Anatomy of a player: Albert Pujolsby Josh Kalk
October 10, 2007
Albert Pujols is one of the best hitters in all of baseball and has been for several years. This should come as no surprise to anyone reading this. But what makes Pujols such a dangerous hitter? Does he feast on fastballs or does he prefer hanging breaking balls? We know he walks a lot, but how much of that is due to him being a patient hitter and how much of that is pitchers pitching around him? We know he doesn't strike out a lot, but is he swinging at strikes, or just making contact with the ball no matter where it is thrown (I like to call this Estrada-esque after Brewers catcher Johnny Estrada)? Up until now it has been hard to tell exactly what Pujols is doing that other hitters aren't. Fortunately, with MLB releasing PITCHf/x data we can peek under the hood and see what makes Albert Pujols tick.
For this study I have used a clustering algorithm to classify what type of pitches every pitcher throws. This algorithm isn't perfect; it still calls a few sliders curveballs, and a few splitters sinkers, but it correctly classifies the majority or them. I'd guess that it is correctly classifying about 95% of pitches thrown right now and probably 99% of four-seam fastballs (which I will refer to as just fastballs from here on). Once these pitches are classified, I can go back and look at things from a hitters perspective and group together all the fastballs a hitter has seen while PITCHf/x was tracking them. This allows me to make nice tables and plots to start breaking down what a hitter is good at and what he struggles with. So let's start with a table for Albert Pujols grouped by the type of pitch that was thrown. Of the 565 at-bats Pujols had during the year, PITCHf/x was only on for 140 of them but that is better than nothing, though a small sample size warning will apply to everything you see below.
Type Number Thrown In Play/Strikeout Batting Average Slugging Average Fastball 466 106 .377 .774 Curveball 81 17 .176 .294 Slider 177 27 .333 .333 Change 123 23 .217 .314
I have left out a few pitch types that had really small statistics but you can see a complete table here. I am calculating the batting average and slugging average using both balls in play and the final pitch if a batter struck out. This is slightly unfair because at least two pitches were needed to get the hitter to a two-strike count, but leaving them out is also unfair as it would not penalize hitters at all for striking out. Hopefully, a modified metric can be found that will reach a happy medium with this data, but for now this is what we will work with.
The first thing that jumps out at me from the table is Pujols absolutely destroyed fastballs while PITCHf/x was watching. Despite this, pitchers threw him almost exactly the league average percent of fastballs, 46%. On the other hand, it seems that Pujols really struggled with the off-speed stuff particularly with his power numbers. Was Pujols swinging at some bad balls or having trouble with balls in the strike zone? To determine that we can plot the location of every offspeed pitch thrown to Pujols and what the result was. I am just going to show you the plot for sliders because there are more sliders than any other off-speed pitch, but the results are pretty similar for the curves and changes.
I have drawn in a strike zone as the rule book defines it, but you can see that some balls off the plate are being called strikes. This is exactly what John Walsh found when he studied the subject. I really should display that strike zone instead since it is the one being called, but there is something aesthetically pleasing to look at the MLB-defined strike zone to me. Anyway, it is now clear that Pujols isn't swinging at very many off-speed pitches out of the zone. In particular, pitchers have thrown a large number of sliders low and away, and Pujols pretty much laid off all of these pitches. The balls in the strike zone tell a different story however. Even the hanging breaking balls that are thrown right in the middle of the plate Pujols is either fouling them off, weakly putting them in play, or just letting them go by.
While I suspect that Pujols' true numbers are a little more even than we are seeing here, my guess is he really is a far better hitter against fastballs than off-speed pitches. Because he isn't swinging at many balls out of the strike zone he should be hitting them well. The thing that makes Pujols different from the many other players that only hit fastballs well is his willingness to let the off-speed pitches go. This is why he is still getting a lot of fastballs to hit, because pitchers are trying to make him chase the bad balls and he just won't do it. This is reflected by how often Pujols gets the count in his favor. While PITCHf/x was watching the pitcher had Pujols 0-1 47 times but Pujols had the pitcher 1-0 57 times. Pitchers are trying to be careful with him and falling behind and then are forced into throwing him the fastball he tees off on.
So how should pitchers be pitching to Pujols? Well despite his relative inability to hit off-speed pitches I don't recommend throwing him curveball after curveball. They are called off-speed pitches for a reason, and I suspect that the change in speed is the key to why these pitches are effective against Pujols. That said, I am throwing him an off-speed pitch to start the at-bat almost every time. A "get me over" change up is likely to reach the catcher's mitt and get you ahead 0-1. From there mix in the fastball but don't throw it in the strike zone. I would recommend up and in to keep him off guard. This is almost completely opposite to how pitchers normally attack hitters. The standard method is to get ahead with the fastball and then throw an off-speed pitch to finish the hitter off. This just isn't going to work against Pujols. While I do think this strategy will help, it is Albert Pujols after all. You can't stop him; you can only hope to contain him.
<< Return to Article