¶43 (concurring) — I am concerned that both the majority and Justice Gordon McCloud’s concurrence make broad statements about when a Petrich19 instruction is required in a multicount case. As the majority quietly observes, this question is not before us. Majority at 223 n.10.1 am also concerned that the concurrence erroneously characterizes the majority as announcing a broader holding than it does. But to me, the path to greater clarity is to state the relevant holding in its simplest terms: Carson’s claim of ineffective assistance of counsel fails because the State made a clear election before the case went to the jury, specifying exactly three separate acts for the three separate counts of first degree child molestation. Relying on the State’s election—as both the majority and Justice Gordon McCloud’s concurrence ultimately do—I conclude that defense counsel’s performance was neither deficient nor prejudicial. On this basis, I concur in the decision to affirm Carson’s conviction.

¶1 — David Carson appeals his conviction on three counts of child molestation in the first degree. He argues on appeal that he received ineffective assistance of counsel because his attorney objected to a Petrich1 instruction proposed by the State. A defendant must overcome a heavy burden in order to prevail on a claim of ineffective assistance; he must demonstrate that his attorney’s performance was professionally deficient and that he suffered prejudice as a result. Carson satisfies neither prong of this standard.

¶2 Defense counsel reasonably concluded that the proposed instruction’s language, tailored as it is for use in single-count cases, would be confusing and potentially prejudicial in Carson’s multicount case. Under these circumstances, Carson cannot establish deficient performance. Moreover, defense counsel’s objection did not prejudice Carson because the prosecution’s closing argument clearly elected the acts on which the State was relying. For these reasons, we affirm.

BACKGROUND

I. Factual Background

¶3 In 2009, Carson, who was homeless at the time, moved into the home of his childhood friend Dustin Halbert. Halbert lived with his fiancée, Tiffany Hagen, and her *211three children, including C.C., who had just turned five years old at the time Carson moved in. Carson watched the children during the day while Halbert and Hagen worked. Carson lived with the family for approximately one year and two months.

¶4 Three months after Carson had moved out of the house, C.C. told Hagen that Carson “tried to put his penis in [C.C.’s] butt.” Hagen called police. Thirteen days later, C.C. underwent a medical examination and was interviewed by child forensic interviewer Cornelia Thomas. The forensic interview lasted just over 40 minutes. A DVD (digital video disk) of the interview was introduced as an exhibit at Carson’s trial. The jury viewed the video of the forensic interview three times: once during Thomas’s testimony at trial and twice during its deliberations after sending requests for the video to the court.

¶5 As recorded in the DVD, C.C. describes three separate incidents of sexual assault by Carson in some detail: one where Carson twisted C.C.’s “business,” C.C.’s term for a penis; and two incidents where Carson attempted to put his “business” in C.C.’s “bottom.” According to C.C.’s statements during the interview, Carson twisted C.C.’s “business” in a bathroom. C.C. also said that Carson attempted to put his “business” in C.C.’s “bottom” while they were in his “mom’s room,” which he also referred to as his “mom and dad’s” room. At some point during this assault, Carson put tape on C.C.’s mouth and used zip ties to tie C.C.’s hands behind his back. In the third incident that C.C. described, Carson attempted to put his “business” into C.C.’s “bottom” in C.C.’s own room after making C.C. look at his Spider-Man blanket.

¶6 The following month, Detective Thomas Catey interviewed Carson at the Pierce County Sheriff’s Department headquarters. During the interview, Carson denied having any sexual contact with C.C. Carson claimed that the allegations were retaliation for Carson’s moving out of their house and leaving them in a financial bind.

*212¶7 The State charged Carson with one count of rape of a child in the first degree and one count of child molestation in the first degree. Eleven days before trial, the State filed an amended information charging three counts of child molestation in the first degree. Because C.C. never provided dates on which each event occurred, each count of the amended information specified a charging period covering the entire time that Carson lived in the same house as C.C.—April 1, 2009 to May 31, 2010:

That DAVID WILLIAM CARSON, in the State of Washington, during the period between the 1st day of April, 2009 and the 31st day of May, 2010, did unlawfully and feloniously, being at least 36 months older than C C, have sexual contact with C C, who is less than 12 years old and not married to the defendant and not in a state registered domestic partnership with the defendant, contrary to RCW 9A[.]44[.]083, and against the peace and dignity of the State of Washington.

II. Trial

¶8 When he testified at trial, C.C. had difficulty recalling the specifics of each incident, often responding with “I forgot” or “I don’t remember” to questions from both attorneys. C.C. recalled a few particulars, such as Carson tying C.C.’s hands and taping his mouth while they were in his parents’ room, but C.C. was unable to confirm or recount most of the details he had described in his videotaped interview 18 months earlier.2 During his cross-examina*213tions of C.C. and other state witnesses, defense counsel largely avoided questions that could elicit details of the specific incidents. When examining his own witnesses, defense counsel focused on Carson’s background, Carson’s responsibilities while he lived with Halbert and Hagen, and the witnesses’ observations of C.C.

