I’ve been following the data long enough that it’s time to make some estimates. These estimates are based on the observation that there is a 12-day lag between positive test results and deaths [Note: later this was updated to 13 days]. Based on the combination of positive tests 12 days earlier and an approximation of naive CFR, I’ve produced daily estimates for 5/1 – 5/12, as shown below. The daily estimates could be off quite a bit, but I think the cumulative estimate will be pretty close.