THE BOOK--Playing The Percentages In Baseball

Would the GB advantage disappear if the Infield fly rule were revoked?  I think they should do away with it.  Why should a pop up be a better result with a man on 1st than a hard ground ball?

Far as I can tell from sites like fangraphs, we only have GB data back through 2002.  So we’re dealing with a pretty small sample size.  (Who knows, maybe James hand-coded GB% for all pitchers well before then, but as far as I know we have less than a decade of public data on it).

Not to mention, even if we had data, there are far, far more high-fastballers than there are GB guys who can K something like 6.5+ per nine IP.  Seriously, how many pitchers have a GB rate above 50% and K more than 6.5 per 9?  Webb, Carpenter, Halladay… I don’t think there’s too many others. 

Couldn’t you say “they all either get hurt or lose the plate” about any type of pitcher?  I feel like that happens to everyone.

Bill is just really, really wrong here. 

Yes, Verlander, Santana, Johnson, and Pedro Martinez all lived off their high fastballs, racking up strikeouts along the way.  That works if you have a big fastball and can get it by hitters. 

On the other hand, we can point to Tim Hudson and Derek Lowe as evidence that the low strikeout, high groundball approach works great.  Obviously, guys like Halladay, Felix (who most certainly does NOT live off the high fastball), and Jimenez are doing just fine with their GB+K combos.

And then there’s the second tier types - Mark Buehrle, Jake Westbrook, Joel Pineiro, Aaron Cook, Paul Maholm… there are tons of GB pitchers who don’t lose command of the strike zone and succeed in spite of a lack of strikeouts. 

Hell, Pineiro simultaneously became an extreme groundball guy and an extreme no walk guy at the same time.

Dave beat me to it, but Felix is an example of a GB pitcher (career 56.3%, has had seasons as high as nearly 61%) who strikes out a lot of guys (8.11 K/9).  He throws the occasional high cheese because he has an upper 90s fastball, but he lives off good stuff lower in the zone - and these numbers show it.  He’s also done it consistently, and there is ZERO evidence that he will not be able to sustain this (absent injury, of course...knock on wood).

I think Felix, even if he’s always pitching down in the zone, can’t be discussed here, because he has one of the fastest FB in the league. 

The focus should really be on guys who have a 89-92 mph fastball, who are getting tons of GB.  Bill doesn’t like those pitchers, and thinks they have a worse career arc than other kinds of pitchers.

It’s possible he is right, but that’s the kind of stuff that needs to be studied.  Are Brown, Halladay, Pettitte, Lowe, etc the exceptions to his rule, or, are they are representative as any other class of pitchers?

If we’re not going to discuss the GB guys with good fastballs, then we can throw out Bill’s examples of the flyball guys - Johnson, Verlander, and Sabathia are all upper end velocity guys too. 

Apples to apples, compare 88-92 MPH sinkerball guys to 88-92 MPH flyball guys, and I would bet my house that the sinkerball guys hold their own.

Don’t forget about Masterson.  He still walks a ton of people, but if he could develop a bit better control he could easily become a successful high K and high GB rate guy that is successful.

If we’re not going to discuss the GB guys with good fastballs, then we can throw out Bill’s examples of the flyball guys - Johnson, Verlander, and Sabathia are all upper end velocity guys too.

Apples to apples, compare 88-92 MPH sinkerball guys to 88-92 MPH flyball guys, and I would bet my house that the sinkerball guys hold their own.

You should be able to do this pretty easily with the data you have at Fangraphs, right?

A word of caution to anyone that tries to do this study - you need to use RA. Actual RA. FIP obviously doesn’t pick up on anything that groundballers do that isn’t tied into home runs. ERA has a ground ball pitcher bias due to the nature of unearned runs. Has to be RA. Absolutely, positively has to.

Yes, definitely RA.

And, David, it’s not so much which class of pitchers is better (clearly, the GB pitchers will eat the FB pitchers for lunch if we are controlling for fastball speed).

The issue on the table is if the James-identified class of pitchers have a worse aging curve than other class of pitchers.

So, if you start with pitchers who give up runs at say less than 90% of league average through age 28, who have a fastball speed of under 92, and at least half their balls in play are groundballs, does THAT group of pitcher age differently than we’d expect of other pitchers (if we only include the 90% and age 28 parameters)?

