<$BlogRSDURL$>
Proceeds from the ads below will be donated to the Bob Wuesthoff scholarship fund.

Saturday, October 08, 2005

Stephen Smith, Your Table Is Ready: Management By Baseball On Angels Strategy

Management By Baseball has a humdinger of a column up about the Angels' offense with runners in scoring position and especially with runners in scoring position with two outs (RISP2).
... [T]he Yankees follow informed baseball wisdom, while the Angels violate it, at least on the offense side.

The truth is not well known. Perhaps I should say I didn't know it before I was given a generous chunk of time with their skipper, Mike Scioscia. Scioscia, in case you doubted it, has little love for sabermetrics types. He's very courteous about it, but he's equally opinionated. It's not that the Angels reject the numbers -- far from it. They use numbers few others do, to their competitive advantage.

...

Scioscia, as it turns out, loves on-base percentage. It's just that he doesn't have it on his roster. The Angels don't reject on-base average as something valuable. They just don't have an abundance of it, and therefore try to distill and preserve every iota of value out of the amount they do have of it. The theory being that if you can get a runner on first in any manner, get the runner into scoring position in any manner, optimal or not. Get runners moving with run-and-hit, steal bases, shake up the defense.

...

The numbers the Angel management are passionate about, and which may be the factor that provides them an affordance for success in lieu of on-base, are batting w/runners in scoring position (RISP) and batting with runners in scoring position with 2 out (RISP2). In these, they excel, compared to the American League and compared to the Yankees. And they especially excel against the way the produce at the plate relative to their other at bats.

...

This won't be a popular thing to say, but the Angel approach on the batting side is pure Moneyball. That is, the economics side of the argument, not the specific attributes the Athletics front office found undervalued. If multiple mid-budget teams are pursuing high-OB guys and steering away from the speedy contact hitters, there will develop an overlay in speedy contact hitters. There will develop and overlay in talent who aren't so speedy but are trainable.
Via BTF.

Comments:
read this as well, as well as some of the discussion about it on Baseball Think Factory. If this is a "new strategy" that the sabermetricians haven't really taken into account, it will be interesting to see how they deal with it. They could blow it off as complete luck (which is done by some of the guys on BTF) or they could attempt to quantify it by looking at certain stats in a different way and making it closer to a Universal Baseball equation. I have to say, though, that the people that blow off the strategy as complete luck have got to considered as not good statisticians until at least some new statistical methods have been tried.
 
My main problem with this is that if there's an actual skill to being successful with RISP, then why can't it be applied to other situations?

If the Angels hit .290 with RISP and .260 without (just making these numbers up), why don't they just hit like they do with RISP all the time?
 
Well, I guess the question really becomes: why should there be any difference between RISP and normal situations? Well, in order to answer that question, you need to know what is different. When runners are in scoring position, the tone of at bats is different, a pitcher changes what pitches he throws, a batter may swing differently, there is more pressure on both of those guys, and the defense will play differently. Basically, the runners have to affect play in some fashion, but it is not known to what extent or exactly how.

My guess is that the Angels organization have tried to discover it (when others have completely ignored it) and have some system they think works. High contact guys are going to tend to be better with RISP than high walk, high K guys (with respect to each player's normal results) just because a strike out is particularly bad with RISP (without, it is just another out), and a walk doesn't really help unless the bases are loaded (whereas walks without RISP get people on so that later hitters can knock them in). Like someone said in the BTF comments, you really have to look at how they do relative to the average change. If every team is boosted by .015 points with RISP, then the Angels improvement in andrew's example would only really be .015 OVER the league average improvement...still likely significant, but not necessarily so ridiculous.
 
RE: He could thank Bud Black, instead of taking credit for some half-cocked, post facto conclusion regarding hitting with RISP, developed when Scioscia and Hatcher were sitting around the office trying to figure out how to justify their existence, and hitting upon some random number that supposedly does it.

