest to Mackay because he did not know Mackay’s exact position and he needed to ensure Mackay’s death. [Citations.] Mumin fired the bullets at Door 2 at . . . a height that was capable of striking an officer standing on the other side of the door. [Citation.] In doing so, Mumin intended to kill the multiple officers he believed to be on the other side of those doors, thereby creating a kill zone.” As Canizales recognized, depending on the evidence, a jury might conclude the actor harbored independent intents to kill each person within the range of his explosive or weapon fire. (See Canizales, supra, 7 Cal.5th at p. 608.) But that is not a concurrent intent case. It does not depend on a finding that the actor intended to kill a primary target and created a kill zone intending to kill everyone else in it to ensure the target’s death. Thus, contrary to the Attorney General’s suggestion, even assuming there was evidence Mumin intended to indiscriminately kill everyone in a particular area, a question we need not decide, such evidence would not have supported instructing the jury on concurrent intent. E. The Error Was Prejudicial as to the Allegation of Attempting to Kill Johnson Having concluded the concurrent intent instruction should not have been given with regard to the Johnson count (count 5), we turn to the question of prejudice and how that question should be evaluated. When a jury has been instructed on both proper and improper theories for conviction, the appropriate standard of prejudice turns on the type of error involved. If the improper theory “is incorrect only because the evidence does not support it” (People v. Aledamat (2019) 8 Cal.5th 1, 7 (Aledamat)), reversal is not required if “a valid ground for the verdict remains, absent an affirmative indication in the record that the verdict actually did rest on the inadequate ground” (People v. Guiton (1993) 4 Cal.4th 1116, 1129). Presentation of a factually inadequate theory constitutes an error “of state law subject to the traditional Watson test . . . .” (Id. at p. 1130.)7 By contrast, a legally inadequate theory is not merely incorrect because it is

7 Under Watson, reversal is not required unless it is reasonably probable that a result more favorable to the appellant would have been reached in the absence of the error. (People v. Watson (1956) 46 Cal.2d 818, 836.) factually wanting but “because it is contrary to law.” (Aledamat, at p. 7.) When a given instruction misstates the law, the more demanding standard of Chapman v. California (1967) 386 U.S. 18, 24, applies, requiring reversal unless the error was harmless beyond a reasonable doubt. “ ‘[L]egal error requires a more stringent standard for prejudice . . . [because] jurors are presumed to be less able to identify and ignore an incorrect statement of law due to their lack of formal legal training. [Citation.] Factual errors, on the other hand, are less likely to be prejudicial because jurors are generally able to evaluate the facts of a case and ignore factually inapplicable theories.’ ” (Aledamat, at p. 8.) Here, the prosecutor attempted to rely on a concurrent intent theory to prove an intent to kill Johnson. The court gave an earlier version of CALCRIM No. 600 but modified it, attempting to incorporate the Canizales factors.8 The modified instruction began by stating that, to prove the commission of attempted murder, the People must prove defendant “took at least one direct but ineffective step toward killing another person” and “intended to kill that person.” The instruction defined “direct step,” then continued: “The defendant must possess the intent to kill a human being. It is not required defendant intended to kill a specific human being.” These general principles were accurately described. The court went on to state, however: “A person may intend to kill a specific victim or victims and at the same time intend to kill everyone in a particular zone of fatal harm or ‘kill zone.’

8 The trial here began in September 2019, three months after our Canizales decision. “In order to convict the defendant of the attempted murder of Officer Luke Johnson, the People must prove that the defendant not only intended to kill the person opening the door [Mackay] but also either intended to kill Officer Luke Johnson or any other officer outside the door attempting to apprehend him or intended to kill everyone within the kill zone. If you have a reasonable doubt whether the defendant intended to kill Officer Luke Johnson or any other officer outside the door attempting to apprehend him, or intended to kill the person opening the door by killing everyone in the kill zone, then you must find defendant not guilty of the attempted murder of Officer Luke Johnson. “To determine whether the defendant intended to create a zone of fatal harm or ‘kill zone’ and the scope of any such zone, consider the circumstances of the offense, such as the type of weapon used, the number of shots fired, the distance between the defendant and the alleged victims, and the proximity of the alleged victims to the primary target. “This theory may only be used to convict the defendant of the attempted murder of Officer Luke Johnson if it is proven that the defendant intended to kill everyone in the zone of fatal harm. It is insufficient that the defendant acted with conscious disregard of the risk that others may be seriously injured or killed by his actions.” (Italics added.) As given, the instruction made an intent to kill Mackay the analytical foundation from which to infer an intent to kill Johnson. Thus it relied exclusively on a concurrent intent analysis.9 The ineffective attempt to modify CALCRIM No. 600 resulted in a confusing amalgam that explicitly linked the Johnson attempted murder to a concurrent intent approach. The given instruction did use the phrase “kill zone,” but it did not otherwise define the term or require a finding that defendant intended to create a kill zone around Mackay in order to ensure his death. As a result, it gave no guidance as to how the jury should determine whether Johnson, or any other officer outside the doors for that matter, was actually in the kill zone, as Canizales requires. (Canizales, supra, 7 Cal.5th at pp. 609– 610.) As discussed, the existence and location of a primary target is important because those facts help define the extent of the kill zone, which is necessary in order for the jury to determine whether the secondary target was within that zone. An inference as to the intent to kill is logically linked to the degree of lethal force the defendant has been proven to exert. It would not be reasonable to conclude the defendant intended to kill someone demonstrably outside the zone he intentionally created.10 On the facts here, because the defendant shot into an undelineated open area, the jury would have been at sea

9 As Canizales noted, there can be other available independent intent theories that do not rely on a concurrent intent analysis. (See Canizales, supra, 7 Cal.5th at p. 608). We have no occasion to discuss those theories here. 10 On this point, CALCRIM No. 600 now provides: “A person may intend to kill a primary target and also [a] secondary target[s] within a zone of fatal harm or ‘kill zone.’ A ‘kill zone’ is an area in which the defendant used lethal force that was designed and intended to kill everyone in the area around the primary target.” (Italics added.) attempting to decide just how far the alleged kill zone extended.11 Canizales held the concurrent intent instruction given there and the state’s argument were legally faulty because the “prosecutor’s definition of the kill zone as an area in which people ‘can get killed’ or are in a ‘zone of fire’ was significantly broader than a proper understanding of [what] the theory permits. Indeed, it essentially equated attempted murder with implied malice murder. [Citation.] Thus, the prosecutor’s argument had the potential to mislead the jury to believe that the mere presence of a purported victim in an area in which he or she could be fatally shot is sufficient for attempted murder liability under the kill zone theory.” (Canizales, supra, 7 Cal.5th at p. 614.)

