A Meandering Examination of Fly Ball Pull Rate, Featuring Stars of the Game and Also Isaac Paredes

This all started because I wanted to write about Isaac Paredes. He’s my kind of player, excellent despite all sorts of warning signs that what he’s doing shouldn’t be working. Advanced metrics and in-person scouting assessments are both quite negative on Paredes, and yet he’s batting .255/.354/.503, good for a 140 wRC+, in mid-September. He’s been one of the most valuable players on one of the best teams in baseball. It’s so weird!

But lo and behold, the exact thing I wanted to write about has already been written. Curse you, Esteban Rivera! Well, not actually, of course. Esteban’s writing is great, and it’s also of particular interest to me because he’s so observant of hitting mechanics. But I can’t exactly write an article about how Paredes’ pull-happy tendencies have helped him keep regression at bay when there’s a better article talking about just that already on the site.

Anyway, you should read Esteban’s article. But I didn’t think I could get away with stopping mine here, so I decided to flip the analysis on its head and look at players who are, for lack of a better way to describe it, anti-Paredesian this year. Sure, it’s useful to have a fly ball chart that looks like this:

Those pulled home runs are hilarious! And he hasn’t hit the ball to the opposite field with authority at all. The joke here is that he keeps pulling the ball so close to the line that he’s getting home runs out of batted balls that would be outs if he hit them anywhere else. But I digress. Not everyone has done this, and the players who come closest to doing the opposite are extremely interesting to me.

As Esteban pointed out, Paredes and Cody Bellinger comfortably top the list of hitters whose actual production has most exceeded their Statcast-expected production on fly balls. That’s for a pretty obvious reason: xwOBA and xSLG and all the rest don’t consider the horizontal angle — “spray angle” if you speak the lingo — of batted balls. Most of Paredes’ homers would be fly outs if they were hit to center field or even to the power alleys, so they don’t rack up huge xNumbers, but they’re home runs in the real world. Hence, the discrepancy.

If you’re looking for the opposite side of the spectrum, then, your first guess would probably be all-fields sluggers who live up the middle or the other way fairly often. And you’d be right, but you could have gotten most of the way there by just looking for players with Junior in their name:

More broadly, all of these guys have tremendous top-end power. They’re also spray hitters; none are in the top half of the league when it comes to pull rate on aerial contact. Look at Tatis’ spray chart on fly balls:

That cluster of outs in deep center is outrageous. Tatis has already hit 12 batted balls that went 375 feet in the air and turned into an out. That’s near the top of the majors. Bobby Witt Jr., to continue the Junior theme, has 16. Acuña has 13 of them. Those well-struck balls that carry to a pesky center fielder really resonate; I can conjure them up in my mind’s eye easily.

So is there something inherent in high-power players that makes them do worse than xwOBA’s naive expectation on fly balls? Is their own power somehow betraying them? That would be a strange takeaway from this, but it’s hard to shake that feeling when you look at the numbers.

I don’t think that’s really the right way to look at it, though. Why constrain ourselves to xwOBA when we can do the investigation ourselves? The real issue here seems to be that the various Juniors are hitting too many balls to straightaway center. Put another way, here are each of those hitters’ wOBA-xwOBA differentials – the “error” between modeled results and actual results – based on which direction they hit their fly balls. I also added Paredes as a contrast:

Curiouser and curiouser. Paredes isn’t just pulling more of his balls, he’s getting far more out of his pulled ones. Let’s zoom in on that segment in particular:

Now I think we’re getting somewhere. Paredes isn’t outperforming these guys on a per-pulled-fly-ball basis; he’s right in the middle. But he’s hitting a ton more of them, and as we know, he’s lofting balls at an angle/speed combination that xwOBA hates but that still go for homers. And wow, what in the world is going on with Chapman?

I think the barrel rate column is really important. The massively powerful hitters we’re interested in are crushing a ton of balls to the pull side. But if you truly smash one, it hardly matters where you hit it. Here’s one of Acuña’s barrels:

I half expected that ball to break down to its constituent atoms. That would have been out of any park in baseball to any part of the field. It’s hard to outperform your xwOBA when xwOBA says “oh, yep, that ball becomes a home run 100% of the time.”

