When we talk about ‘stuff,’ most often we’re talking about velocity. Or so posits Jack Moore at FanGraphs as he talks about Josh Johnson’s recent resurgence.

But even though we know from Mike Fast’s pre-Astros research that every mile per hour above 89 leads to 0.2 fewer runs allowed for the pitcher, that’s sort of boring. I mean, all we have to do is sort for velocity on the leaderboards and discover, hey, these guys are all good. Maybe you can an extra gold star to Jeff Samardzija, Matt Moore and Jason Hammel for discovering that they are in the top ten for fastball velocity, but then the velocities start to bunch up around 90-91 pretty fast.

How about STUFF. You know, filthy, nasty dirty stuff. Like the stuff that Yu Darvish, Stephen Strasburg and Aroldis Chapman own. That stuff.

Before he moved to a front office, Jeremy Greenhouse used to pursue this stuff. He used velocity, horizontal movement and vertical movement to create a run value for nastiness, and then he called it StuffRV. When he ran the numbers in 2009, A.J. Burnett, Felix Hernandez, Zack Greinke, Edwin Jackson and Ubaldo Jimenez made up the top five. Roy Halladay, Matt Garza and Brandon Morrow were in the top ten. Braden Looper, Livan Hernandez and an end-of-career Greg Madux brought up the rear. Stephen Strasburg was the only starter in the top 25 when Greenhouse ran the StuffRV numbers in 2010, as the top of the list featured relievers like Mariano Rivera, Henry Rodriguez, Daniel Bard and Brian Wilson.

That’s nice, but he didn’t give us the tools to run these numbers on our own. We’re still left wondering how to judge stuff best as we sort through the wire in our leagues.

Thanks to recent work by Adam Foster of Project Prospect, we have a single number to focus on: vertical movement. In a (free) guest post to Baseball Prospectus (that was linked in the URL Weaver post on this site), Foster buries the lead: vertical movement has about the same correlation with swinging strike rate as velocity does. Less vertical movement means more swinging strikes. We’ll call it rise here — because it’s the opposite of sink — but it just means that the ball doesn’t drop as much as other fastballs. It’s only ‘rise’ optically, since there’s gravity and all that.

Using vertical movement may not be as effective as using StuffRV, but it will produce easy results for fantasy managers here in 2012. Sticking with the fastball, since it is the most-thrown pitch in baseball, we can now create a handy table to try and identify some under-rated starting pitchers. Buy low time for any of these qualified starters in the top 25 for fastball ‘rise’?

ERA SIERA vFA (pfx) FA-Z (pfx) K% SwStr%
Brian Matusz 4.82 4.58 90.8 12.7 17.0% 7.4%
Clayton Kershaw 2.65 3.2 93 12.5 23.3% 10.0%
Wei-Yin Chen 3.68 4.18 90.5 12.2 17.5% 9.1%
J.A. Happ 4.54 3.65 90.4 12.2 23.2% 9.3%
Drew Smyly 3.96 3.57 91.2 11.4 21.8% 9.8%
Matt Moore 4.59 3.86 94.2 11.3 23.4% 11.8%
Cole Hamels 2.93 2.85 91 11.2 26.1% 12.9%
Carl Pavano 6 4.29 86.2 11.1 12.3% 6.1%
James McDonald 2.39 3.25 91.8 11 25.1% 9.4%
Brandon Beachy 1.98 3.96 91 10.9 20.5% 7.8%
Brandon Morrow 3.01 3.64 92.9 10.7 21.9% 8.5%
Bruce Chen 4.44 4.05 87.2 10.7 16.9% 7.1%
Jered Weaver 2.61 3.5 88.4 10.6 21.4% 6.8%
Justin Verlander 2.69 3.04 94.6 10.6 25.6% 11.5%
Cliff Lee 3.18 2.53 91.2 10.6 27.4% 8.3%
Jake Arrieta 6.32 3.65 93.3 10.6 20.5% 7.5%
Colby Lewis 3.13 3.41 88.1 10.5 20.7% 8.5%
Wandy Rodriguez 3.27 3.87 89.4 10.5 15.6% 7.6%
Phil Hughes 4.76 3.73 92.4 10.4 21.2% 7.9%
Bud Norris 4.81 3.36 91.9 10.3 25.1% 11.0%
Mat Latos 4.85 3.99 92.9 10.3 19.8% 8.6%
Tim Lincecum 6 3.92 90.2 10.3 23.4% 11.0%
Jason Hammel 3.22 3.47 93.6 10.1 22.7% 9.7%
Tommy Milone 3.77 4.3 87.1 10 14.7% 8.3%
James Shields 4.06 3.22 91.7 10 22.3% 10.4%