¶9 The prosecution’s closing argument focused exclusively on the three incidents that C.C. described in his videotaped interview. The prosecutor explicitly told the jury that those incidents were the only acts the State wished the jury to focus on for the purposes of its deliberations. The prosecutor then proceeded to walk the jury through the specifics of these three incidents, which correlated with the three described in C.C.’s interview.

¶10 In the defense’s closing argument, defense counsel argued that the allegations were entirely fabricated. He argued that true sources of the allegations against Carson were Halbert and Hagen, who concocted the claims of sexual abuse as retaliation for Carson’s leaving them in a financial bind. During the relatively short time that defense counsel spent discussing the specific allegations against Carson, he focused almost entirely on the video of C.C.’s interview with Thomas and urged the jury to view the video. He argued that the allegations against his client were completely false, the by-product of coaching by C.C.’s parents combined with embellishment by C.C. himself.

¶11 Prior to closing arguments, the parties discussed the State’s proposed Petrich instruction, which read:

The State alleges that the defendant committed acts of Child Molestation in the First Degree against C.C. on multiple occasions. To convict the defendant on any count of Child Molestation in the First Degree, one particular act of Child Molestation in the First Degree must be proved beyond a reasonable *214doubt, and you must unanimously agree as to which act has been proved. You need not unanimously agree that the defendant committed all the acts of Child Molestation in the First Degree.

¶ 12 The possibility of a Petrich instruction had first been raised by the trial court the week before trial. Defense counsel had then suggested that the instruction might not be necessary because the evidence showed “three separate and distinct, identifiable acts.” On the second day of trial, the parties discussed their proposed instructions with the court, including the State’s Petrich instruction. Defense counsel argued that the instruction was unnecessary because the evidence pointed to “three separate and distinct incidents, one in [C.C.’s] room, one in his mother’s room, and one in the bathroom.”3 Defense counsel argued that reading the Petrich instruction could cause jury confusion:

I think [the Petrich instruction] can be confusing in a case where you have specific instances alleged, where three acts potentially are alleged . . . and three acts are charged. If you read that instruction, it confuses the jury into thinking, well, if you agree that one act happened, then you must agree that all of them happened. And that concerns me.

¶13 The following day, after both parties rested, defense counsel specifically objected to the State’s proposed Petrich instruction as confusing. The State responded that it had not elected to tie specific acts to specific counts “in the Information,” noting that each of the counts alleged the same charging period were “due in large part to [C.C.’s] inability to specifically say on this day and at this time” each incident occurred. The trial court sustained defense counsel’s objection and declined to give the instruction.

*215¶14 The jury returned verdicts of guilty on all three counts.

III. Appeal

¶15 Carson appealed, asserting that his trial counsel’s objection to the Petrich instruction constituted ineffective assistance of counsel.4 The Court of Appeals affirmed. State v. Carson, 179 Wn. App. 961, 320 P.3d 185 (2014). The court held that trial counsel’s objection to the Petrich instruction was a reasonable trial tactic and thus did not constitute ineffective assistance. The court quoted extensively from the record, noting defense counsel’s concerns that the Petrich instruction could confuse the jury. The Court of Appeals further noted that this objection comported with the defense’s broader trial strategy, in which defense counsel “avoided discussing specific incidents in his closing argument and argued instead that CC’s testimony and statements were so muddled, inconsistent, and confusing that they created a reasonable doubt about whether Carson had committed any of the acts or the charged crimes . . . .” Id. at 979.

¶16 In dissent, Judge Worswick asserted that defense counsel’s objection to the Petrich instruction was based on a belief that Petrich was inapplicable in multicount cases. Id. at 982-84. She concluded that this objection was based “on an erroneous view of the law” and thus cannot “be characterized as a legitimate trial tactic.” Id. at 984.

STANDARD OF REVIEW

¶17 Ineffective assistance of counsel is a fact-based determination, and we review the entire record in deter*216mining whether a defendant received effective representation at trial. State v. Rhoads, 35 Wn. App. 339, 342, 666 P.2d 400 (1983); State v. Grier, 171 Wn.2d 17, 33, 246 P.3d 1260 (2011). The defendant bears the burden of establishing both “that counsel’s performance was deficient” and that “the deficient performance prejudiced the defense.” Strickland v. Washington, 466 U.S. 668, 687, 104 S. Ct. 2052, 80 L. Ed. 2d 674 (1984). Reviewing courts must be highly deferential to counsel’s performance and “should recognize that counsel is strongly presumed to have rendered adequate assistance and made all significant decisions in the exercise of reasonable professional judgment.” Id. at 690.

