If we assume that TreeTime is perfectly calibrated, then the
maximum value that Chronumental could achieve (oracularly predicting exactly the correct date in every
case) would be 90%, since TreeTime outputs 90% confidence intervals.
on reflection doesn't make sense. Chronumental could exactly re-implement the TreeTime algo and get 100%! Probably the reviewers will flag this but if not fix it anyway.
I was re-reading the paper and:
on reflection doesn't make sense. Chronumental could exactly re-implement the TreeTime algo and get 100%! Probably the reviewers will flag this but if not fix it anyway.