I have an idea for bat tracking data

I was in Hawaii last weekend, enjoying a nice vacation relaxing from the end of baseball season, when I found myself thinking about interception points. Weird? Overly obsessed with baseball? perhaps. But it seemed to me that a kid in the pool was hilariously waving a Wiffle ball almost the entire time, spraying it “foul” every time. “Oh, look, the next Luis Arez,” I thought, and went back to drinking my umbrella drink. But it stuck with me, and when I got home, a database query emerged from my mind fully formed like Athena after Zeus had a headache.

Where is the best place to make contact with the ball? It depends on who is swinging. Statcast measures the point of contact of each swing relative to the hitter’s center of gravity, and the data clearly shows that there are many paths to success. This always gets in the way when I’m looking at swing path data. But that little kid gave me an idea when he hit the best swing I’d seen all day, his wiffle ball line drive that would have elicited screams at the left-field foul line (he hit left-handed). Because he’s so late in his normal swing, his sweet spot is also slightly later. What if I looked for clues about swing timing based on the batter’s own swing?

I counted every batter in 2025 who had 300 or more at-bats (either foul balls or balls in play). For each hitter, I compiled a summary tally of all their results and then divided their shots into three groups: deepest point of contact, middle point of contact, and farthest forward point of contact. You can think of it as late, on time, and early, and adjust it based on that player’s swing. The later you start your swing, the more you “let it move” and the deeper your contact point becomes relative to your center of mass. The earlier you start, the further “in front” you are and the further you reach.

Without this data normalization, I wouldn’t be able to determine the heads or tails of the touchpoint data. But after defining each swing relative to the player’s own average, things start to make more sense. I measured each swing’s wOBA, xwOBA, exit velocity, hard hit rate, and launch angle (just in case), relative to each player’s overall production. I then weighted each batter based on the number of times they made contact (I ignored swings and misses, which seemed like a different category to me), and the aggregated data made sense:

Let’s go through this table to make sure everything is tracking. Later swings, the ones with the deepest contact point, produce bat speeds that are slower than the average for a given hitter. This makes sense, since there’s plenty of evidence that swings speed up over time. But the actual results were basically average. My interpretation is that this bucket is filled with mistakes and flukes, shots that fool defenders. The last-minute jab, the grounders that went well beyond the batter’s normal spray pattern, the general weirdness. That’s how you get a wOBA higher than xwOBA; for some reason, fill it with balls that frustrate the other team’s expectations.

Reaching the middle barrel we had the best contact. In these swings, hitters do exactly what they want to do, making contact around their average point. A 40 point increase in wOBA is huge and, well, crap. Hitters perform better on time, and on-time is very well defined as it relates to a player’s average swing. In my opinion, the reason the hitters don’t perform well here is that the xwOBA direction is agnostic, and many of the hardest-hit pitches in this group are back to the middle, where the wall is further away.

Finally, “leading” has the worst results. You may be wondering what this has to do with the launch angle revolution, and the fact that fly balls are so popular right now. But again, we’re talking about every swing as it relates to the hitter. Those who uplift and celebrate? Most of their swings are already in front, so their “in front” bucket is when they catch the ball early, even for them. This is why, despite a longer swing, the ball speed is slower and the overall results are poorer.

Another way of saying the same thing: When a hitter makes contact with the ball further away from the average point of contact, contact becomes more variable. The emission angle standard deviation increases significantly. The catastrophic rate dropped sharply, more so than you would expect based solely on changes in average exit speeds. Relative to their overall tendencies, hitters hit more pop-ups and more ground balls.

None of this should shock you. If you’ve swung a bat enough times in your life, you can probably imagine it. In order to turn this into useful data, I thought I’d look specifically at a few players to show how different people’s swing behavior differs.

A batsman who excels up front? Consider Mike Trout:

When Trout arrived late, he real Late. When he was still early, his results were outstanding. In other words, “early” looks no different to him than being on time. My interpretation of this result is that, for whatever reason, Trout’s swing started late, so even his swing in deep third was still filled with solid contact and good results. Another way to put it: Too many of his swings are late. Look at how disastrous the first bucket is compared to the average. He should probably try to minimize those swings, even if it means being on top more often.

For comparison, look at Juan Soto:

This is the distribution of a player who clearly understands his swing. The center of his swing distribution, where he swings on time? He is smoking. When he’s early or late, he doesn’t perform well. But he was on time. He’s Juan Soto! When he’s on time, he weaponizes his incredible hitting ability into his best swing. That’s what you expect from him, right? If you’re curious, other hitters with similarly shaped distributions are Shohei Ohtani, Corey Seager, Nick Kurtz, Ronald Acuña Jr., and Austin Riley, as well as Ty France, because baseball can be weird sometimes.

While looking for the ultimate source, I learned something interesting, which is that batters do more damage when they arrive late than expected. This group is where the pull outliers are. This is Mookie Betts:

Isaac Paredes:

And Francisco Lindor:

If Trout starts his swing too late, then these guys all start their swing too early. Yes, that’s exactly what they are doing! If you intentionally stay in front as often as possible to gain power, your swing will be front-facing even by your standards Way Too far forward. Therefore, your “best” contact will be at the back, or at least certainly not at the front. That’s clearly a trade-off these guys are making, and they’re doing it on purpose. Still, it’s cool to see an expression like this.

My research is still in its early stages. I’m not sure this trichotomy is the best way to look at things. There are many alternatives, ranging from other partitioned quantities to continuous distributions. I’m also not sure which statistics are most useful for this. Should I include bat speed? Do I still need to standardize by location? If you have any thoughts please feel free to chime in as I’m still debating this myself.

Again, I don’t think these tables are the best way to visualize things. Maybe some graphical representation would be better, but there are almost too many variables for me to understand. Another question: I don’t know how useful this data is from a predictive, systemic perspective. It’s definitely cool and that’s why I’m writing this, but it’s not really the end state.

In other words, welcome to the offseason. There will be some interesting signings and some interesting trades. ZiPS team-by-team predictions are always a read. Prospect list will start soon. Hall of Fame season falls in time for the holidays. Visiting relatives at home and discussing baseball through the ages! But there’s still a lot of space to fill, and I like to fill it with ongoing research. If you have any suggestions on what I can do, please let me know. I think it’s already cool, and I think it can get even better. Oh yes, this is data.