Originally Posted by Dan
OK, I can see that, definitely. Actually wouldn't casey be one of those low IsoD guys? Also, how do you explain a guy like Dunn, or Rob Deer, or Dave Kingman who have low BA with high HRs?
Well, Deer and Kingman had problem chasing balls out of the zone as well. But they had really bad contact rates when they swung at good pitches. During his best years, Sosa got his contact rate up as well as increasing his discipline.
With Dunn, there's a few things going on. Certainly his contact rate when he chooses to swing isn't superb. However, the bigger problem is that Dunn is among the worst in baseball at taking strikes. You could say he's too selective for his own good. It's not that he's losing hits by watching the ball on the corner. It's that he's getting himself in to two strikes counts so often that his poor contact rate gets translated in to strikeouts. It hurts a lot more to whiff on a 2-2 pitch than on a 1-1 pitch.
I would argue that this is why Dunn's batting average is so low as well. If you could put the ball in play in every at bat and get a standard distribution of batted ball types, you'd hit roughly .300 every year, depending on your luck with fielders. But if when you swing, you have a tendency to miss instead of put the ball in play, the BABCM (Batting average on Balls in the Catcher's Mitt) is .000, thus dropping your average. And if you tend to wait until you have two strikes to swing, you're not going to get as many swings as other guys.
Now, because Dunn is so utterly productive when he manages to get the ball play, and because he's often walking when he doesn't put the ball in play, he's still very valuable. But his low average is directly attributed to his perhaps too disciplined approach and his poor contact rate when he does swing.
Consider that when Sammy hit .300, either his power was off the charts or his strikeout rate was a good deal lower than we see with Dunn. However, we can look at Dunn's monthly splits to get a good glimpse of this.
BA K/PA ISO
Apr .261 .314 .250
May .252 .339 .321
June .287 .238 .330
July .239 .257 .216
Aug .276 .200 .345
Sept .274 .200 .258
K/PA has a somewhat strong negative correlation with BA (r = -.56). That is, within this tiny data set, as Dunn struck out more, he hit for a lower average.
ISO has a somewhat strong positive correlation with BA (r = .63). That is, within this tiny data set, as Dunn hit for more power, he also hit for more average.
His worst BA month, July, was one in which he hit for little power and had a medium K/rate (for Dunn). For fun, I created a single variable by taking each of the average K/PA minus the observed (so that positive values are good) and the observed ISO minus the average. This gets you a single number combing the two. If you regress that against batting average, you find that those two things combined have an r-squared of .6229 (r = .79). I did the same thing simply ranking the months on each variable and taking the average rank versus the batting average rank and you get basically the same thing.
So simply put, most of the month to month (or year to year) variation in batting average can be attributed to strikeout rate and power fluctuations when you do make contact. This is why Albert Pujols hits .330 regularly and Bonds hit .340 plus during his roided phase.