Archive for February, 2008

A Critique of the Clemens Report

In today’s NY Times Keeping Score column, several Penn economists are critical of “the Roger Clemens report”. In particular, the authors feel that comparing Clemens to three excellent pitchers who excelled late in their careers is a dodgy tactic.

A better approach to this problem involves comparing the career trajectories of all highly durable starting pitchers. We have analyzed the progress of Clemens as well as all 31 other pitchers since 1968 who started at least 10 games in at least 15 seasons, and pitched at least 3,000 innings. For two common pitching statistics, earned run average and walks-plus-hits per innings pitched, we fitted a smooth curve to all the data from these 31 pitchers and compared it with those for Clemens’s career.

Relative to this larger comparison group, Clemens’s second act is unusual. The other pitchers in this durable group usually improve steadily early in their careers, peaking at around age 30. Then a slow decline sets in as they reach their mid-30s.

Clemens follows a far different path. The arc of Clemens’s career is upside down: his performance declines as he enters his late 20s and improves into his mid-30s and 40s.

Clemens Aging Comparison

I have a few comments. First, the reason his career is upside down in terms of WHIP is because of his walks. Clemens’s career pattern in preventing walks is bizarre, as I previously documented, but walks are not the thing I think of when I think or performance gains from steroids.

Walk% Above Avg.

On the other hand, his strikeout performance did decline with age, which means that the decline in walks likely was not the product of being able to pitch more in the zone.

Strikeout% above Avg.

The authors are also critical of the Clemens Report for using ERA because “a pitcher’s E.R.A. is affected by factors, like defense, that have nothing to do with his pitching.” But, using WHIP doesn’t solve this, because all hits except home runs involve the defense.

As to the comparison with other pitchers who excelled at advanced ages, this is not meant to demonstrate Clemens’s performance path was expected. Clemens’s aging pattern is certainly atypical. See Cy Morong’s evaluation of an aging cohort that includes Clemens. What these comparisons do convey is that such performances have been observed before, and therefore just because Clemens performed in this way doesn’t mean that steroid use is the only possible explanation. Clemens aging pattern is not the norm, but it is also not odd enough to prove anything.

Finally, I’d like to reiterate that Clemens’s most suspicious late-career spike (2004-2006) occurred at a time when McNamee was working for Clemens but had no knowledge of steroid use. It would be odd that he would get the drugs from McNamee in 1998, 2000, and 2001; but then find another source. And given that I don’t find McNamee to be a compelling witness, I am inclined to believe Clemens pitched clean until I see some better forensic evidence.

The authors, for whom I have great respect, bring something interesting to the table, but I don’t interpret the findings to be an indictment Clemens. Although, I agree with them that the Clemens report is not a compelling document. I think it was unnecessary, and it has served as a focal point for criticism rather than to exonerate the pitcher.

Addendum: I received an e-mail from Randy Hendricks explaining some of the impact that the Clemens report did have, particularly in dispelling the common misperception that Clemens was done in 1996. I guess I was a bit narrow-minded in my interpretation, since I already knew this. Thus, I was incorrect in calling the report unnecessary.

McNamee’s New Evidence

As the Clemens-McNamee affair grows more bizarre by the day, I think that Brian McNamee’s submission of new physical evidence has some serious implications that impact McNamee more than Clemens. According to McNamee’s attorney, his client decided to reveal the new evidence to federal prosecutors in early January.

McNamee’s lawyers say their client lost any lingering loyalty to Clemens after a news conference in Houston on Jan. 7 in which Clemens played a tape his lawyer had secretly recorded of a telephone call between Clemens and McNamee.

The next day, Jan. 8, McNamee brought the evidence to his lawyers, they said. Two days later, they handed it to federal authorities in New York, who are now having the items tested.

“They obviously were not happy about it,” Ward said of the authorities’ reaction to the delay. “But they understood certain omissions that he made, and they’re happy they have it in their possession.”

McNamee’s lawyers said the physical evidence dates from 2001 and 2002. The photograph that includes the beer can, they said, reflects injections McNamee gave Clemens, although they said there could also be used needles from other persons stored in the can.