ANALYSIS

¶18 Carson fails to establish either deficient performance or prejudice. Defense counsel did not perform defi-ciently because his objection to the State’s proposed Petrich instruction was a reasonable trial tactic with the legitimate goal of avoiding jury confusion. That objection was both tactically and strategically reasonable. Carson’s ineffective assistance claim also fails to satisfy the prejudice prong of the Strickland standard because the State elected the three acts on which it was relying, thus rendering a Petrich instruction unnecessary and eliminating any possible prejudice that could have resulted from defense counsel’s objection to such an instruction. We therefore reject Carson’s ineffective assistance claim and affirm.

I. Background on the Petrich Instruction

¶19 In the trial that gave rise to the Petrich instruction, the State presented evidence of “numerous incidents of sexual contact [that] were described in varying detail.” Petrich, 101 Wn.2d at 568. Despite presenting at least four distinct incidents and evidence indicating an unknown number of additional incidents, the State charged Petrich with only one count each of indecent liberties and statutory rape. At the close of evidence, the defense moved to compel *217the State to elect the specific act on which it relied for conviction. The court denied the motion, and the jury convicted the defendant of both statutory rape and indecent liberties.

¶20 This court held that the trial court’s denial of the defendant’s motion to compel election deprived the defendant of his right to a unanimous verdict, setting forth what became known as the Petrich rule:

When the evidence indicates that several distinct criminal acts have been committed, but defendant is charged with only one count of criminal conduct, jury unanimity must be protected. . . . The State may, in its discretion, elect the act upon which it will rely for conviction. Alternatively, if the jury is instructed that all 12 jurors must agree that the same underlying criminal act has been proved beyond a reasonable doubt, a unanimous verdict on one criminal act will be assured.

Id. at 572.

¶21 The Petrich instruction was later incorporated into the Washington Pattern Jury Instructions. 11 Washington Practice: Washington Pattern Jury Instructions: Criminal 4.25, at 110-12 (3d ed. 2008) (WPIC).5 The WPIC’s note on use for the Petrich instruction states that the instruction should be used “when the evidence indicates that several distinct criminal acts have been committed, but the defendant is charged with only one count of criminal conduct.” See id. at 110 note on use. The WPIC instruction reflects the single-count roots of Petrich, instructing the jury that it must find that “one particular act” was “proved beyond a reasonable doubt” and “must unanimously agree as to which act has been proved.” Id. at 110.

*218II. Deficient Performance

¶22 Defense counsel’s objection to the State’s proposed Petrich instruction did not constitute deficient performance. Our analysis begins with the strong presumption that counsel’s performance was reasonable. State v. Kyllo, 166 Wn.2d 856, 862, 215 P.3d 177 (2009). “When counsel’s conduct can be characterized as legitimate trial strategy or tactics, performance is not deficient.” Id. at 863. This presumption can be overcome if the defendant can establish that “ ‘there is no conceivable legitimate tactic explaining counsel’s performance.’ ” Grier, 171 Wn.2d at 33 (quoting State v. Reichenbach, 153 Wn.2d 126, 130, 101 P.3d 80 (2004)).

¶23 The presumption of effective representation imposes on the defendant the burden on appeal to “show in the record the absence of legitimate strategic or tactical reasons supporting the challenged conduct by counsel.” State v. McFarland, 127 Wn.2d 322, 336, 899 P.2d 1251 (1995). Carson fails to meet that burden in this case because the record demonstrates that defense counsel’s decision to object to the Petrich instruction was due to concerns about the potential for jury confusion. Counsel’s concerns were reasonable, and acting on them was a legitimate trial tactic. His performance, therefore, was not deficient.

A. Counsel’s objection was reasonable because the proposed instruction was confusing and potentially prejudicial

¶24 The proposed instruction that the State offered was a word-for-word copy of the model Petrich instruction that appears in the Washington Pattern Jury Instructions. As defense counsel noted, that instruction’s statement that “one particular act... must be proved beyond a reasonable doubt” made little sense in Carson’s case because Carson was charged with three separate counts of child molestation. The confusion was exacerbated by the final sentence of *219the instruction, which would have informed the jury that it “need not unanimously agree that the defendant committed all the acts of Child Molestation in the First Degree.” As defense counsel argued, “If you read that instruction, it confuses the jury into thinking, well, if you agree that one act happened, then you must agree that all of them happened.” At best, then, the instruction would have been confusing and it would have failed to ensure that the jury relied on a separate act for each count. At worst, the instruction would have been highly prejudicial to Carson.6 The record firmly demonstrates that these concerns were the animating force behind defense counsel’s objection to a Petrich instruction.