11 In In re Sambrano (2022) 79 Cal.App.5th 724, the court suggested a kill zone instruction could be reduced to a single sentence along these lines: “If, having considered all the circumstances of the attack, you find beyond a reasonable doubt that the defendant intended to kill by killing everyone in the area in which was located, then you may infer, but are not required to infer, that the defendant intended to kill everyone in that area.” (Id. at p. 734, fn. 3.) This suggested instruction contravenes Canizales because it omits the very factors that case directed the jury should be told to consider. (See Canizales, supra, 7 Cal.5th at p. 607 [describing the contours and limits of the concurrent intent theory and relevant circumstances the jury should consider].) As importantly, it fails to require the second Canizales point essential to support the intent inference: a finding that the attempted murder victim was located in the kill zone. In re Sambrano, supra, 79 Cal.App.5th 724 is disapproved to the extent it conflicts with Canizales as well as the views expressed herein. The inadequacy of the instruction here was similarly exacerbated by the prosecution’s closing argument. The district attorney urged defendant knew the police were after him and intended to kill Mackay, the officer opening the door to his hideout. As to Johnson, the prosecutor admitted: “I can’t prove the defendant knew Detective Johnson was standing in the exact position that he was. But the evidence supports, based on the defendant’s action, that the defendant, when he fired those three rounds through the door, intended to kill the person opening the door and intended to kill anyone near him attempting to apprehend him.” Those could be legally sufficient theories, depending on the evidence, but they are not concurrent intent theories. That explanation only relied on the defendant’s direct intents to kill independent of an identified primary target: an intent to kill Mackay, the person opening the door; and an intent to kill anybody else attempting to apprehend him. Only then did the district attorney seem to shift her focus to the kill zone. “There is no doubt that Detective Johnson was in that kill zone. That the defendant, when he hid himself behind those three doors, fired those three rounds through the doors, that he attempted to kill not only the officer opening the door, but every single officer who was near him, who was there to apprehend him after he was located.” (Italics added.) The prosecutor’s argument did not properly characterize a kill zone as being the area around Mackay; the extent of which was to be evaluated in terms of the number of shots fired, the distance between defendant and the alleged secondary victim, and the distance between a secondary victim and Mackay. Instead she described an amorphous area that encompassed any officer outside the doors who was there to arrest him. She did not argue that Mumin intended to kill Mackay as his primary target and, inferentially, intended to kill every secondary target around him to ensure Mackay died. The argument falls short for the same reasons the court’s instructions missed the mark. So we are left with the following conclusions. The concurrent intent instruction should not have been given because it was factually unsupported. The instruction that was given was also not legally sound, nor was the prosecution’s argument on that point. Just as in Canizales, “there is a reasonable likelihood that the jury understood the kill zone instruction in a legally impermissible manner.” (Canizales, supra, 7 Cal.5th at p. 614.) Accordingly we apply the Chapman standard, which permits an affirmance only when, “after examining the entire cause, including the evidence, and considering all relevant circumstances, [the court] determines the error was harmless beyond a reasonable doubt.” (Aledamat, supra, 8 Cal.5th at p. 13.) Applying that standard, we cannot conclude the error was harmless beyond a reasonable doubt as to the attempted murder count involving Johnson. As Bland and Canizales make clear (see discussion, ante, at pp. 8−9) the concurrent intent theory is an alternative way to prove the required intent to kill a surviving victim. Here the prosecutor acknowledged there was no evidence defendant targeted Johnson directly. As a result, she tried to prove the required intent to kill him by relying on the alternative concurrent intent theory. The Attorney General suggests “there was overwhelming evidence Mumin intended to kill the multiple officers he believed were outside the community room,” including Johnson, because “Mumin needed to kill all [the] officers he believed were present — and not just the one opening the door — to escape.” The suggestion is not well taken. Even granting that a reasonable jury could have so concluded, “a reviewing court may hold the error harmless where it would be impossible, based on the evidence, for a jury to make the findings reflected in its verdict without also making the findings that would support a valid theory of liability.” (Lopez, supra, 14 Cal.5th at p. 568; see Aledamat, supra, 8 Cal.5th at p. 15.) Under the present circumstances, the jury’s verdict does not necessarily reflect the jury made the factual inference urged by the People. The gun and magazines recovered in the community room showed defendant had 28 more rounds at his disposal after he fired only three. The conclusions the jury could have drawn from that evidence are several, some of which are more inculpatory than others. Ultimately, it is not for us to reweigh the evidence in hindsight. The concurring opinion acknowledges that, in an appropriate factual scenario, juries may validly rely on the concurrent intent theory to convict a defendant of the attempted murder of a secondary victim. (See conc. opn., post, at pp. 2–3.) However, the concurrence suggests we “heed the lessons of experience” and no longer actually instruct juries on that theory, instead leaving the matter to be explained by counsels’ argument. (Id. at p. 18; see also id. at p. 3.) The concurrence asserts that 46 of 70 Court of Appeal decisions released after Canizales have cited that decision in finding instructional error. (See id. at p. 10; see also conc. opn., post, Appen. A.) Yet the concurrence also acknowledges that 42 of those 46 cases involved trials predating Canizales. (Id. at p. 10.) Canizales explicitly noted that earlier decisions “articulating the kill zone theory are incomplete to the extent that they do not require a jury to consider the circumstances of the offense in determining the application of the kill zone or imply that a jury need not find a defendant intended to kill everyone in the kill zone as a means of killing the primary target . . . .” (Canizales, supra, 7 Cal.5th at p. 607, fn. 5.) Rather than showing the concurrent intent theory is itself unworkable, these cases demonstrate that, in reviewing cases tried before Canizales, the Courts of Appeal have properly applied its clarification and heeded its caution that “there will be relatively few cases in which the theory will be applicable and an instruction appropriate.” (Id. at p. 608.) Even among the four cited cases tried after Canizales that found error, one involved “an earlier version of CALCRIM No. 600 that was identical to that given in Canizales . . .” which failed to “define a kill zone in terms of a primary target at all.”12 Another case, like Rayford, misapplied the standard of review and improperly substituted its evaluation of what inferences were reasonable, thus usurping the province of the jury.13 The legal landscape that has emerged after Canizales to date hardly justifies jettisoning a concurrent intent instruction. The lesson of experience is that the theory is a complex one that must be employed with care and explained with precision. More fundamentally, the concurrence does not persuasively explain how failing to instruct the jury would solve the problems the concurrence articulates. Although prior cases have suggested the concurrent intent theory, when applicable, does not require a jury instruction (see, e.g., Bland, supra, 28 Cal.4th at p. 331, fn. 6), Canizales cautioned “the potential for the misapplication of the kill zone theory, as evidenced by prior appellate cases, illustrates the importance of more clearly

12 People v. Anderson (Nov. 2, 2022, A162395) 2022 WL 16630809, at p. 8 [nonpub. opn]. 13 See People v. Brown (May 26, 2023, B309004) 2023 WL 3672919, at p. 10 [nonpub. opn]. defining the kill zone theory in future cases.” (Canizales, supra, 7 Cal.5th at p. 606.) Yet the concurring opinion would leave this clarification to the parties’ jury arguments. While the arguments of counsel can be of assistance, it is the instructions of the court that are binding on the jurors. Indeed, the standard jury instruction informs them that “[y]ou must follow the law as I explain it to you, even if you disagree with it. If you believe that the attorneys’ comments on the law conflict with my instructions, you must follow my instructions.” (CALCRIM No. 200 [Duties of Judge and Jury]; see also Penal Code section 1126.)14 Further, as the present case demonstrates, even attorneys sometimes misdescribe the concurrent intent theory. Declining to provide correct instructions would expand, not limit, the potential for misapplication of the concept. When determining whether to instruct on concurrent intent at all, the court serves an important gatekeeping function. Consideration of whether the theory is properly presented by the facts of the case, and thus support an instruction, should be considered at the hearing during which instructions are requested by counsel and settled by the court. The concurrence correctly acknowledges that a theory of liability where a defendant intended to indiscriminately kill everyone in an area without targeting any particular person “is distinct” from a concurrent intent theory. (Conc. opn., post, at p. 21.) The foundation for the concurrent intent theory is that

14 That provision states: “In a trial for any offense, questions of law are to be decided by the court, and questions of fact by the jury. Although the jury has the power to find a general verdict, which includes questions of law as well as of fact, they are bound, nevertheless, to receive as law what is laid down as such by the court.” the use of a particularly lethal method of attack to kill a targeted individual can support an inference that the attacker intended to kill the primary target and concurrently intended to kill everyone near the target to ensure the target’s death. The distinction is important. A standard attempted murder instruction might sufficiently explain the intent to kill requirement under the indiscriminate killer scenario. Yet, as Canizales explained, a jury, under certain circumstances, may require more guidance as to how an intent to kill secondary targets relates to a defendant’s intent to kill an identified, primary target. Defendant contends he also suffered prejudice with respect to count 4, the attempted murder count involving Mackay. He suggests the court’s attempted murder instruction only explained the concurrent intent theory and “said nothing about what the jury had to find to convict Mumin on Count 4 involving Mackay . . . .” Not so. The prosecutor’s sole theory of liability as to Mackay was that defendant shot at the door intending to kill the person opening it because defendant knew that person had to be one of the officers who had been pursuing him. Those straightforward inferences do not employ a concurrent intent analysis, nor did the district attorney do so when arguing in support of those charges. Not only did the prosecutor limit her concurrent intent argument to the count involving Johnson, the instruction itself only mentioned Johnson when describing that concept. In light of the evidence, the given instruction, and the jury argument, defendant could not have been prejudiced as to the count involving Mackay. The attempted murder conviction relating to the attempted murder of Mackay in count 4 is affirmed. For a contrary conclusion, defendant cites a note sent by the jury which stated: “We all agreed that the defendant fired at police officers intending to hit one or more of them and had officer died, he’d be guilty of murder. [¶] 1) Is that sufficient to determine intent to kill? [¶] 2) Is intent judged solely on evidence presented or can we speculate about the defendant[’]s state of mind?” The court responded to the first question: “It is not my role to tell you whether the evidence is sufficient to prove intent to kill. It is your role as jurors to determine whether the evidence establishes intent to kill. [¶] See Cal Crim 600, 225.” As to the second question, the court admonished jurors not to “speculate or guess” and told them that they “may draw reasonable inferences from the evidence presented.” The jury’s question asked about an intent to kill but referenced neither Mackay nor a concurrent intent theory. Defendant’s suggestion the jury necessarily applied that theory to Mackay is unfounded. In sum, we conclude the court’s erroneous concurrent intent instruction was prejudicial as to the Johnson attempted murder count.15 However, there was no concurrent intent instruction given as to Mackay, and the other instructions on the Mackay allegations were proper.16

15 We express no view regarding whether double jeopardy principles bar a retrial for the attempted murder of Johnson. (See Burks v. United States (1978) 437 U.S. 1, 10–18.) 16 Amici curiae Innocence Rights of Orange County and the Office of the State Public Defender (OSPD) both urge we should abolish the concurrent intent theory. Neither party has addressed the argument and it was not asserted in the trial court. As this issue is not properly before us, we decline to address it. III. DISPOSITION The judgment of the Court of Appeal affirming the conviction on count 5 is reversed. In all other particulars the judgment is affirmed. The matter is remanded to the Court of Appeal with directions to remand the matter to the trial court for further proceedings as may be necessary.

