Discussion about this post

User's avatar
Pete Wung's avatar

Interesting perspective. I don't come at statistics from a performance perspective but from a statistical perspective. I uses statistics in volleyball as a snapshot of a match at that point in time with those teams. Or as descriptive statistics. Anything that happens after that snapshot veers into the inferential statistics domain. From the greater scheme of viewing statistics, sports statistics has too many variable, the stats don't usually fall neatly into the probability distribution shape, and there are too many conditionals involved that the stats need to be collected as Bayesian distributions, and there just aren't enough samples to gain any knowledge about the statistical significance. Everything you said makes sense, but that all comes from trying to draw inference from a single set of data points, not really advised. But you can't keep coaches from jumping into that habit, because that is what we do. I have to try very hard to look at the descriptive statistics and stay away from drawing too much inference.

Expand full comment

No posts

Ready for more?