In what amounts to a new disclosure, they said the photograph of the unused needles and steroids specifically dated to 2002, when McNamee said he was no longer giving Clemens injections. But they said Clemens nevertheless gave McNamee the unused drug items in October of that year because he did not want to take them on a flight to Houston, implying that Clemens might still have been using steroids or H.G.H. in 2002 even without McNamee’s involvement.

This clearly violates the terms of his plea agreement with federal prosecutors to be truthful, because he stated that he had no knowledge of steroid use by Clemens after 2001. But, that is not the only reason why I think this new evidence is interesting.

McNamee was linked to Kirk Rodomski by checks written to Radomski in 2003 and 2004. McNamee claimed that these purchases were for non-baseball clients. This claim becomes a little less believable now, and I will not be surprised if prosecutors do go after McNamee for distributing steroids. His lies severely undermine his credibility as a government witness.

In the end, the physical evidence may do more damage to McNamee than Clemens. Forensic experts have largely dismissed the relevance of the syringes due to dating, chain of custody, and tampering issues. The evidence serves only to bring more questions about McNamee’s motives and believability.

And something tells me that this isn’t the last twist in the case.

Update: And here is that twist.

Brian McNamee told congressional investigators Thursday that he injected Roger Clemens’ wife with human growth hormone before she appeared with the pitcher in Sports Illustrated’s swimsuit issue in 2003, according to a Washington source.

That didn’t take long.

My Anger Boils Over

I’ve written quite a bit on the Gwinnett Braves stadium as details about the deal have leaked out over the past few weeks. In today’s Atlanta Journal-Constitution, I condense my analysis to an 800-word Op-Ed to share with my fellow metro Atlanta residents.

Like many baseball fans living in metropolitan Atlanta, I was delighted to hear that the Braves would be moving their Triple A franchise to Gwinnett County. However, since learning the financial terms of the deal, my enthusiasm has waned. The arrangement essentially transfers Gwinnett residents’ hard-earned tax dollars to the Braves’ corporate owner, Liberty Media. It is inexcusable that the Gwinnett County Commission would approve such an agreement without vetting the proposal publicly.

After examining the details, it is easy to see why government officials were so eager to keep this deal behind closed doors until it was too late. It may be a sweetheart deal for Liberty’s corporate shareholders, but it is fiscally irresponsible and morally reprehensible.

Now, if I lived in Gwinnett County, then I would really be mad.

If you are coming to the site for the first time, you can read all of my analysis of the Gwinnett stadium deal here. To read the posts sequentially, go to the bottom of the page and scroll up.

How Did Clemens Age Relative to Other Pitchers?

This is the question Cy Morong attempts to answer on his blog Cybermetrics—what a good blog name. Cy uses a sample of pitchers who pitched a minimum of 500 innings in the five years before before their fortieth birthdays and 200 innings in the following five years. He uses ratios of pre- and post-40 performances to evaluate the change in several statistics.

He finds:

In general, Clemens does much better at the older age. But he is not always first or close to first in the rankings of the various stats.

Jonah Keri makes a brief comparison between Clemens and six other pitchers using ERA+.

Even among the greats, Clemens stands out. He appeared to be declining in his thirties, but put up a monster season at age 34, posting a 221 ERA+. It’s not uncommon, though, for power pitchers to take big steps well into their thirties—the real divergence came later. After two decent seasons at ages 39 and 40, Clemens won a Cy Young Award in 2004 and posted an outrageous 1.87 ERA in 2005, for a 226 ERA+ that’s the thirteenth-best of all time. The huge spike at such an age was unprecedented in major-league history.

For some reason he calls out my analysis in a rather bizarre way.

Nor is he cleared by research from the blog Sabernomics, which found little evidence of jumps in performance after the injections McNamee alleges. The problem with this analysis is there’s little information on the effects of PEDs on pitchers. They can increase muscle mass and thus fastball speed; some also believe they speed injury recoveries and allow one to work out more often, increasing durability. Clemens could have taken PEDs preemptively—preventing falloffs rather than triggering spikes—or taken doses too small to have an effect. And, of course, he could’ve taken PEDs on other occasions.