¶25 It is true, as Judge Worswick noted in her Court of Appeals dissent, that the State’s proposed instruction was “an accurate statement of applicable law.” Carson, 179 Wn. App. at 984. But an accurate statement of law can be confusing when it is applied to circumstances different from those that existed when the statement of law was first made. Here, the specific language of the Petrich instruction was designed for single-count cases and is confusing when read in a multicount case. While the State’s proposed instruction was an accurate statement of the law, defense counsel may reasonably have feared that reading it would confuse the jury to his client’s detriment in Carson’s particular case. For these reasons, the objection was a legitimate trial tactic.

¶26 Defense counsel’s objection also advanced the defense’s broader trial strategy. On the record, defense counsel explicitly stated that his “theory of the case” was that *220Carson “left [C.C.’s parents] in a bind and they suggested” the details of the sexual misconduct to C.C. As the Court of Appeals noted, defense counsel’s summation “avoided discussing specific incidents . .. and argued instead that CC’s testimony and statements were so muddled, inconsistent, and confusing that they created a reasonable doubt about whether Carson had committed any of the acts or the charged crimes.” Id. at 979.

¶27 Because the defense’s theory focused on the credibility of C.C. and his testimony, defense counsel may have concluded any parsing out of the individual acts would have undercut his strategy by planting the idea in the jury’s mind that C.C. might have told the truth about at least some of what happened. At best, then, the Petrich instruction was irrelevant to the defense’s broader trial strategy; at worst, it could have actively undercut that strategy. This further underscores the reasonableness of defense counsel’s decision to object to the reading of the State’s proposed instruction.

¶28 The concurrence asserts that Carson’s attorney should not have objected to the modified Petrich instruction because such an instruction would not have been inconsistent with his primary strategy of painting the charges as wholly false.7 We must resist the temptation to substitute our own personal judgment for that of Carson’s attorney because “it is all too easy for a court, examining counsel’s defense after it has proved unsuccessful, to conclude that a particular act or omission of counsel was unreasonable.” Strickland, 466 U.S. at 689. Thus, the deficient performance inquiry does not permit us to decide what we believe would have been the ideal strategy and then declare an attorney’s performance deficient for failing to follow that strategy. On *221the contrary, counsel’s performance is adequate as long as his challenged decisions “can be characterized as legitimate trial strategy or tactics.” Kyllo, 166 Wn.2d at 863 (emphasis added). The decision to object to the State’s proposed Petrich instruction plainly can be so characterized. Even assuming the objection would have been, as the concurrence suggests, “not inconsistent with” trial counsel’s primary strategy (concurrence (Gordon McCloud, J.) at 235), many trial advocacy experts recommend that attorneys eschew alternative arguments before a jury,, which may view the presentation of an alternative argument as a sign that the attorney believes his first argument is weak:

[A] difficult issue is whether an advocate should advance alternative theories that are not inconsistent—for example, defending a contract suit by claiming duress and also arguing that the defendant’s performance did not constitute a breach of the contract. Because of the cost of overtrying, advancing the second defense may be viewed by the jury as revealing the attorney’s lack of confidence in his first defense. Here again, most experienced attorneys advise, as a general rule, against advancing multiple theories to the jury, even if such theories are not inconsistent.

ROBERT H. Klonoff & Paul L. Colby, Winning Jury Trials: Trial Tactics and Sponsorship Strategies 39-40 (3d ed. 2007). Indeed, we have previously held that an all-or-nothing approach, though risky, is a reasonable trial strategy. Grier, 171 Wn.2d at 42. Here, defense counsel may reasonably have wished to avoid an argument that could be summarized as “[a] 11 the allegations are false—and even if they are not all false, you have to agree on which allegations are true” or “Carson did not abuse C.C.—and even if he did, he did not do it three separate times.”

¶29 Because objecting to the proposed Petrich instruction in Carson’s case can be characterized as part of a legitimate trial strategy, defense counsel did not perform deficiently when he objected to the instruction and instead focused on portraying the allegations against his client as wholly false.

*222 B. Counsel’s objection was not based on a misunderstanding of law

¶30 We reject Carson’s argument, echoing the Court of Appeals dissent, that defense counsel’s objection to the Petrich instruction was per se unreasonable because it was “based on misunderstandings of the law.” Carson, 179 Wn. App. at 984-85 (Worswick, C.J., dissenting). The day before he objected, defense counsel stated that he did not think that a Petrich instruction was necessary in Carson’s specific case because the evidence revealed “three separate and distinct incidents, one in [C.C.’s] room, one in his mother’s room, and one in the bathroom.” Defense counsel appears to be correct in this regard—we have never specifically held that Petrich must be read in cases where there is exact congruence between the number of incidents described in the evidence and the number of criminal counts charged.8

