The Straightest Fastballs
We do have leaderboards for this sort of thing, but the problem is that if you’re looking at horizontal movement, PITCHf/x data for lefties and righties is very different. For example, the average four-seam fastball for a lefty goes 91 mph, but breaks +5.5 inches. The average four-seam fastball for a righty goes 92.2 mph and breaks -4.2 inches. So if you sorted one way or the other for horizontal fastball movement, you would just be getting the best lefties and righties.
So I did the work for you. Voila, the straightest fastballs in the game. It may be the least important facet of a fastball, behind location and velocity, but it’s still a facet we don’t talk about much.
Let’s do righties first:
| Pitcher | AVG(Pfx_x) | Velocity |
|---|---|---|
| Josh Fields | 1.72 | 94.08 |
| Tim Lincecum | 1.58 | 89.77 |
| Grant Balfour | 0.23 | 91.45 |
| Stephen Fife | 0.02 | 88.78 |
| Chad Bettis | -0.06 | 93.15 |
| David Robertson | -0.24 | 93.03 |
| Mike Bolsinger | -0.39 | 88.27 |
| Jarred Cosart | -0.47 | 93.70 |
| Anthony Bass | -0.50 | 94.06 |
| Sonny Gray | -0.53 | 93.15 |
| Brandon Workman | -0.54 | 90.80 |
| Chris Archer | -0.58 | 93.91 |
| AJ Ramos | -0.59 | 91.30 |
| Greg Holland | -0.63 | 95.76 |
| Preston Claiborne | -0.68 | 91.03 |
| Tommy Kahnle | -0.73 | 94.37 |
| John Axford | -0.74 | 94.01 |
| Shawn Tolleson | -0.74 | 91.58 |
| Jered Weaver | -0.74 | 86.20 |
| Dane De La Rosa | -0.98 | 89.78 |
Will you look at that. Tim Lincecum not only has no fastball command, but he owns the second-straightest fastball in baseball. That’s how you get the home runs. And for the largest part, this is a list full of pitchers that have bad fastballs, or fastballs that seem worse than their velocity. Anthony Bass throws 94! It’s straight. John Axford has homeritis — looks like he has the toxic command plus straight fastball problem that Lincecum has had since 2010, when his horizontal movement on his fastball died and never came back.
There are ways to beat the rap. It looks like velocity and command are those ways. Greg Holland is obviously all about the velocity. David Robertson? Great command for the last three years. Sonny Gray? Good command and good velocity — but will it hold when it’s only command and not great velocity? Jered Weaver suggests that command is more important than velocity when it comes to longevity with a straight fastball. Good news for Gray, bad news for Chris Archer and Jarred Cosart. There will be a perfect time to sell those guys in dynasties, and it might be soon. (Dunno why you’d own Cosart, but if you can sell him…)
On to the lefties.
| Pitcher | AVG(Pfx_x) | Velocity |
|---|---|---|
| Clayton Kershaw | 0.51 | 92.88 |
| Xavier Cedeno | 0.83 | 91.21 |
| Joe Savery | 1.90 | 90.95 |
| Mark Buehrle | 2.16 | 83.54 |
| Wade LeBlanc | 2.44 | 85.41 |
| Joe Thatcher | 2.46 | 85.85 |
| Brett Cecil | 2.50 | 91.67 |
| Drew Smyly | 2.51 | 89.87 |
| John Lannan | 2.56 | 87.91 |
| Zac Rosscup | 2.91 | 90.95 |
| Ryan Rowland-Smith | 2.98 | 87.25 |
| Charlie Leesman | 2.99 | 86.52 |
| Matt Thornton | 3.04 | 94.07 |
| Vidal Nuno | 3.10 | 88.93 |
| Jose Quintana | 3.12 | 91.11 |
| Tim Collins | 3.14 | 92.36 |
| Nick Hagadone | 3.16 | 93.84 |
| Jeff Beliveau | 3.21 | 89.50 |
| Lucas Luetge | 3.23 | 90.55 |
| Josh Edgin | 3.26 | 92.61 |
| Sean Marshall | 3.26 | 87.45 |
| Jaime Garcia | 3.27 | 90.46 |
| Felix Doubront | 3.28 | 89.93 |
| Justin Marks | 3.34 | 89.97 |
| Neal Cotts | 3.50 | 91.29 |
| Tyler Skaggs | 3.56 | 91.67 |
| Franklin Morales | 3.69 | 91.07 |
| Will Smith | 3.74 | 93.06 |
| Danny Duffy | 3.80 | 93.68 |
I wouldn’t be too worried about Clayton Kershaw being at the top of this list. We know he has great command of the pitch, and it still has good velocity. Mark Buehrle shows you how you can get by with a straight lefty fastball and great command. Drew Smyly doesn’t have a ton of velocity to begin with, but his natural command looks like it might be good. If it’s more league average than strong, though, he may not have a ton of shelf life. Jose Quintana has somehow been less than the sum of his parts, but he does get a 8% whiff rate on his fastball. He must have some deception. Jaime Garcia throws his sinker much more often and now it makes sense. Felix Doubront has no velocity, no command, and no movement on his four-seamer. Huh.
