After Fair Trial Analysis reported that Thomas Zeigler would probably be acquitted in a retrial with newly discovered DNA evidence, the State asked the Florida Supreme Court to strike FTA’s amicus brief.
The case and FTA’s findings
Zeigler was convicted of four murders committed at his Winter Garden furniture store on Christmas Eve 1975. The victims were his wife, her parents, and Charles Mays. Zeigler, who was also shot, maintained that he encountered a robbery and fought for his life. Prosecutors argued that he planned the killings and staged the crime scene.
Decades later, advanced DNA testing produced evidence that the defense argued undermined important parts of the prosecution’s account. Zeigler sought a new trial under Florida’s Jones standard, which requires newly discovered evidence to be sufficiently consequential that it would probably produce an acquittal on retrial.
FTA studied that question through a randomized survey experiment involving 241 jury-qualified Florida adults, with statistical weights adjusting the sample toward Orange County’s demographic profile. Participants considered different versions of the evidence, including the original trial evidence and a retrial incorporating the DNA findings. Barry Edwards and FTA presented the findings in an amicus brief filed July 9, 2026, offering the research to assist the Florida Supreme Court’s assessment of whether acquittal would be probable.
Using a model of jury selection and deliberation, FTA estimated an 86.3% probability of jury acquittal, with a reported margin of error of 8.2 percentage points. Including additional investigation evidence increased the estimated probability to 87.6%.
The State’s objections
The State’s motion challenged both the propriety of the submission and the research supporting it. The State argued that the brief introduced case-specific information that had not been presented to the trial court. It also questioned the adequacy of FTA’s evidence summaries and methodology, including whether the analysis accounted for jury deliberation.
The State relied on Hildwin v. State, a decision affirming the exclusion of a particular mock-trial study offered to demonstrate that newly discovered evidence would probably produce an acquittal.
FTA’s response
FTA’s response argues that the brief addresses an issue already central to the appeal: whether Zeigler would probably be acquitted in a retrial with the newly discovered evidence.
The survey data were collected after the trial court proceedings. They provide an empirical resource for analyzing the existing legal question, rather than introducing a new claim or purporting to establish who committed the murders. The Court retains responsibility for applying the governing standard.
FTA also distinguishes its research from the mock-trial study discussed in Hildwin. FTA measured individual verdict preferences through standardized survey conditions and then translated those preferences into jury verdict probabilities. That distinction matters. Observing the outcomes of one or two mock juries supplies very little information about the probability of acquittal across possible juries. FTA’s approach uses hundreds of individual responses to estimate population preferences, quantify statistical uncertainty, and calculate corresponding verdict probabilities.
The analysis accounted for selection and deliberation through a mathematical model.
The response further argues that the state’s criticisms of trial summaries used on the surveys should be specific and open to examination. If a summary omits important evidence, identify the omission. If a model rests on an inappropriate assumption, explain the assumption and its consequences. A substantiated criticism can inform the weight assigned to the research. FTA argues that disagreement with its conclusions does not justify excluding the entire brief from consideration.
Comparing methods by the same standards
The dispute raises a question extending beyond this case: how should courts evaluate the reliability, transparency, and accuracy of predictions about what a jury would do? Those considerations should guide the assessment of empirical research—and the comparison between that research and the subjective predictions courts ordinarily make.
When a court decides that newly discovered evidence would—or would not—probably produce an acquittal, it makes a prediction about a hypothetical jury. A judge may explain that prediction thoughtfully and at length. But without an explicit estimation method, the prediction remains a subjective judgment. Other judges may reach different conclusions from the same record, with no common procedure for measuring the difference or assessing whose prediction is more accurate.
FTA’s method exposes its reasoning to scrutiny:
- Reliability: Standardized materials and documented procedures allow researchers to repeat the analysis and examine whether results remain stable.
- Transparency: The survey instrument, data, code, and model assumptions allow others to inspect how the estimates were produced.
- Accuracy: Specific predictions can be evaluated against relevant evidence, tested under alternative assumptions, and revised when weaknesses are identified.
Transparency does not guarantee accuracy. A reproducible calculation can still rest on flawed inputs or assumptions. But disclosure makes those problems identifiable and correctable. A general intuition offers much less opportunity for comparable testing. In Zeigler’s case, for example, the State asserts that Zeigler would probably be convicted in a retrial with new evidence, but its conclusion is not based on data or any specific method of analysis. Subjective assessments of acquittal probabilities escape criticism not because they are reliable, but because they offer nothing to evaluate or criticize.
Courts should evaluate the competing approaches by consistent standards. When a person’s entitlement to a new trial turns on what a jury would probably do, the quality of that prediction deserves scrutiny.