Thoughts on Pulling the Ball

There have been two things somewhat stuck in my head that feel contradictory but maybe aren’t. One is that hitters should pull the ball more to get to more power. The other is that “using all fields” is a time-honored phrase for hitters who benefit by spraying the ball around the park.

I am on the record being a fan of pulling the ball. Earlier this year, I shared these numbers on the advantage of pulling the ball in the air:

Added context, league wide from 2023-24, on why “pulled” matters:

wOBA/xwOBA on FB/LD 95+: .787/.794

wOBA/xwOBA on PULLED FB/LD 95+: 1.070/.882

When you pull ball hard it’s WAY better than just hitting hard. Pulling a lot is good way to make contact play up and beat xwOBA https://t.co/EiEVv21feP

— Chad Young (@chadyoung) May 9, 2024

And if you go back to my pre-season excitement about Ke’Bryan Hayes, it was because he was lifting and pulling:

239) Ke’Bryan Hayes $5 me

Three highest flyball rate months of career were May, Aug, Sept ’23. All over 40%. Fourth highest was Apr ’23 at 36.8%. never over 35.1% before. Never pulled >17.4% of FB in month. July/Sept ’23 were 47.4% and 34.3%. I’m excited. Woulda gone to $10.

— Chad Young (@chadyoung) December 10, 2023

Now, that turned out to be misguided excitement, but that’s because he has failed to pull and lift the ball this year – he pulled it in April and June and lifted it in May, but hasn’t done both at the same time. Plus he’s hitting the ball less hard!

Which gets me back to that first tweet and my thoughts on Hayes. There is a recipe implied in the data in that first tweet: you need to hit the ball hard (the “95+ mph” in the tweet), elevate the ball (data focuses on fly balls and line drives), and then if you pull the ball, you can beat your xwOBA.

Eno Sarris covered some of this in a recent piece on “lucky” power, focusing on Steven Kwan and his Guardians teammates. In that, he wrote:

One thing that’s changed for both Kwan and all the Guardians? They’re pulling their fly balls more now. They were second to last in how many fly balls they pulled last year, and this year they are third in that statistic. Pulled fly balls and pulled Barrels go further than opposite field contact. The slugging percentage to the pull side in the air across the league is 1.388, to the opposite field in the air it’s .286. That’s a massive difference.

In discussing this in the Ottoneu Slack community, someone called out that SLG to the pull side might be higher and pulled balls might go further simply because they are hit hard more often. And that sent me down a bit of a rabbit hole: what do wOBA, xwOBA and batted ball distance look like for pulled vs. opposite field contact, if you control for the quality of contact (exit velocity and launch angle)?

Let’s start with a simple table. Each cell in this table represents a range of launch angle and exit velocity, and displays the pull wOBA for that range minus the opposite wOBA for that range. Big positive numbers (which show in green) mean pulled balls perform a lot better. Large negative numbers (red) indicate ranges where opposite field contact performed better. All data is 2022-2024.

For an example of what this means, if you go to the cell under LA = 26-30 and EV = 96-100, you see .876. That means that for all balls hit with a launch angle between 26 and 30 and with an exit velocity between 96 and 100 (both ranges are inclusive), the wOBA on pulled balls was .876 higher than the wOBA on opposite field contact. Specifically, wOBA on pulled balls is 1.471 vs. .595 going the other way. The empty cell bottom right is because there were not balls hit the other way at that launch angle with that exit velocity.

Most of that table matches what we would expect – a whole lot of green, signifying that hitters got better results when they pulled the ball. But the far left and the top are a little different. You get a decent amount of red.

Baseball Savant defines a line drive as anything between 10 and 25 degrees of launch angle, so that far left column is mostly grounders. Not balls beat into the dirt, but the kind of ball that might become a one-hopper or something. That second column are mostly low line-drives.

Lower launch angles and lower velocities lead to better outcomes if you go the other way. As exit velocity increases, the value of pulling the ball increases. The reason, as Eno noted, has to do with how far the ball travels. The table below shows the same ranges, but in this case the cells show pulled batted ball distance minus opposite field batted ball distance. Positive green numbers are pulled balls going further. Negative red numbers are pulled balls not traveling as far.

This was maybe the most interesting thing to me – at low launch angles, going the other way INCREASES batted ball distance. This is even true for cases where pulling the ball still leads to better results. For example, if you hit a ball the other way at 16-20 degrees and 101-105 mph, it will travel about 14 feet further, but those balls have a wOBA .141 lower than balls hit the other way. Perhaps, in this case, the added distance is just more time for the ball to hang up and get caught? The average distance to the pull side is 314 vs. 328 the other way, so the extra distance is only very rarely going to result in a home run. Only 3 of 791 batted balls in that zone over the last three years have found the seats. Instead, the added distance just makes them less likely to find grass.

Having looked at this, I started to share what I was finding and someone asked if maybe the pulled balls just had a higher xwOBA. Yes, they were performing better but that was because, even within a 5 mph and 5 degree of LA range, they were hit with more power at better launch angles.

This table is a bit more confusing. The cells represent the same ranges but the number in each cell represents how much more pulled balls outperform xwOBA than balls hit the other way. The formula is:

[Pulled wOBA – Pulled xwOBA] – [Oppo wOBA – Oppo xwOBA]

To help make this more clear, let’s go back to that 26-30, 96-100 cell, which shows .846 on this table. That .846 is calculated from:

[Pulled wOBA – Pulled xwOBA] – [Oppo wOBA – Oppo xwOBA] = [1.471-.685]-[.595-.655] = .786 – (-.060) = .846

So pulled balls in that range did have a slightly better xwOBA than opposite field balls in that range (.685 vs. .655), but the difference in results is because of how much those pulled balls outperformed their xwOBA, while balls the other way underperformed xwOBA.

Again, you can see a pattern consistent with the rest of the tables: when you elevate the ball and hit it hard, you want to pull it – hitters get much better results when they pull the ball, even if we control for EV and LA. But if you are not elevating the ball you actually want to go the other way. And if you are not hitting the ball hard, you lose any benefit from pulling the ball and may, in fact, benefit from going the other way.

If you could somehow teach a hitter to use all fields when they make weak contact or go the other way on low line drives, while pulling the ball anytime they hit it high and hard, you could create a hitter that regularly beats xwOBA. I am not sure you can actually teach a hitter to do that, but there may be some hitters who do that naturally. And it can definitely help you evaluate players.

In a piece over on PitcherList, Mark Steubinger noted that Jonathan India has benefited from using all fields. And India has been pulling the ball less, to better results, which feels counter to the narrative on pulling the ball to get to more power. But, as Steubinger notes, “one of the few knocks you can have on India’s game is that he doesn’t hit the ball particularly hard.” He also doesn’t elevate the ball much. And the tables above tell us that, for that player, pulling the ball more will hurt more than it will help.

On the other hand, I wrote yesterday about Lawrence Butler and his lack of pulled balls as a problem.

Of 331 hitters with 150 PA or more, Butler has the 28th highest oppo% and the 50th lowest pull%. His hard hit balls aren’t as valuable as they seem, because hard-hit pulled balls are more valuable than balls hit hard the other way.

Butler hits the ball a lot harder than India and has a much higher average launch angle, with far more fly balls. Given that, it isn’t a surprise his wOBA is lower than his xwOBA – he is regularly hitting hard-hit balls in the air the other way, which have historically underperformed Statcast’s xwOBA.

Finding these patterns in hitters profiles can help us identify how we should interpret their wOBA vs. xwOBA.