I let this list go a little longer because it had some interesting names at the bottom. Tyler Skaggs and Danny Duffy are dynasty assets right now, but will it hold if Skaggs sees his command suffer, or when Duffy’s velocity inevitably declines? With Danny Duffy not sporting a single pitch type with plus whiff rates and a history of bad command, there’s not a lot to get excited about long-term. Despite the short-term results.
Straight fastballs: if you’ve got one, you better have command and velocity or at least one. Otherwise, bending is better.
UPDATE: Here are the straightest most-thrown fastballs (min 300) —
| Pitcher | AVG(Pfx_x) | AVG(start_speed) |
|---|---|---|
| Josh Fields | 1.72 | 94.08 |
| Tim Lincecum | 1.58 | 89.77 |
| Jarred Cosart | -0.47 | 93.70 |
| Sonny Gray | -0.53 | 93.15 |
| Thomas Kahnle | -0.73 | 94.37 |
| John Axford | -0.74 | 94.01 |
| Shawn Tolleson | -0.74 | 91.58 |
| Tyler Thornburg | -1.13 | 93.54 |
| Yovani Gallardo | -1.25 | 90.96 |
| Brad Peacock | -1.44 | 92.14 |
| Jake Odorizzi | -1.59 | 90.34 |
| John Lackey | -1.61 | 92.01 |
| Chris Young | -1.73 | 85.30 |
| Zack Greinke | -1.75 | 91.16 |
| Anibal Sanchez | -1.99 | 91.81 |
| Josh Tomlin | -2.22 | 88.70 |
| Yu Darvish | -2.27 | 92.49 |
| Garrett Richards | -2.30 | 96.04 |
| Trevor Rosenthal | -2.42 | 96.35 |
| Chris Tillman | -2.43 | 90.51 |
With a phone full of pictures of pitchers' fingers, strange beers, and his two toddler sons, Eno Sarris can be found at the ballpark or a brewery most days. Read him here, writing about the A's or Giants at The Athletic, or about beer at October. Follow him on Twitter @enosarris if you can handle the sandwiches and inanity.
Is there any evidence (besides common sense) that horizontal movement is negatively correlated with probability that a fastball will be hit for a home run?
I’m thinking a logistic regression with pfx_x as the independent variable and 0 or 1 (for whether the pitch was hit for a home run) as the dependent variable, limited to fastballs (and perhaps also limited to pitches that were swung on).
Should have linked this, and will put it in now. Movement is the least important facet. I still found the lists interesting. This writer came to the same conclusion as I did, but with more numbers: velocity then location then movement. http://www.fangraphs.com/community/what-is-more-important-for-a-fastball-velocity-location-or-movement-2/
I found the lists really interesting too. Thanks for digging up the article man…
Thanks for the work! Would love to see it taken further, like breaking down where these pitches went in the zone (to back up the “better command” argument for the pitchers named) and what happens to these pitches (similar to that earlier article on examining the result of pitches down the middle). By combining the location and result, perhaps we would get a sense of who is actually better command and who is getting lucky? Final piece would be year-to-year correlation/aging curve to see if this is a skill pitchers normally improve? Common sense answers to all of this but FG is about putting numbers to test common thought. Plus, lots of Archer hate on FG continues (despite improving FIP, xFIP, SIERRA) so would be nice to see what the numbers give a 25yr old of chances of improving command or movement.