In case I was unclear, Scioscia never took credit for devising it; I get the sense that the organization (perhaps Black & Scioscia included, perhaps before they arrived) worked it out several years back since they've been developing and training players against that for some years. This is basically counter to the way Scioscia batted himself; he went deeper into counts as a rule to walk more frequently than league averages -- it would be abnormal, though not impossible, for someone to install a set of tactics that went against their own pattern that succeeded reasonably well.

RE: if there's an actual skill to being successful with RISP, then why can't it be applied to other situations?

What josh said. Put a runner on first, a pitcher's pitch selection will be different. Make the count 2-0, change what the pitcher throws and where. A reasonably intelligent batter will have a better (still imperfect, just better) incremental chance of guessing correctly, and in turn this gives the batter a better (still imperfect, just better) incremental chance of doing something good with the pitch. So teaching about this situation *might* help a batter in other situations, but probably not -- just as learning to bunt well has helped some batters become better at their swinging, but usually it has no effect.
 
3-0 Hit and run and Molina at the plate.

Doubling stealing with Vlad up.

Letting Erstad have 600 at bats.

I know these are isolated incidents, but if this is what I remember, who knows what else he's done. Time to start justifying his existence.
 
Jeff Angus -- thanks for stopping by.

Andrew -- regardless of what Matt says, Scioscia does employ some stupid tactics from time to time. That said, a double steal in front of Vlad isn't necessarily a bad idea; he's principally a spray hitter, so it's more likely he'll get a hit than something over the fence.

Matt -- your response to Andrew sounds exactly like the kind of ex post facto reasoning Steve was referring to upthread. The problem here is that there have been so many attempts to find clutch hitting and all of them have failed. Sure, you can find guys who will give you some ghostly vanishing trace every so often, but nobody's been able to prove it. Show me some mathematical proof that this stuff translates to runs scored and that it's a true ability. Steve is right to be skeptical, but let us not dismiss it out of hand, either.
 
RE: a double steal in front of Vlad isn't necessarily a bad idea; he's principally a spray hitter, so it's more likely he'll get a hit than something over the fence.

Yes, I should add that in talking about having runners be aggressive and sending them on specific pitches, Scioscia said during the interview, "You’re going to get a hell of a lot of singles with a guy on first. No matter which team, you’re going to get a lot more singles than home runs. If you can get that guy to third instead of to second that’s a lot better statistical position to be in. If you can create more of those situations, you’re going to have more runs on the bottom line."

The 1996 season of Rob Deer is the only exception I can find, though I'm not sure that would be an indication I would bank on as normal.
 
Jesus, I guess my lousy writing skills have gotten me in trouble again.

Scioscia is peddling snake oil. He's looking at numbers and trying to come up with explanations for them.

Yes, in the absence of statistical proof. Maybe he has it, and we haven't seen it. But until he or someone else does, this amounts to the kind of "I have found a truly wonderful proof of this, but this interview is too short for me to explain it".

Matt -- I didn't mean you agreed with Scioscia's every move; we've been through a whole season together, so you'd think I knew your disagreements with Mike by now. And again I didn't mean to imply you were talking about clutch hitting, only that Scioscia seems to be and believing in same as an ability, thus flying in the face of everything we know about "clutch" as an ability.
 
I wouldn't be so quick with the "everything we know" card; there is some well-researched disagreement in the stathead community about whether certain players (Gary Sheffield, for example) consistently add value with their clutch hitting.

Yes, but every single study that I've read that makes any comment to this effect has to lard it with so many caveats it may as well say, "no clutch for you".

And I would think that, regardless of belief, a manager *should* be figuring out ways to optimize situational performance, even if that means blowing smoke up everyone's asses.

Now this I can get behind.

On one hand I guess I envy the conviction that all clutch is BS & any emphasis on Productive Outs is foolish.

When the PO people can show that their nonsense has a positive and strong correlation with runs scored, I will believe them. As it stands, it doesn't, and therefore I laugh at the superior intellect. (Or something.)
 

Post a Comment

Note: Only a member of this blog may post a comment.



Newer›  ‹Older
This page is powered by Blogger. Isn't yours?
Google

WWW 6-4-2