Patience and Position

“If at first you don’t succeed, try, try again. Then quit. There’s no point in being a damn fool about it.” – W.C. Fields (1880-1946)

I received plenty of excellent feedback on my article last week on Caribbean players, and I’d like to thank everyone who took the time to a) read what I wrote and b) respond so intelligently. Several of the interesting responses I received I’ve written about and investigated on my blog. These included one particularly thought-provoking theory arguing that perhaps Caribbean players develop more plate discipline over time as they’re exposed to more advanced coaching, and another that a platoon advantage for Caribbean players resulting from their more frequent ability to switch-hit might contribute to their decreased propensity to walk. As it turns out, neither seems to be supported by the data. For example, the following graph shows the walk rates of Caribbeans versus non-Caribbeans by age, and as you can see, the gap does not shrink over time.

Although neither idea panned out, both are far from implausible and hadn’t even occurred to me.

A third possible explanation for the differing walk rates proposed by several smart cookies including Nate Silver and John Walsh is “position bias.” As I showed in the article last week (and what most people know from observation) Caribbean players man the middle infield positions of second base and shortstop more frequently than do non-Caribbeans. In fact, in the last decade even as the Caribbean, Central and South America were becoming more thoroughly scouted, Caribbeans logged 37% of their games at those two positions while non-Caribbeans did so just 19% of the time.

This disparity, so the argument goes, might explain the overall difference in walk rate since middle infielders as a group possess lower walk rates. In other words, the difference in walk rates between the two groups might be explained not by region of birth but simply by the greater percentage of players from the Caribbean who are middle infielders. I have to admit that this idea did occur to me just before the previous article was published but I did not investigate it until the emails starting coming in.

One way to look at this theory would be to weight the totals for Caribbeans and non-Caribbeans by position. However, since the Lahman database does not segment offensive statistics by position, I took a simpler approach and calculated the rates for all Caribbean and non-Caribbean players who played over 100 games at any position in a season. The results in table and then graph form follow.

Non-Caribbean   BB/PA   SO/PA     AVG    SLUG     OBP     OPS
1B              0.107   0.139   0.280   0.463   0.358   0.821
2B              0.086   0.110   0.268   0.378   0.333   0.710
SS              0.079   0.119   0.262   0.369   0.321   0.691
3B              0.095   0.135   0.271   0.431   0.342   0.773
C               0.088   0.132   0.261   0.402   0.328   0.730
OF              0.094   0.142   0.279   0.444   0.348   0.792
DH              0.107   0.145   0.277   0.458   0.356   0.814


Caribbean       BB/PA   SO/PA     AVG    SLUG     OBP     OPS
1B              0.094   0.152   0.292   0.485   0.360   0.845
2B              0.073   0.111   0.279   0.388   0.333   0.721
SS              0.064   0.118   0.260   0.358   0.309   0.667
3B              0.071   0.141   0.271   0.443   0.324   0.767
C               0.068   0.136   0.272   0.418   0.323   0.741
OF              0.083   0.146   0.285   0.457   0.345   0.803
DH              0.110   0.151   0.283   0.477   0.363   0.840

What you can see from this is that indeed non-Caribbean middle infielders posses lower walk rates than do outfielders, first baseman, third baseman, designated hitters and even slightly lower than catchers. These lower rates also apply to Caribbean players at every position. However, you’ll also note that at every position except DH Caribbean players have a lower walk rate than non-Caribbeans. The differences (excluding DH) range from -.010 (OF) to -.023 (3B) for Caribbean players, while the overall difference shown in the previous article was -.016.

If there were a positional bias in play we would expect to see Caribbeans and non-Caribbeans with roughly the same rate at each position. That we don’t indicates that the fact that Caribbeans more often play middle infield positions does not fully explain the overall gap. Some quick estimates taking the positional bias into account indicate that the overall gap in the period 1996-2004 would shrink by about half if we weighted the walk rate by position. Another good-looking theory (partially) spoiled by some unsightly facts.

Interestingly, Caribbean catchers and third baseman had lower walk rates than second baseman and you can see that the ranking of OPS generally is the inverse of the defensive spectrum with the exception being that first base outpaces DH. You’ll also note that batting average, slugging percentage etc. are generally higher than average in the table above since the table includes only players who played in more than 100 games in a season, and therefore automatically excludes most players who performed below average.

What are we to conclude then? Regardless of the reasons that might include culture, training, temperament, or selection it would appear that Caribbean hitters do indeed take to heart the old adage that “you don’t walk off the island.”