Overthinking It: Bryce Harper’s Brain is a Neural Net Processor

There’s something weird and wonderful about the distribution of pitches seen by this season’s two most exciting under-21-year-olds. David Golebiewski pointed it out the other day using information from Inside Edge, and I wanted to see whether PITCHf/x data would show the same thing. It did. That was the only excuse I needed to write about Bryce Harper and Mike Trout, neither of whom can make a move on a baseball field without leaving a trail of article topics behind him.

Here is a list of the hitters who’ve seen the lowest percentage of fastballs this season, combining four- and two-seamers and not counting cutters (classifications courtesy of Harry Pavlidis):

Bryce Harper is seeing fewer fastballs than any other hitter. Think about that. Pitchers have collectively decided that it’s a better idea to throw fastballs to every established slugger than it is to throw fastballs to Harper, a 19-year-old rookie with four career home runs. Maybe it’s because his reputation as baseball’s next best player preceded him, or maybe it’s because the scouting reports said he could catch up to anything (or that he'd struggle to lay off breaking balls, though according to Kevin Goldstein, there wasn't any widespread concern that that would be a weakness). Josh Hamilton is five spots below Harper on the list. Hamilton is a former MVP who is Paul Konerko’s batting average away from leading the AL in the traditional Triple Crown stats and has hit as many home runs in a single game this season as Harper has hit in 30, and pitchers have thought, “Yeah, I want to try to sneak a fastball past this guy” more often when facing him than they have when facing Harper.

That’s surprising, but it’s not shocking, since it’s hard to be shocked by anything Harper does. This is shocking, at least to me:

That’s the list of hitters who’ve seen the highest percentage of fastballs this season. For the most part, the hitters who see the most fastballs are the ones who can’t hit fastballs over the fence. The highest 2011 home run total of any player in the top 10 in fastball percentage is six (a two-way tie between Callaspo and Jeter). Mike Trout is right next to Jamey Carroll on this list, which is probably the only list in the world on which those two players would appear side by side.

Mike Trout is not a hitter who can’t hit a fastball over the fence, but he’s being pitched like one, which is extremely strange. Trout’s 135 plate appearances as a 19-year-old didn’t go quite as smoothly as his first 134 plate appearances as a 20-year-old. Maybe major-league pitchers remember last season’s stats but haven’t read any prospect rankings, in which case they might be thinking, “Okay, this guy is roughly as good as Robert Andino” when Trout comes to the plate. Maybe they stick comic books inside their advance reports and read them while they pretend to prepare. Or maybe they’re terrified of putting anyone on ahead of the middle of the Angels’ order, which could help explain why Izturis and Callaspo, who often hit behind Trout, are two of the few hitters who’ve seen more fastballs than he has.

Regardless of the reason, I’m going to go out on a limb and guess it isn’t going to last. Trout’s slugging percentage is .521, which is higher than Harper’s. He won’t always outslug Harper, but he might always be able to slug .521. There aren’t any other .500 sluggers on that list, and once the league learns that that’s what Trout is, he won’t be on it either.

Before long, Trout will start seeing fewer fastballs. When he does, we shouldn’t forget that for the first 100-plus plate appearances of his age-20 season—which might also be a Rookie of the Year season—pitchers pitched Mike Trout like they pitched a 34-year-old Juan Pierre, the shambling corpse of Chone Figgins, and Jamey Carroll, who hasn’t homered in almost four years.

Because I wasn’t sure what to make of all this, I asked a pro scout with an AL team if he could explain why Harper has gotten so much more respect from opposing pitchers. His theory makes a lot of sense:

Blind conjecture: People hear or think “Bryce Harper” and they immediately think, “power that we’ve been hearing about since he was 12.” People hear Trout, they think, “speed, defense, hit.” I think his pop is underrated.

What the scout is describing sounds a lot like the association fallacy. Trout bats leadoff, and he plays a fine center field, and he steals a lot of bases. Players who do those things typically don’t hit for a lot of power, so when pitchers see him doing those things, they might mistakenly assume that he isn’t a threat to hit for extra bases. In most cases, that wouldn’t be the worst generalization to make. In Trout’s case, though, it’s a bad one, because anything most players can do, Trout can do better. That’s what makes him Mike Trout.

There’s one more thing I want to mention about Bryce Harper’s first month.

Take a look at these two tables. The top table tells us how Harper hit each pitch type from his arrival at the end of April through the first half of May. The bottom table gives us the same breakdown for the second half of May. Both samples just over 60 plate appearances. “Breaking” combines sliders and curves. “Fastball” combines fastballs and sinkers. “Offspeed” is everything else: changeups, cutters, splitters, knuckleballs, screwballs, and probably that thing Vicente Padilla throws. BAcon and SLGcon are batting average and slugging percentage, respectively, on pitches put in play.

The first thing to notice is that Harper has made pitchers work much harder in the second sample of 60-something plate appearances, in part by chasing fewer pitches outside of the zone. The second thing to notice—which I put in pretty colors so you couldn’t not notice it—is that Harper seems to have adapted to big-league breaking balls after about two weeks. Some guys go entire careers without figuring out how to hit breaking balls. If the book against Harper was that he was vulnerable to breaking balls, the book can be burned. He’s now a breaking ball-resistant strain of superstar.

When I was a freshman in college, only a little younger than Harper is now, I played Super Smash Bros. for Nintendo 64 every day, which I assumed was what my orientation advisor meant by being social and seeking out new experiences. Gradually, my dormant grammar school Smash Bros. skills came back, and I got really good at using Donkey Kong’s down aerial attack.* None of my friends knew how to avoid it, and for a while, it couldn’t be beaten. Naturally, I used it all the time. And the more I used it, the more often they recognized the move as soon as the animation started. Once that happened, they could not only avoid it, but exploit it. From that point on, it was just another tactic that worked some of the time, but not all of the time. After a few more fruitless years of trying and failing to bring DK’s down aerial attack back, I graduated with an English degree.

*I know that it’s called a down aerial attack because there’s a wiki just for Super Smash Bros., and the down aerial attack has its own page. Anyway.

If Harper has learned to hit breaking balls, then pitchers have lost their down aerial attack, and they’ll have to try something else besides refusing to throw him fastballs. These are the sort of adjustments and counter-adjustments we see every young hitter make, but because Harper is so talented, he might be making them at an accelerated pace.

These are very small samples we’re dealing with. In his first couple weeks in the majors, Harper went hitless against breaking balls, and in his second couple weeks in the majors, he destroyed them. That doesn’t mean he was a terrible breaking ball hitter before, or that he’s an incredible one now. It might just mean his luck turned, or that he got hot and could be about to go cold again. But it might also mean that Harper is a T-800.

The more contact Harper has with humans—particularly humans who throw baseballs—the more he learns. Now we'll see whether the league can learn anything about him.

Harry Pavlidis provided research assistance for this article.