That is, does including the extra information of average to below-average fastball, and above average groundball give us more information?

@8/Colin Wyers: If that’s the case, I would suggest using tRA. I’m hoping that Fangraphs starts carrying it again; it’s my favorite DIP stat. It’s like the wOBA of pitching.

Well, tRA has some implicit assumptions - for instance, that the value of an additional ground ball is linear. Does that work for extreme ground ball pitchers? I dunno. I can give you my guess (probably) but why worry about it?

The problems tRA is meant to solve are problems in assigning value to individual pitchers, not measuring populations of pitchers. When doing a population study, I think the right idea is to be as unobtrusive as possible when measuring, and let the data be expressive.

Colin is right-on here.  We either way to use runs allowed, or sum of WPA/LI*boLI (which is similar to what runs allowed is).  If we were talking about relievers, I’d favor WPA/LI*boLI.  For starting pitchers, it’s pretty much the same thing as runs allowed.

We don’t want to give static values to various events, as if a Roy Halladay single is the same as any other pitcher, nor do we want to presume that the sequencing of events is a non-existant skill.

I really hope that someone does a study on this. I would love to see the results.

I disagree that you should compare the 88-92 MPH sinkerballers to the population of all pitchers.  If Bill wants to say that he doesn’t trust pitchers with fringe stuff as much as guys with good stuff, well, yeah. 

His claim appears to be that he’d take an equivalent 88-92 MPH flyball guy over an 88-92 MPH sinkerball guy.  It’s Tim Hudson vs Jered Weaver, essentially. 

I think you have to just throw out all of stuff guys, both GB and FB variety, and that requires not using a generic aging curve.

Among starting pitchers in the PITCHf/x era, here’s who’s lived most at the top of the zone with a rising fastball:
Roy Oswalt
John Maine
Pedro Martinez
Daisuke Matsuzaka
Wandy Rodriguez
Oliver Perez
Tim Redding
Jonathan Sanchez
Randy Wolf
Javier Vazquez
Josh Beckett
Rich Harden
Clayton Kershaw
Scott Kazmir
Max Scherzer

And here’s who’s lived most toward the bottom of the zone with a sinking fastball:
Derek Lowe
Aaron Cook
Chris Volstad
Brandon Webb
Chein-Ming Wang
Jake Westbrook
Roy Halladay
Jamie Moyer
Nick Blackburn
Trevor Cahill
Kyle Kendrick
Jon Garland
Braden Looper
Zach Duke
Joel Pineiro

I just checked BB-Ref to check my memory, and it’s right: Clemens has a slightly above average career GB/FB ratio.

In any event, if I had one of those guys on my staff, I wouldn’t complain one bit.

Oh, for whoever does the study: Bill is also saying the groundballers have a higher chance of injury. If this is so, this would be the first time I’ve heard of a verifiable class of pitchers who are more injury-prone than another class (aside from those pitchers who have been hurt in the past being more injury-prone than those that have not). That might be an even bigger finding.

Off the top of my head, Dave, the 88-92 fly ball guys are likely to have more Ks than the 88-92 GB guys, and I would expect them to be better pitchers over all - more pop ups, fewer HR/FB.

Why?

Because - and again, this is just rampant speculation on my part - I expect those guys to have an above-average curveball.

Some of the other names that were mentioned by Bill James or others fall in between, as follows, listed from high heat to low sinkers:

Jered Weaver ~20th percentile
Johan Santana ~25th percentile
Mark Buehrle ~40th percentile
Justin Verlander ~45th percentile
Randy Johnson ~55th percentile
C.C. Sabathia ~65th percentile
Tim Hudson ~75th percentile
Felix Hernandez ~80th percentile
Andy Pettitte ~80th percentile
Adam Wainwright ~85th percentile
Paul Maholm ~90th percentile
Chris Carpenter ~90th percentile

Off the top of my head, Dave, the 88-92 fly ball guys are likely to have more Ks than the 88-92 GB guys, and I would expect them to be better pitchers over all - more pop ups, fewer HR/FB.