I am surprised Gausman isn’t here. I read recently on Fangraphs (I believe) and another site that his FB is really rather straight. Can you confirm that he wasnt missed for some reason?
Gausman might fall below whatever minimum IP the author used for the tables.
No IP limits, you see all the relievers in there. I can run it with a count to avoid the rare four-seamers, I’ll do that now. Gausman still has a straight four-seam fastball (-3), but it’s not as straight as these.
Do you have any proof that straight fastballs are less effective than tailing ones.
In the case of weaver and kershaw the lack of side spin gives their fastball more rise therefore making it harder to hit. Why do you assume fastball horizontal movement is good?
Not to mention that being the most extreme one way or the other shouldn’t really be considered “straight”. If batters are accustomed to having pitches tail a couple inches, then somebody like Kershaw who is far away from the average wouldn’t really be the easiest to square up. Likewise, the “worst” sinking fastball in terms of vertical drop could also be called the best rising fastball. Not technically rising obviously, but having apparent rise relative to what is expected. Pitch classification is a fickle thing. The best cutter is the worst two-seamer, etc.
Great piece. Not encouraging stats for me. I own Smyly and Duffy in a 12 team AL only league, and I own both Smyly and Quintana in a 20 team league.
Quintana is definitely doing something right, whether it’s deception or command or great pitch sequencing.
It seems important to note that at least Gray and Weaver (cherry picked big names) don’t throw the 4 seamer all that often, at only 35.8% and 16.9%, respectively. Instead, they throw the 2 seamer which has different break characteristics. It might be interesting to look at pitchers that feature both pitches prominently and how the sinking two seamer potentially keeps batters off guard to the straighter four seamer.
I question the value of the listing without also considering vertical break and the pitchers’ other fastball offerings. For example, Cosart throws a cut fastball (94 mph avg. velocity) at the same percentages as his 4 seam fastball. Cosart’s cut fastball has no horizontal movement, but breaks downward (at times broadcasters will mistake it for a downward breaking slider). Some pitch f/x data bases will misclassify Cosart’s cutter as a 4 seam fastball, because of the velocity, which might be troublesome for this type of comparison.
Remember when Tanaka was said to have too straight of a fastball to succeed?
Echoing what others have commented on, I don’t see that the lists are helpful without knowing whether a straight fastball is on average a bad thing. My sense was that a pitch that breaks differently than the league normal was a good thing, so if batters customarily see fastballs break to a moderate extent, perhaps a straight fastball differs from batters’ expectations.
Gray’s velocity is back to where it was in 2013 (93.2 MPH 2013, 93.0 MPH in 2014), and it has been ticking upwards since the season started, so I don’t think he’s a real velocity concern now.
One thing that’s interesting about that most-thrown list: I know for the knuckleball, velocity and movement are negatively correlated.
i’m not going to be selling archer. there’d be too many buyers to choose from.
To be fair, a big part of his success this year has been moving away from the straight four-seamer to his better two-seamer.
“Will you look at that. Tim Lincecum not only has no fastball command, but he owns the second-straightest fastball in baseball.”
I might be misunderstanding this entire post, but doesn’t Grant Balfour have the second-straightest, at 0.23? And doesn’t Stephen Fife have the straightest fastball of all, at 0.02? The table at the top is sorted from positive to negative, so Tim Lincecum is listed at second at 1.58, but a value closer to 0 is straighter, right?
Also, about Lincecum (and Balfour and Fields), how is that their break is the other direction? Isn’t that interesting? I’m confused
Eno, another insightful post about pitching. Thanks and keep it coming! Woudl this be a factor at all in why some guys consistently over/under achieve their fip/xfip?
David Robertson’s fastball has never been an issue with him even when he was famous for walking everyone. His long stride and release point has been the bigger factor in keeping him from getting hurt. Him learning the cutter and improving his command has just been icing on the cake