Dan Lambert Doesn't Think Bayesian Analysis Helps When It Comes to History

Timothy and Lydia McGrew use a Bayesian analysis (or inference) to establish the resurrection of Jesus by a percentage of over 99%. Richard Swinburne uses it to establish that if God exists then Jesus was bodily resurrected from the grave by a percentage of 97%. However, by definition, even any given mundane (or ordinary, as opposed to an extraordinary) historical event is a unique one-of-a-kind event. And the events leading up to it, and follow it, are also unique one-of-a-kind events. So when looking to see if any given event took place in the past the historian must calculate the probability of the whole series of events, the ones that came before and after the event in question, even though all of them are unique one-of-a-kind events. How does one assign probabilities to them all when they are all dependent on each other? In one sense they are all improbable for that reason. To see this from a different perspective, consider an observer before any event takes place. All events in the future have an extremely low probability to them. No one could have predicted yesterday that I would be typing these very words today, you see. In any case, here is Lambert on Facebook taking about this subject:

Atheist philosopher Louis Antony agrees with Lambert. In any case, theists cannot even agree among themselves how to establish what happened in the past. Until they can agree with this do we even have to consider the case they make for the resurrection of Jesus?