NIST-affiliated statisticians and an Innocence Project attorney debated whether likelihood ratios should be used in court, stressing model uncertainty, the need for robust data and disclosure, and warning that courts may exclude or undermine such evidence without transparency.