Revisiting PFL

[kucoach7]

After many years of analyzing data from this god-forsaken game and producing, honestly, very little in the way of useful results, I've come to believe that there are some situations where the game has thresholds. These sorts of discontinuous relationships mess with the regressions and correlations I normally produce. Recently, I reviewed my first article on PFL viewtopic.php?f=7&t=4993 and wondered if the lack of a strong correlation may have been caused by one of these threshold/discontinuous relationships.



This graph groups PFLs for all signed players this year and plots the average fouls per minute for five point buckets. To me, it looks like there are some thresholds here. If your PFL is below 30, you foul considerably more than if your PFL is in the 30-60 range. If your PFL is 90 or above, you don't foul. This makes me feel like our PFL training is probably okay as is. You really only need to get PFLs to the low 30s before it probably doesn't affect playing time. The average player with a PFL of 35 can play 35 minutes per game and average fewer than 3.5 fouls per game.

[IamQuailman]

Good stuff Todd!

[kucoach7]

Adding this picture that better illustrates what I am talking about.

[NOLa.]

Solid analysis! I always look forward to yours and garbageman’s threads like these.


Sent from my iPhone using Tapatalk

[ballsohard]

Ku, please claim 5 points