Batting average by count and pitch type: fact and fallacy.

Many baseball coaches, sportswriters, and television announcers have commented on the fact that batting averages are low with two strikes or high with less than two strikes. For example, Thomas Boswell, in an article he wrote criticizing Ted Williams' theory that you should take the first pitch,

noted that "an analysis of nearly 100 of the top hitters in baseball suggests that the best [hitters] bat about 70 points higher and slug 130 points higher" on 0-0. (1) From this Boswell claims that batters "should be swinging at a lot more first pitches, that Ted Williams was a "dumbo," and that "most hitters are idiots." Boswell defends his attack on Ted Williams by writing "Sorry, big guy ... you made the call, so you've got to take the fall."

Thomas Boswell is not alone. During both the 2000 and 2001 World Series, Tim McCarver, broadcasting for Fox Sports, used a graphic showing Major League hitters bat over 100 points lower with two strikes. Mr. McCarver was surprised by the poor performance of major league hitters in two-strike situations.

Andrew Torrez, in his book Off Base: New Insights into an Old Game, (2) notes that the "average major leaguer hit .187 with two strikes, which is more reminiscent of the average pitcher." Torrez goes on to argue that taking 1-1 makes sense for a "free-swinger" because if he would have swung at a pitch out of the zone, the count would have gone 1-2, "where he's almost certain to be out." Torrez suggests that this subtle insight separates good managers from mediocre ones.

Finally, USA Today noted, "All pitchers, virtually without exception, hold hitters to a lower batting average when the ball-strike count favors the pitcher. Batting averages tend to be about 140 points lower when the ball is put in play on a pitcher's count, as compared to a hitter's county." (3)

We recently tested whether these same trends hold in college baseball by calculating the batting average (AVG) and slugging percentage (SLG) by count for Stanford Baseball (both Stanford and opponents batting) from 1998 through 2001. (4) These results are illustrated in Figure 1.

[FIGURE 1 OMITTED]

As you can see, both SLG and AVG are much lower with two strikes. For example, with less than two strikes, AVG and SLG are around .353 and .563, respectively. With two strikes, AVG and SLG drop to .183 and .276, respectively. In other words, batters hit 170 points lower and slugged 287 points lower with two strikes than they did with less than two strikes. The overall AVG and SLG for our data set was .273 and .429, respectively. Therefore, batters hit 83 points higher and slugged 134 points higher on 0-0. That is amazing! Or, is it?