It’s good that we know that velocity is a part of this mythical stuff. Brian Matusz, Wei-Yin Chen, Bruce Chen and J.A. Happ take a ding for their lack of gas even if they’ve got nice rise. Then you have your no-duhs, the guys that have velocity and rise like Clayton Kershaw, Cole Hamels, Jered Weaver, Justin Verlander and Cliff Lee. Everyone likes them.

There are a few established ace-types that are struggling right now that look better in this light. They make for more obvious buy-lows, and you could hang your hat on this analysis if you needed to go after Tim Lincecum, Mat Latos or James Shields. You might throw Matt Moore in that category, too. They all have enough velocity and positive z-movement to be great. Lincecum is having a control problem, of course, and Latos seems to be missing something, but here they look better.

Maybe the most interesting names are the less established guys on this list. Drew Smyly has enough gas and vertical movement to believe in his above-average swinging strike rate. He could really pitch to his SIERA going forward, and might be on your wire now. Jake Arrietta just had a great game Wednesday night against a National League opponent, but you can see the pieces are there — some of the best rise and velocity on this list and a good SIERA to boot. James McDonald and Phil Hughes have pitched a little longer, but now there’s a reason to believe in their hot starts so far this year.

The search for a stuff metric that’s easy for us laymen to calculate, or is housed on a major statistical website, that search will just have to go on. In the meantime, we can use velocity and vertical movement to try and identify undervalued pitchers.

Good luck hunting.

[This post has been updated to reflect the fact that the larger numbers here mean less sink (more 'rise'). I confirmed with Adam Foster that less sink means more swinging strikes. The players and the analysis remain the same, it's just that their particular advantage in vertical movement comes from their lack of drop (and not from having MORE drop). Most likely, more drop means more ground balls. Dave Allen wrote a great post about how balls down in the zone lead to ground balls and up in the zone, they lead to strikeouts, which seems to work in tandem with Foster's findings. Be careful about the pfxz numbers -- on some sites (like FanGraphs) the higher number means less sink. On other sites, they use a negative scale to reflect gravity's work (check Adam Wainwright's Brooks Baseball page and see both the regular and gravity-adjusted graphs there), and there a less negative number means less sink. Either way, it seems that the 'bigger' number means more swinging strikes when it comes to fastballs according to Foster's findings. ]

Comments (4)

  1. I am puzzled by your terminology. Is FA-Z the vertical movement as measured by the PITCHf/x system? The table seems to be in descending order by this parameter. It a large number better than a small number? The reason I ask is that in the text you refer to “sink”, suggesting that that is a good thing. But a pitcher with “sink” would have a small vertical movement. Hopefully you can see my confusion.

    • Z is measuring vertical movement, FA represents fastball. It’s impossible for a fastball to actually rise, so any measurement is measuring sink.

      • I won’t try to parse the semantics of that. Instead, let me go back to my question and try to clarify. I realize that the PITCHf/x parameter known as “pfx_z” measures the vertical movement, with gravity removed. A positive value means it moves upward relative to a gravity-only pitch. Are the numbers in the FA-Z column actually negative numbers, so that 12.7 is really -12.7, which means it falls 12.7″ more than it would under gravity alone.

        My impression has always been that a sinking fastball has a positive pfx_z, but a smaller one than a non-sinking fastball. A fastball is almost always thrown with backspin. More backspin means a larger pfx_z; less backspin means a smaller pfx_z. In my simple way of thinking, a sinker is a fastball with less backspin (but still some backspin) than a regular 4S fastball.

        Now finally for my original question: Are you arguing that a larger FA-Z, as listed in the table, is “better” than one with a smaller FA-Z? Does larger represent more sink or less sink?

Leave a Reply

Your email address will not be published. Required fields are marked *