¶ 82. (dissenting). This trial was a contest of the credibility of the only two witnesses to the street encounter: the defendant and the complaining witness. This was a classic instance of "he-said-he-said."
¶ 83. As the prosecutor correctly argued to the jury, "Basically what this case comes down to is credibility. Is Clarence Price [the complaining witness] more credible or is Thomas Mayo [the defendant] more credible?"
¶ 84. The defendant and the complaining witness both testified. They agreed that there was an altercation. They disagreed, however, about what happened during the encounter. The defendant asserted that it was a drug deal gone bad. The complaining witness asserted that it was an unprovoked robbery. Each man *684had a felony record. The jury had no independent confirmation of either man's competing account of the encounter.
¶ 85. Credibility determinations are for the jury. The problem is that the credibility determinations in the present case were made in a trial awash with errors and improper conduct by the defense counsel and by the prosecuting attorney, as the majority opinion carefully explains.
¶ 86. I agree with the majority opinion that the multiple errors of both the defense counsel and the prosecutor must be viewed cumulatively, that is in the aggregate, to determine whether the errors were "so serious as to deprive the defendant of a fair trial, a trial whose result is reliable."1
¶ 87. The majority opinion and the State conclude that the errors and improper conduct were not sufficient to warrant a new trial under any of the three doctrines that protect a fair trial against trial errors: plain error (Wis. Stat. § 901.03(4)), interest of justice (Wis. Stat. § 751.06), or prejudicial error (Wis. Stat. § 805.18). I disagree with the majority opinion and conclude that the cumulative errors affected the entire evidentiary picture and directly and substantially affected the credibility contest.
¶ 88. The errors the defendant and majority opinion identify can be grouped into three overlapping and interrelated categories: (I) ineffective assistance of defense counsel; (II) improper statements by the prosecutor; and (III) improper admission of hearsay testimony.
¶ 89. The majority opinion concludes that there was deficient defense counsel performance and errone*685ous prosecutorial conduct. Although I might have concluded that the defense counsel and prosecutor erred in more ways than the majority opinion identifies, for purposes of this dissent, I accept the majority opinion's conclusions about what conduct was erroneous. The majority opinion also concludes that the hearsay testimony was admissible under the excited utterance exception to the hearsay rule. Majority op., ¶ 54.1 accept this conclusion for purposes of this dissent. In light of the other errors the majority opinion identifies, the hearsay ruling and other instances in which the majority opinion finds no error are ultimately irrelevant in my analysis.
¶ 90. In contrast to the majority opinion, I conclude that the cumulative, interrelated errors (as identified by the majority opinion) were pervasive and so infected the jury's credibility determination that the aggregate errors undermine confidence in the verdict. A new trial is warranted regardless of which test is used and regardless of whether the State or the defendant has the burden of proof of prejudicial error.2
H-1
¶ 91. The majority opinion concludes that various aspects of defense counsel's conduct fell below the objective constitutional standard for effective assistance of counsel in several respects, all of which amount to defense counsel's complete failure to conduct any *686independent investigation (but were not prejudicial). Defense counsel's deficient performance as acknowledged by the majority opinion is as follows:
(1) Defense counsel failed to conduct any independent investigation and failed to interview any potential witnesses, relying completely and solely on the police reports and defense counsel's conversations with the defendant. Majority op., ¶¶ 3, 57, 59, 62.
(2) Subsumed under defense counsel's failure to investigate is the defense counsel's failure to obtain the complaining witness's sworn testimony at the preliminary hearing, which might have been used to impeach the complaining witness.
(3) Had defense counsel investigated he would have been in a position to present possible corroboration testimony from an additional witness, McMorris. Majority op., ¶ 62.
¶ 92. The Sixth Amendment right "to have the Assistance of Counsel" is the right to effective assistance of counsel.3 Effective assistance of counsel requires "an adequate investigation of the facts of the case, consideration of viable theories, and development of evidence to support those theories."4 Since Strickland v. Washington, 466 U.S. 668 (1984), the seminal case on ineffective assistance of counsel, the United States Supreme Court has become more exacting in analyzing investigatory failures.5
*687¶ 93. The type and intensity of the investigation defense counsel conducts varies from case to case, and the court generally defers to counsel's judgment.6 Nevertheless, this court has held that defense "counsel's lack of any significant independent investigation falls outside of this wide spectrum" of representation and preparation.7 The United States Supreme Court has similarly declared that "if counsel entirely fails to subject the prosecution's case to meaningful adversarial testing, then there has been a denial of Sixth Amendment rights that makes the adversary process itself presumptively unreliable." United States v. Cronic, 466 U.S. 648, 659 (1984).