Sure - that’s the trade-off.  Pitch up in the zone for strikeouts or down in the zone for groundballs.  Every pitcher makes that choice.  Some go one way, others the other way.  Some start off as one type of pitcher, then change course mid-career (Pineiro being a prime example). 

I can’t think of any reason why the guys who choose to pitch up in the zone would have better curveballs, other than if you’re guessing that guys who throw high fastballs do so to create separation from their low breaking balls, but I haven’t seen any evidence that’s the case, nor have I ever even heard it suggested anecdotally.

Oh, for whoever does the study: Bill is also saying the groundballers have a higher chance of injury. If this is so, this would be the first time I’ve heard of a verifiable class of pitchers who are more injury-prone than another class (aside from those pitchers who have been hurt in the past being more injury-prone than those that have not). That might be an even bigger finding.

Could this be associated with the idea that most ground ball pitchers tend to throw sliders rather than curveballs? Since sliders have more injury history associated with them...?

Just a thought that crossed my mind while reading this.

Other than if you’re guessing that guys who throw high fastballs do so to create separation from their low breaking balls, but I haven’t seen any evidence that’s the case, nor have I ever even heard it suggested anecdotally.

Well there is this:

http://www.hardballtimes.com/main/article/pitch-sequence-high-fastball-then-curveball/

I can’t think of any reason why the guys who choose to pitch up in the zone would have better curveballs

That isn’t what I said, though. I said guys with pedestrian fastballs and high FB rates. Because, let’s face it - they’re in the majors for SOME reason, right?

All of the pitchers listed in the original post except for Randy Johnson had/have very effective changeups, or split-fingered fastballs. I’m not sure if that has anything to do with his reasoning regarding better career arcs, but many of the ground ball type-pitchers like Wainwright use a cutter in combination with the fastball as opposed to a changeup or split. If a guy throws 93 plus miles per hour and as he ages lose a few miles per hour on the fastball, the split or change should still be effective with a decent difference in mph between the two pitches. If a guy throwing 90 mph lose a few miles per hour on his fastball, there becomes very little difference between the sinking fastball and the cutter and he becomes Brian Bannister.

It could be I have no idea what I’m talking about, but the best I can figure on the injury thing is that it may take somewhat screwed up mechanics to make a ball sink.  Something funky about the way you release the ball makes it sink other than just the grip of the ball.  This may especially be true for someone whose sinking action is nastier than someone else’s.  But it also makes sense that to make a ball rise or at least go up in the zone more than normal, you may also have to have somewhat screwed up mechanics.

The way I’ve always seen it is that alot of sinkerballers open up too early and somewhat short arm the ball and thus make it sink more than normal.  They follow through too much, too soon.  They rush the delivery too much.  Meanwhile, pitchers who consistently throw up in the zone don’t always follow through that well, if at all.  They delay the delivery too long.  Sinkerballers release it with their shoulder and hips wide the hell open while high cheese guys release it too closed.  Sinkerballers seem to fight their elbows while high cheesers seem to fight their shoulder.  Just because of the mechanics of the release that causes the resultant movement, the sinkerballer is more likely to cause stress to his elbow, meanwhile the high cheese guy is likely to cause more stress to his shoulder.  Pitchers seem to recover better from Tommy John than they do torn labrums and torn rotator cuffs.

"turning the ball over” (pronation) to make it sink, or sink and run, can be stressful on the elbow.

I think what Bill James is saying is that “once Jeff Suppan loses his groundball rate, he’s through”.

To me, it’s obvious. The only reason why guys like Suppan or Pineiro remained in the league was because they were able to get grounders without walking guys. If they can’t do that, they can’t pitch. They’re nowhere near as talented as wainwright, carpenter, etc.

Guys that “turn to” inducing groundballs as their “primary goal” are doing so because it’s the only way they can remain in the league.

There aren’t very many (if any) “average stuff” groundballers that have long, successful careers.

RE: #25

I don’t think Bill James is comparing Suppan or Pineiro to those pitchers though. Anybody would rather have the pitchers he listed when compared to Suppan and Pineiro, so if that is what he is saying, he’s not really saying anything. He seems to be saying that if you have two statistically similar pitchers, but one is effective up in the strike zone and the other is a ground ball pitcher, he will take the high strike zone guy because that pitcher will have a better career and remain effective longer.