CORRIGAN, J.

We Concur: JENKINS, Acting C. J.

KRUGER, J.

GROBAN, J.

ROBIE, J.*

__________________________ * Associate Justice of the Court of Appeal, Third Appellate District, assigned by the Chief Justice pursuant to article VI, section 6 of the California Constitution. Concurring Opinion by Justice Liu

Today’s opinion addresses “the proper standard of review when a defendant challenges a court’s decision to instruct on a concurrent intent, or ‘kill zone,’ theory as applied to an allegation of attempted murder.” (Maj. opn., ante, at p. 2.) I agree that in this case, “although the Court of Appeal applied the proper standard, it erroneously concluded that sufficient evidence supported the giving of a concurrent intent instruction” (ibid.) and that this error “was prejudicial as to [one] attempted murder count” (id. at p. 44). I write separately to focus on a threshold issue that would render consideration of the proper standard of review unnecessary: the viability of continuing to instruct juries on the “so-called kill zone theory.” (People v. Canizales (2019) 7 Cal.5th 591, 596 (Canizales).) Under this theory, “a defendant may be convicted of the attempted murder of an individual who was not the defendant’s primary target.” (Canizales, supra, 7 Cal.5th at p. 596.) A jury may convict “only when the jury finds that: (1) the circumstances of the defendant’s attack on a primary target, including the type and extent of force the defendant used, are such that the only reasonable inference is that the defendant intended to create a zone of fatal harm — that is, an area in which the defendant intended to kill everyone present to ensure the primary target’s death — around the primary target and (2) the alleged attempted murder victim who was not the primary target was located within that zone of harm. Taken together, such evidence will support a finding that the defendant harbored the requisite specific intent to kill both the primary target and everyone within the zone of fatal harm.” (Id. at pp. 596–597.) In Canizales, we observed that, “[a]s past cases reveal, there is a substantial potential that the kill zone theory may be improperly applied.” (Canizales, supra, 7 Cal.5th at p. 597.) As a result, “trial courts must be extremely careful in determining when to permit the jury to rely upon the kill zone theory.” (Ibid.) We anticipated “there will be relatively few cases in which the theory will be applicable and an instruction appropriate.” (Id. at p. 608.) We cautioned trial courts to “tread carefully when the prosecution proposes to rely on such a theory” and ensure that any instruction is sufficiently supported by the evidence. (Ibid.) Noting that “the potential for misapplication . . . remains troubling,” we provided a detailed clarification of the inquiry and relevant evidentiary considerations. (Id. at p. 607.) As this case shows, there continues to be confusion as to the proper application of this theory. Before and after Canizales, the kill zone instruction has been given where it should not have been. When we first recognized that “the fact [a] person desires to kill a particular target does not preclude finding that the person also, concurrently, intended to kill others” (People v. Bland (2002) 28 Cal.4th 313, 329 (Bland)), we did not set out to create a new doctrine for proving attempted murder. To the contrary, we said concurrent intent is “not a legal doctrine requiring special jury instructions. . . . Rather, it is simply a reasonable inference the jury may draw in a given case: a primary intent to kill a specific target does not rule out a concurrent intent to kill others.” (Id. at p. 331, fn. 6.) Nevertheless, since Bland, special jury instructions have been developed and the theory used in a range of cases resulting in confusion, misapplication, and prejudicial errors, all while adding little if any analytical value. Today’s opinion again attempts to clarify the theory, but I am doubtful it will lessen the need for error correction by appellate courts. I suggest we yield to experience and abandon the “kill zone theory” as a distinct theory of attempted murder. Doing so would not eliminate the concept of concurrent intent; it would simply clarify that concurrent intent to kill is an inference the jury may draw from the totality of circumstances in attempted murder cases with multiple victims, not a distinct theory warranting a separate instruction. I. Although we have repeatedly used the term “kill zone” in prior cases, it was not coined by us. We adopted it from an out- of-state case, which discussed a hypothetical illustrating the concept of concurrent intent. (Bland, supra, 28 Cal.4th at pp. 329–330.) The “kill zone” shorthand obscures that concurrent intent, first embraced in Bland, is a simple idea. Bland was an attempted murder case with multiple victims. The defendant and a fellow gang member approached a car containing three individuals, identified the driver as a rival gang member, then shot “into the vehicle” and “fired a flurry of bullets” as it drove away. (Bland, supra, 28 Cal.4th at pp. 318, 331.) The driver died, and the two passengers — both of whom, “it appears, were not gang members” — were injured. (Ibid.) The jury convicted Bland of attempted murder of the passengers. (Id. at p. 318.) At trial, the jury was given a “ ‘transferred intent’ ” instruction. (Id. at p. 319.) We examined whether “ ‘a specific intent to kill’ ” could be transferred from the intended victim (the driver) to the unintended victims (the passengers). (Id. at p. 326; see id. at p. 317.) We held that it could not: “[T]he doctrine of transferred intent does not apply to attempted murder.” (Id. at p. 331.) Unlike murder, attempted murder requires “the intent to kill”: the attempt crime “sanctions what the person intended to do but did not accomplish, not unintended and unaccomplished potential consequences.” (Id. at p. 327.) As a result, “[s]omeone who intends to kill only one person and attempts unsuccessfully to do so, is guilty of the attempted murder of the intended victim, but not of others.” (Id. at p. 328.) A defendant’s “mental state must be examined as to each alleged attempted murder victim.” (Ibid.; see id. at p. 331.) At the same time, we recognized that “a person who shoots at a group of people” may still “be punished for the actions toward everyone in the group even if that person primarily targeted only one of them.” (Bland, supra, 28 Cal.4th at p. 329.) This is because “the fact the person desires to kill a particular target does not preclude finding that the person also, concurrently, intended to kill others.” (Ibid.) We quoted the following explanation of concurrent intent from the Maryland Court of Appeals: “ ‘The intent is concurrent . . . when the nature and scope of the attack, while directed at a primary victim, are such that we can conclude the perpetrator intended to ensure harm to the primary victim by harming everyone in that victim’s vicinity. . . . [C]onsider a defendant who intends to kill A and, in order to ensure A’s death, drives by a group consisting of A, B, and C, and attacks the group with automatic weapon fire or an explosive device devastating enough to kill everyone in the group. The defendant has intentionally created a “kill zone” to ensure the death of his primary victim, and the trier of fact may reasonably infer from the method employed an intent to kill others concurrent with the intent to kill the primary victim. When the defendant escalated his mode of attack from a single bullet aimed at A’s head to a hail of bullets or an explosive device, the factfinder can infer that, whether or not the defendant succeeded in killing A, the defendant concurrently intended to kill everyone in A’s immediate vicinity to ensure A’s death. The defendant’s intent need not be transferred from A to B, because although the defendant’s goal was to kill A, his intent to kill B was also direct; it was concurrent with his intent to kill A.’ ” (Id. at pp. 329–330, quoting Ford v. State (Md. 1993) 625 A.2d 984, 1000–1001.) We observed that this understanding of intent was congruent with California case law. (Bland, at p. 330, citing People v. Vang (2001) 87 Cal.App.4th 554.) Bland did not change or add to ordinary principles of attempted murder liability. Every attempted murder conviction requires “ ‘the specific intent to kill and the commission of a direct but ineffectual act toward accomplishing the intended killing.’ ” (Canizales, supra, 7 Cal.5th at p. 602.) And when “a single act is charged as an attempt on the lives of two or more persons, the intent to kill element must be examined independently as to each alleged attempted murder victim.” (Ibid.) Bland simply applied these principles to a specific factual scenario involving multiple victims. It recognized that in certain cases where the defendant intended to kill a particular person, that intent “does not preclude” the jury from finding that the defendant “concurrently” intended to kill others. (Bland, supra, 28 Cal.4th at p. 329.) Instead, the jury may reasonably infer the intent to kill others based on circumstantial evidence of the crime — the “ ‘method employed’ ” or the “ ‘mode of attack.’ ” (Id. at p. 330.) Such inference is not novel or unusual: Because “direct evidence of intent to kill is rare,” intent to kill “ordinarily . . . must be inferred from the statements and actions of the defendant and the circumstances surrounding the crime.” (Canizales, at p. 602.) Concurrent intent does not establish an alternative route to attempted murder liability. “ ‘[D]irect’ ” intent to kill must be proven as to each victim, and the jury may infer such intent from the totality of circumstances. (Bland, supra, 28 Cal.4th at p. 330.) In this regard, concurrent intent is unlike transferred intent, which employs a “legal fiction” to achieve a policy of criminal liability. (People v. Scott (1996) 14 Cal.4th 544, 551.) And unlike transferred intent, concurrent intent is “not a legal doctrine requiring special jury instructions.” (Bland, at p. 331, fn. 6.) “Rather, it is simply a reasonable inference the jury may draw in a given case: a primary intent to kill a specific target does not rule out a concurrent intent to kill others.” (Ibid.) Although the “reasonable inference” of concurrent intent has subsequently been referred to as the “kill zone theory,” cases after Bland have not strayed from its basic understanding. We have referenced the kill zone theory in three cases where it did not apply to the facts at hand. In People v. Smith (2005) 37 Cal.4th 733 (Smith), the defendant fired a single bullet at a woman and her child, both of whom were in his “direct line of fire.” (Id. at p. 745.) We rejected application of the kill zone theory, characterizing it as “simply recogniz[ing]” that under certain circumstances “a rational jury could conclude beyond a reasonable doubt that [a defendant] intended to kill not only his targeted victim, but also all others he knew were in the zone of fatal harm.” (Id. at p. 746.) We noted that it did not “preclude a conclusion” that “two convictions of attempted murder” could be supported “under the totality of the circumstances shown by the evidence” at hand. (Id. at p. 745.) This discussion reaffirmed that the kill zone theory is but one framework for inferring an intent to kill. It is not the only possible framework in multiple victim cases; it applies to a narrow set of circumstances involving “lethal force designed and intended to kill everyone in an area around the targeted victim . . . as the means of accomplishing the killing of that victim.” (Id. at p. 746.) In People v. Stone (2009) 46 Cal.4th 131 (Stone), we reiterated Bland’s explanation “that if a person targets one particular person, under some facts a jury could find the person also, concurrently, intended to kill — and thus was guilty of the attempted murder of — other, nontargeted, persons.” (Stone, at p. 137.) In finding this theory inapplicable to a defendant who fired a single shot into a group of 10 people, we again understood the kill zone theory as one potential inference of intent to kill in cases that involve a “primary target.” (Id. at p. 140; see id. at p. 138, 140–141.) We distinguished this inference from the inference behind “a person who intends to kill” but “has no specific target in mind.” (Id. at p. 140.) Such “[a]n indiscriminate would-be killer” can also satisfy the intent to kill requirement, but this involves a different factual scenario and inference. (Ibid.; see Canizales, supra, 7 Cal.5th at p. 608 [describing Stone as involving an “evidentiary bas[i]s, other than the kill zone theory, on which a fact finder can infer an intent to kill for purposes of attempted murder liability that do[es] not depend on a showing that the defendant had a primary target”].) Similarly, in People v. Perez (2010) 50 Cal.4th 222 (Perez), we rejected the defendant’s contention that all “single-bullet cases involving more than one potential attempted murder victim must be analyzed under a kill zone rationale.” (Id. at p. 232.) We found Bland “not controlling on the[] facts” because its “theory of multiple attempted murder is necessarily defined by the nature and scope of the attack,” and “firing a single shot from a moving car at a distance of 60 feet at the group of eight individuals” did not fit within this framework. (Perez, at p. 232.) Thus, in each of these cases describing and distinguishing the kill zone theory, we retained the same basic understanding that it “is simply a reasonable inference the jury may draw in a given case.” (Bland, supra, 28 Cal.4th at p. 331, fn. 6.) We recently summarized these origins in Canizales and again made clear that “the so-called kill zone theory” was an application of ordinary principles of attempted murder liability to specific facts. (Canizales, supra, 7 Cal.5th at p. 596.) We stated that this theory “permits a jury to infer a defendant’s intent to kill an alleged attempted murder victim from circumstantial evidence (the circumstances of the defendant’s attack on a primary target).” (Id. at p. 597.) Looking to “circumstantial evidence to support a permissive inference regarding a defendant’s intent . . . is not unusual. As we have described on many occasions, intent to kill often must be inferred from circumstantial evidence surrounding the crime.” (Id. at p. 606.) For that reason, our articulation that the jury must find that the “only reasonable inference is that the defendant intended to create a zone of fatal harm” is based on combining CALCRIM No. 225, the standard jury instruction on circumstantial evidence, with the factual scenario that kill zone theory addresses. (Canizales, at p. 597.) Given this background, it is easy to see why a special instruction is never necessary. (Bland, supra, 28 Cal.4th at p. 331, fn. 6; accord, Stone, supra, 46 Cal.4th at p. 137; Smith, supra, 37 Cal.4th at p. 746.) The kill zone theory recognizes one way a jury might infer intent to kill in a specific factual scenario, but it is not a new legal doctrine nor does it change the elements of attempted murder. As one Court of Appeal recognized, “If the evidence supports a reasonable inference that, as a means of killing the primary target, the defendant specifically intended to kill every single person in the area in which the primary target was located, then the prosecutor can make that argument and the jury can draw that inference without the aid of a kill zone instruction — the ordinary instructions on attempted murder will provide all of the necessary legal tools.” (People v. McCloud (2012) 211 Cal.App.4th 788, 803 (McCloud).) Indeed, the kill zone instruction is an anomaly; no other scenario- specific inference of intent to kill appears to come with a specialized instruction.

