State v. Dabate: Fitbit Data Admissible Under Porter Without Proprietary Algorithm Proof; Oral Expert Opinions Must Be Disclosed Under Practice Book § 40-11

I. Introduction

In State v. Dabate, 351 Conn. 428 (officially released March 11, 2025), the Connecticut Supreme Court affirmed the murder conviction of Richard G. Dabate for the shooting death of his wife, Connie Dabate, rejecting a broad array of trial-error claims. The case arose from the state’s theory that the defendant killed his wife in their Ellington home and staged the scene to mimic a home invasion, including self-inflicted injuries and a fabricated narrative to police.

The appeal presented four major issues: (1) whether multiple alleged instances of prosecutorial impropriety required a new trial or, alternatively, warranted reversal under the court’s supervisory authority; (2) whether Fitbit movement data was admissible as scientific evidence under State v. Porter; (3) whether the defendant’s hospital interview required Miranda v. Arizona warnings because it was custodial; and (4) whether an undisclosed expert opinion constituted a disclosure violation under Brady v. Maryland or the Practice Book.

II. Summary of the Opinion

  • Prosecutorial impropriety: The court found four improprieties—(i) violating a court order by repeating the inflammatory “mini Cheshire” question; (ii) asking about an alleged plan to kill the victim on a Vermont trip without an adequate good-faith foundation; (iii) telling jurors the defendant was counting on them to be “gullible,” “lazy,” and “unintelligent”; and (iv) failing to disclose a treating physician’s anticipated expert opinion that injuries appeared self-inflicted in violation of Practice Book § 40-11 (a). Nonetheless, applying State v. Williams, the court held there was no reasonable likelihood the verdict would have differed absent the improprieties.
  • Supervisory authority: The court declined to reverse as a sanction; the improprieties did not rise to the “rare case” affecting the perceived fairness of the system.
  • Fitbit evidence: The trial court did not abuse its discretion admitting Fitbit data after a Porter hearing; an expert may establish reliability through validation studies and general acceptance without explaining the device’s proprietary internal algorithm.
  • Miranda: The hospital interview (before the suppressed “third room” portion) was noncustodial; a reasonable person would have felt at liberty to terminate the interview, and the setting did not create the “police-dominated” coercive pressures requiring Miranda warnings.

III. Analysis

A. Prosecutorial Impropriety

The court reiterated the familiar two-step framework from State v. Coney: determine first whether impropriety occurred, and second whether it deprived the defendant of a fair trial. Even absent individual prejudice, the court assesses whether the cumulative effect rendered the trial fundamentally unfair, consistent with State v. Weatherspoon and State v. Fauci.

1. Precedents Cited (and How They Shaped the Court’s Treatment)

  • State v. Ortiz and State v. O'Brien-Veader: Used to frame when violating a trial court’s evidentiary order becomes prosecutorial impropriety and why the clarity of the order matters. The court distinguished close-to-the-line advocacy from unambiguous, repeated defiance.
  • State v. Martinez: Cited for the proposition that an order must be unambiguous—sometimes clarified by colloquy—before a breach becomes impropriety. Here, the “Cheshire” prohibition was unmistakable once the court instructed to remove “that one word.”
  • State v. Singh and State v. Albino: Singh forbids asking a witness to label other testimony as “a lie, mistaken or wrong,” or arguing acquittal requires finding other witnesses lied. Albino illustrates a “stacking” pattern that can violate Singh in totality. Dabate distinguished both: the prosecutor’s questions were largely interrupted by objections, were framed as impeachment by contradiction rather than forced credibility calls, and were not used to argue “acquit only if they lied.”
  • State v. Barnes, State v. Diaz, State v. Annulli, and State v. Payne: These cases provided the governing rule that cross-examination may probe misconduct-related matters if there is a good-faith factual basis and the use is impeachment, not propensity. The “Vermont trip” question failed for want of an adequate foundation, and Payne was invoked to show impropriety where the prosecutor suggests wrongdoing without evidentiary support.
  • State v. Ciullo and State v. Rowe: Ciullo supplied the prosecutor’s heightened duty to stay tethered to evidence; Rowe reaffirmed generous latitude in closing argument. Dabate balanced these principles to find some rhetoric acceptable (“truth…under attack,” though “unnecessarily close to impropriety”) but condemned juror-insult rhetoric.
  • State v. Pouncey and State v. Singh (summation context): Cited for the danger of inflaming passions and diverting jurors from deciding based on evidence. The “gullible/lazy/unintelligent” remarks were deemed improperly inflammatory—especially because they came in rebuttal, preventing response.
  • Doyle v. Ohio, State v. Patrick M., State v. Angel T., and State v. Coccomo: Used to distinguish permissible impeachment by prearrest, pre-Miranda silence (Patrick M.; Doyle’s limits) from improper insinuations based on retaining counsel (Angel T.). Dabate treated the “newspaper article” question as pre-Miranda silence impeachment rather than penalizing counsel-retention.
  • Brady v. Maryland, Strickler v. Greene, and State v. Ortiz: Framed the three Brady elements and the “favorable to the accused” requirement. The undisclosed physician opinion was inculpatory, so not Brady material.
  • Practice Book § 40-11 (a), State v. Festo, State v. Jackson, Taylor v. Illinois, and State v. Hargett: Provided the rule-based disclosure regime aimed at preventing surprise and preserving meaningful defense preparation. Dabate treated the nondisclosure as a prosecutorial impropriety under the rules of practice even though it was not a Brady violation.
  • State v. Williams and State v. Sinclair: Supplied the six-factor test for whether impropriety likely affected the verdict (invitation, severity, frequency, centrality, curative measures, strength of case). Williams served as the decisive framework for rejecting reversal.
  • State v. King, State v. Elson, State v. Glenn, State v. Payne, and State v. Pouncey (supervisory authority): Used to emphasize supervisory reversal is extraordinary, aimed at systemic integrity, and requires balancing prejudice, practical burdens, and other sanctions.

