Nobody Actually Hits Good Pitching

During Game 3 of the ALCS, Jeff Francoeur noted that 10 of the 12 pitchers who had surrendered postseason home runs to Alex Bregman were All-Stars, eliciting a comment about how Bregman hits well against good pitching. That’s undoubtedly a fun fact, but the notion has been rattling around my brain for the last couple weeks. I’ve definitely heard stories about players who hit good pitching, but I’ve never heard of anyone looking at splits based on the quality of the pitcher on the mound.

The main problem with that idea is a logical one. Break down any batter’s performance into its constituent parts, and you’ve entered a zero sum game. If you’re at your best against great pitchers, that means you’re hitting worse against everyone else. It’s hard for me to imagine that there are many players who fare better against Justin Verlander than they do against, well, anyone on the Nationals.

There could be players who are less bad than average, but if I had my choice of superpowers, I’m not sure that’s the one I’d pick. It would definitely help out in the playoffs, but over the course of the season, batters see a lot more average and poor arms than they do great ones. I’d rather perform well the majority of the time.

To get altogether too cute with things, if there are players who hit better against the best pitching, then that’s a weird reverse split and we should take advantage of it. We can platoon them with players who do better against poor pitching. Player A plays against frontline starters, and Player B starts against the back of the rotation, then sits back down when the high-leverage relievers come in.

Like I said, this has been rattling around my brain for a while. So in order to dislodge it, I did some research. I identified the best pitchers in the league by FIP for each of the last two years, minimum 40 innings. Our top 50 represents the top 6% of all pitchers and the top 12% of pitchers who threw at least 40 innings. In 2021 and earlier, those percentages are slightly lower, as the number of pitchers used in a season has been climbing.

I went to Baseball Savant and pulled the wOBA of every batter in the league when facing that year’s top 50. For the rest of this article, I’ll be talking about that number, as well as a player’s wOBA against all pitchers, and the difference between the two. For the sake of simplicity, I’ll call them top 50 wOBA, overall wOBA, and wOBA difference. I’m about to type wOBA so many times; please prepare yourself for some serious semantic satiation.

Here is a graph of the 59 players who had at least 40 plate appearances against top 50 pitching in both 2021 and ’22:

There wasn’t a single player with a positive wOBA difference two years in a row, though J.P. Crawford is oh so close over there in the upper right corner of the blob. The 2021 and ’22 wOBA differences have a correlation coefficient of -.20. In other words, if you have a high wOBA difference one year, it doesn’t mean you’ll do it again next year; it means you’re likely to get a visit from the regression monster.

I pulled the same numbers for the 2018 and ’19 seasons, and for the ’17 and ’18 seasons, and made two more scatterplot blobs. They had positive but tiny correlation coefficients of .1 and .06. At this point, I was ready to say that running a better wOBA against top pitchers wasn’t a repeatable skill.

I was’t thrilled, however, with the small sample of players and the small amount of plate appearances it took to qualify, so I went back to Baseball Savant and pulled top 50 wOBA for every batter in the last 10 years (I set a 20-inning minimum in 2020). It took a while.

In the last 10 years, 550 players made at least 100 plate appearances against top-50 pitching. That’s our sample. Before we really dig into the numbers, here are the best hitters of the past 10 years against top-50 pitching:

Well, we’ve got some of the best hitters of the past 10 years. We’ve also got some other dudes. Most telling, I think, is that the superstars are all at or below their overall wOBA, whereas the surprises are over-performing theirs.

Now that we’ve had a chance to admire Juan Soto, we can dig into the math. As you’d expect, batters generally did much worse when facing the best pitching in the league: 95.1% of players were worse, 4.2% were better, and 0.5%, or three players, were exactly the same.

Against top-50 pitching, wOBA dropped by an average of .057, or 17.9%. The median change was nearly identical. To put that in perspective, in 2022, 57 points of wOBA was the difference between Marcus Semien and Victor Robles. Here’s a quick graph to give you the shape of the distribution. Each number along the x-axis represents wOBA changes within 20 points. (For example, the bucket for +20 includes increases of 11–30 points of wOBA.)

As you can see, the distribution is centered around our average loss of 57 points, and it’s extremely symmetrical. Just in time for the holiday season, you can turn this graph upside down to make a lovely, slightly overlarge menorah.

Here are the players with the biggest wOBA increases:

Suffice it to say that this list no longer contains a Juan Soto or a Mookie Betts. More importantly, nobody has a wOBA difference big enough that it makes too much of a difference. There goes our weird platoon plan.

Next, I wanted to check whether good hitters fare better against good pitching. Here’s a scatterplot of overall wOBA and wOBA difference:

What you’re looking at is a blob with a very slight slope. That’s because there’s a small negative correlation between a player’s overall wOBA and their wOBA difference (-.24 going by actual wOBA change, and -.09 going by percentage wOBA change). In other words, the better the hitter, the steeper the drop-off when they face an ace.

Before we run too far with that conclusion, keep in mind that there’s likely some regression the mean baked in here. More importantly, better players tend to hit more often, so they’re less likely to be outliers. Here’s what happens when we start tightening up our sample size:

That’s a pretty clear trend. The players get better every time we increase the plate appearance requirement, but the wOBA difference holds fast at -18%. The list of players with at least 500 plate appearances includes players like Freddie Freeman, Paul Goldschmidt Mike Trout, and Bryce Harper, and even they fare no better against top-50 pitching.

If our hypothesis was that the best hitters find a higher gear when they’re facing great pitching, then we can officially put it out to pasture. Of the 100 players with the most top-50 plate appearances, only one had a positive wOBA difference. Elvis Andrus, please claim your prize.

The last factor is one that I speculated upon earlier. Players just don’t face elite pitching often enough for it to have an outsized effect on their overall performance. This year, Shohei Ohtani put up an atrocious .229 top-50 wOBA, but because only 57 of his 666 plate appearances came against top 50 pitching, he still put up a .370 overall wOBA, putting him in the 96th percentile.

Comparing overall wOBA to top 50 wOBA is admittedly a crude approach. I’d be interested in taking another look if and when I have the means to analyze the data in a more nuanced way. I don’t want to rule out this line of thinking out completely, even if it’s only good for bringing to light the occasional fun fact.

We have all sorts of splits these days. We break down hitters against handedness, pitch type, velocity, and vertical approach angle. The list is certain to keep on growing. For now at least, “he hits good pitching” is unlikely to become one of our go-to metrics.

Oh, and what about Bregman, the guy who got all this started? Well, he’s definitely not better against top-50 pitching, but he is just a tiny bit less bad than average. His wOBA difference is -.047, or -13%, which puts him right where you want to be: on the good edge of the blob.