¶ 94. This court has approved of and adhered to Standard 4-4.1 of the ABA Standards for Criminal Justice, The Defense Function (approved draft 1971, 2d ed. 1982, 3d ed. 1993), which imposes an unequivocal duty on defense counsel to conduct "a prompt investigation of the circumstances of the case and explore all avenues leading to facts relevant to the merits of the case ...." This duty to investigate exists "regardless of the accused's admission or statements to defense counsel of facts constituting guilt or the accused's stated desire to plead guilty." For cases referring to the duty-to-investigate standard with approval, see, e.g., State v. Love, 2005 WI 116, ¶ 40, 284 Wis. 2d 111, 700 N.W.2d 62; State v. Pitsch, 124 Wis. 2d 628, 638, 369 N.W.2d 711 *688(1985); State v. Felton, 110 Wis. 2d 485, 501, 329 N.W.2d 161 (1983); Roe v. State, 95 Wis. 2d 226, 238-39, 290 N.W.2d 291 (1980); State v. Bowie, 92 Wis. 2d 192, 206, 284 N.W.2d 613 (1979); State v. Harper, 57 Wis. 2d 543, 552-53, 205 N.W.2d 1 (1973).
¶ 95. The record in the present case is clear: There was absolutely no investigation on behalf of the defendant. Defense counsel conceded this point. Majority op., ¶ 57. The majority opinion acknowledges this point. Majority op., ¶¶ 3, 57, 59, 62. This deficiency, however, was not a matter of trial strategy. A lawyer cannot make a rational decision about a defense without first investigating. All of defense counsel's errors in failing to investigate were clear error.
¶ 96. We now turn to the question whether the defense counsel's numerous errors in failing to investigate the case require a new trial. The majority opinion subjected the claims of ineffective assistance of counsel to a cumulative error analysis, examining whether, when taken as a whole, defense counsel's deficiencies prejudiced the defendant. Majority op., ¶ 64. I agree with the cumulative error approach; I disagree with the majority's conclusion that the defendant was not prejudiced.
¶ 97. In the present case, defense counsel's total failure to investigate may be viewed as effectively denying the defendant assistance of counsel. Under this analysis, no prejudice need be shown, and no harmless error test need be applied.8 The defendant should get the relief sought: a new trial.
*689¶ 98. If, however, the Strickland prejudicial error standard is applied to defense counsel's failure to investigate in the present case, the focus of the inquiry regarding whether counsel's constitutionally deficient performance is prejudicial is not on the outcome of the trial, but on "the reliability of the proceedings."9 ¶ 99. The State's "he-said-he-said" case was weak. Defense counsel's failure to investigate affected the entire evidentiary picture and directly and substantially affected the credibility contest between the defendant and the complaining witness. Under the total circumstances of this case, defense counsel's deficient performance undermined the reliability of the proceedings. Defense counsel's failure to investigate therefore constitutes prejudicial error in the present case.
II
¶ 100. Although the defendant is, m my opinion, entitled to a new trial on the basis of the errors of defense counsel alone, I turn next to discuss the other errors that infected the trial.
¶ 101. In addition to the errors of the defense counsel, the majority opinion concludes that the prosecutor also committed errors. The prosecutor's errors are described in the majority opinion as follows:
(1) In her opening statement the prosecutor infringed upon the defendant's federal (Fifth Amendment) and state (Art. I, § 8) constitutional right to remain silent by calling attention to the fact that the *690defendant remained silent after being accused of robbery. Majority op., ¶¶ 3, 46.10
(2) In questioning the State’s witnesses diming the case-in-chief, the prosecutor repeatedly infringed upon the defendant's federal (Fifth Amendment) and state (Art. I, § 8) constitutional right to remain silent by calling attention to the fact that the defendant remained silent after being accused of robbery. Majority op., ¶¶ 3, 46.11
(3) During closing argument, the prosecutor made improper comments regarding the role of defense counsel when she said that it was defense counsel's job to "get his client off the hook," and "not to see justice done but to see that his client was acquitted." Majority op., ¶¶ 3, 42, 43.
(4) During closing argument, the prosecutor improperly told jurors that her job was to examine police reports, to decide whether to file charges, and to decide whether to dismiss charges if she thought the charges were unjust or untrue. The prosecutor's comments can be found at majority op., ¶ 16. The prosecutor's reference to her quasi-judicial task of filing charges amounted to the improper discussion by counsel of materials not in evidence and therefore exceeded the scope of proper argument. Majority op., ¶ 41.
*691¶ 102. These prosecutorial errors affected the jury's view of the defendant and his version of the incident. By stressing the defendant's silence in the face of accusations, the prosecutor wanted the jury to infer that the defendant is guilty: an innocent person would have denied the charge.