2. Legal Reasoning

The court’s impropriety analysis is notable for drawing a firm boundary between (a) ordinary trial-level objectionable questions (often cured by sustained objections) and (b) true improprieties that undermine trial fairness or the court’s authority.

  • Unambiguous order violation (“Cheshire”): The court rejected the state’s “heat of cross-examination” explanation. Given the “singular meaning” of “Cheshire” in Connecticut criminal parlance, repeating the word after the court ordered its removal was treated as intentional and inflammatory.
  • Foundation for misconduct insinuation (Vermont trip): The prosecutor’s question suggested a plan to kill on a specific trip without establishing a good-faith basis tied to that time frame; earlier internet searches months before did not supply a “temporal connection.”
  • Closing rhetoric targeting jurors: The court condemned the move from urging common sense to effectively telling jurors that accepting the defense would prove they were stupid or oath-breaking. The rebuttal posture magnified the risk because the defense could not respond.
  • Discovery: Brady vs. Practice Book: The court separated constitutional disclosure (Brady) from rule-based disclosure (Practice Book § 40-11 (a)). Although inculpatory expert testimony is not Brady material, it can still be “material to the preparation of the defense” under § 40-11 (a), and the state may not evade obligations by leaving anticipated expert opinions unwritten.
  • No due process reversal under Williams: Applying the Williams factors, the court emphasized: the improprieties were not invited; they were infrequent in a five-week trial with extensive evidence; targeted curative measures were promptly applied; struck evidence did not benefit the state; and the state’s case (electronic/physical evidence, motive, and contradictions) was “very strong.”

3. Impact

  • For prosecutors: Dabate is a pointed warning that inflammatory analogies to infamous crimes (here, “Cheshire”) are not merely “rhetorical excess” but can be treated as defiance of court authority; and insinuations of uncharged plans require a demonstrable good-faith foundation.
  • For discovery practice: The decision strengthens the practical reach of Practice Book § 40-11 (a) by treating an undisclosed oral expert opinion intended for use in the state’s case-in-chief as a disclosure impropriety, and by expressly cautioning that non-memorialization risks due process concerns.
  • For appellate review: The opinion reinforces that Connecticut will not use supervisory authority as an “end run” around Williams unless the conduct is of the utmost seriousness for systemwide legitimacy, not merely troubling in the individual case.

B. Fitbit Evidence and the Porter Gatekeeping Function

1. Precedents Cited

  • State v. Porter (and Daubert principles as described via Lynch v. State): The court applied the flexible reliability/relevance inquiry—testing, peer review, error rate, general acceptance, expert credentials, objectivity, and extrajudicial development.
  • Lorraine v. Markel American Ins. Co.: Cited by the trial court for authentication principles concerning electronically stored information.
  • United States v. Morgan, United States v. Chiaradio, and State v. Foreman: Cited to reject the claim that the expert must explain a proprietary algorithm or source code to establish reliability. Reliability can be established through validation, testing, and expert familiarity with output and methodology.
  • State v. Burch (Wisconsin): Noted (in a “but cf.” posture) as an example of admitting Fitbit data even without expert testimony; Connecticut did not decide that broader issue because admission was proper after a Porter hearing.

