Looking at How Pitches Per Plate Appearance Affect Rays Batter Stats
Recently, my good friend Jason Collette looked at how pitches per plate appearance affects batting average. Great minds think alike as I was also looking at PPA that morning with the idea of how "triple slash" statistics (Batting Average (BA), On-Base Percentage (OBP), and Slugging Percentage (SLG)) move in relation to batters taking more or less pitches. There's a Rays tie-in that I will delve into further, but for now, let's take a look at the outcomes of this research. Using Fangraphs, I pulled all total batter statistics from 2002 through 2011 for batters with 1,000 or more plate appearances so that we can look at BA, OBP, SLG, and even better wOBA which ties together all three. This yielded 550 players ranging in ability from Jeff Mathis (.253 wOBA*) to Barry Bonds (.475 wOBA*). The asterisk, in this instance, stands for "regressed." Don't get too hung up on the math, but know that all batters have had 220 PAs of league average wOBA (.329 over this period) added in so that we can reduce some of the noise inherent in even larger samples such as these. You can cherrypick through and see guys that have very high or low numbers seemingly with no relationship to PPA. To look at the aggregate a bit more easily I performed several simple regressions between PPA and the slashline statistics plus wOBA. (All coefficients are statistically significant at less than 1%)The mean shows the average for all players while the Coefficients are what we're most interested in here. As Collette pointed to, the more pitches you see in a given plate appearance the less likely you are to get a hit. This runs converse to OBP, SLG, and wOBA however. What these say is that for each full pitch a batter sees per PA you would expect a drop in BA of .015, but an increase in OBP of .052, SLG of .052, and an increase in wOBA of .048. Because batters are not likely to add a full PPA I thought it would be illustrative to take a look at how the average batter over this sample would look with fractions of an extra pitch added or subtracted:The average batter would be expected to see his line move to .260/.362/.450 with a wOBA of .353 if he was able to add half a pitch per plate appearance. Likewise, if a batter took three tenths of a pitch less then he'd see his line move to .272/.320/.409 with a wOBA of .315. Yes a batter might have a higher batting average, but he would make a worse player in the process!How does this tie in with the Rays? Well, it's no secret that the Rays have had a very disciplined approach at the plate over the tenure of Derek Shelton's time as the hitting coach under Joe Maddon. Fans get frustrated when they see batters watch grooved fastballs or take a borderline pitch that ends up called a strike, particularly when it sends the batter back to the dugout. No doubt it's a frustrating outcome, but baseball is a game of frustrating outcomes requiring tempered expectations and a hope to manage probabilities.So far we have seen how these stats change when compared to an average batter. The next step is to see how each of these scenarios would look compared to the Rays player's established lines. For this we will only be looking at wOBA as it represents the best of the slashline categories. Also, we'll be using the wOBA* statistic, because several of these players don't have all that many plate appearances. This allows us to get a better idea of their true talent level instead of the noise of a smaller sample. Note that for extremely low plate appearance totals then you're looking at basically the league average altered very little by the minuscule samples of the player, specifically Canzler, Lobaton, Guyer, and Chirinos. Here we can see each player's wOBA* if they were to progressivley take one tenth of a pitch more than they have over their careers. Bear in mind that Brignac and Molina are the only guys that have shown a lesser than average propensity to take pitches over their careers as pointed out by Collette. Since the Rays are already taking more pitches than the average Cub (or Angel or Ranger) then I think the next table is more indicative. You may want to rethink your position if you're one of those Rays fans that bemoans the team working counts or watching pitches. There is a reason that the hitting coach, manager, general manager, etc... have stuck with this philosophy. I applaud them for dealing with the ugly wart of bad batting average knowing that getting on base and extra base hits are more important.