Thursday, November 17, 2005

A neat little observation over at U.S.S. Mariner:

"One of the neat little things I’ve discovered recently that I’m shocked I didn’t know earlier is that on a team basis, runs scored are basically equal to team OPS. Seriously, it’s that simple. If you have a .750 OPS, you’ll score somewhere in the neighborhood of 750 runs. It’s not exactly 1:1 (the actual factor for 2005 was 99.2), but it’s darn close. How close? Last year, the M’s had a .703 OPS and scored 699 runs. That’s pinpoint accuracy from what should be a rough tool. And, it’s right for every team in baseball within 10 percent. The other factors besides OPS that go into run scoring (baserunning, clutch hitting, random luck) account for 10 percent variance on either side. So, this hypothetical .750 OPS team would be expected to score 750 runs, with the actual range being 675-825. We think they’ll score around 750, but we know they won’t score less than 675 or more than 825. It might seem like a big range, but dealing with realms of possibility rather than trying to predict what will happen is a much better way to get an actual view of likely outcomes."

Let's test this with the NL Central:

Cards, 805 runs, .762 OPS
Astros, 693, .730
Cubs, 703, .764
Brewers, 726, .754
Reds, 820, .785
Pirates, 680, .723

Well, it's not that close, but if you run a test for linear regression on that data, you get a correlation coefficient of 0.66. This means that 66% of the variation in team's runs scored can be 'explained' by differences in team OPS. Kinda wordy, but here's the short of it - OPS is an excellent predictor of runs scored. I admit that it's not as great a predictor as I thought (at least for the NL Central - I might run the data for all of baseball later), but it's still probably the best we have.

Kinda neat, huh? With that in mind, let's look at a projected '06 Astros lineup and my best guess at their OPS:

The No-Moves Team:

Ausmus - .650
Bagwell - .800
Biggio - .750
Everett - .700
Ensberg - . 900
Lane - .850
Taveras - .700
Berkman - .900
Pitcher - .400

Team OPS (I know this ignores the bench, but let's go with it for now): .739

I think those projections are reasonable; obviously, Bagwell is the hard one to predict, but I'm calling for Ensberg and Berkman to still be excellent, Lane to improve, and Ausmus / Everett / Taveras to all still pretty much suck.

The Trade-for-Estrada and sign Nomar-to-a-short-deal Team:

Estrada - .750
Bagwell - .800
Biggio - .750
Nomah - .800
Ensberg - .900
Lane - .850
Taveras -.700
Berkman -.900
Pitcher - .400

Mmm. We're just a trade for Estrada and a 2 year / 12 million dollar deal to Nomar away from a team OPS of .761. This might not seem like a huge difference, but it could be, particularly since our pitching cannot possibly be as good as last year's. 22 more runs Anyway, something to think about. This lineup is also much more balanced, with no Ausmus / Everett / P to kill rallies like last year.

Oh, and I don't understand why we let the Reds claim Mike Burns off waivers. That's two guys we lost this year(Todd Self and Burns) that would have helped the Astros a lot at the league minimum. Bad moves.

1 comment:

Jason said...

Interesting discussion of OPS, although I disagree with the changes to the roster that would raise OPS, but I've written enough about that.

I hope you aren't losing too much sleep about Burns and Self. They are good players, but they aren't exactly irreplaceable. I think someone at OWA said it best:

I think what the Astros are saying is that all of these players are very replacable. Burns is okay but not stellar. Gipson is fairly long in the tooth and not an offensive powerhouse. Self has shown limited power for a guy whose "natural" position is first base. I'm sure all are well liked and that in a perfect world the Astros would like to retain their services, but sometimes it's a bit of a cattle call when it comes to personnel decisions. They got cut out of the herd.