Alright, you guys asked for it (or at least two of you did). Here are responses to the comments on “Eureka!”
Hubert, here’s the graph without the two outliers (the Texas and Iowa games). The trend is still pretty distinct.
Randy, the R-squared of the original trendline is 0.189. I’m not sure the R-squared is all that important, though. The argument is not that pace predicts MSU’s turnover % on its own. The point is that the relationship runs inverse to what you’d normally expect–meaning MSU may be different than most teams in that playing at a higher pace actually tends to make them execute more efficiently on offense. There are no doubt other predictive factors for TO%, though; opponent’s defensive TO% is the obvious one.
Here’s Wisconsin’s graph. The Badgers seem to protect the ball better when playing at a slower pace.
And here’s Indiana’s graph. No clear trend.