To this, I say, “huh?” First, I didn’t look at spikes in performance after use to find evidence of use. I looked declines in his performance prior to use at specific times alleged by Brian McNamee. The idea was that Clemens would be more tempted to use at times he appeared to pitch poorly. I found that these were odd times to use. Second, I also noted a general aging trend in his performance. I noted the same post-40 spike that Keri finds, but I interpret it quite differently. What I find interesting about this spike is that it occurred during a time when Clemens still trained with McNamee, yet McNamee was not aware of any steroid use at the time and MLB was conducting steroid testing. While it is possible that Clemens would get his steroids from another source, I find it odd that Clemens would go to others for steroids when McNamee had already allegedly helped him use them.

Update: I corrected my initial misspelling of Jonah Keri’s name.

The Anatomy of the Gwinnett Stadium Deal

We are now learning more details about how Gwinnett County plans to pay for the Triple-A Braves new stadium. The county is still pushing the line that this is a sound business deal that won’t be a burden to taxpayers.

County Commissioner Bert Nasuti, who started the movement to bring baseball to the county, said the stadium financing is based on highly conservative estimates and is about as risk-free as it could be.

“We wouldn’t be doing it otherwise,” he said.

Really, Bert? Let’s take a look at look at the financing plan as it is listed in todays AJC (Gwinnett edition).


* Land: $5 million
* Stadium: $40 million
* Total: $45 million

The land has already been paid for, but the stadium has yet to be built. And given its need to be built quickly, it is likely that the the $40 million price tag will go up. The underestimation of construction costs is common for publicly-financed stadiums, and the county is responsible for all cost overruns. It is possible that the construction will come in under budget, but I doubt it.

Next, let’s look at the revenues.

* County recreation fund $12 million
* Revenue bonds (borrowing) $33 million

The $12 million is gone. What is lost from these funds is the opportunity cost of what these funds could have been used for: park improvements, special programs, security, etc. Whether taxpayers of Gwinnett preferred these things or a Braves stadium is an open question. It’s too bad the County Commission didn’t let the voters have a say.

Paying off the debt is where the county truly becomes creative. In some areas, I believe the county is being realistic, but in others the projections are outrageously optimistic.


* Stadium rent $250,000
* Ticket fees $400,000
* Parking fees $200,000
* Car rental tax $600,000
* Naming rights $500,000
* GCVB contribution $400,000
* TOTAL $2,350,000

In regard to rent, ticket fees, and parking, I think these estimates are accurate.

The Braves are required to pay $250,000 a year in rent for the first five years, after that time the rent payment will rise with the consumer price index. Assuming an average CPI growth of 3%, these averages out to about $350,000 year over the next 30 years.

The ticket fees come from a $1 fee per ticket that the Braves must pay to the county for every ticket sold. $400,000 is the minimum the team must pay the county, and given the size of the stadium, it is not likely that attendance will generate much more than this.

The Braves must split net parking revenues with the county. Assuming an average attendance between 6,000 and 7,0000, one car per three fans, and a $3 parking fee, $200,000 seems to be a reasonable estimate of the county’s parking share.

Next, let’s jump to the naming rights, the last source of funding that does not require taxpayer assistance. Gwinnett can sell the rights to the stadium, but it must pay the Braves $350,000 annually from this revenue. Should naming rights sell for more in the future, the county must pay a higher share to the Braves. The county is projecting that it will earn $500,000 a year ($850,000 total — $350,000 to the Braves).

Two experts interviewed for the story state that they feel the projection is reasonable.

Adam Zimmerman, executive vice president of the Marietta sports management firm Career Sports and Entertainment, believes Gwinnett should be able to leverage the county’s wealthy population and the stadium’s relationship to the major-league Atlanta Braves to score a healthy deal. After all, he said, the stadium’s sponsor would be attaching its name to the next generation of major-league talent in Atlanta.

“It’s almost a catch-the-rising-star platform that a corporation could own,” Zimmerman said. “There’s something that could really be had from owning that gateway to the majors.”

