Sabermetrician in Exile

For whatever it is worth, in his seminal book on feature writing, Jon Franklin says he absolutely hates the whole flashback technique in a narrative feature. Actually said that a feature writer - a great one - might be able to pull it off once a career.

I don't know for sure that I agree with that, but then again, I don't have the first feature writing Pulitzer handed it sitting on my mantel, either, like Franklin does. He thinks it's a cheat that doesn't actually work. Although I even see a lot of movies do it now as a framing device - like "Saving Private Ryan," "Michael Clayton," "(500) Days of Summer," "Walk the Line," etc., etc.

 

Rob Neyer on McCracken, the piece and his own role in the popularization of DIPS:

http://espn.go.com/blog/sweetspot/post/_/id/6835/voros-mccracken-changed-the-game

 

Sounds more like Rob Neyer on Rob Neyer.

I'm not completely sold on DIPS.

I believe that it's the best way for us laymen to predict pitcher performance, because we lack a lot of the nitty-gritty mechanical details that pitching coaches and organizations do. But I'm actually pretty unconvinced that the ebb and flow is due to "luck." I think it is also evidence of the constant cat-and-mouse game between advance scouts and pitching, and hitting coaches.

 

It doesn't seem unusual to you that the cat-and-mouse game is so perfectly balanced that the fluctuations perfectly mirror what we'd expect from luck?

 

It doesn't perfectly mirror it, though.

Even the Passan article talked about how they've had to adjust the theory for ground ball pitchers, for example.

This is a concept I admittedly need to read more about. But to me, it seems logical that if you are giving up a lot of home runs, for example, you may also be giving up a lot of shots in the gap or through the infield holes.

Wonder if there is any correlation between home run rates and BABIP?

If McCracken's theory is true, then every second that pitching coaches are doing anything other than counseling pitchers on how to strike guys out and avoid walks and home runs is a second wasted.

 

Yes, they've found adjustments to it. Knuckleballers, for example, and GB/FB ratios. But once you make those adjustments, it's back to looking completely random. It's counterintuitive, like you said, but it's there.

The correlation between HR rates and BABIP is the opposite of what you'd think. High HR pitchers will tend to be fly-ball pitchers, and fly-ball pitchers will tend to have a slightly lower BABIP than ground-ball pitchers. Basically, ground balls are more likely to be hits than fly balls, and fly balls are more likely to be home runs.

 

McCracken should have sought or eventually settled for a writing platform like Neyer and Keith Law. Whether you appreciate their wonkery or not, those in this species have an ability to write and express their ideas well. Law has an outstanding blog that often seems to center more on cooking than baseball. McCracken tried innovative thinking and it took traction only within the inner circle. If not for Lewis, most would not know of his contribution. Have you ever heard even a vaguely mainstream reference to DIPS? Nobody knows the name of the poor sap who invented the paper clip, either.

And while I have never been one to shout down the use of a fat joke, there are a few references to his obesity that suggest it is an impediment to his finding meaningful work in baseball. There are no fat football coach "geniuses" on the unemployment lines.

 

Right, but I'm not comparing a GB pitcher to a FB pitcher. I'm saying, let's say that one year Cole Hamels gives up 1.2 HR/9 and the next year he gives up 0.7 HR/9. Taking the HR at-bats out of the equation, does a standard individual pitcher's BABIP fluctuate according to his HR/9 rate? Does his 2B/9 rate track with his HR rate?

You watch games. How often have you seen a guy getting lit up, where a home run is followed by an on the-nose single followed by another home run followed by a line drive in the gap?

A guy loses velocity, and it makes sense to me that not only will his K/9 rate be affected, but BABIP, too, because hard hit balls are more likely to be hits. They have to be.

 

What sort of good pitch doesn't fit into one of those two goals? Pitching to weak contact is avoiding walks and home runs. Pitching to miss bats is pitching for strikeouts.

 

No to the former, yes to the latter.

I've also seen guys get torched and have a series of hard-hit balls go right at fielders. And I've seen guys get dinked to death by balls that keep finding holes. "They have to be" just doesn't cut it, because it's true that they don't.

Here's a relevant article on the subject:

http://www.baseballprospectus.com/article.php?articleid=10113

It's very counterintuitive, but the best and worst pitchers give up line drives at the same rate.

A pitcher who gives up a lot of home runs isn't doing so because he gives up a lot of line drives. He does so because he's a fly-ball pitcher who gives up a lot of balls in play. Give MLB hitters a lot of chances to hit balls in the air toward the outfield, and a lot of them will end up on the other side of the fence, line drive or not.

 

It is our tendency as human beings to look for patterns and logic, even when there is none.

I don't know nearly enough about the mechanics of DIPS/FIP to have a strong opinion on them, but because all 30 major league organizations have adjusted their player analysis operations to account for this theory, I tend to believe that my eyes are probably lying to me when they say that a pattern exists.

As the story said, other than a few minor adjustments (for knuckleballers and for GB/FB pitchers) no one yet has been able to crack a serious flaw in McCracken's theory. A lot smarter people than me have tried to find that flaw, to no avail.

 

Humans are notoriously vulnerable to "patterns" in truly random data. If you're interested in a light read devoted to this subject, I highly recommend How We Know What Isn't So by Thomas Gilovich.