Wednesday, October 31, 2007

The power of Scott Boras

OK, one more post roughly related to Alex Rodriguez and his contract--

Tyler Cowen at Marginal Revolution wonders how Scott Boras might be able to command higher prices for his clients than other agents do. J. C. Bradbury is skeptical of this power. Such skepticism is to be expected from economists, who would be surprised to see a single actor fundamentally change the dynamics of a competitive market as Boras is supposed to do, but I don't think you can seriously dispute that Boras has fundamentally shifted prices at times, especially in the amateur draft.

Cowen lists some mechanisms by which Boras might beat his market, but he neglects what I consider the most interesting possibility: that Boras actually makes his players better. This recent story in ESPN the Magazine describes the ways in which Boras tries to increase the skill and durability of his players. An ability to increase durability seems plausible to me, and if it seems plausible to owners, it may well cause them to pay more for Boras's clients. In the case of A-Rod, durability is a crucial factor, arguably the crucial factor, even a decisive one: if he stays healthy, he will almost certainly become the home run king. I can easily imagine Boras's longtime management of Rodriguez's training regimen being worth millions of dollars to a team.

Monday, October 29, 2007

A-Rod's Contract Again

I usually agree with Rob Neyer, but I don't agree with this blog post, where Neyer contends that Alex Rodriguez isn't worth the money he's now earning or will earn. As evidence that A-Rod is overpaid, Neyer cites Nate Silver's study of the first years of A-Rod's present contract.

I see three problems with Neyer's case.

First, and most trivially, Neyer cites the contracts of Rodriguez, Mike Hampton, and Manny Ramirez as regrettable decisions by irrational owners: "All three franchises, within just a few years, regretted those deals. Terribly regretted those deals." Sure, Hampton's deal was a disaster, but doesn't it seem a little nuts to criticize Boston for the Ramirez deal in the very week that the team wins its second World Series? I mean, I discount the meaning of postseason performance as much as anyone, but it's hard right now to imagine a better way for Boston to have spent that cash.

Second, the Neyer/Silver argument may be outdated. Neyer doesn't account for the increasing revenues in MLB. The increases may not be enough to change the big picture, but they need to be accounted for.

But the more fundamental problem with Neyer's argument is that he's making a case about a market that simply doesn't exist. Baseball owners don't get to sign free agents on the basis of Silver's calculations of their value. The asking price of free agents is (give or take) the amount of the richest competing offer plus a little bit. In such a market, top free agents will always and necessarily command more than their demonstrable value, while top young players in the present salary structure receive less. The owner who offers the Silver-Neyer price for free agents simply won't sign any of them. The rational price is above the median assessment of a player's demonstrable value. That is, the right price is what Neyer would wrongly call an irrational one.

The interesting quirk of this situation, however, is A-Rod's act of opting out of his present contract, which costs the Yankees $23 million. Avoiding that loss should be worth a lot to the Yankees; they could rationally pay, say, $20 million more than A-Rod's free agent price to extend him. The fact that A-Rod appears to be turning down a contract extension means a) he's bluffing, b) he really doesn't want to play for the Yankees anymore, or c) he and the Yankees are each betting on evaluating the free agent market better than the other. I'm guessing A-Rod wins that bet.

Friday, October 26, 2007

Undefeated seasons and aligning incentives

The Sports Guy has written recently about the relative probabilities of going undefeated in the NFL and in a given fantasy football league. Simmons skips the obvious historical approach--getting the a fantasy stats service to tell him how many teams go undefeated and comparing the incidence with the NFL's history--but he offers good reasons for thinking the undefeated fantasy season the rarer achievement. I'll add a couple of thoughts about the role of incentives in the comparison.

Many of Simmons's points boil down to the simple fact that fantasy results are hard to control due to misaligned incentives. If the Patriots are winning by three touchdowns and your fantasy team needs Tom Brady to through for two more, you're out of luck because Brady doesn't care what you need. His incentives are different from yours. Incidentally, this scenario demonstrates why I think fantasy baseball is a better pretend sport than fantasy football: in baseball, Manny Ramirez is going to try to hit well whenever he comes to the plate. His incentives are aligned with his fantasy owners' because there's no way to run out a clock.

(Side note: the latest Nobel prize in economics was awarded for work on mechanism designs that maximize incentive alignments. Here is one explanation of the work.)

OK, so the point is that misaligned incentives make fantasy football tougher to control. But there's also a contrary influence of incentives. In most fantasy football leagues, every team is trying to win a given year's championship. In the NFL, some teams are trying to win the Superbowl, but many of them are looking at least partly to the future, some are in full rebuilding mode, and a few are coasting along on low salaries to soak up guaranteed profits through revenue sharing. Therefore, the NFL is guaranteed to have unbalanced resources, with a handful of really good teams standing in the way of any undefeated season. It would be much easier to sweep a league that disbanded every team each year.

How do these variously misaligned incentives shake out to answer Simmons's question? I don't know. I'd love to see some data.

