Stat Relevance - Revisited

Just over a year ago, I ran a study in search of the statistics which are most helpful in identifying a good player.  I took my final Top 30 player list for 2007 and charted where each finished in 13 categories.  The results of that study can be found here.

This year I took my season-ending Top 70 list and charted the top 30 players whose stats are available at LPGA.com.  Since Ji-Yai Shin and Stacy Lewis aren't listed, I substituted Brittany Lang and Juli Inkster.  I added my own creations (Total Driving and Total Putting) to the chart for curiosity's sake.  Then I tallied the number of players who finished in the Top 30 of each category, as well as the "Top 20" players' (the actual top 21 minus Shin) appearances in the Top 20 of each category.  Keep in mind I'm only measuring the stat's ability to identify an upper echelon player.  I'm not saying which stats are more or less meaningful - each of these stats has significant meaning in its own context.  Using those totals, here is how I rank the relevancy of each statistic for the 2008 season:

As I explained last year, the top three are not surprising since my rating system is directly based on them.  I'm comfortable with that conflict-of-interest because the question here is about the statistics themselves, not about the accuracy of my Top 30.  If anyone has a problem with that conflict, tell me which of those Top 30 players to leave out and who should replace them and if the exchanges are reasonable, I will rerun the study.

Birdies made the most impressive jump from last year (up from #7), especially since it surpassed Birdie Percentage.  I was convinced that Birdie % was the more valuable but perhaps that thinking is flawed.  GIR dove from #5 to T8 as only 15 of the Top 30 and 13 of the Top 20 finished high in my rankings.  I would imagine that shortfall to be a one-year aberration (GIR is typically very good at identifying a quality player) but will keep this example in mind when I re-test Stat Relevance in the future.

Total Driving and Total Putting didn't come in as high as I might have hoped, but I must point out that they do score as more relevant than all but one of the stats they ostensibly stand-in for (PPGIR).  I explained why Victories ranks so low last year but will re-state the gist of it - since Louise Friberg and Leta Lindley won tournaments in 2008 while Na Yeon Choi and Suzann Pettersen did not, does that mean Friberg and Lindley are better players or had better seasons than Choi and Pettersen?

A few words about the last-place finish of Driving Accuracy...It Wasn't Even Close.  Only five of the Top 30 and one of the Top 20 (Eun-Hee Ji ranked 17^th) clocked in with this category.  Sand Saves had twice the numbers (twelve and two) that Accuracy did.  Does Accuracy mean nothing?  Like GIR, part of this is probably a one-year aberration.  A lot of it is player-specific - a longer hitter doesn't really care much when she misses the fairway unless she is OB, blocked by a tree or submerged.  Usually she'll have a 7-iron or less in hand and at most courses, she'll still be able to spin that shot enough to hold the green.  The shorter hitters don't have that luxury and for them, Driving Accuracy is a very important statistic.  The difference between players' lengths is primarily club-head speed - since the shorter player generates less speed, she puts less spin on the ball and the rough penalizes her a lot more, not to mention the fact that she's using a less-lofted club out of that rough.  Frankly, the low ranking of Driving Accuracy is just another indicator that length with all clubs in general is an even more important factor towards good play than I already thought.