July 11, 2008
Lies, Damned Lies
Running Up the White Flag
We've become far more sophisticated over the years about performing financial analysis on trades. We now understand, for instance, that there is a very measurable financial benefit to a team when it makes the playoffs. We also understand that the way you create long-term value in baseball is by employing players who are still in their reserve-clause or arbitration seasons, and paying them a fraction of their market value. It is the interweaving of these two factors that creates most of the trade market.
On Tuesday, as you're probably aware, the A's traded Rich Harden and Chad Gaudin to the Cubs for four prospects. If we look at this deal from a MORP perspective, it undoubtedly looks like a positive for the A's. Harden is under contract only for this season and then with a (cheap-looking) team option for next one, while Gaudin has a couple of years left before he hits free agency.
Based on some relatively ad hoc adjustments from their PECOTA forecasts at the beginning of the year, I'd guess that Harden is worth something like $8 million above his contracted value between this year and next, and Gaudin probably about the same between now and 2010, discounting what he's likely to make in arbitration. The Cubs will probably also get a free agent compensation pick when (and if) Harden departs, which is worth another several million dollars. So there's roughly $20 million in value coming to the Cubs, which should be boosted somewhat because they're in the midst of playoff contention.
By contrast, the MORPs of the four players the A's acquired add up to about $77 million over the lifetime of their reserve and arbitration seasons. About 40 percent of that total will eventually need to be paid for in arbitration, leaving around $45 million of net value coming in to the A's.
In a case like this, though, sometimes I think we can lose the forest for the trees. Oakland did something else in this transaction: they punted on a promising season. At the time they made this deal, the A's had roughly a 40 percent chance of making the playoffs according to the Playoff Odds Report. While that probability doesn't go completely to zero because of this transaction, it will certainly take a big hit, especially as the A's are downgrading precisely when the other contenders in the league are upgrading
More important than the reality may be the perception. Their local press was pretty bad, with Ray Ratto invoking the specter of the White Sox' White Flag Trade of 1997. There's something of an implicit social contract between teams and their fans: a ballclub doesn't have to compete every year, but it does have to look like it's trying.
I can think of only two recent transactions that compare to this. The dataset is small because usually teams in Oakland's position are looking to add talent rather than subtract it. The first is the White Flag deal itself, and the second was the Indians' trade of Bartolo Colon to the Expos in 2002, at which point they were just a game out of first place. Were those teams especially punished by their fans after making those transactions?
The White Flag Trade
First, let's simply look at the White Sox' per-game attendance patterns since they moved into what was then New Comiskey Park in 1991:
Year W-L Attendance 1991 87-75 36,224 1992 86-76 33,101 1993* 94-68 31,865 1994 67-46 30,042 1995 68-76 22,204 1996 85-77 20,696 1997 80-81 23,165<-White Flag 1998 80-82 17,069 1999 75-86 16,529 2000* 95-67 24,047 2001 83-79 21,805 2002 81-81 20,703 2003 86-76 23,945 2004 83-79 23,834 2005* 99-63 28,924 2006 90-72 36,511 2007 72-90 33,141 *: Playoff Year
The two years in the immediate aftermath of the White Flag trade, 1998 and 1999, were the Sox' two worst seasons since crossing the street and moving into their new digs. However, the White Sox have generally been a strong team, and in both '98 and '99 they were marginal. Was their attendance lower simply because they were a worse club? Or were the numbers especially low because the fans were frustrated with the White Flag Deal?
We can get at this question by means of a fairly simple regression analysis, which accounts for the White Sox' win percentage in both the current season and the season previous, as well as a couple of one-time events that might have affected attendance: their honeymoon period following their move to their new ballpark in 1991, the 1994-95 strike, and their 2005 World Series victory. The regression model predicts what attendance "should" be each season based on these variables; we can then compare that against the actual result.
In general, the regression model does quite well for itself; the White Sox have always relied heavily on walk-up sales, and have attendance which is fairly predictable. However, we do see that attendance was below the regression line in 1998 and 1999. In '98, the Sox drew about 3,000 fewer fans per game than we would have projected based on their won-loss record, and in '99, they drew about 1,500 fewer fans. Once they reached the playoffs in 2000-with Keith Foulke, who had been acquired in the White Flag, being a big part of that-fans forgot about the transaction, and the effects were moot.
Losing 3,000 fans per game translates to about 250,000 over an 81-game home schedule. Assuming an average ticket price of $20, that translates to about $5 million in lost income. Then the White Sox were another 125,000 or so fans below their expectation in 1999, which is another $2.5 million in losses. If fans were upset with the club, there were probably also losses in the form of decreased concessions and merchandise sales, and lower TV ratings. So, conservatively speaking, we can probably say that the White Flag trade cost them at least $10 million in fan disgruntlement.
The Colon Deal
Now, let's do the same thing with the Indians, dating back to their move into Jacobs Field in 1994:
Indians' attendance has generally been a little harder to predict, in part because you have the complicating factor of a new park and a strike coming online at the same time; our model just gets confused and drops both variables. Nevertheless, one thing that's been notable is that while attendance was close to its predicted levels in 2003 and 2004, it did not bounce back to anywhere near its former marks once the team got better again. Whether that's because the Colon trade occurred, I don't think we can say. There may simply have been some sort of especially long-term honeymoon effect from Jacobs Field opening, which had worn off by then. Perhaps the marginal economic conditions in Ohio over the past few years have played a role. In general, the Indians have not drawn all that well for a team with their degree of success, and the effect seems to have originated at about the time of the deal.
That is not to say that these were bad deals for the Indians and Sox. On the contrary, the Indians got Cliff Lee out of theirs, who has been as good this year as Colon ever was, while the White Sox' quickly rebuilt themselves into a strong club and won a World Series. But teams do need to be aware that there may be some medium-run damage to their brand when they make a dump trade during a season that they're performing well.