Date: Tue, 15 Feb 2000 07:27:22 -0600
Reply-To: Texas Rangers MLB Discussion <RANGERS@TAMU.EDU>
Sender: Texas Rangers MLB Discussion <RANGERS@TAMU.EDU>
From: Beau Sharbrough <beau@CONNECT.NET>
Subject: Melvin: Bang for the buck

GaryC can speak for himself on this issue -- I won't pretend to do so.

However, I can recognize it when a lawyer says that he wants to avoid an issue entirely for reasons that have nothing to do with that issue. As Sam Irvin said, when they've got the law, pound the evidence. When they've got the evidence, pound the law. When they've got the law and the evidence, pound the table. I can't tell for sure, Jamey, whether you're pounding the law or the table here. I'm sure that you're avoiding a discussion of the evidence.

Let me ask as nicely as I can for us to try to have a discussion on the meaning of these measurements. There are lies, damned lies, and statistics. It's clear to me that they can be used to reach darn near any conclusion that a guy wants to.

WHAT THAT GUY SAID

This particular statistic says that when you compare wins to payroll over the last few years, the Rangers come out right in the middle of the pack. The author of the report wanted to interpret that number as a measure of the ability of the GM and the on-the-field-manager to get the most bang for the buck.

It makes some sense, having noticed that for the most part teams with high payrolls get into the playoffs and teams with low payrolls don't -- that we've witnessed the revival of the first and second division in major league baseball. So to statistically correlate wins with payroll is interesting.

Before I get too carried away, I want to grant to everyone that the correlation wasn't all that great. If you've seen regression analyses before, correlation numbers go from -1 to +1. A correlation of .5 isn't the kind of close correlation that would make me feel like I'd plugged right into the oracle of delphi or anything.

But the increase in correlation over the period interested me. The point that the writer made was that from year to year, spending increasingly explains on the field success. I wonder if there is some truth in it, and what causes bring about this effect. Understanding an effect doesn't automatically lead me to understanding of the cause. Weather is a shining example of this.

Last gripe about stats -- why use a linear regression? Why broadcast the fact that you're stuck in linear thinking? How about non-linear regression? Something that says that the rich get richer. Oh well, stats are useful, but nothing like an encounter with a burning bush that talks.

DOES IT MAKE ANY SENSE TO US?

So I think that I can understand Jamey's reluctance to paint Melvin's portrait based solely on this regression analysis. Okay, that's fair. Statistics have a way of making it look like Junior is a better player than Barry Bonds and I know that's wrong <g>. But what does this stat tell us?

It tells us that compared to the other teams, Texas is not winning more games with cheaper players. We don't have the young players who have arrived and made a big splash like they do at Montreal and Cincy. We don't have Andruw Jones getting about $10,000,000 a year less than his market value. I'm sure I can tell why Atlanta is so well rated -- they've just about got the cheapest offense in baseball -- only the Falcons were worse this year -- and the money that they spent on pitchers has paid off well for them.

Let's look at our Rangers. Do we have any bargains who have helped us win a lot of games in the past few years? Let's see, Johnny Oates playing a young player every day, who would that be? Tatis? We've had a high-priced first baseman, catcher, shortstop, right fielder, and relief pitcher. Our starters have been making less money than Atlanta's, but we're spending a lot more on everyday players than they are, and without their great rotation, we don't get the bang for the buck that they get. I know that we spent more at just about every single everyday position than Atlanta last year.

And the Yankees? There wasn't more than ten games difference in the number of wins purchased for NYY and TEX. Does the list think that since the Yanks are so close to the predicted values that Torre isn't a good manager and that Bull Watson wasn't a good GM?

All these numbers are saying is that we haven't made deals that got us cheap players that beat expensive players. It is clear from looking at this analysis that if we were to put five rookies making near the league minimum and win about 75 games that this stat would say that Melvin-Oates were really getting their money's worth, like Montreal (who won 365 games in 5 years). Maybe they'd be, but the fans wouldn't.

Jamey's analysis of baseball usually is an educated look at the tools of the players and the way that they perform when it's time to step up. I enjoy it. I think that statistics don't measure that, and I think that's part of the reason that Jamey rejects them as the sole measure of a man.

I like them because they help to reveal insight, not because they reduce a man to a result of a formula. And the stats in this case say that Melvin hasn't gotten any more major league wins per dollar of payroll over the last five years than the major league average. I can accept that. I'm certain that it's true.

I believe, and I think that Jamey does, too, that Doug Melvin's biggest success since coming to Texas has not been seen in tBiA, but instead has been seen in OKC and Tulsa and Port Charlotte. It's hard to make big changes in a major league team overnight.

I've hard from a talking burning bush that starting this year, we're going to begin to see the future that Doug Melvin brought to the Rangers -- when several young players will play everyday. Over the next few seasons, we'll see some young arms like we haven't seen here since 1986. If you run that same regression analysis for the next five years that Wm Woodward ran for the last five, the Rangers will probably make Melvin and whoever he gets to replace Laiste look pretty good.

I don't know if Gary and Jamey will be looking at that regression analysis and agreeing at that point. I know that people will be wondering how we won so many games without spending more. I see a system that has as much potential to make stars as Montreal has been showing. And hey, guys, we're in the first division!

Geez, where did THAT post come from?

GJsneaker@AOL.COM wrote:

> GaryC ripped Doug Melvin and told me I vastly overrate him, and supported his
> position as follows:
> [snipped here]
> Gary, your posts are always insightful. Please take your shot against Doug
> Melvin using something other than a sabermetric analysis. I can't believe
> you actually base your position (that I overrate Melvin) on a *formula.*
>
> Jamey