¶31 But more importantly, this was not the reason that defense counsel cited when he objected the following day. As discussed above, the record demonstrates that the potential for jury confusion was the driving force behind defense counsel’s objection to the proposed Petrich instruction. *223When asked if he was objecting to the proposed Petrich instruction, defense counsel’s response was “I think it is confusing, yes.” Concerns that the jury could be “confused” or “misled” dominated defense counsel’s explanations of why he did not want a Petrich instruction read to the jury.9 Thus, defense counsel’s objection was not “based on” his understanding (or misunderstanding) of applicable law and his performance could not have been rendered deficient by any misconceptions he may have harbored regarding that law. For these reasons, even if we were to adopt a per se rule that misunderstandings of law automatically render an attorney’s performance deficient, it would not rescue Carson’s ineffective assistance claim.

¶32 Moreover, no decision of this court has addressed the circumstances of Carson’s trial: a defendant charged with n (which, in Carson’s case, is equal to three) counts, evidence tending to show only n incidents that meet the elements of the crime, and the prosecution identifies the acts on which the charges are based.10 Certainly, defense counsel believed *224that the evidence showed only three distinct incidents—a point that he made each time he discussed the possibility of a Petrich instruction, including the discussion that occurred after the State completed its case in chief. See 1 Report of Proceedings (RP) at 6; 3 RP at 335; 4 RP at 408. Reading the Petrich instruction in such a multicount case could be prejudicial to the defendant, as defense counsel noted: the instruction’s reference to “one particular act” combined with the disclaimer that the jury “need not unanimously agree that the defendant committed all the acts” might confuse a jury into thinking that they may convict the defendant on all counts as long as they unanimously agree on at least one criminal act.

¶33 The WPICs do not provide an example of a multicount variation on the Petrich instruction. As previously noted, the WPIC note on use for the Petrich instruction advises courts to “[u]se this instruction . . . when the evidence indicates that several distinct criminal acts have been committed, but the defendant is charged with only one count of criminal conduct.” WPIC 4.25 at 110 note on use (emphasis added). A sentence in the comment on the pattern instruction states that “[i]f the instruction is being modified for multiple counts, then the instruction needs to clearly require unanimity for one particular act for each count chargedId. at 113 cmt. (citing State v. Watkins, 136 Wn. App. 240, 148 P.3d 1112 (2006)). The comment does not, however, provide an example of a suitable instruction.11

*225¶34 Against that background, one can hardly blame defense counsel for choosing to avoid the Petrich instruction rather than attempting an ad hoc rewrite of the instruction, particularly given that the distinctions between the various incidents were irrelevant to the defense’s theory of the case. Accordingly, it was reasonable for defense counsel to object to the Petrich instruction. For these reasons, Carson’s trial counsel did not perform deficiently. We reject his ineffective assistance claim on that basis.

III. Prejudice

¶35 Given the State’s clear election of three acts in its closing argument, Carson was not prejudiced by the absence of a Petrich instruction. Because we are reviewing only defense counsel’s performance rather than the trial court’s failure to give a Petrich instruction, we use the Strickland standard for ineffective assistance claims and require the defendant to demonstrate that his attorney’s deficient performance prejudiced him. State v. Robinson, 138 Wn.2d 753, 765-69, 982 P.2d 590 (1999). We use the Strickland standard even where the defendant claims that his attorney’s deficient performance affected his constitu*226tional rights. Id. at 765-66; State v. West, 139 Wn.2d 37, 42, 983 P.2d 617 (1999).12

¶36 Under Strickland, we do not presume prejudice unless the trial “loses its character as a confrontation between adversaries,” State v. Webbe, 122 Wn. App. 683, 694-95, 94 P.3d 994 (2004), as when there is (1) a complete denial of the assistance of counsel, (2) the State’s interference with counsel’s assistance, or (3) an actual conflict of interest. See id.; Strickland, 466 U.S. at 692; In re Pers. Restraint of Davis, 152 Wn.2d 647, 674, 101 P.3d 1 (2004). Strickland explained that a presumption of prejudice makes sense in these narrow circumstances because “[prejudice ... is so likely that case-by-case inquiry into prejudice is not worth the cost” and that with respect to conflicts of interest, “counsel breaches the duty of loyalty, perhaps the most basic of counsel’s duties.” Strickland, 466 U.S. at 692.13

*227¶37 None of Strickland’s presumptively prejudicial circumstances exist here. Consequently, under Strickland’s standard of review, Carson bears the burden of establishing that “there is a reasonable probability that, but for counsel’s deficient performance, the outcome of the proceedings would have been different." Kyllo, 166 Wn.2d at 862. Here, far from demonstrating prejudice, the circumstances of Carson’s trial strongly suggest that the absence of a Petrich instruction was harmless because the State, in its closing argument, elected the acts on which it was relying.14

¶38 Petrich’s multiple acts instruction applies only when the State fails to “elect the act,upon which it will rely for conviction.” Petrich, 101 Wn.2d at 572; see also WPIC 4.25 at 110 note on use (“If there is evidence of multiple distinct occurrences of the crime, but the prosecution elects to rely upon a specific occurrence to support a conviction, then this Petrich instruction should not be used.”). For an election to be effective, “either the State must tell the jury which act to rely on in its deliberations or the court must instruct the jury to agree on a specific criminal act.” State v. Kitchen, 110 Wn.2d 403, 409, 756 P.2d 105 (1988).