Wednesday, October 24, 2007

The Rockies' momentum

I've loved watching the Rockies win game after game lately. I have not loved media speculation about whether the layoff between the team's last playoff game and the World Series will break their momentum. Momentum plays a big role in commentary about sports, but statistical analyses have found consistently that independence generally trumps momentum. Thomas Gilovich's wonderful debunking of the basketball "hot hand" in How We Know What Isn't So is a great early example.

The streak has demonstrated that the Rockies are a better team than almost anybody imagined a couple of months ago. They have higher real and Pythagorean winning percentages, and their terrific pitching staff seems to have been solidified by mid-season shifts in the roster and role assignments. The team will now face its toughest opponent and toughest arena. To think of momentum or its dissipation as the central issue of the Series is a distraction. The Series will be determined by the quality of the teams and the quirks of short-series baseball, not whether the Rockies have continued to please the God of Momentum.

Tuesday, October 23, 2007

Peer effects in golf

Tim Harford posts about a new piece of economics research finding no peer effects in professional golf tournaments. That is, according to the researchers, golfers aren't affected by the quality of their playing partners in tournaments.

I've long been skeptical of peer effects in golf. But the general finding doesn't (as far as I know) address the key specific question: does being paired with Tiger Woods on Sunday hurt other golfers? Just about everyone thinks so, but I'm skeptical because there's a simple explanation for the appearance of a Tiger Effect.

That explanation is this: intimidation aside, Tiger is the best golfer in the world. Therefore, if he and another player are contending to win a tournament, the other player's performance is by definition more of an aberration than Tiger's. (Any player who is tied or nearly tied with Tiger on Sunday has overachieved relative to Tiger.) Therefore, if Tiger and his playing partners do what we would normally expect of them, they would create the sense of Tiger intimidating the other players into Sunday collapses. Utterly ordinary expected performances would create the same Tiger Effect that golf analysts and fans now perceive.

Obviously, this reasoning does not disprove a real Tiger Effect. But any test of the effect should account for this explanation, the fact that courses generally get harder on Sunday, and other reasons why Tiger's playing partners may not be wilting but simply finding their level on Sundays.

Thursday, October 18, 2007

A guy who ought to know

ESPN the Magazine leads off an article in its issue of October 22 thusly:

Recent history says the team popping the bubbly this season won't be the best record. So is it all luck? The Mag's Buster Olney asks a guy who ought to know: Braves pitcher John Smoltz.

What amazes me about this and many similar formulations is that writers seem to go out of their way to say, in essence, "This is a question that fall squarely in the province of statistical analysis rather than observation"--in this case, to frame the question as one of the relationship between probability and uncertainty--and then tumble directly into personal anecdote.

In next month's magazine, I hope to see the same logic in the other direction:

What does it feel like to take the mound with your team's season on the line and tens of thousands of hostile fans burying you in boos? The Mag asks a guy who ought to know: Louisiana Tech Professor James J. Cochran.

Sunday, October 14, 2007

Livan Hernandez and myths of the postseason

I've been a Giants fan for a long time: my first sentence was "Go Giants, beat Reds." Therefore, I remember all too well the lesson I learned in 2002: when you ask the gods to do something, you'd better watch our for ironic compliance.

In 2002, I had been arguing for years that there was no reason to think that Barry Bonds, then widely regarded as a postseason choker, was a different player under pressure. Here is a post I wrote in 1997 to that effect; there were many others. In 2002, Barry got his big chance to play in the postseason again, and he was ridiculously great: excellent in two playoff series, then about as good as anyone had ever been in a World Series: .471/.700/1.294.

While the gods granted my request to demonstrate that we should base postseason expectations on regular season performance, however, they also had Livan Hernandez prove the point in the other direction. Hernandez then enjoyed a reputation as a tremendous postseason pitcher, and indeed, he had pitched a two brilliant playoff games early in his career. But he had been declining as a pitcher, and though his postseason W-L record and ERA had held up, his supporting statistics had collapsed; the Livan pitching for the Giants was clearly not the Livan of 1997. In 2002, Livan pitched a solid game in the first series, a very shaky but lucky game in the second (6.1 IP, 10 baserunners, 0 strikeouts, 2 ER), and two utterly disastrous games in the World Series.

He pitched less than six innings total in the two games and gave up nine earned runs. Surely, thought I, this is the end of his reputation as a postseason force.

But it wasn't. As the Diamondbacks entered this postseason, the talk started again: don't pay attention to the regular season numbers, we heard, because Livan has another gear in October!

Indeed, Hernandez had a good postseason W-L record, but that was more a function of luck and run support than excellence: his regular season and postseason ERAs were nearly identical. And the best part of that postseason record came a full decade ago, when he was a much better pitcher in all situations than he is now.

But none of this stops's Mark Simon from saying, in a blurb that can't be linked directly, that Livan is "one of baseball's best postseason pitchers":

It will be up to one of baseball's best postseason pitchers to try to cool off the Rockies in Game 3 of the NLCS on Sunday, with Livan Hernandez trying to get the Diamondbacks a desperately needed victory. Hernandez doesn't exactly have the best history at Coors Field, but he's been known to dial it up a notch when it counts.