II.

In Canizales, we offered an extensive discussion of the kill zone theory in an effort to “guard[] against the potential misapplication of the theory” as “evidenced by prior appellate cases.” (Canizales, supra, 7 Cal.5th at pp. 606, 607.) We developed a two-part inquiry for when the theory “may properly be applied.” (Id. at p. 607.) We provided examples of “circumstances of the offense” the jury should consider — “the type of weapon used, the number of shots fired (where a firearm is used), the distance between the defendant and the alleged victims, and the proximity of the alleged victims to the primary target.” (Ibid.) We also clarified that evidence a defendant “acted with only conscious disregard of the risk of serious injury or death for those around a primary target does not satisfy the kill zone theory” and that “evidence of a primary target is required.” (Id. at pp. 607, 608.) We cautioned that the kill zone theory should apply to “relatively few cases.” (Id. at p. 608.) Since we filed Canizales four years ago, 70 appellate cases have cited it in evaluating kill zone instructions given by trial courts. In nearly two-thirds of these cases (46 out of 70), the Court of Appeal found error involving the instruction, with 37 cases finding reversible error. (See appen. A, post.) Most of these cases involved kill zone instructions given by trial courts before we decided Canizales. But even among the five cases involving kill zone instructions given after Canizales, reviewing courts have found error with the instruction in four of those five cases, including the case now before us. (See appen. A, post.) Today’s opinion asserts that these case statistics demonstrate that reviewing courts have “properly applied [Canizales’s] clarification and heeded its caution,” and that some instructions given after Canizales reflect errors addressed in Canizales and today’s opinion. (Maj. opn., ante, at p. 41.) Apart from the merits of each decision, these cases show the striking number of kill zone instructions given and the frequency of error. That many of them involved instructions given prior to Canizales is of little comfort. Canizales did not represent a major departure from prior law. It did not change the underlying concept of concurrent intent, and as today’s opinion recognizes, it did not change the standard for determining when to provide a kill zone instruction (maj. opn., ante, at p. 18). Before and after Canizales, our colleagues on the Courts of Appeal have observed that giving a kill zone instruction is not uncommon and often leads to error. (See In re Sambrano (2022) 79 Cal.App.5th 724, 734 (Sambrano) [“[I]t is not clear why it would ever be prudent to give such an instruction. It appears easy to commit error by instructing the jury on the kill zone theory, but it is literally impossible to err by declining to do so.”], fn. omitted; People v. Medina (2019) 33 Cal.App.5th 146, 156 (Medina) [“A kill zone instruction is never required, and as numerous appellate cases attest, giving such an instruction can often lead to error.”].) Many cases have shown how the kill zone theory can be mistakenly applied to facts that do not support it. Because the theory “is used to support an inference that the defendant concurrently intended to kill a nontargeted victim, . . . evidence of a primary target is required.” (Canizales, supra, 7 Cal.5th at p. 608.) Yet instructions are given in cases with no primary target. In Sambrano, for example, the defendant and two others drove into rival gang territory; the defendant, who was driving, stopped the car after “repeatedly driving past a group of people gathered outside a house,” and his two passengers “began shooting.” (Sambrano, supra, 79 Cal.App.5th at p. 728.) He was convicted of six counts of attempted murder; two alleged victims were struck by bullets outside the house, and four alleged victims who were not hit were inside the house. (Id. at p. 729.) The trial court instructed the jury on the kill zone theory, “stating that Sambrano could be found guilty of attempted murder if he created a kill zone and intended to kill everyone within that zone.” (Id. at p. 732, fn. omitted.) But there was “no evidence that any person at the gathering in front of the . . . house was the primary target of defendants’ attack.” (Id. at p. 734.) There was “no evidence that Sambrano knew or recognized anyone at the gathering or that the . . . house had been defendants’ planned destination” and “no evidence that any particular person at the gathering did or said anything that might have led defendants to target a particular person when the shooting commenced.” (Ibid.) Nor did the prosecutor’s argument establish a particular target; instead, the prosecutor suggested that the kill zone theory could apply “even if there is no primary target.” (Id. at p. 735.) But without evidence of a primary target, no kill zone instruction can be given. (See Canizales, supra, 7 Cal.5th at p. 609 [noting the “correct” observation that “ ‘[w]ithout a primary target, there cannot be concurrent intent because there is no primary intent to kill as to which the intent to kill others could be concurrent’ ”].) The Sambrano court found the kill zone theory “categorically inapplicable” on the facts and the error prejudicial, vacating the convictions. (Sambrano, at p. 734; see id. at p. 736.) Other cases reflect the same error. (See, e.g., People v. Thompkins (2020) 50 Cal.App.5th 365, 394, disapproved on another ground in In re Lopez (2023) 14 Cal. 5th 562, 584 [reasoning that kill zone instruction should not have been given where defendant fired 10 shots into a restaurant and the “prosecution never attempted to identify any particular target victim or victims”]; People v. Mariscal (2020) 47 Cal.App.5th 129, 139 [holding that “giving the [kill zone] instruction was error” because there was no “intended victim” where defendant “had no prior interaction” with the group of men sitting together on bleachers and did not “know them or have any reason to attack any one of them more than any of the others”]; Medina, supra, 33 Cal.App.5th at pp. 149, 150, 156 [concluding that “it was error” to provide kill zone instruction where “there was no evidence the defendants here had a primary target” when they drove through rival gang territory, crashed a car, “and started shooting at bystanders”].) Even when the kill zone theory appears to map onto the facts of a case, the instruction might still be erroneously given because the evidence is insufficient to support such an inference. In Canizales, the defendant “attacked his target by firing five bullets from a nine-millimeter handgun at a distance of either 100 or 160 feet away,” and “the attack occurred at a block party on a wide city street,” not an “area or structure from which victims would have limited means of escape.” (Canizales, supra, 7 Cal.5th at p. 611; see ibid. [noting that “bullets were ‘going everywhere’ ” and “a target . . . immediately ran down a city street after the first shot was fired”].) We held that the instruction should not have been given because “the evidence concerning the circumstances of the attack (including the type and extent of force used . . . ) was not sufficient to support a reasonable inference that defendants intended to create a zone of fatal harm around a primary target.” (Id. at p. 610.) This type of error is also not uncommon. (See, e.g., In re Lisea (2022) 73 Cal.App.5th 1041, 1056 [concluding that the kill zone instruction was “erroneously worded” and had “virtually no evidentiary support” because “only three to six shots were fired from a smaller caliber weapon, from a vehicle moving away from the crowd, into a public area and with no evidence of the shots being fired from close range”]; People v. Booker (2020) 58 Cal.App.5th 482, 500 [concluding there was insufficient evidence to support a kill zone instruction where defendant shot at a car with two individuals inside because the defendant “fired a total of three to seven shots directed at the front driver’s side,” the shots “were directed at [the primary target] at close range,” “there were no bullet holes . . . that would have reflected a spray of bullets,” there was no “evidence any bullets reached . . . where [the nontarget victim] was sitting,” and “there was no evidence suggesting [the defendant] used a rapid-firing semiautomatic or automatic weapon”], fn. omitted.) Misapplication of the theory to a case with insufficient evidentiary support reveals a more basic problem. In some instances, the kill zone theory may be used as an end run around proving the requisite intent to kill in