Wednesday, November 02, 2005

Baker Column Disproves Theory of Intelligent Design

Scientists worry evolution theory next to fall

There’s a lot of talk this week about Geoff Baker’s claim that the White Sox World Series win has single-handedly “disproved the Theory of Moneyball.”

It’s the sort of Morganesque statement that hardened stats junkies have lived with for years, and riffed with mockery or faux-aggravation.

Although Baker’s column is provocatively titled, he is too distracted or lazy to make anything but an inchoate argument. After eight-hundred words we’ve learned that a “champagne-doused” Ozzie Guillen thinks “Kenny Williams did a tremendous job,” but we still don’t know:
• What Baker claims the “Theory of Moneyball” is,
• Why the evidence has disproved its thesis, or
• Whether Geoff Baker has his own testable theory of baseball, other than his hackneyed claim that “there is more than one way to build a winner on a limited budget and that skills like bunting, stealing bases and defence all have their place.”

Our friends over at AN were already busy re-iterating why steals, small ball and defence aren’t the energizing, moisturizing, tantalizing, romanticizing, surprising, her-prizing, revitalizing tonic that their adherents claim.

We're going to spend our time goading Baker into a do-over of his column, in which he faces up to the following:

1) The regular season counts. The A’s are the second-most decorated American League team in history, but their playoff ambitions have been frustrated in recent years. But as I pointed out in September, they are also one of only six teams to reach 500 wins since 2000, and the only team other than the Yankees to never have an off-year (never ending the year 10 games or farther back of a playoff spot).

2) The A’s consistently win more games, for less money. While poorly-spent money will not buy a championship, in most sports a wealthy owner can buy success. An owner with his own sports network can parlay the on-field success into a self-sustaining franchise, so he doesn’t lose his shirt in the process.

But the rest of baseball needs to operate like a normal business: spending can’t exceed income for any reasonable period of time, and investment has to produce results.

A lot of GM’s do a lousy job of this. And a lot of owners don’t seem to care, because owning a sports team is a glamorous method of frittering away wealth – like opening a boutique restaurant, or backing a Broadway production.

The A’s are the Warren Buffet of baseball franchises.

On the field, they know the team’s job is to win games (which takes us back to the debate about how to win – and into topics like steals, small ball and defence ).

Off the field, they know the team is supposed to win those games with the minimum expenditure required, and without sacrificing the present or the future.

Over the past six years, the Yankees and the A’s have posted nearly identical regular season records. The Yankees did it for $782 million in payroll, the A’s did it for $238 million. Baker derisively sniffs that “the Oakland formula supposedly lessened the financial burden.” Well yes, Geoff, it did: to the tune of $544 million.

How did Chicago fare over this period? They spent about $100 million more than the A’s, won 60 fewer games, and made the playoffs once, in the year where they spiked their payroll to $75 million, or $20 million more than the A’s. But that one playoff visit was enough.

Does anyone other than Geoff Baker think that a well-run business produces a mediocre product five years out of six, and banks everything on that sixth year turning out well? It sounds more akin to a strategy of reverse-Russian roulette.

3) He’s attacked a straw man. Baker never states the Moneyball theory that he claims has been disproved. Is there no testable hypothesis?

Instead, Baker comes at the proposition sideways. We read that “Moneyball's formula dictates…seeking out "undervalued" diamonds in the rough,” and that’s as close as we get to understanding what the argument is that Baker thinks he’s making.

A better fighter would have succinctly stated Moneyball’s best case, and offered relevant evidence to disprove it (Hint: repeating Jermaine Dye’s observation that “pitching and defence wins” does not count as evidence).

4) The so-called revisionism is really originalism. Forgetting that Michael Lewis came to the project as a business writer, Baker objects to the unattributed contention that Moneyball is about “finding ‘value’ in a changing marketplace” rather than the advocacy of certain methods (such as OBP) over others (like steals).

Baker has missed the point entirely.

If the team’s job is to win games, then management must first understand how games are won – which requires knowledge of how the different elements of performance contribute to creating or preventing runs. Statistical tools help us understand how the different elements contribute to the whole, and how players are likely to perform in the future based on what we know of the past.

The second element involves assembling these pieces in the most financially efficient way.

Any decision a manager makes reflects the dynamic interaction between the marketplace where the pieces are available, and the different ways of assembling run production and run prevention (were you to offer Billy Beane a cheap base-stealer who succeeds more than 75% of the time, he’d snap him up).

The best of the insights in Moneyball, and its predecessors, showed us that some existing measurement tools are flawed, leading managers to mis-value the contribution a player can make towards winning.

