Forums

[Chris R]

Those Ken Pom stats really have no relationship to the Penn game. He is averaging stats through 1/2 seasons and full seasons. We are discussing a 40 minute game. Pomeroy also cannot and does not plot the 3pt FG% of open looks vs contested looks. This is the entire walnut right there. Not every 3pt attempt is weighted equally in degree of difficulty.

He also admits "there are games where good shooters get a bunch of open looks and they make more than a third of their attempts."

Team 3pt FG% is about 1/2 open looks and 1/2 contested looks. But it means nothing when a team is given open looks 85% of the time in a single game. If Penn got the same looks from the arc in all their previous games as they did today, their 3pt FG% would be much higher than it is now and we'd be talking about how they came close to their average.

Averages are based on constant variables over long periods of time. But each game has its own subset of variables that may be decidedly unique -- and unique not by chance but for completely explainable reasons. And any single game is a very short window where statistical averages have less meaning.

Being wide open for 3pters in the same offense against the same defense in the same game is not a lottery. Its execution. In fact its not an average at all; its unusual and fits in its own statistic bucket.

Pomeroy also talks about good 3pt shooting teams who are defended can take a step back and take an uncontested 3pter, but again it doesnt account for bad defense in any single game where a good shooter is given an uncontested look directly on the arc.

Statistics are handy, but like anything they can be detrimental to reality if you get overly consumed by them. RPI can do this if you dont understand it and aren't careful.