Humble Pie

I was surprised by both the result and performance Sunday in Perth, and understand why there has been a pretty robust response to it amongst fan media, on social media, and even from Brendan Rodgers. In particular, the first 75-80 minutes were awful, and believe that characterization is fairly well supported by a lot of the underlying game data, which I am sure Alan will share within the forthcoming Stat Pack.

When events unfold counter to what my analysis was heading into a game, or any timeframe of games, the first thing I like to do is revisit my priors. Was the analysis heading into the game missing something important, or did I weight some of the factors differently than how events proceeded to unfold?

As for priors, this column was dedicated to highlighting the potential, if not likelihood, of performance levels ratcheting down as the jeopardy in games ended. Here is an updated version of the graphic from that column:

This season’s jeopardy vs non-jeopardy samples are now at 26 vs 6 games, but we can see that the rate of dropoff per this metric has not been unusual. I share in the sense that the last six games certainly feel worse, and believe there is some statistical evidence to help explain that as well.