¶ 103. In her closing statement the prosecutor painted an ugly portrait of defense counsel relentlessly focused on procuring a "not guilty" verdict at all costs. The logical implication of the prosecutor's remarks was don't trust defense counsel, don't trust the evidence defense counsel offered, and don't trust the defendant.
¶ 104. In contrast, the prosecutor presented herself as prosecuting only cases in which the complaining witness offered a reliable rendition of events. The prosecutor wanted the jury to infer that she would not have brought this case if she didn't believe the complaining witness and that because the prosecutor believes the complaining witness, the jury should as well.
¶ 105. In short, the prosecutor's improper conduct was designed to and did attack the credibility of the defendant and underscored the complaining witness's credibility. In a case like the instant one, where the credibility of the defendant and the complaining witness determines guilt,, the prosecutor's errors were grave indeed.
¶ 106. The prosecutor's improper conduct was exacerbated by the defense counsel's deficient performance in failing to investigate, in failing to present a corroborating witness, in failing to procure the testimony of the complaining witness at the preliminary examination to use for impeachment purposes, and in failing to object when the prosecutor erred.
¶ 107. When I consider the centrality of the credibility contest and the defense counsel's and the *692prosecutor's errors in the aggregate, I am troubled, as was the Ninth Circuit Court of Appeals in a similar case, by the possible cumulative effect of those errors which go to the credibility of the defendant and the complaining witness. In the context of the present case and considering the entire record, I cannot conclude "that a balkanized, issue-by-issue harmless error review would be very enlightening in determining whether the [defendant was] prejudiced by the errors." United States v. Wallace, 848 F.2d 1464, 1476 (9th Cir. 1988).
¶ 108. I conclude that under the proper application of the cumulative error test, the multiple errors of the defense counsel and the prosecutor undermine confidence in the outcome of the trial and denied the defendant of his right to a fair trial.
¶ 109. For the reasons set forth, I conclude that a new trial is warranted, and therefore I dissent.
Majority op., ¶ 33 (quoting Strickland v. Washington, 466 U.S. 668, 687 (1984), ¶ 64.
For discussions of alternative ways of stating harmless error tests and of different tests for different errors, see, e.g., State v. Hale, 2005 WI 7, ¶ 60, n.9, ¶¶ 79-85, 277 Wis. 2d 593, 691 N.W.2d 637 (Abrahamson, C.J., concurring); id., ¶¶ 85-90 (Wilcox, J. concurring); id., ¶¶ 100-18 (Butler, J., concurring); State v. Harvey, 2002 WI 93, ¶¶ 50-54, 254 Wis. 2d 442, 647 N.W.2d 189 (Crooks, J., concurring); id., ¶¶ 55-77 (Abrahamson, C.J., dissenting).
Strickland v. Washington, 466 U.S. 668, 686 (1984) (quoting McMann v. Richardson, 397 U.S. 759, 771 n.14 (1970)).
Henderson v. Sargent, 926 F.2d 706, 711 (8th Cir. 1991).
Jenny Roberts, Too Little, Too Late: Ineffective Assistance of Counsel, The Duty to Investigate, and Pretrial Discovery in Criminal Cases, 31 Fordham Urb. L. J. 1097, 1110-11, 1114-15 (2004).
Strickland v. Washington, 466 U.S. 668, 689-91 (1984) (recognizing the "wide latitude counsel must have in making tactical decisions" and that "strategic choices made after less than complete investigation" may be reasonable if "reasonable professional judgments support the limitations on investigation").
State v. Thiel, 2003 WI 111, ¶ 45, 264 Wis. 2d 571, 665 N.W.2d 305.
See, e.g., King v. State, 810 P.2d 119, 123 (Wyo. 1991) (presuming prejudice when defense counsel failed to secure trial testimony of two potential eyewitnesses).
State v. Pitsch, 124 Wis. 2d 628, 642, 369 N.W.2d 711 (1985), quoted with approval in State v. Thiel, 2003 WI 111, ¶ 20, 264 Wis. 2d 571, 665 N.W.2d 305. See also majority op., ¶ 64.
At oral argument the State conceded that the prosecutor's remarks regarding the defendant's pr e-Miranda silence during her opening statement were improper. Majority op., ¶ 40.
The majority opinion concludes that defense counsel's failure to object to the improper comments of the prosecutor during closing argument and the prosecutor's use of pr e-Miranda silence was acceptable defense trial strategy. Majority op., ¶ 63.
At oral argument the State conceded that the prosecutor's remarks regarding the defendant's pr e-Miranda silence during direct examination of the State's witnesses were improper. Majority op., ¶ 40.