2. Legal Reasoning

Dabate’s key evidentiary contribution is the court’s acceptance that the reliability of wearable-device step data may be established through validation studies, peer-reviewed research, and general acceptance—without requiring the proponent to explain proprietary internal algorithms. The court treated the algorithm argument as an over-demanding “black box” objection inconsistent with how courts assess many modern technologies.

3. Impact

  • Wearables as mainstream scientific evidence: The decision will likely accelerate the routine use of wearable metrics (steps, activity timelines) in criminal timelines, so long as the state can present a qualified expert and validation evidence.
  • Defense strategy shift: Challenges may increasingly turn from “you can’t explain the algorithm” to device-specific issues (placement, user behavior, syncing gaps, firmware/version differences), or to cross-examination on limits and error rates—i.e., weight rather than admissibility.

C. Miranda Custody in a Hospital Interview

1. Precedents Cited

  • Miranda v. Arizona and State v. Mangual: Provided the objective two-step custody inquiry—freedom to terminate/leave, plus whether the environment presents coercive pressures akin to station-house interrogation.
  • State v. Jackson: Supplied hospital-specific custody considerations (medical restraints vs. police restraint, whether police exploit treatment constraints, isolation from family, indicia of arrest).
  • State v. Garrison (Conn. 2024): Recent guidance for hospital interviews; Dabate used it to support a noncustodial conclusion even where multiple officers are present, emphasizing tone, lack of arrest indicia, ability to interact with staff/family, and voluntariness.
  • State v. Brandon and State v. Greenfield: Used to underscore that explicit “free to leave” language is not always required, and acquiescence/consensual tone can undermine custody claims.

2. Legal Reasoning

Dabate emphasizes that long duration alone does not create custody when the defendant is not restrained by police, the door is not locked, medical staff come and go, the atmosphere is conversational, and the suspect behaves as though he is voluntarily giving an account as a victim/witness. The “police-dominated” quality central to Miranda was not present for the admitted portion of the hospital interview.

3. Impact

  • Hospital interviews remain fact-intensive: Dabate reinforces that medical immobility or IV attachment is not itself custody; the question is whether police convert treatment constraints into coercive detention.
  • Careful line-drawing over time: The court acknowledged that tone can shift during extended interactions; notably, the trial court suppressed the later “third room” statements after invocation of counsel, illustrating how admissibility can change mid-encounter even within one continuous episode.

IV. Complex Concepts Simplified

  • Prosecutorial impropriety vs. reversal: A prosecutor can act improperly, but a conviction is reversed only if the misconduct likely affected the verdict (the Williams factors).
  • Supervisory authority: Even if a trial was fair enough to satisfy due process, an appellate court can reverse to protect system integrity—but only in rare, exceptionally serious circumstances.
  • State v. Singh rule: Prosecutors generally may not ask a defendant to say another witness “lied” or to argue that acquittal requires finding other witnesses lied; credibility is for the jury and the burden never shifts.
  • Brady v. Maryland vs. Practice Book § 40-11: Brady requires disclosure of evidence favorable to the defense (exculpatory or impeachment). Practice Book discovery can be broader, requiring disclosure of expert reports/statements “material to the preparation of the defense” or intended for use in the state’s case-in-chief—even if inculpatory.
  • Porter hearing: Connecticut’s test for scientific evidence reliability and relevance; the court assesses the method’s validity, not whether the judge agrees with the conclusion.
  • Miranda custody: Not every police interview requires warnings—only those in a custodial setting with coercive pressures similar to a station-house interrogation.

V. Conclusion

State v. Dabate is a consequential Connecticut decision in three domains. First, it reaffirms that multiple prosecutorial improprieties—though “troubling” and condemned “in strong and unqualified terms”— do not mandate reversal absent a reasonable likelihood of a different verdict under State v. Williams, and that supervisory reversal remains extraordinary. Second, it signals the growing judicial comfort with modern wearable-device evidence: Fitbit step data can satisfy Porter through validation research and general acceptance without access to proprietary algorithms. Third, it continues the court’s recent hospital-interview jurisprudence by applying Mangual, Jackson, and Garrison to hold that a lengthy hospital interview may still be noncustodial when police do not create station-house-like coercion.

The opinion’s most practically important compliance lesson may be its treatment of discovery: even non-Brady, inculpatory expert opinions intended for use in the case-in-chief can trigger disclosure duties under Practice Book § 40-11 (a), and strategic non-memorialization is perilous in light of the court’s due process cautions.