Fig. 2.

(A.) A linear pool predictive density made in June, 2020 over dates for when a COVID-19 therapy will show a significant survival benefit in a randomized clinical trial enrolling more than 200 patients. (B.) Linear pool predictive densities made in Aug., 2020 over dates when a SARS-CoV-2 vaccine will show a significant survival benefit in a randomized control trial enrolling more than 200 patients across three different viral platforms. Below each density is the $10\mathrm{th}$, $25\mathrm{th}$, $50\mathrm{th}$ (median), $75\mathrm{th}$, and $90\mathrm{th}$ percentile. True values, if available, are represented as a filled circle. The linear pool was uncertain when a vaccine would show a survival benefit, assigning 80% confidence intervals that spanned close to two years for when a COVID-19 therapy would show a benefit.