¶39 We have never held that the State’s election of an act must be ratified by the court or incorporated into the charging document or jury instructions in order to be effective. On the contrary, and as our use of the phrase “tell the jury” in Kitchen suggests, id. (emphasis added), an election can be made by the prosecuting attorney in a verbal statement to the jury as long as the prosecution “clearly identified] the act upon which” the charge in question is based. State v. Thompson, 169 Wn. App. 436, 474-75, 290 P.3d 996 (2012) (“[b]ecause the State clearly identified the act upon which the sexual motivation allegation was based” *228in its closing argument, “no unanimity instruction was necessary”); compare, e.g., id., and In re Pers. Restraint of Delgado, 160 Wn. App. 898, 902, 261 P.3d 899 (2011) (prosecutor “ ‘clearly elected ... the criminal acts associated with the two counts during its closing arguments’ ” (quoting State v. Delgado, noted at 139 Wn. App. 1068, 2007 WL 2085344, at *4, 2007 Wash. App. LEXIS 2098, at *10)), with State v. Williams, 136 Wn. App. 486, 497, 150 P.3d 111 (2007) (no clear election in closing argument where prosecutor “emphasized” one act over others but did not “expressly elect to rely only on” one act “in seeking the conviction”).15

¶40 Here, the evidence tended to show only three separate acts16 and an election arguably was not even necessary. But even if we were to assume arguendo that the evidence could be construed as pointing to additional acts, the State, in its closing argument, clearly and explicitly elected the three acts on which it was relying for conviction: the incident where C.C.’s “business was twisted ... in the bathroom” and the incidents where Carson “tried to put his penis in [C.C.’s] bottom” in C.C.’s room and in his parents’ room. The State specifically disclaimed its intention to rely on any other instances where Carson may have attempted to put his penis in C.C.’s bottom, stating that it was only “focusing on” the incidents in C.C.’s room and Hagen’s room and telling the jury that those were the only incidents that the State “would like you to focus on for the purposes of your deliberations.”

*229¶41 This suffices to constitute an election. The prosecutor, unlike the court, cannot actually instruct the jury to ignore particular testimony or evidence. By specifying exactly three instances of sexual misconduct and disclaiming the State’s intention to rely on other acts, the State effectively elected the acts on which the State sought a conviction.17 Because Petrich instructions need not be read in cases where the State makes an election, no Petrich instruction was necessary in Carson’s trial and the absence of such an instruction did not prejudice Carson.18 Thus, even if the defense attorney’s objection constituted deficient performance, Carson’s ineffective assistance claim would fail.

CONCLUSION

¶42 Carson has established neither deficient performance nor prejudice. We therefore reject his ineffective assistance claim and affirm.

¶44 (concurring) — Defense counsel objected to a Petrich20 instruction in this case for two reasons. First, he voiced a legitimate concern that the instruction—as proposed by the State—could confuse the jury into thinking that if the evidence supported a conviction on one count, it necessarily supported convictions on all three counts. 3 Report of Proceedings (RP) at 336. Second, he asserted—incorrectly—that no unanimity instruction is necessary in a multicount case. He explained that “[generally, when you read the comments to the Petrich instruction, it doesn’t apply ... to multi[ ]count cases:. .. it’s when you have one count but you have like six possible acts that could have accounted for [it].” 4 RP at 405 (emphasis added). *231He supported this argument by stating that he had “looked at the comments [to the pattern jury Petrich instruction] that indicate, as I read it, that [the instruction] should only be used where you’re alleging one count but multiple acts.” Id. at 408 (citing 11 Washington Practice: Washington Pattern Jury Instructions: Criminal 4.25, at 110-12 (3d ed. 2008) (WPIQ).

¶45 The majority holds that these were good reasons showing legitimate defense strategy because a Petrich instruction is always unnecessary when the number of acts alleged is equal to the number of counts charged. See, e.g., majority at 222 (“Defense counsel appears to be correct [that a Petrich instruction was not necessary]—we have never specifically held that Petrich must be read in cases where there is exact congruence between the number of incidents described in the evidence and the number of criminal counts charged.” (emphasis added)).

