Part I: Curveball Velocity, Location, or Movement: What is more important?

The curveball is often used as an ‘out’ pitch. This implies either it is difficult to hit or is often taken for a called strike. I was interested in exploring both of those possibilities, and as such, I have decided to present research addressing both. Part I, presented below, addresses the questions of how difficult the curveball is to hit and what makes it difficult to hit.

Earlier this week, I shared some research about the relative importance of velocity, location, and movement with respects to major league fastballs. The approaches I used to answer the curveball problem were very similar to the approaches I described previously. Again, I used the 2011 MLB season as my dataset, and included only pitches to right handed batters. Since curveballs are thrown far less frequently than fastballs, this time I included both right and left handed pitchers to increase my sample size. Another reason I wanted to include lefties is I wanted to know if the direction of the horizontal break mattered.

Is a curveball more difficult to hit than a fastball?

In the 2011 MLB season, right-handed batters hit 1,749 home runs off fastballs and swung and missed on 16,622. In other words, once a batter had decided he was going to take a swing, if the pitch was a fastball he was nearly 10x more likely to swing and miss than hit a home run (9.5 times to be more exact).

Also in 2011, right-handed batters hit 178 home runs off curveballs and swung and missed on 3,532. This translated into the batter being nearly 20x more likely to miss on the curveball once they decided to swing (19.8 times to be more exact). So, if we simply compare the likelihood of swinging and missing versus hitting a home run between fastballs and curveballs, we can say that in this respect, curveballs are twice as difficult to hit than fastballs.

Therefore, the answer to this question is a resounding, YES!

What makes a curveball difficult to hit?

Not all curveballs are alike. They can be thrown at varying velocities, locations, and with varying amounts of movement. This section will look at each variable to try and see what the relationship is between each variable and difficulty to hit. Difficulty to hit will again be measured using a ratio between swinging strikes (SS) and home runs (HR). A higher ratio means more difficult to hit.

In terms of velocity, Figure 1, it appears as though curveballs thrown over a range of 70-82 mph are about 20x more likely to be swinging strikes (ie. average). Only when a curveball is thrown around 84 mph is it considerably more likely to be a swinging strike (60x). This means that it probably doesn’t matter how hard a pitcher throws his curveball (between 70-82 mph), unless he can somehow throw it 84 or more and still get enough movement on it to still be classified a curve.

Figure 1 – Swinging strike (SS) to Home run (HR) rate broken down by velocity

Much like with fastballs, location again seems to have the biggest impact on how difficult a curveball is to hit, Figure 2. Curveballs on both the inside and outside part of the plate are considerably more difficult to hit, with inside curveballs (SS:HR = 51) about 7x more difficult than curveballs down the middle (SS:HR = 7)  and outside curveballs (SS:HR = 108) about 14x more difficult. In terms of vertical location, curveballs thrown in the bottom of the zone were 130x more likely to be swinging strikes than home runs. Curveballs left up in the zone (aka hanging curveballs) are only 3x more likely to be swinging strikes. To me, this emphasizes the point of how dangerous a hanging curveball really is. For every three hanging curveballs missed, one is hit for a home run.

Figure 2 – Swinging strike (SS) to Home run (HR) rate broken down by horizontal and vertical location. For the horizontal direction, zero indicates the middle of home plate and negative numbers are inside to a right-handed hitter. For the vertical direction, zero is the ground.

Movement also seems to have an effect on how difficult a curveball is to hit, but not nearly to the same extent as location, Figure 3. What I found interesting here is that curveballs that only broke 2 inches either in or out to right handed batters were most likely to be swung on and missed. Intuitively, I would expect no horizontal movement to be easiest to hit, with difficulty increasing as absolute horizontal movement increased. This was not the case. Instead, it looks like in terms of horizontal movement, a below average curveball (2 inches of movement) is slightly more difficult to hit than an average or above average curve. The exception here is of course a terrific curveball with about 12 inches of tailing movement is also more difficult to hit. In terms of vertical movement, it didn’t seem to matter much, unless the curveball dropped by at least 10 inches.

Figure 3 – Swinging strike (SS) to Home run (HR) rate broken down by horizontal and vertical movement. For the horizontal direction, negative numbers are inside to a right-handed hitter.

To summarize this bit of research on curveball velocity, location, and movement, I would say that it appears as though location is far and away the most important thing when it comes to determining what makes a curveball difficult to hit. Curveballs down in the zone are much more difficult to hit. Curveballs up in the zone get hit out at an alarming rate. Curveballs can be effective both on the inside and outside parts of the plate. Curveballs with more movement are not necessarily more difficult to hit. My theory here is, perhaps curveballs with more movement are more difficult to locate. And since location is shown to be extremely important, maybe not being able to locate is what makes curveballs with more movement somewhat less effective.

Finally, I know me saying that a curveball has to stay down to be effective isn’t news to anybody. But, I think putting some numbers to show just how important it is to keep curveballs down emphasizes this point even more so. Also, this shows a pitcher doesn’t need a great curveball to be effective, he just needs a curveball he can locate.

Next up in Part II, I can plan to look at curveballs from a called strike point of view to see what types of curveballs batters are more likely to take for called strikes. Stay tuned!

Newest Most Voted
Inline Feedbacks
View all comments
10 years ago

Cool stuff. A few questions

Are you sure you aren’t reading too much into the 84 mph velocity issue? Is there any chance some of those may actually be misclassified sliders?

Also, did you consider summarizing for all pitchers and using a multiple regression to determine which effects (and interactions) are significant? For horizonatal location / movement, you could use the absolute values to account for differences between left- and right-handers.

If you used a multiple regression, you could also use relative importance metrics to tease out which terms are not just statistically significant but are actually meaningful, which could be interesting to see as well.

I like what you’ve done here and hope to see more soon.

10 years ago

Nice research done here. Short question: where did you get data about pitch types?
For instance here at Fansgraph i take some random batter, here goes his play log for 2011:

There is no data about pitches. I can only have it for the whole game:

But that’s not enough to analyze! The fact he batted 79% fastballs 2011-09-28 @LAA doesn’t say anything about HRs and swinging strikes among those. I’m totally lost here. Maybe that stats is available at fangraphs+?

I would much appreciate if you could clarify this.

james wilson
10 years ago

The right hander’s more horizontally breaking curve is effective to right handed hitters when well located, and meat when thrown to left hander’s–it has no intimidation effect and is coming right into their power. Which is why you do not see it thrown much to lefties. The vertical curve is effective to either side because a vertically breaking pitch is not in the plane of a normal swing and is therefore awkward to hit, unless it is inside and can be golfed. Which as you note it often is, about 140 yards.

Still and all, a pitcher may make his living with 60-90% fastballs, but the same could never be said of curveballs. Familiarity will eventually get you creamed.

9 years ago


I’d think that either 1) using absolute value of horizontal movement or 2) separating pitchers out by handedness should go a long way to solving the multiple regression problem.

9 years ago

Wondering why you have completely ignored called strikes. Seems like half the power of the pitch might come from there and yet success might not stem from the same factors as it does for swinging strikes.

9 years ago

… because that is Part II… my apologies…