Offseason Notes: A Graph of Pitcher Height vs. GB%
Today’s edition of Offseason Notes is a graph.
Because I’m painfully disatisfied with the present, I spend a lot of my time thinking about the future. And because I spend a lot of my time thinking about the future, I spend a lot of time thinking about baseball prospects. And because I spend a lot of time thinking about baseball prospects, I spend a lot of time reading scouting reports of baseball prospects — which form of literature (i.e. the scouting report) I consider more important than most of what else is available in the entire Western Canon.
In reading scouting reports, I freqently see it suggested — as regards pitchers — that height is important because it allows a pitcher to throw on a “downhill plane.” Frequently, in those cases where a pitcher throws on a downfill plane, it is also suggested that doing so will allow said pitcher to induce ground balls (and, in turn, prevent home runs) with more frequency.
I do not intend to dispute the logic of this reasoning — nor to suggest at all that this is a statement made by every author of the literary form known as the scouting report — however, knowing that right-hander Chris Young both (a) is 6’10″ and (b) has a career ground-ball rate in the high-20%s, I grew curious.
Accordingly, I looked at both the the height and ground-ball rate of every pitcher with more than 150 batters faced (i.e. the sample threshold at which ground-ball rate becomes reliable), 2002-11.
Here is the result:
As one will note by the r-squared, there is basically no connection whatsoever between height and ground-ball rate (and some internet browsing reveals that David Gassko reached a similar conclusion at the Hardball Times in 2006). Of course, not every pitcher has the same angle of release: Tim Lincecum, for example, throws almost straight over the top, while Justin Masterson is the rare starter to throw from a sidearm angle. Generally speaking, however, while there might be some advantage to pitcher height and the ability to throw on a downhill plane, it isn’t showing up in ground-ball rates.
Thanks to Mr. Jeff Zimmerman for running the sweet query that provided the above date.
1
Could a similar analysis be done with height of pitch release? Not sure how reliable Pitch F/X data is w/ regard to that.
Agree. Release point would be a better indicator.
Also, I wonder if the data could be controlled for pitch-type. In other words, look exclusively at fastball and fastball release points. Off-speed pitches (at least I would conjecture) tend to be more flyball prone. Gotta remove/normalize pitch mix as a variable.
Uh, that’s not what R^2 means at all. R^2 is the fraction of the variation explained in GB% by height. If you want to conclude that height makes no difference, you’d first want to check to see if the coefficient on ‘x’ is statistically different from zero. Which it may well be. But a large R^2 would be very strange to see here, because there are lots of other things that account for GB% and you’d need to start including those to get a large R^2.
That’s ignoring the fact that maybe taller pitchers have an easier road to the majors because they can induce more ground balls with lower-quality ‘stuff’ than shorter pitchers. If that’s true, and I have no idea whether it is or not, the whole exercise is meaningless because you have an endogeneity problem, and your estimated coefficient is significantly biased, and only a lower bound for the effect of height.
I agree, you can’t make that conclusion, especially when you don’t control for other variables that affect ground ball percentage.
I typed up a more detailed response earlier, but it didn’t go through when I hit submit. :(
Andrew, while I agree that Carson doesn’t use the best language to describe the effect he’s seeing, his graph still supports the point of this article. He is not trying to show that tall people and short people in general induce the same amount of ground balls. He is trying to dispel the notion that tall pitching prospects have higher ground ball rates than short pitching prospects.
The fact that there are so many other variables that account for GB% is exactly what he showed. Height alone is not a good predictor of future ground ball rates.
Furthermore, he shouldn’t have any sample selection bias because he’s sampling the whole population. Well, technically, he’s sampling the MLB pitchers instead of pitching prospects, but I’d be surprised if the two plots were vastly different.
The selection bias would occur years before pitchers make the majors. If it’s easier to get grown balls when you’re taller, taller pitchers will be more likely to be good enough when they’re 16 to draw the attention that, one day, if they’re lucky, will let them into this sample. But of course even shorter pitchers who make it to the big leagues will be able to induce grounders almost as well as their taller brethren, BECAUSE OTHERWISE THEY WOULDN’T MAKE IT TO THE MAJORS.
Anybody know who the 5’6″ pitcher is? Tim Collins?
Might be Danny Ray Herrera. He’s puny.