¶46 I respectfully disagree. If defense counsel had done his background research, he would have discovered that our court has held—and the WPICs confirm—that a Petrich-type unanimity instruction should be given in multicount cases and that it should be given in tried-and-true language that our court has quoted and the WPIC committee has endorsed. I therefore agree with the dissent in the Court of Appeals that trial counsel’s performance was deficient.

¶47 However, I also agree with the majority that the deficient performance did not cause prejudice. I therefore concur.

I. The Majority Incorrectly Concludes That a Unanimity Instruction Is Unnecessary in a Multicount Case

¶48 As the majority notes, the Petrich instruction originated in a case in which the State charged only one count each of two crimes and supported each count with multiple alleged acts. Majority at 216-17 (quoting Petrich, 101 Wn.2d at 572). But Petrich has outgrown its single-count roots. Now, it also applies where multiple acts provide the bases *232for multiple counts of the same crime. 11 WPIC 4.25, at 113 cmt. (“If the instruction is being modified for multiple counts, then the instruction needs to clearly require unanimity for one particular act. for each count charged.” (citing State v. Watkins, 136 Wn. App. 240, 148 P.3d 1112 (2006))); State v. Vander Houwen, 163 Wn.2d 25, 37-39, 177 P.3d 93 (2008) (defendant entitled to unanimity instruction where he was charged with 10 counts of waste of wildlife and 10 counts of killing game out. of season based on evidence that defendant shot 10 different elk).

¶49 This extension of Petrich is unsurprising: a Petrich instruction serves just as compelling a purpose in a multicount case as it does in a single-count case. Vander Houwen, 163 Wn.2d at 38 (“Because the State did not articulate which cause number went with each elk, it is impossible to ensure that all of the jurors voted to convict based on the same two dead elk.”); State v. Kitchen, 110 Wn.2d 403, 411, 756 P.2d 105 (1988) (Petrich ensures jury unanimity “as to which act or incident constitutes the crime” (emphasis added)).

¶50 Contrary to the majority’s assertion, the WPIC does “provide an example of a multicount variation on the Petrich instruction.” Majority at 224. The WPIC comment cites to State v. Hayes, in which Division One held the following to be a proper multicount Petrich instruction:

“A separate crime is charged in each count. You must decide each count separately. Your verdict on one count should not control your verdict on any other count.

“To convict [for each count, you must find as an element]:

. . That on or about the 1st day of July, 1990 through the 31st day of May, 1992, but an occasion separate and distinct from that charged in [the remaining counts], the defendant had sexual intercourse with [the victim];

“. . . For you to return a verdict of guilty all twelve jurors must agree that the same incident of sexual intercourse has been proved beyond a reasonable doubt.”

*23381 Wn. App. 425, 431 n.9, 914 P.2d 788 (1996) (emphasis added) (second and fourth alterations in original) (citations omitted) (quoting jury instructions).

¶51 The WPIC comments also cite State v. Borsheim, 140 Wn. App. 357, 165 P.3d 417 (2007), another multicount case. WPIC 4.25, at 113 cmt. Borsheim held that the following instruction satisfies the jury unanimity requirement in a multicount case:

“There are allegations that the Defendant committed acts of rape of [a] child on multiple occasions. To convict the Defendant, one or more particular acts must be proved beyond a reasonable doubt and you must unanimously agree as to which act or acts have been proved beyond a reasonable doubt. You need not unanimously agree that all the acts have been proved beyond a reasonable doubt.”

Borsheim, 140 Wn. App. at 364 (quoting jury instructions).

¶52 In light of these two examples, defense counsel did not have to choose between “avoid [ing] the Petrich instruction” and “attempting an ad hoc rewrite of the instruction.” Majority at 225. Instead, he could have read the authority cited in the WPIC comment and requested an instruction based on the examples it provided. I respectfully disagree with the majority’s conclusion that trial counsel’s performance reflected a legitimate tactical decision. Instead, trial counsel’s failure to conduct this background research constituted deficient performance. See State v. Kyllo, 166 Wn.2d 856, 865-69, 215 P.3d 177 (2009) (defense counsel’s request for a particular WPIC constituted deficient performance where research would have shown at least three decisions and a statute revealing that the instruction misstated the law).

¶53 The majority also makes the sweeping general assertion that where the number of counts charged is the same as the number of acts alleged, that congruence makes Petrich inapplicable:

[N]o decision of this court has addressed the circumstances of Carson’s trial: a defendant charged with n (which, in Carson’s *234case, is equal to three) counts [and] evidence tending to show only n incidents that meet the elements of the crime .... Reading the Petrich instruction in such a multicount case could be prejudicial to the defendant ....

Majority at 223-24 (emphasis added).

