I wanted to take a methodical look at each specific play in the ALDS to determine the magnitude of the impact each one had on the Detroit Tigers’ eventual win. I thought about using WPA – Win Probability Added – but, while that would tell us how important each event was to the outcome an idividual game, I wanted a better look at how each event swung the advantage of the series (in real time).
I came up with Series Probability Added (I’m not sure if anyone has done this or not, it’s probably not unique to me), in which I combined WPA for in-game probability (using FanGraphs’ play logs) with the assumption of 50-50 odds for all future games.
It turns out that pretty much all the plays that felt huge at the time were actually really, really huge. Page through to see a description of each play, the series probability swing, and a video highlight.