How Important Is Pitch Efficiency?
In my season review, I mentioned the surprising drop in efficiency for Brandon Beachy, actually throwing more pitches per plate appearance this past season compared to 2011. With the increase in success, I was wondering how pitch efficiency affects performance. I created correlation tables for starting pitchers who threw at least 60 IP in a season the past three seasons. Here are the correlations with respect to P/PA:
Apr. 20, 2012; Phoenix, AZ, USA; Atlanta Braves starting pitcher Brandon Beachy (37) throws during the first inning against the Arizona Diamondbacks at Chase Field. Mandatory Credit: Matt Kartozian-US PRESSWIRE
| 2012 | 2011 | 2010 | Mean | StDev | |
| ERA | -0.14 | -0.08 | -0.17 | -0.13 | 0.05 |
| FIP | -0.08 | -0.14 | -0.20 | -0.14 | 0.06 |
| xFIP | -0.04 | -0.14 | -0.12 | -0.10 | 0.05 |
| BABIP | -0.18 | -0.09 | -0.04 | -0.11 | 0.07 |
| LOB% | 0.19 | 0.19 | 0.17 | 0.18 | 0.01 |
| LD% | 0.09 | 0.03 | 0.04 | 0.05 | 0.03 |
| GB% | -0.32 | -0.39 | -0.19 | -0.30 | 0.10 |
| IFFB% | 0.24 | 0.26 | 0.17 | 0.23 | 0.05 |
| HR/FB | -0.09 | -0.03 | -0.21 | -0.11 | 0.09 |
| K% | 0.50 | 0.60 | 0.53 | 0.55 | 0.05 |
| BB% | 0.46 | 0.43 | 0.54 | 0.48 | 0.06 |
| Swing% | -0.28 | -0.17 | -0.17 | -0.20 | 0.06 |
| Cont% | -0.42 | -0.46 | -0.44 | -0.44 | 0.02 |
| FBv | 0.29 | 0.35 | 0.33 | 0.32 | 0.03 |
| IP/GS | -0.16 | -0.16 | -0.14 | -0.15 | 0.02 |
As you can see, the correlations are not particularly strong, but most of them have an effect on performance. After giving it some thought, most of these relationships should make sense, as the LD% was the only surprising one to me, but that is the weakest correlation on the board. There are some covariance issues, most notably between K% and GB%. In general, the lack of sign change (positive to negative, or vice versa) shows the consistency of the relationships.
Starting with the obvious trends, more pitchers per PA means less contact and swinging, leading to more K’s and walks. It would also seem correct to assume harder throwers use more pitches to get through hitters. BB and K have been the staple of finding inefficiency, so I can’t imagine anyone arguing with these results.
The home run trend was smaller than I thought it would be, since more pitches means more two-strike counts, which should be harder to hit for power. The popup rate is not surprising, since the two-strike swings are more defensive and easier to get jammed. The inverse relationship with GB% is probably due to strikeout pitchers being flyball pitchers more than batters hitting more flyballs late in the count. The line drive trend was the surprising one to me, but I guess the defensive swings create more liners, though they’re likely weaker than normal. Putting all of that together, the BABIP trends down a bit, which results in a higher strand rate (LOB%).
In the overall performance, more pitches per batter faced leans towards a lower ERA, FIP, and xFIP, though it’s not particularly strong. The increased K’s and decreased homers outweigh the increased walks. The caveat here is the length of outings. The range of the P/PA values is 3.25 to 4.25, so if the pitcher on average gets through the order three times, you’re talking about a maximum of 27 more pitches.
Let’s say we have Pitcher A who throws 3.5 P/PA and Pitcher B who throws 4 P/PA. They each average 104 pitches a game, which leaves pitcher A facing 28.5 hitters a game, while B faces 25 a game. Since the league average OBP is around .320, you’d expect one of A’s extra hitters to get on base, leaving him getting 2.5 outs more per start. However, the expected ERA for A is about .4 runs higher than B, so would you rather have a pitcher throw 190 innings with a 3.60 ERA or 215 innings with a 4.00 ERA? I don’t see a difference between the two, so pitch efficiency may not be as big of a deal as we make it out to be. I would be more concerned about Beachy’s drop in K’s coinciding with the rise in P/PA. That is probably more of a problem than the pitch count itself.