attempted murder cases with multiple victims, particularly where the evidence of such intent is weak. In McCloud, for example, the defendants “fired 10 shots from a semiautomatic handgun at a party at which over 400 people were present. Three bullets struck three victims, killing two and injuring the third. The seven remaining bullets hit no one.” (McCloud, supra, 211 Cal.App.4th at pp. 790–791.) One of the defendants was convicted of second-degree murder for the two deaths and 46 counts of attempted murder based on a kill zone theory. (Id. at p. 791.) The record contained “no evidence that [the defendants] intended to kill 46 people with 10 bullets” nor any “evidence that it would have been possible for them to kill 46 people with 10 bullets (given the type of ammunition and firearm they used), or that they believed or had reason to believe it was possible.” (Id. at pp. 799–800, fn. omitted.) As a result, the Court of Appeal concluded the instruction was erroneously given. It further found the error prejudicial and concluded, in a sufficiency of evidence review, that “the evidence is sufficient to support only eight attempted murder convictions, because 10 shots were fired but two of them killed victims . . . , for which [the defendant] was separately convicted and punished.” (Id. at p. 807.) Erroneous use of the kill zone theory in McCloud resulted in 38 attempted murder convictions that were legally insufficient, even “ ‘view[ing] the evidence in the light most favorable to the prosecution, and presum[ing] in support of the judgment the existence of every fact the trier could reasonably deduce from the evidence.’ ” (McCloud, supra, 211 Cal.App.4th at p. 805.) Unlike the other counts subject to reversal due to instructional error, these counts could not “be retried on remand.” (Ibid.) It is not clear how many other cases of insufficiently supported kill zone instructions involve insufficiently supported attempted murder convictions. But this potential for excessive charges and invalid convictions is a troubling consequence of the instruction. Separate and apart from misapplication of the kill zone theory to the facts of a case, the standard language for the instruction poses its own concerns. The kill zone instruction is part of the attempted murder instruction, CALCRIM No. 600. For one thing, the moniker “kill zone” is itself problematic. It is difficult to imagine how a juror tasked with evaluating the guilt of someone accused of intending to kill “primary” and “secondary” “targets” by creating a “kill zone” could view the defendant through neutral eyes. (CALCRIM No. 600.) One court has said there is “nothing argumentative in this instruction” because it is analogous to terms in other instructions such as “ ‘flight,’ ” “ ‘suppress[ion] of evidence,’ ” and “ ‘consciousness of guilt.’ ” (People v. Campos (2007) 156 Cal.App.4th 1228, 1244.) But there is an important difference between describing a defendant as having “fled [or tried to flee]” (CALCRIM No. 372) or engaged in “[s]uppression” of evidence (CALCRIM No. 371), which are objective facts, and instructing the jury to determine whether “the defendant intended to create a ‘kill zone’ and the scope of such a zone” (CALCRIM No. 600), which is a legal construct. By asking whether the defendant created a “kill zone,” an inflammatory phrase suggesting mass violence, the latter is more susceptible to inviting the jury to see the defendant as dangerous and biasing the jury toward guilt. Further, the kill zone instruction runs the risk of confusing the jury or presenting the theory in a legally impermissible manner. This lengthy instruction, as excerpted from CALCRIM No. 600, is reprinted in appendix B, post. It was reworked after Canizales found it “should be revised to better describe the contours and limits of the kill zone theory as we have laid them out.” (Canizales, supra, 7 Cal.5th at p. 609.) It now addresses Canizales’s concerns that the prior version failed to define “ ‘kill zone’ ” beyond “ ‘particular zone of harm’ ” and failed to “direct the jury to consider evidence regarding the circumstances of defendants’ attack.” (Id. at p. 613.) But the instruction, which today’s opinion characterizes as providing “more guidance” to a jury (maj. opn., ante, at p. 43), still contains language that may result in confusion or error. For example, the instruction presents the kill zone theory as an alternative theory of attempted murder liability by informing the jury that the prosecutor must prove either intent to kill the nontarget victim or intent to kill everyone within the kill zone — when in fact the latter (intent to kill everyone within the kill zone) is how a jury finds the former (intent to kill the nontarget victim). The instruction’s formulation appears to imply, incorrectly, that an intent to kill everyone within the kill zone is different from or negates the requirement to find an intent to kill each alleged victim. (See Canizales, supra, 7 Cal.5th at p. 597.) In addition, the instruction does not say the defendant must “intend[] to kill everyone in the kill zone as a means of killing the primary target” (id. at p. 607, fn. 5, italics added) until the very last sentence. Elsewhere, it refers to the defendant “intend[ing] to kill everyone in the area around the primary target” (CALCRIM No. 600) without saying that the defendant does so to kill the primary target. Even if CALCRIM No. 600 were again revised to address these concerns, I question whether the game is worth the candle. The previous revision led to a more convoluted and lengthy instruction that still appears inaccurate. The fact that courts, prosecutors, and jury instruction authors continue to have trouble properly stating and applying the kill zone theory suggests that a reasonable juror is unlikely to fare better. More fundamentally, further revision of the jury instruction does not address the concern that a special instruction unnecessarily highlights the kill zone theory in jurors’ minds. Although the bench notes to CALCRIM No. 600 recognize that an instruction is “not required” and is “for the court to use at its discretion,” the instruction is accompanied by a directive to “Give when kill zone theory applies.” And because the instruction is long and complex, far more so than the rest of the attempted murder instruction, it may draw attention to the kill zone theory beyond what is warranted to inform the jury of “simply a reasonable inference [it] may draw in a given case.” (Bland, supra, 28 Cal.4th at p. 331, fn. 6.) Because the potential for error in employing the kill zone instruction has proven to be substantial, and because the instruction itself is unnecessary and confusing, I see no reason to retain the instruction or embellish it with more clarifications and admonitions. Today’s opinion again emphasizes the importance of “mak[ing] sure that [the jury] receives proper instruction on [the kill zone] theory and only in the narrow and particularized circumstances to which it may legitimately be applied” (maj. opn., ante, at p. 26), echoing similar language in Canizales. (See Canizales, supra, 7 Cal.5th at p. 608 [“We emphasize that going forward trial courts must exercise caution when determining whether to permit the jury to rely upon the kill zone theory.”].) It reasons that the kill zone theory “is a complex one that must be employed with care and explained with precision,” and that a lack of an instruction “would expand, not limit, the potential for misapplication.” (Maj. opn., ante, at pp. 41, 42.) But we have said repeatedly that concurrent intent is “not a legal doctrine requiring special jury instructions.” (Bland, supra, 28 Cal.4th at p. 331, fn. 6; accord, Stone, supra, 46 Cal.4th at p. 137; Smith, supra, 37 Cal.4th at p. 746.) The concept itself is simple. (Ante, at pp. 3–9.) And where counsel misstates or misapplies it, the court can and should provide a correction. (Maj. opn., ante, at p. 42 [“the court serves an important gatekeeping function”].) At bottom, it is the existence of a distinct kill zone instruction, not the complexity of the underlying concept, that has proven to be problematic. The availability of a kill zone instruction, so labeled and singled out as a distinct theory of attempted murder, has led to overreliance, and such overreliance cannot be solved with a more precise instruction. Other than further exhortation that the instruction should be given sparingly, today’s opinion offers little if any guidance to reduce errors. I think it is time to heed the lessons of experience. The more prudent course is to abandon a separate kill zone instruction.

