A second concern about my point guard study (http://basketball-statistics.com/howtheshootingabilitiesofpointguardsaffectoffenses.html), besides the complications of true shot percentage (http://basketball-statistics.com/tryingtounderstandthepointguardresultsparti.html), is the possibility that the relationship between three-point shooting in point guards and increased team Offensive Ratings is a matter of correlation and not causation.  Perhaps, the theory goes, a point guard shoots a high percentage from long distance because the team is good at offense, rather than his three-point shooting being a reason why the offense is successful. 

There are number of ways to get some clues to see if this is a problem.  A simple method that I have chosen is using the point guards’ career three-point percentages rather than the current season’s percentages.  The career percentage should be a more accurate indicator of whether or not the player is truly a good shooter from long range and not just a product of offense.  Of course, if a player has spent his entire career with one team the results may be less useful, but nevertheless we push forward:

As you can see, instead of using the cutoff points of 30% and 40%, I changed them to 32% and 38%.  I did this because there are fewer players that have career percentages as extreme.  22.73% of home lineups featured a point guard with a career three-point percentage greater than 38%, and 33.95% featured a player with a percentage lower than 32%.  On the road, those numbers were 23.06% and 34.07%, respectively. 

The results are practically identical to those of the original study.  I think we can safely say that using career three-point percentage instead of current percentage does not really make a difference.

Copyright © 2009 Basketball-Statistics.com