I thought downhill plane isn’t about inducing ground balls as much as it is about inducing swinging strikes. I vaguely recall a very good BP article that looked at swinging strike % versus vertical velocity in the strike zone. Essentially, the quicker the ball ‘drops’ thru the zone, the higher the swingnig strike rate.
Cheers to the author for actually being cognizant of the fact that an extremely low correlation (r^2) means that there is very little that can be inferred from the data. Wayyy too many Fangraphs articles making judgements based on tiny correlation values.
I read a study some place in the back of beyond that found that the release point of tall pitchers was closer to the plate and the ball time in flight was thus shorter and optically to hitters appeared to be faster by 1-2 mph.
What we observed is that pitches released closer to home plate get to the plate faster. While tall pitchers generally tend to extend further towards home plate, pitcher height is only part of the equation.
Josh Orenstein
TrackMan, Inc.
http://sportsillustrated.cnn.com/2011/writers/tom_verducci/04/12/fastballs.trackman/index.html
I still think it makes sense that taller MLB pitchers should have the greatest potential for GB%. But this does not mean that all MLB pitchers pitch in a way that maximizes their assumed GB potential according to height. There is a selective sample issue with this method, but what if you worked backwards and looked at the top 30 GB% hurlers and checked their average height? Wouldn’t this provide some evidence to confirm or reject the hypothesis?
Top 30 GB% 2007-2011 (min 400 IP)
1. Brandon Webb 6’3″
2. Tim Hudson 6’1″
3. Derek Lowe 6’1″
4. Aaroon Cook 6’4″
5. Fausto Carmona (Roberto Heredia) 6’4″
6. Jake Westbrook 6’3″
7. Justin Masterson 6’6″
8. Charlie Morton 6’5″
9. Ricky Romero 6’0″
10. John Lannan 6’4″
11. Felix Hernandez 6’3″
12. Joel Pinero 6’0″
13. Trevor Cahill 6’4″
14. Roy Halladay 6’6″
15. Paul Maholm 6’2″
16. Jason Marquis 6’1″
17. Rick Porcello 6’5″
18. Chris Carpenter 6’6″
19. Chris Volstad 6’8″
20. Ubaldo Jimenez 6’4″
21. C.J. Wilson 6’1″
22. Sean Marshall 6’7″
23. R.A. Dickey 6’2″
24. Clay Buchholz 6’3″
25. Zach Duke 6’2″
26. Mike Pelfrey 6’7″
27. Dontrelle Willis 6’4″
28. Adam Wainwright 6’7″
29. Hiroki Kuroda 6’1″
30. Manny Parra 6’3″
The average height is 6’3.6″ (75.6 inches). That’s a pretty tall bunch of dudes. But MLB pitchers tend to be… the average height is just around 6’3″. So what’s that telling us? And are things skewed just because we’re looking at 400+ innings 2007-2011 (taller pitchers are more likely to stick for 400 innings, better pitchers are more likely to stick for 400 innings and better pitchers are more likely to get ground balls)?
Let’s compare to a random sample of 30 others from this group (every 5th pitcher adjusted by Pace from PitchF/X “Plate Discipline”)
1. Tim Wakefield 6’2″
2. Dallas Braden 6’1″
3. Anibal Sanchez 6’0″
4. Gio Gonzalez 5’11″
5. Jake Peavy 6’1″
6. Barry Zito 6’2″
7. Brett Myers 6’4″
8. Colby Lewis 6’4″
9. Randy Wolf 6’0″
10. Braden Looper 6’3″
11. Justin Verlander 6’5″
12. David Price 6’6″
13. Clayton Kershaw 6’3″
14. Max Scherzer 6’3″
15. Mat Latos 6’6″
16. Jered Weaver 6’7″
17. Zack Greinke 6’2″
18. Andy Pettitte 6’5″
19. Micah Owings 6’5″
20. Kyle Kendrick 6’3″
21. Ian Kennedy 6’0″
22. Jason Hammel 6’6″
23. Doug Davis 6’4″
24. Jeff Suppan 6’2″
25. Javier Vazquez 6’2″
26. Phil Hughes 6’5″
27. Jorge De La Rosa 6’1″
28. Vicente Padilla 6’0″
29. Edinson Volquez 6’0″
30. Josh Becket 6’5″
Average height is 6’2.9″ (74.9 inches).
So again, like the graph, it points to a tiny positive relationship.