Quote:
Originally Posted by sprague
That is enough games to stop a "small sample size" issue for wins or stats
|
This isn't true just because you state it to be so
The current 162 game season isn' a large enough sample for some stats - largely defensive ones.
Then there's the fact that teams and players change over the course of the season
There's also a lot of non-understanding of what a sample is. A sample is a measurement of a portion of some whole.
So a full season can be a sample or it can be a population (the whole).
If a team was 100-62 during a season, that's not a sample, it's the measurement of their wins and losses during the season. Their record in April, that's a sample of the season. Or their record in day games. Or their record in every fifth game. Mind you, these may not be good, or, random samples.
But 162 can be a sample. It's typically used that way when trying to determine "true talent level". True talent can't be measured directly. It can only be measured in samples. Maybe 162 games. Or 120. Or whatever.
We can say with 100% accuracy that the Cardinals won more games than the Brewers.
We can't say with 100% certainty that the Cardinals were truly better than the Brewers. It's not a large enough sample (and our measurements aren't precise enough).
We can say, with very high certainty, that Tim Anderson hit .335 in 2019.
We cannot say with certainty that Tim Anderson was a true talent .335 hitter - he, almost certainly, was not.