III.

In this case, defendant Ahmed Mumin was convicted of two attempted murder charges involving alleged victims James Mackay and Luke Johnson. The jury was instructed with a modified version of CALCRIM No. 600, which provided the elements of attempted murder and a kill zone instruction. An examination of the facts shows again how a kill zone instruction can be misapplied. Officers Mackay and Johnson were searching an apartment complex where Mumin was present. (Maj. opn., ante, at pp. 2–3.) From the outside, they “approached a building with four adjacent doors leading to a community room.” (Id. at p. 3.) The doors were closed and Mackay “went to Door 1,” which was the door “farthest to the right.” (Ibid.) Johnson “provided cover” and was “positioned about 25 feet away from Door 1 and to the left of it.” (Ibid.) He was “generally in line with Door 2,” the adjacent door, “and back some distance from the plane of the doorways.” (Ibid.) “Mackay stood by the right hinges of Door 1 and reached across to operate its handle.” (Ibid.) As he “opened the door slightly,” Mumin, who had been hiding inside the community room, “fired once through the opening and twice through the closed Door 2. Neither officer was hit.” (Ibid.) As a threshold matter, it is questionable whether Mackay can properly be considered a “primary target.” (Canizales, supra, 7 Cal.5th at p. 608.) At trial, the prosecutor argued in closing that Mumin “intended to kill the person opening the door and intended to kill anyone near him attempting to apprehend him.” The Attorney General presents a similar argument before us, reasoning that Mumin “had a motive to kill any officer who attempted to take him into custody, not just one officer in particular.” These constructions suggest that the theory of intent here was an intent to kill any approaching officers, which happened to include Mackay, rather than an intent to target Mackay as the officer opening the door and to kill others nearby to ensure his death. To the extent the prosecutor’s theory was that Mumin “believe[d] he’s been cornered” and thus “form[ed] the intent to kill the officers that are trying to apprehend him,” it does not posit a primary target necessary for application of the kill zone theory. Even if Mackay could be considered a primary target, the instruction was still erroneous. Under a kill zone theory, the question for the jury is whether Mumin “intended to create a zone of fatal harm” around Mackay “in which [he] intended to kill everyone present to ensure [Mackay’s] death” and whether Johnson “was located within that zone.” (Canizales, supra, 7 Cal.5th at p. 597.) Any finding that Mumin intended to kill Mackay would not preclude a finding that he also intended to kill Johnson; the intents would arise at the same time. But no such reasonable inference can be supported by these circumstances. In essence, Mumin fired three shots through closed doors into an open area where two officers were located. As today’s opinion notes, this is far from the type of “ ‘flurry’ ” or “ ‘hail’ ” of bullets into confined spaces that have characterized previous kill zone cases. (Maj. opn., ante, at p. 29.) Moreover, because Johnson was “at least 25 feet away and off to the side,” significantly greater force than three bullets would have been “necessary to demonstrate an intent to create a lethal zone encompassing” both Mackay and Johnson. (Id. at p. 30.) As a result, the evidence was insufficient for a jury to reasonably infer that Mumin “acted to create a kill zone around Mackay in order to ensure his death.” (Id. at p. 31.) Furthermore, even if it were reasonable for a jury to infer from Mumin’s three shots that he “intentionally created a zone of fatal harm” around Mackay, a jury could not reasonably infer that such a zone encompassed Johnson. (Canizales, supra, 7 Cal.5th at p. 610.) Johnson was standing 25 feet away from the door Mackay opened. As the prosecutor recognized in closing argument, Johnson specifically positioned himself outside of what he considered to be the range of gunfire; he stood “slightly to the left” to “avoid the fatal funnel.” Taken together, the facts here suggest a scenario in which Mumin intended to kill Mackay “and simply ran the risk that others might be killed,” which is insufficient to sustain an attempted murder conviction as to Johnson. (Maj. opn., ante, at p. 32.) As in other cases with insufficiently supported kill zone instructions, I question whether the error here reveals that the evidence was simply insufficient to support a reasonable inference of Mumin’s intent to kill Johnson. As today’s opinion notes, the prosecutor’s argument confused concurrent intent with intent to “ ‘kill the person opening the door and . . . anyone near him attempting to apprehend him.’ ” (Maj. opn., ante, at p. 38; see id. at pp. 38–39.) While the latter “could be legally sufficient” to support an attempted murder conviction, it is “not concurrent intent.” (Id. at p. 38.) Similarly, the Attorney General here argues that “Mumin intended to kill the multiple officers he believed to be on the other side of those doors” — a suggestion that “Mumin intended to indiscriminately kill everyone in a particular area” (id. at p. 33). But even under this rationale, which is distinct from a kill zone theory, “the intent to kill element must be examined independently as to each alleged murder victim.” (Canizales, supra, 7 Cal.5th at p. 602.) The circumstances that inform our conclusion as to the erroneous kill zone instruction include Mumin’s firing “three shots from a handgun,” the “openness of the space” he fired into, the fact that “neither detective was hit,” and the distance between Johnson and Mackay. (Maj. opn., ante, at pp. 28, 30; see also id. at p. 31.) These also appear to be reasons the evidence in this case is legally insufficient to support a reasonable inference that Mumin had an indiscriminate intent to kill any approaching officers that encompassed or equated to an intent to kill Johnson. The basic attempted murder instruction would have crystallized this issue by asking the jury to consider whether “the defendant took at least one direct but ineffective step toward killing another person,” and “the defendant intended to kill that person.” The totality of circumstances would have informed the jury’s determination of intent to kill. Instead, the added kill zone instruction resulted in the jury considering whether Mumin “either intended to kill Officer Luke Johnson or any other officer outside the door attempting to apprehend him” or “intended to kill everyone within the kill zone.” These inquiries are relevant to Mumin’s intent to kill Johnson, but the instruction as given tended to allow a generalized intent (to kill officers) to supplant the individualized intent (to kill Johnson) required for attempted murder. This misdirection is suggested by a note from the jury asking whether its agreement “that the defendant fired at police officers intending to hit one or more of them and had officer died, he’d be guilty of murder” was “sufficient to determine intent to kill.” The ultimate question submitted to the jury appeared to be lost in the lengthy and confusing verbiage of the kill zone instruction. Whether it is because the instruction was insufficiently supported by evidence or because the evidence on intent to kill Johnson was insufficient to support any attempted murder conviction, the conclusion remains that “a concurrent intent instruction should not have been given.” (Maj. opn., ante, at p. 27.) This case, like the other examples above, shows the recurring misapplication of the kill zone theory and further confirms why eliminating the instruction is the best path forward. Doing so does not prevent a prosecutor from pursuing a concurrent intent argument in an appropriate case; the jury can always reach the inference underlying the theory based on conventional attempted murder principles.

