Saturday, August 7, 2010

Don't Blame the Freshman Forward

-You have to make your free throws to win close games.
-You have to be aggressive for the officials to give you the benefit of the doubt.
-Senior players are more likely to excel in pressure situations.

There is some truth to all these statements. Teams need to make shots to win and on average players get better over time. But do these factors help teams over-achieve? Does experience and free throw success contribute to a team winning more games than their point differential would indicate?

The short answer is yes. Experience and free throw shooting are correlated with Ken Pomeroy’s measure of “Luck”. Teams that have these qualities are more likely to win more games than their Pythagorean ranking predicts.

But the next three pictures should suggest that statistics are often misleading. These pictures plot Pomeroy’s Luck rating for teams over the last seven years relative to the team’s FT%, the team’s FTA/FGA, and the team’s Experience level. (Over this time period, the unluckiest team was Florida Atlantic in 2004, while the luckiest team was Wagner in 2008. Wagner won 5 overtime games in 2008 and won a number of other close games as well.)

Figure 1 confirms what a regression analysis shows - a team’s free throw percentage is not a statistically significant predictor of Luck. There is a minimal positive correlation, but as the wide scatter plot shows, the relationship is virtually meaningless. There are bad free throw shooting teams that are lucky and good free throw shooting teams that are unlucky.

The trend line for free throw attempts is a little steeper than the trend line for free throw percentage, but as the scatter plot reveals, the relationship is not very strong. Playing aggressive and taking the ball to the basket often fails. But aggressive teams do tend to have a little more luck on average.

Finally, we come to the role of experience. I am certainly not going to convince you based on the scatter plot that there is a relationship between the variables. But I can tell you that the positive correlation is statistically significant at the 1% level. And if you want to visualize the reason, I think it is hidden in the upper left-hand corner of the plot. There is not a single inexperienced team with a luck rating over 0.1.

There is no magic formula for winning close games. Sometimes veteran players make their free throws; sometimes they do not.
But this should be a bit empowering. I recall when Illinois could not make their free throws in 2008 and skidded towards the "unluckiest team of the year" award. There was a sense that there was nothing they could do to win close games. The numbers suggest otherwise. Don't blame the freshman forward if he misses a free throw that costs you the close game. There were plenty of other plays that made a difference too.