Jeffrey Grill, who has helped build naming-rights deals as a partner with the law firm of Pillsbury Winthrop Shaw Pittman in Washington. Stadiums in heavily-populated, affluent areas have the best chance of getting a big contract, he said.

“Gwinnett County has been one of the fastest-growing counties in the country for the past 20 years. I would believe that the annual value of the naming rights should be significantly in excess of the national average and that $850,000 per year is a possibility,” he said.

I won’t dispute the qualitative assessments of these gentlemen. There is no disputing that Gwinnett is a growing area that many corporations would like to be associated with. However, that doesn’t mean any number is possible. The quantitative evidence is not favorable, as I say in the story.

The anticipated $850,000 deal would be the second-highest naming rights deal for a Triple-A team behind Chuchanski Park in Fresno ($1.1 million) and ahead of Raley Field in Sacramento ($750,000) and PGE Park in Portland ($710,000). No other naming rights deal comes close to these. The top-three stadiums are on the more-expensive west coast and serve as the host to the top-level baseball club in a major metropolitan area (not a suburb)—two of these cities host NBA teams. The Fresno and Portland facilities host other sports teams (minor league soccer and college football). Everyone knows where Fresno and Portland are. Have people outside of Georgia ever heard of Gwinnett County? I find it hard to believe that Gwinnett is comparable to these other markets.

The average Triple-A naming rights deal is $338,000. The most recent stadium naming rights deal for a Triple-A stadium occurred in Lackawanna County, PA for the Scranton/Wilkes-Barre Yankees, for which PNC will pay $365,000 per year. I think Gwinnett will be lucky to get $450,000 from any naming rights deal, which nets the county $100,000—$400,000 less than the county anticipates. If the county doesn’t get that money from naming rights, it will get it from taxpayers.

This brings me to the only official tax that Gwinnett county officials admit to. Though they operate on the erroneous assumption that the tax will not be paid by its own citizens, it is most certainly will be.

“It takes the burden off the local taxpayers,” Commissioner Bert Nasuti said of the car rental tax, which is also in place in Atlanta to fund improvements made to Philips Arena in the late 1990s. “The reason you do that is so you don’t raise taxes.

I cannot comment on how much revenue that the tax will raise, but it will most certainly be paid Gwinnett County residents—the people who rent cars in Gwinnett County—not outsiders lured in by the existence of a Triple-A team.

As for the last revenue item, the Gwinnett Convention and Visitors Bureau will contribute $400,000. This revenue will mainly come from the Arena at Gwinnett Center, another taxpayer-funded facility. Even if that facility is earning a profit—and I doubt that it is—the excess revenue going to the Braves stadium still represents a cost to Gwinnett taxpayers. Instead of funding roads or schools, the revenue now subsidizes a stadium that hands over a large portion of its earnings to a sports franchise. Using revenues from a taxpayer-funded arena to pay for a baseball stadium is like plugging a power strip into itself to generate power.

Thus, when we add up the tax liability to Gwinnett citizens to cover the $2.35 million annual debt payments, taxpayers are left to cover $1.4 million—approximately 60% of the debt.

When the deal was announced, county administrator Jock Connell stated, “We anticipate it paying for itself from day one.” This plan doesn’t come close to meeting this standard, and it is a shame that county officials have been so deceptive in presenting their plan to the public.

Also in the AJC, as a sidebar to the story, there is a story on my discussion of the stadium deal on Sabernomics. Thanks to Michael Pearson for covering the story so well, so as to make my job easier.

What Are the Best Super Bowl Squares?

If you play a Super Bowl squares game at your Super Bowl party, here are some tips for for purchasing your squares from an old post by Doug Drinen.

So, to recap, here is the official Doug Drinen plan for making friends at your Super Bowl party.

1. As soon as the squares are drawn, find the people with the 70 and 07 squares. Offer them $3 for their square. If they accept, tell them that they just traded an expected payoff of $3.80 for a mere $3.00. Call them suckers.

2. Find the guy with 22 and let him know that since the merger in 1970 there has only been one single 22 game (it was the Dolphins and Bills in week 13 of this year). Call him a sucker too.

Doug recently updated his analysis using quarter-by-quarter scores, and you can read it here.