this is the bad side of stats... you can have all sorts of sophisticated, elegant algorythms and advanced metrics, but it doesn't mean it tells you anything meaningful.
this is something related to "clutch". in reality, all this is doing is created an extremely small sample size with insanely high volatility, especially in any given year, and likely not enough in the course of their career in an amount of time to make it useful before they are an old has-been.
i had no idea what that was, either :P
just like it's a horrible idea to go spend wildly on some guy that performed well in the playoffs, even if multiple years. the truth will out. he simply had a good 2-weeks, which can happen at anytime over 162 games, and means nothing if it happens at the end of the year as far as how good tha tplayer is or is not.
Last edited by NoOne; 10-23-2019 at 09:53 PM.
|