IV.

Finally, it should be noted that eliminating the kill zone instruction does not affect a larger analytical quandary implicated by this case: the circumstances under which multiple attempted murder convictions may be supported. Mumin argues that no kill zone instruction can be supported here because he was unaware that “Johnson or any other officer was anywhere outside the community room besides the person opening [the door].” He reasons that he “necessarily could not have created or intended to create a ‘zone of fatal harm’ . . . around Mackay” with the intent to kill everyone in it if he thought no one else was present. The Attorney General counters that such knowledge “is not required where the circumstances of the attack indicate a deliberate intent to take the lives of others in a location where people may reasonably be expected to be present.” He reasons that a “belief that multiple officers were on the other side” is sufficient. Today’s opinion does not address this contention, instead concluding that the kill zone instruction was insufficiently supported based on other factors. (Maj. opn., ante, at pp. 27–33.) I do not purport to resolve it either. I note only that a defendant’s knowledge of the presence of others is related to a more general difficulty of determining how many attempted murder convictions may be supported in any given multiple victim case. Consider this example from Bland: “ ‘a defendant who intends to kill A and, in order to ensure A’s death, drives by a group consisting of A, B, and C, and attacks the group with automatic weapon fire.’ ” (Bland, supra, 28 Cal.4th at p. 330.) In this hypothetical, the defendant could be liable for the attempted murder of B and C under a kill zone theory. But if the group consisted of even more individuals, how might the jury determine for which ones the defendant is liable for attempted murder? Is it everyone in the group, or only those the defendant can see? Does it depend on the number of shots fired or the number of injuries sustained? These concerns are not unique to kill zone cases. Although a defendant who shoots at the group with intent to kill but without a primary target is not subject to a kill zone analysis, the same questions arise. In Stone, a case not involving the kill zone theory, we recognized that “difficulties can arise . . . regarding how many attempted murder convictions are permissible” and did not premise that difficulty on the underlying theory of intent. (Stone, supra, 46 Cal.4th at p. 140; see id. at p. 141 [declining to address the issue because the defendant was charged with only one attempted murder count].) To be sure, some of the factors that might inform this inquiry are circumstances that a kill zone instruction asks the jury to consider: the type of weapon used, the number of shots fired, the distance between the defendant and the victim, the distance between the nontarget victim and the primary target. (See Canizales, supra, 7 Cal.5th at p. 607.) But these are examples of “circumstantial evidence to support a permissive inference regarding a defendant’s intent,” and such evidence can be examined in any attempted murder case, with no need for a special instruction. (Id. at p. 606.) We examined evidence of this sort in Perez, where the defendant was convicted of eight counts of attempted murder after he “fired a single bullet at a distance of 60 feet, from a car going 10 to 15 miles per hour, at a group of seven peace officers and a civilian who were standing less than 15 feet apart from one another in a dimly lit parking lot late on the night in question. . . . The bullet hit one officer in the hand, nearly severing his finger, but killed no one.” (Perez, supra, 50 Cal.4th at p. 224.) We reversed seven of the eight convictions, reasoning that “a rational trier of fact could find that . . . [the defendant] acted with intent to kill someone in the group he fired upon,” but not all of them. (Id. at p. 230.) We noted that “there is no evidence that defendant knew or specifically targeted any particular individual or individuals in the group of officers he fired upon,” no “evidence that he specifically intended to kill two or more persons with the single shot,” and “no evidence defendant specifically intended to kill two or more persons in the group but was only thwarted from firing off the required additional shots by circumstances beyond his control.” (Id. at pp. 230–231, fns. omitted.) In sum, questions related to how these factors inform attempted murder convictions are not unique to kill zone cases. (See Perez, supra, 50 Cal.4th at p. 232 [rejecting an argument that a kill zone theory applied].) Elimination of the instruction does not affect this inquiry one way or the other. V. Since our recognition of the kill zone theory, there has been little if any upside to specially instructing juries on it. The case law shows continuing confusion as to when and how the theory applies, resulting in judicial inefficiencies and, with troubling regularity, erroneous convictions. Eliminating the kill zone instruction does not eliminate the underlying concept of concurrent intent. Concurrent intent is simply a reasonable inference the jury can make as to a defendant’s intent to kill in a specific factual context involving multiple victims. (See Bland, supra, 28 Cal.4th at pp. 329–330.) As with any other inference, the prosecutor can argue for the inference based on circumstantial evidence, and the jury can draw inferences as the evidence warrants. There is no need for a separate and convoluted instruction, especially one with such high potential for error. I would vacate the conviction for attempted murder of Johnson and attached enhancements, and I would otherwise affirm the judgment.

LIU, J.

I Concur:

EVANS, J. APPENDIX A

Court of Appeal Decisions Evaluating Kill Zone Instructions Since People v. Canizales (2019) 7 Cal.5th 591

An asterisk (*) denotes that the court found error involving the giving of a kill zone instruction. A double asterisk (**) denotes that the error resulted in the reversal or vacatur of an attempted murder conviction. In some cases, the court declined to reach a conclusion as to error and instead concluded that any error was harmless. These cases have not been denoted as cases involving error. (See, e.g., People v. Stelly (Aug. 16, 2021, A157142) [nonpub. opn.] 2021 WL 3615764; People v. Ruiz (Nov. 25, 2019, F076231) [nonpub. opn.] 2019 WL 6271799.) A dagger (†) denotes that the kill zone instruction was given after Canizales. This list does not include cases where the court evaluated whether there was sufficient evidence to support an attempted murder conviction under a kill zone theory, as opposed to evaluating whether there was error involving the giving of a kill zone instruction. (See, e.g., People v. Lazo (Nov. 2, 2022, B304615) [nonpub opn.] 2022 WL 16630910; People v. George (Jan. 11, 2021, E072299) [nonpub opn.] 2021 WL 82315.)

1. People v. Mason (Aug. 15, 2019, B283892) (nonpub. opn.) 2019 WL 3822003 2. People v. Salvador Espinoza (Aug. 15, 2019, B288107) (nonpub. opn.) 2019 WL 3821795 3. People v. Dorantes (Sept. 3, 2019, B289777) (nonpub. opn.) 2019 WL 4164803** 4. People v. Guardado (Oct. 2, 2019, B284144) (nonpub. opn.) 2019 WL 4855111** 5. People v. Galstyan (Nov. 4, 2019, B279947) (nonpub. opn.) 2019 WL 5689840 6. People v. Goins (Nov. 12, 2019, B281831) (nonpub. opn.) 2019 WL 5884387 7. People v. Gray (Nov. 21, 2019, B282321) (nonpub. opn.) 2019 WL 6206257 8. People v. Harris (Nov. 21, 2019, D075379) (nonpub. opn.) 2019 WL 6208343 9. People v. Singh (Nov. 22, 2019, E067985) (nonpub. opn.) 2019 WL 6242187** 10. People v. Garcia (Nov. 25, 2019, B259708) (nonpub. opn.) 2019 WL 6269807 11. People v. Ruiz (Nov. 25, 2019, F076231) (nonpub. opn.) 2019 WL 6271799 12. People v. Anderson (Dec. 12, 2019, B251527) (nonpub. opn.) 2019 WL 6768776* 13. People v. Warner (Dec. 16, 2019, C077711), review denied and opinion ordered nonpublished March 25, 2020,

