Shawn Brody wrote an outstanding article last month, titled “Bryce Harper is locked in”. What he showed was that, in the midst of Harper’s stellar opening month, he had advanced beyond simply distinguishing between strikes and balls. Harper has additionally been swinging more often at strikes that are in his “happy zone” (the inside portion of the plate where Harper thrives) and letting strikes go that are more difficult for him to drive.
The piece blew my mind. It’s an amazing skill that Harper is displaying, and requires an extraordinary level of pitch recognition and self-awareness, even by the heightened standards of major-league hitters. And I wanted to know: who else can do the same? This article is my attempt to answer that, by quantifying that Harper-ian state of being “locked in” that Shawn wrote about.
The first step was identifying each player’s “happy zone,” the parts of the strike zone where they do the most damage. I used Baseball Savant to search for all batted balls, as broken into the nine parts of the strike zone, and calculated each player’s wOBA on contact (wOBAcon) for each of those zones. I next used Savant to calculate each player’s swing rate, as broken down by strike zone. (I used 2016 data, to ensure that there was enough of a sample to trust each batter’s numbers, even when broken down in this fashion.)
My goal in creating this metric was not to make patient hitters look better than aggressive hitters, or even to make hitters who can distinguish between strikes and balls look better than those who can’t. The only thing I’m interested in, at this moment, is the ability (or inability) of hitters to distinguish between good strikes and bad strikes, based on their own characteristics. With that in mind, the next thing I did was scale those zone-specific wOBAcons to each player’s overall wOBAcon in the strike zone, and did the same with the swing rates. So a wOBAcon+ (this is as long as the names will get, I promise) of 1.00 means that the player’s performance on contact in that part of the strike zone is exactly as good as his performance on contact in the strike zone as a whole, while a wOBAcon+ of .80 means his performance is 20 percent worse, and a wOBAcon+ of 1.20 means his performance is 20 percent better.
The final step was to compare a player’s wOBAcon+ and swing rate+ for each portion of the strike zone, by taking the difference between the two and adding those differences across each zone. A smaller number indicates a closer match between the relative rate at which a player swings inside the zone and the relative rate at which that player performs on contact inside the zone. A figure of 0.000 would indicate a perfect match — e.g., a player who swings 30 percent more often at up-and-in pitches, and performs 30 percent better on pitches in that zone, and lines up equally across every other part of the strike zone as well. That total figure, which I’m calling KNOWTHYSELF, is the end result of this process.
Now, we’re almost to the good part, but first, a couple of disclaimers. One, it’s misleading to think of any of the things I’m comparing as independent of each other, or stable. The foundation of KNOWTHYSELF is built on sand, albeit sand that I think is stable enough to result in something interesting and useful. A player’s wOBAcon depends not just on their own abilities, but on how opposing pitchers approach them, and that in turn depends on how the batter swings. Someone who swings less often at low-and-away strikes than up-and-in strikes might see more pitches to that low-and-away part of the zone, and that could in turn change their wOBAcon. It’s circular, to a degree. Like I said, I think that degree is small enough, and progress around the circle slow enough, to make it useful, but it’s important to mention that.
Two, it’s also not clear that having this kind of intra-zone discrimination is always going to be a good thing. I’m not doing any kind of calculation of the cost to a player of a taken strike, and in some situations (e.g., every two-strike count), it’s obviously correct to swing at any pitch inside the strike zone, even if it’s not a pitch that the batter thinks they can hit well. I hope this reflects something real and reflective of hitting ability or wisdom, but it could be nothing at all.
Enough with the intro; time for a leaderboard. The following are the batters with the lowest KNOWTHYSELF scores:
KNOWTHYSELF Leaders, 2016
There is a lot to break down in this table. The first thing to note is that this doesn’t seem to match up with our expectations of traditional plate discipline. The player in first place, by a large margin, is Adam Jones, who in 2016 had a 60.6 percent swing rate and just a 5.8 percent walk rate. But inside the strike zone, Jones did a better job than anyone else of taking advantage of his own abilities:
Some of the other names in the top ten are similarly aggressive hitters, including Jones’s teammate on the Orioles, Jonathan Schoop (2016 walk rate of 3.2 percent). But several of the players on the list are well-known as discerning batters, suggesting that there (unsurprisingly) may be some overlap between the ability to tell a strike from a ball and to tell a good strike from a bad strike. Corey Seager and Brandon Belt both fall into that category, but its standard bearer is probably Freddie Freeman, who comes in at number two on the leaderboard:
Freeman has always been excellent, but like Harper, he’s in the midst of an ultra-breakout kind of season, hitting .342/.463/.750, good for a 206 wRC+. Speaking of Harper, the Nationals slugger was in the middle of the pack in his down year last season, ranking 77th of 174 qualifiers.
And finally, what about the bottom of the leaderboard? Here are the players with the ten highest KNOWTHYSELF scores:
KNOWTHYSELF Laggards, 2016
Another strange mix, featuring great hitters who had down years (Jose Bautista, Ryan Zimmerman), light-hitting low-power types (Martin Prado, Brock Holt), and most other player types you can imagine.
So what’s the takeaway? KNOWTHYSELF doesn’t seem to correlate particularly well with general hitting ability, nor does it seem to be concentrated among players with great batting eyes. That could mean it’s revealing a vast trove of information that was previously hidden, or it could mean that it’s reflecting randomness over any kind of real process. Both can be true in differing parts, but I’m leaning more toward the latter.
That said, I’m still going to track this for 2017, because I’d like to dig into it a bit more before I dismiss it outright. The idea of a player operating with such a high level of hitting skill and instinct so as to be able to distinguish between individual parts of the strike zone is fascinating to me. I sure hope it’s a real thing, and that this made-up metric accurately estimates it.