Here’s another way of looking at it, this time ignoring our six-player cohort. Different types of batted balls have different sensitivities to horizontal angle. Bucketing isn’t exactly scientific, but it’s an easy way to show data in a table, so that’s what I did here. Take a look at the actual results of fly balls hit at various speeds to various parts of the park:

This is the crux of what’s going on. In the middle velocity bands, well-struck but not crushed fly balls, which way you hit it matters a ton. If you truly demolish it, any direction will do. A different way of stating what’s going right for Paredes – and wrong for our band of sluggers – is that if you make solid but not overwhelming contact, you’d be well-served to pull it.

Alright, then, another chart. Here’s the batted ball distribution on fly balls hit between 95 and 105 mph:

Right, that’s basically what I expected. Paredes is comically ahead of the pack when it comes to maxing out that type of contact. We kind of already knew this, though. I think a better question is whether we should expect the rest of the group to keep putting up similar power numbers to what they’re doing now, or whether we should expect some kind of reversion towards the mean stemming from improved production on these mid-velocity fly balls.

To do this, I ran a straightforward test. I looked at pull rate on fly balls hit between 95 and 105 mph. I took every season starting in 2015 and looked for consecutive seasons where a player hit at least 50 fly balls in that exit velocity range. The results are encouraging. 31.6% of the variation in year-two pull rate can be explained by variation in year-one pull rate. These are for small samples, too – the average number of batted balls per player-season was only 58. You’d expect a good amount of noise, and there’s still a strong correlation between pulling the ball in one year and doing it again in the next year.

Here’s another way of looking at it: the league average pull rate on these 95-105 mph fly balls is right around 32%. The top 10% of my back-to-back 50-batted ball seasons set pulled their fly balls 51.2% of the time in year one and 50% of the time in year two. The bottom 10% pulled theirs 13.4% of the time in year one and 23.5% in year two. That second one is also heavily influenced by a very weird pair of seasons by David Ortiz – a 15.1% pull rate on these medium fly balls in 2015 followed by 46.3% in 2016. So those big swings do happen, but less frequently than you’d expect if it were pure random chance.

Of course, we don’t have to limit ourselves to a single year of data when trying to predict the future. Paredes didn’t suddenly start pulling the ball this year. He’s consistent; he had a 74.2% pull rate on them last year, and he’s at 65.1% this year (without enough to qualify for my 50-batted-ball minimum in either, to be fair). We don’t have to treat this year like it exists in a vacuum.

Between 2015 and 2021, 64 players hit at least 200 fly balls between 95 and 105 mph. I looked at how those hitters’ pull rates in that segment compared to their pull rates over the past two years (minimum 50 batted balls). That left me 34 players – a lot of the guys in that first sample are out of the league or have been injured recently. And the r-squared here is a whopping 62.4%. In other words, if you know that a guy historically pulls his medium-hit contact, it’s a good bet that he’ll continue to do so. Nolan Arenado, Marcus Semien, and José Ramírez pulled the ball a lot from 2015 to 2021 and still do it now. Mike Trout, Freddie Freeman, and Nick Castellanos rarely went to the pull side in the past, and they still rarely do. It’s an inherent feature of swing shape, in my opinion, and has a lot to do with why Paredes’ xwOBA numbers look so weird.

Back to the six players this article is ostensibly about: Are they all doomed to float lazy fly balls to the center fielder for the rest of their lives instead of gracefully tucking them over the pull-side wall? In a word: yes. Take a look at those six and compare them to Paredes:

There you have it. Paredes is doing something they aren’t. To be clear, I don’t think he’ll keep up his preposterous pace; he’s even outperforming them on pulled fly balls in that 95-105 mph band, and that seems likely to come out in the wash long-term. But he’s truly doing something valuable, and these other hitters aren’t. That’s not to say that what they’re doing is wrong. If I were coaching Acuña or Soto, I’d tell them to keep doing what they’re doing, because it’s clearly working. But part of the cost of their approach is that they don’t max out the value of their middling contact. It hasn’t stopped them from being excellent hitters, but it does mean that if you peruse a leaderboard, you should keep in mind this unspoken but inherent weakness.