S260341

14. People v. Garcia (Dec. 18, 2019, C066714) (nonpub. opn.) 2019 WL 6888452** 15. People v. Rios (Dec. 20, 2019, F074350) (nonpub. opn.) 2019 WL 6975115 16. People v. Esquivel (Dec. 23, 2019, B269545) (nonpub. opn.) 2019 WL 7046538** 17. In re Rayford (2020) 50 Cal.App.5th 754** 18. People v. Booker (2020) 58 Cal.App.5th 482** 19. People v. Cardenas (2020) 53 Cal.App.5th 102** 20. People v. Mariscal (2020) 47 Cal.App.5th 129* 21. People v. Thompkins (2020) 50 Cal.App.5th 365** 22. People v. Escobar (Jan. 10, 2020, B259309) (nonpub. opn.) 2020 WL 112664* 23. People v. Kennedy (Jan. 15, 2020, B264661) (nonpub. opn.) 2020 WL 218756* 24. People v. Torres (Jan. 17, 2020, C087086) (nonpub. opn.) 2020 WL 255068 25. People v. Ratcliffe (Feb. 11, 2020, E063690) (nonpub. opn.) 2020 WL 634410* 26. People v. Casique (Feb. 21, 2020, B284945) (nonpub. opn.) 2020 WL 858137** 27. People v. Granados (Feb. 25, 2020, B257627) (nonpub. opn.) 2020 WL 896844 28. People v. Stone (Mar. 2, 2020, B293532) (nonpub. opn.) 2020 WL 994144 29. People v. Gomez (Mar. 4, 2020, B293727) (nonpub. opn.) 2020 WL 1041611** 30. People v. King (Mar. 18, 2020, E070384) (nonpub. opn.) 2020 WL 1284895* 31. People v. Melson (April 1, 2020, B292679) (nonpub. opn.) 2020 WL 1545707** 32. People v. Mays (April 3, 2020, B291995) (nonpub. opn.) 2020 WL 1648660** 33. People v. Miranda (April 8, 2020, B266817) (nonpub. opn.) 2020 WL 1698391** 34. People v. Gonzalez (April 10, 2020, B296206) (nonpub. opn.) 2020 WL 1815073** 35. People v. Williams (April 27, 2020, B259888) (nonpub. opn.) 2020 WL 1983064 36. People v. Alvarado (May 1, 2020, H045500) (nonpub. opn.) 2020 WL 2092478** 37. People v. Sanders (May 1, 2020, B295960) (nonpub. opn.) 2020 WL 2110306** 38. People v. Vivero (June 8, 2020, C086268) (nonpub. opn.) 2020 WL 3046066 39. People v. Riberal (Sept. 29, 2020, C077018) (nonpub. opn.) 2020 WL 5793209 40. People v. Henson (Oct. 4, 2020, C084770) (nonpub. opn.) 2020 WL 6054127** 41. People v. Quiroz (Oct. 16, 2020, E069820) (nonpub. opn.) 2020 WL 6110984** 42. People v. Dominguez (2021) 66 Cal.App.5th 163** 43. People v. Morales (2021) 67 Cal.App.5th 326 44. People v. Mumin (2021) 68 Cal.App.5th 36 — Although the Court of Appeal found no error in the giving of a kill zone instruction, today’s decision reverses that finding and vacates an attempted murder conviction.**† 45. In re Bruno-Martinez (Feb. 18, 2021, C091819) (nonpub. opn.) 2021 WL 631981* 46. People v. Reyes (April 5, 2021, B301357) (nonpub. opn.) 2021 WL 1248216 47. In re Evans (April 30, 2021, B281093) (nonpub. opn.) 2021

WL 1711631**

48. People v. Montanez (May 3, 2021, C083092) (nonpub. opn.) 2021 WL 1730252 49. People v. Gonzalez (May 12, 2021, C089973) (nonpub. opn.) 2021 WL 1956474** 50. People v. Brown (May 21, 2021, C089252) (nonpub. opn.) 2021 WL 2024911** 51. People v. Josue Sanchez (May 28, 2021, B302549) (nonpub. opn.) 2021 WL 2176486 52. People v. Oliver (July 1, 2021, B307225) (nonpub. opn.) 2021 WL 2701376† 53. People v. Morris (Aug. 11, 2021, D076312) (nonpub. opn.) 2021 WL 3523405** 54. People v. Stelly (Aug. 16, 2021, A157142) (nonpub. opn.) 2021 WL 3615764 55. In re Sirypangno (Oct. 4, 2021, D078705) (nonpub. opn.) 2021 WL 4785924** 56. People v. Sanchez-Gomez (Oct. 15, 2021, A156198) (nonpub. opn.) 2021 WL 4807976** 57. People v. Aguilar (Dec. 9, 2021, F077784) (nonpub. opn.) 2021 WL 5832887** 58. In re Lisea (2022) 73 Cal.App.5th 1041** 59. In re Sambrano (2022) 79 Cal.App.5th 724** 60. People v. Lee (2022) 81 Cal.App.5th 232* 61. People v. Perez (2022) 78 Cal.App.5th 192** 62. People v. Brown (Feb. 2, 2022, G060395) (nonpub. opn.) 2022 WL 522503 63. People v. Fields (April 25, 2022, C068047) (nonpub. opn.) 2022 WL 1210474** 64. People v. Cerda (July 5, 2022, B232572) (nonpub. opn.) 2022 WL 2436942 65. In re Milam (Aug. 4, 2022, B312401) (nonpub. opn.) 2022

WL 3097295**

66. People v. Anderson (Nov. 2, 2022, A162395) (nonpub. opn.) 2022 WL 16630809**† 67. In re Hurtado (Feb. 9, 2023, B320947) (nonpub. opn.) 2023

WL 1852252**

68. People v. Avalos (Mar. 24, 2023, F082849) (nonpub. opn.) 2023 WL 2620905*† 69. People v. Brown (May 26, 2023, B309004) (nonpub. opn.) 2023 WL 3672919**† 70. People v. Trujillo (June 7, 2023, F081571) (nonpub. opn.) 2023 WL 3857586** APPENDIX B Standard Kill Zone Instruction, Excerpted from CALCRIM No. 600

“ “[A person may intend to kill a primary target and also [a] secondary target[s] within a zone of fatal harm or ‘kill zone.’ A ‘kill zone’ is an area in which the defendant used lethal force that was designed and intended to kill everyone in the area around the primary target. “In order to convict the defendant of the attempted murder of _____ , the People must prove that the defendant not only intended to kill _____ but also either intended to kill _____ , or intended to kill everyone within the kill zone. “In determining whether the defendant intended to kill _____ , the People must prove that (1) the only reasonable conclusion from the defendant’s use of lethal force, is that the defendant intended to create a kill zone; and (2) _____ was located within the kill zone. “In determining whether the defendant intended to create a ‘kill zone’ and the scope of such a zone, you should consider all of the circumstances including, but not limited to, the following: “[• The type of weapon used(;/.)] “[• The number of shots fired(;/.)] “[• The distance between the defendant and _____ (;/.)] “[• The distance between _____ and the primary target.] “If you have a reasonable doubt whether the defendant intended to kill _____ or intended to kill _____ by killing everyone in the kill zone, then you must find the defendant not guilty of the attempted murder of _____ .]” (CALCRIM No. 600.)

2 See next page for addresses and telephone numbers for counsel who argued in Supreme Court.

Name of Opinion People v. Mumin __________________________________________________________

Procedural Posture (see XX below) Original Appeal Original Proceeding Review Granted (published) XX 68 Cal.App.5th 36 Review Granted (unpublished) Rehearing Granted __________________________________________________________

Opinion No. S271049 Date Filed: August 17, 2023 __________________________________________________________

Court: Superior County: San Diego Judge: Kenneth K. So __________________________________________________________

Counsel:

Raymond M. DiGuiseppe, under appointment by the Supreme Court, for Defendant and Appellant.

Annee Della Donna for Innocence Rights of Orange County as Amicus Curiae on behalf of Defendant and Appellant.

Mary K. McComb, State Public Defender, and Elizabeth H. Eng, Deputy State Public Defender, for the Office of the State Public Defender as Amicus Curiae on behalf of Defendant and Appellant.

Rob Bonta, Attorney General, Matthew Rodriquez, Acting Attorney General, Lance E. Winters, Chief Assistant Attorney General, Julie L. Garland and Charles C. Ragland, Assistant Attorneys General, Steve Oetting, Arlene A. Sevidal, Collette Cavalier and Minh U. Le, Deputy Attorneys General, for Plaintiff and Respondent. Counsel who argued in Supreme Court (not intended for publication with opinion):

Raymond M. DiGuiseppe Attorney at Law P.O. Box 10790 Southport, NC 28461 (910) 713-8804

Elizabeth H. Eng Deputy State Public Defender 770 L Street, Suite 1000 Sacramento, CA 95814 (916) 322-2676

Minh U. Le Deputy Attorney General 600 West Broadway, Suite 1800 San Diego, CA 92101 (619) 738-9055