Anything can happen in a small sample, but this is a myth that deserves to die.

Saturday, October 13, 2007

Sports Guy gets his stathead on

In response to a reader question about teams moving to a short rotation in the baseball playoffs, Bill Simmons writes,

SG: Graham, that's a fantastic question. I don't have an answer for you. The three-day rest thing only seems to work when you don't have another choice (like the Red Sox in 2004, for example). If it's a conscious decision, the results always seem to be brutal. But I have another question: Why is everyone always so confident that sinkerballers are better on three days rest? People just spout this out like it's a foregone conclusion -- oh, yeah, it's fine when Wang pitches on three days rest, he's a sinkerballer. It is? Who said? Do we have scientific proof that it's better for any pitcher (even someone with a specialty pitch like the sinkerball) to be more tired than less tired? I'm dying for them to tackle this on "MythBusters."

That's a great response from Simmons; I want only to add that this case offers an excellent demonstration of the power of the optical revolution in baseball analysis. We suddenly have the power to know exactly what happens to sinkerballers after three or four days of rest, to see whether any effect correlates with the amount of sink on pitches, and to do all of this with a direct measurement of the pitches' break instead of inferring that measurement from fly ball/ground ball ratios. Amazing.

Tuesday, October 9, 2007

A-Rod's contract

Everybody seems to acknowledge that if Alex Rodriguez opts out of his current contract to negotiate a new one, he'll make more money. In other words, he is currently underpriced. Every story about his contract also includes the fact that the Rangers are scheduled to pay the Yankees $3M/year of that contract if A-Rod sticks with it. I haven't seen any story draw this conclusion: the Rangers made a trade that has them paying the Yankees to employ baseball's best player at a bargain price.

February 2004 is not that long ago, after all: "Texas will pay $67 million of the $179 million left on Rodriguez's $252 million, 10-year contract, the most cash included in a trade in major league history." Well played, Rangers! No wonder you're so good!

Monday, October 8, 2007

Shining when the lights are brightest

I was going to write recently about Livan Hernandez's undeserved reputation for postseason excellence, but I just heard an even crazier example. Indians GM Mark Shapiro just said on the Baseball Today podcast that Kenny Lofton is a proven postseason performer who shines when the lights are brightest and whatnot.

No. The only good thing you can say about Lofton in the postseason is that he's gotten to the playoffs a number of times. In fact, I hypothesize that any player who makes the playoffs with a bunch of different teams and manages not to be memorably awful will gain a reputation for clutch postseason play.

But here (scroll down) is the real story of Lofton's playoff performance: 19 series over 11 years, with a solid sample of 360 at-bats, producing at a clip of .253/.323/.353 including the current hot streak. He's been putrid.

Sunday, October 7, 2007

The last day of the regular season

The extraordinary lack of close series in the playoff so far make me recall one of the many components of the great drama of the end of the NL regular season: Jayson Stark wrote for that the Philadelphia crowd gave the visiting scoreboard a standing ovation when it posted the Marlins' seven-run first inning against the Mets. On that Sunday afternoon, two crowds--those in Philadelphia and Colorado--had the extraordinary pleasure of watching the loss of the key rival go up on the scoreboard while the home team in front of them won the crucial last game. That's a tough combination to beat.

Torre's job

Many media reports today say that George Steinbrenner will fire Joe Torre if the Yankees lose their current series against the Indians. I dislike the Yankees anyway, for most of the usual reasons, but such stories add fuel to the fire. The result of a short series in baseball conveys almost no meaningful information about a manager's performance, especially when the manager in question has experienced great postseason success and disappointment by turns. I can think of a few legitimate reasons to fire Torre, but Stenbrenner's would be an especially stupid one.

Tuesday, October 2, 2007

Shape stats and consolidation stats

As has been widely noted in the sports media, Jimmy Rollins last weekend became only the fourth Major League Baseball player to join something called the 20-20-20-20 club: at least twenty doubles, triples, home runs, and stolen bases in a season. I think that's great. I've liked Rollins since I followed him through the Phillies system, including catching him alongside Pat Burrell at the wonderful minor league park in Reading, and the 20-20-20-20 thing offers a quick narrative-in-a-number describing the kind of player Rollins has become.

I would contend, however, that joining this quirky "club" should have no bearing at all on Rollins's ranking among players, including his candidacy for the MVP award. The 20-20-20-20 number is what I'll call a shape stat: it describes the shape of Rollins's production, the way his value manifested itself, rather than the amount of value he produced. Homers are worth more than triples, which are worth more than doubles, and all of those are much more valuable than stolen bases. 20 homers and 20 steals is not as valuable as, say, 27 homers and 2 steals; the former totals are just less common. For evaluative purposes, we need exactly the tools that most sportswriters gleefully ignore, the ones such as VORP that assign informed weights to each of these statistics and then sum up the player's performance.

For the purposes of description and narrative, I've got no problem with shape stats. But if we're talking about the MVP or other ways of ranking players, setting aside the shape stats is a good starting point for any serious discussion.