In today’s environment, that might translate into going after batters with high OBP, or pitchers whose DIPS are better than their peers, even if their win-loss records looks lousy. But that is not because OBP or DIPS are the only way to build a team: it is because their contribution to winning is being undervalued by the market.

Since the performance behind those statistics isn’t very glamorous to baseball writers like Baker, they are likely to remain undervalued for a long-time to come. If that helps confuse Geoff Baker further, or drives another owner into his camp, fans of the Oakland A’s will have much to celebrate.

1 Comments:

At 2:28 AM, Blogger andy grabia said...

Awesome post, Avi.

In an old Cosby Sweater post, I crunched some numbers comparing the A's with the Sox and Yankees. I was reading Faithful at the time, and some of Stephen King's sentiments were driving me crazy. Baker's argument--or lack thereof--is driving me postal. The points I made are essentially along the same lines as yours, but I would like to reiterate them, if I may. I haven't updated them to include 2005 numbers.

" The A’s cranked out 91 wins in the regular season last year. The BoSox? 98. That is a 7 win difference over the regular season. But two points need to be made here:

1) What did those 98 wins cost the RedSox? $127, 298,500. The cost of 91 wins to the A’s? $59, 425, 667. Those 7 extra wins cost the RedSox over $60 million, almost 9 million a game. Sure, they won the World Series this year, but the year before, with almost the exact same payroll? Knocked out in the ALCS by the Yanks, one round later than the A’s.

2) How many of those 98 wins were against quality teams? The A’s play in the toughest division in baseball, battling for a playoff berth 18 times a year against the Angels, Rangers and Mariners. Win percentages last year: .568, .549, .389 (trust me, the Mariners will be better this year.) The BoSox play 18 against the Yankees, Orioles, Blue Jays and Devil Rays (trust me, the O's, Jays and D-Rays won't be better this year). Win percentages last year: .623, .481, .435, .416.

So let’s cut the bullshit here. Is it better to win the World Series than not? Sure. Is it worth an extra $60 million a year? Only John Henry knows. But there can be no doubt that the A’s do more with less money than any team in the majors, with the exception of the Twins ($53 million payroll, 92 wins in '04). They do it in a better division, and they have done it for five years running. Don't believe me? The MLB page goes back 4 years, and here are some more stats:


Oakland A's2000- 91 Wins, 70 Losses, $32 million
2001- 102 Wins, 60 Losses, $34 million
2002- 103 Wins, 59 Losses, $40 million
2003- 96 Wins, 66 Losses, $50 million
2004- 91 Wins, 71 Losses, $59 million

*Salary Source- USA Today


Boston Red Sox
2000- 85 Wins, 77 Losses, $81 million
2001- 82 Wins, 79 Losses, $110 million
2002- 93 Wins, 69 Losses, $108 million
2003- 95 Wins, 67 Losses, $100 million
2004- 98 Wins, 64 Losses, $127 million

*Salary Source- USA Today

The only year the BoSox had a better regular season than the A's was 2004. And the A's always operate with 1/2 to 1/3 of the salary. The A's made the playoffs every one of those years except for 2004. The BoSox only made them in 2003 and 2004. The A's have won 8 playoff games and 0 series since 2000. The Red Sox had only 6 playoff game wins and 1 series win up until this year; of course the World Series victory altered those numbers greatly. Here is a nice breakdown of the A's versus Yankees, as well, for those who want to match the A's up to the envy of the sporting world. They do just fine here, as well.

The A's haven’t won the World Series in that time, but it took the Sox 86 years between wins, so we can let that slide. The fact of the matter is that the A's are THE model sports franchise, and have been since Beane entered the scene. What is ironic to me is that the key to the Sox’ success this year was a Billy Beane-like roster, albeit with a Yankee-like payroll(Again, the irony is ripe. Who’s the Evil Empire now?). Half the bloody Sox lineup last year were players created in Bill James’ laboratory. Ortiz, Youkilis, Mueller, Bellhorn, Foulke. The small ball “fundamentals” that King and O’Nan cherish so much? See Exhibit A: Tony La Russa , the St. Louis Cardinals, and a World Series snooze-fest."

I concede that the White Sox won this year, but that is ONE year. I could as easily argue that teams with good defence and solid pitching DIDN"T win the World Series this year. And it isn't like Moneyballers don't like pitching and defence. Many, including Peter Gammons, have argued that in fact people like Bean and DePodesta had moved beyond stats like OBP this year, focusing instead on defensive statistics.

I could go on and on, but my rage is too much at this point and time.

 

Post a Comment

<< Home