Quote:
Originally Posted by butterfly
|
OK. So what we learn from this is that using publicly available counting stats, we can't make any useful predictions about goalies.
But that doesn't mean goaltending is just random noise. It means that it is dependent upon variables that are not captured in the data available for that study.
By the way, I have to seriously question their use of GSAx as far back as 2007. I certainly had never heard of that stat then, and it was not in common use until the last few years. So I have to cast a skeptical eye on the input data from the earlier years of the sample. Was the metric for expected goals exactly the same in 2007 as in 2021? Are we, in fact, comparing apples with apples? Impossible to say without a lot more information than the article gives.