¶54 Again, I respectfully disagree. Congruence between the number of acts alleged and the number of guilty verdicts returned is certainly relevant to the prejudice inquiry in this case. But congruence between acts alleged and counts charged makes no difference to the initial question presented here: whether counsel reasonably refused a unanimity instruction in the first place. Without a unanimity instruction, the defendant has no assurance that individual jurors will convict based on the same single act for each count. This is equally true whether the number of acts alleged does or does not equal the number of counts charged.

¶55 The majority is incorrect when it asserts that “we have never ‘specifically held’ that a Petrich instruction must be read” in cases where “there is exact congruence between the acts alleged and the counts charged.” Majority at 222 n.8. In Vander Houwen, the state found 10 slain elk in the defendant’s apple orchard and charged him with 10 counts each of waste of wildlife and unlawful big game hunting—10 counts of each charge for 10 alleged acts of shooting an elk. 163 Wn.2d at 31, 38. Because there was neither a unanimity instruction nor an election, and because the jury returned guilty verdicts on only two counts of unlawful killing, we reversed those two convictions. Id. at 38-39. We explained that because the State failed to specify which elk were the bases for which counts, it was “impossible to ensure that all of the jurors voted to convict based on the same two dead elk.” Id. at 38. We thus specifically held that a Petrich instruction is constitutionally required in a case where there is exact congruence between the acts alleged and the counts charged.

*235II. The Majority Incorrectly Holds That Defense Counsel’s Objection to the Unanimity Instruction Was a Legitimate Trial Strategy

¶56 The majority also concludes that defense counsel objected to the jury unanimity instruction because this “advanced the defense’s broader trial strategy.” According to the majority, this strategy was to “ ‘avoid[ ] discussing specific incidents ... and argue [ ] instead that [all of] CC’s testimony [lacked credibility].’ ” Majority at 219-20 (second alteration in original) (quoting State v. Carson, 179 Wn. App. 961, 979, 320 P.3d 185 (2014)).

¶57 This seems like an overly formalistic assessment of how a criminal defense lawyer’s theory of the case works.

¶58 First, the majority’s theory ignores the presence of instruction 3, which advised the jurors that they “must decide each count separately” and that their “verdict on one count should not control [their] verdict on any other count.” Clerk’s Papers at 63. It is unclear why defense counsel would have objected to a unanimity instruction on the basis that it distinguished between the charges while ignoring an instruction that explicitly required the jury to make just such a distinction.

¶59 Second, the majority is incorrect in asserting that a unanimity instruction would have been inconsistent with the general defense strategy of attacking the accuser’s credibility. Carson argued that the jury should not convict on any count at all; the unanimity instruction would have told the jury that it also could not convict on any single count unless every juror unanimously agreed on the incident forming the basis for that count. Success on the first theory depends on discrediting the victim’s testimony as to all three alleged incidents; success on the second theory depends on discrediting the victim’s testimony as to any one, two, or three of those same incidents. The second argument is not inconsistent with the first. In fact, they are completely consistent: they both depend on discrediting the *236victim. Far from contradicting the general defense, the unanimity instruction would have added an extra layer of protection for the defendant’s constitutional rights.

¶60 As noted above, defense counsel did express a legitimate concern about confusing the jury—he feared that in a multicount case, the standard Petrich instruction might confuse the jury into thinking that if it convicted on one count, it must convict on all counts. See majority at 219-20. But defense counsel could have addressed this confusion without also jeopardizing his client’s right to a unanimous verdict. He could have requested an instruction like the one in Hayes, which tells the jurors that they can convict on any one count only if they unanimously agree that the defendant committed the act supporting that count on “ ‘an occasion separate and distinct from that charged in [the remaining counts].’ ” 81 Wn. App. at 431 n.9 (emphasis added) (alteration in original). Our cases clearly support the giving of such an instruction in cases involving not just single counts but also multiple counts. See State v. Noltie, 116 Wn.2d 831, 846, 809 P.2d 190 (1991); Vander Houwen, 163 Wn.2d at 37-38.

¶61 For these reasons, I would hold that defense counsel’s objection to a jury unanimity instruction constituted deficient performance. I agree with the dissent in the Court of Appeals that “basing a decision on an erroneous view of the law can [not] be characterized as a legitimate trial tactic.” Carson, 179 Wn. App. at 984.

CONCLUSION

¶62 I agree with the majority’s holding on prejudice, and I would affirm solely on that basis. But I disagree with the majority’s conclusion that a Petrich instruction is unnecessary in a multicount case, I disagree with the majority’s implication that a unanimity instruction is unnecessary where the number of acts alleged equals the number of counts *237charged, and I disagree with the majority’s conclusion that a unanimity instruction is inconsistent with the defense of general denial. For these reasons, I respectfully concur.