What's Roger Clemens Worth?
by David Gassko
June 02, 2006
He finally made his decision. After keeping his suitors in limbo for months, teasing the Astros, Red Sox, Yankees, and Rangers with carefully placed rumors and visits for months, Roger Clemens finally decided to keep with the status quo and return for a third season with Houston.
Itís a decision that makes sense. Clemens lives near Houston, so he can stay close to home like has always wanted to. He has a chance to play with his son Koby, first in a couple days at Single-A Lexington, and possibly in the major leagues at the end of the season. And Houston is really the only place where Clemens could have gone without a huge hullabaloo. Switching teams would have been cause for press conferences, great media coverage, and lots of questions. Here, Rog-ah basically just took a two-month sabbatical.
The money ain't bad eitheró$22 million over a full season, which will be about twelve-and-a-quarter over the three-and-a-half months Clemens plays. But here's the question: Is he worth it? $22 million is a lot of money; can the Astros expect to make back their expenses, or were they too generous in trying to woo Clemens back?
Let's look at this from all possible angles. First, let's find out if Roger will be worth the money coming his way simply through his regular season performance. Second, let's figure out his effect on Houston's postseason chances. Third, let's see how many extra fans he'll put in the seats.
Roger Clemens is pretty good. Okay, that's an understatement. Roger Clemens is the greatest pitcher of the modern era. That is most certainly the truth. Roger Clemens is the greatest pitcher of all-time. Quite possibly. But what can we expect from him as a 43-year old? My stock answer would be, "There have been few successful 43 year-old pitchers in baseball history, and almost no starters. Even Nolan Ryan threw 35 less innings and saw a quarter-run rise in his ERA at 43."
Of course, Clemens is no ordinary pitcher. Clemens has in fact improved as he has aged, leaving quite the conundrum for those trying to project his statistics. Is he really as good as he has looked the past two years? Or will the other shoe drop? There's no simple answer to that question, and lucky for you, there's no complex answer either.
Clemens has shown some human traits over the past few years, with significant drops in performance as the season wears along. With 4,700 major league innings on his arm, that's to be expected, but the difference between a fresh Clemens and the tired version is staggering. His post All-Star break ERA has been three-quarters of a point higher than his ERA prior to the Midsummer Classic. Here's a breakdown of his month-by-month ERAs between 2003-05:
Staggering, huh, that his ERA is almost a point-and-a-half higher in September than it is in April? But by sitting out the first two-and-a-half months of the year, Clemens hopes to avoid the downfall that he has met each of the past few years.
The Astros hope that Clemens' first start will come on June 22, which would put him on pace for 18 starts. How has he done in his first 18 starts of the season these past few years? Pretty darn good:
Year ERA Innings
2003 3.75 117.67
2004 2.62 116.67
2005 1.48 122.00
Average 2.60 119.00
If Clemens again drops his ERA by over a point, he might just allow negative runs. Yes, he's great, but maybe not that great. Let's say instead that he pitches to his three-year average. It's a simplistic way of looking at things, but probably not too far from the truth. His innings totals have been remarkably stable each of the past three seasons, and once we adjust for aging, regression to the mean, etc., etc. I bet we would end up with a projected ERA of close to 2.60.
So how much would that performance be worth? Well, an average pitcher might put up a 4.30 ERA in those innings, making Clemens around 22.5 runs better than average, or around 2.2 wins. How much is that worth? Well, the average pitcher gets around $5 million a year, which is like $2.8 million when adjusted for Clemens' playing time, and the average free agent gets around $3.5 million per marginal win, which puts Roger at $7.7 million above average, or $10.5 million in total value.
Clemens will actually be paid $12.25 million, and though that's a huge amount for three-and-a-half months of work, he may just be worth it. That $10.5 million number isn't far less, and there are still facets we have yet to take into account.
For those with short memories, let me remind you that the Astros last year, with basically the same roster as this season, with Clemens anchoring the rotation, played in the World Series. They didn't play particularly well, but the games were hard-fought, and hey, just being there put them in the top 7% of all major league teams.
Without Clemens, the Astros have compiled a .500 record (through Wednesday night), with an underlying performance even worse than that. The following is a chart detailing the predicted runs scored and allowed for every National League team, along with their true records, predicted records, and projected seasonal records:
Note that the Astros are projected for all of 78 wins. Worse still, that number sounds about right. When SG of the Replacement Level Yankee Blog ran his preseason projection blowout, simulating 2,000 seasons with each of three different systems, he found that the Astros were projected to win Ö 77 games. But how does Roger Clemens change that expectation?Code:Team Games Wins Losses BsR BsRA pWin pLoss sWin sLoss ARI 52 30 22 268 234 29 23 92 70 ATL 53 28 25 267 259 27 26 84 78 CHN 52 20 32 202 255 20 32 63 99 CIN 53 29 24 274 277 26 27 83 79 COL 53 27 26 240 232 27 26 83 79 FLA 51 17 34 235 258 23 28 68 94 HOU 54 27 27 251 269 25 29 78 84 LAN 53 30 23 281 243 30 23 92 70 MIL 53 26 27 273 272 27 26 81 81 WAS 54 22 32 243 262 25 29 72 90 NYN 52 32 20 271 226 31 21 97 65 PHI 52 27 25 265 270 26 26 81 81 PIT 53 19 34 255 277 24 29 69 93 STL 53 34 19 262 224 30 23 96 66 SD 53 28 25 228 232 26 27 82 80 SF 53 27 26 237 235 27 26 82 80
Clemens will probably replace Fernando Nieve, who is sporting a 5.36 ERA in 47 IP, averaging less than 5.5 IP every time out. With the Astros bullpen ERA at 4.72, that means the 'Stros are expected to allow 5.11 earned runs every time Nieve starts. With Clemens, that number will go down almost 2 full runs, to 3.17. That's a difference of about three-and-a-half wins, which would bring the Astros to 81-82 wins.
How many wins does it take to get the Wild Card? The five-year average in the National League is 92; 91 over the past three years. Let's just split the difference here and say that the Astros need to win 10 more games than expected to have a good shot at making the playoffs. What is the likelihood of that?
I don't want to get too technical, so will you take my word that it's about 2.68%? If you want the math behind this, I'll simply say that I calculated the binomial random variance we can expect in a team's won/loss record over the remainder of the season. Without Clemens, there chances were 0.47%.
Let's say the average playoff team plays eight game in the playoffs. So Clemens just got the Astros an extra .18 playoff games, maybe a little more because they have a better chance of actually winning in the postseason with Rog-ah anchoring the rotation. So let's say he's giving the 'Stros an extra two-tenths of a playoff game; how much is that worth?
Assuming that Houston would sell out its playoff games, one game would add $1.1 million in ticket revenue alone. Forbes tells us that the Astros made 2.34 times more money last year in overall revenue than from gate receipts, so following that math, Houston would be getting $2.55 million extra for each playoff game, making Clemens' playoff worth a little of $500,000.
In short, Clemens improves the Astros' playoff chances, but not by much. Incorporating his playoff-bonus worth, though, Roger's value goes up from $10.5 million to $11 million.
Effect on Attendance
Darren Rovell estimates that the Clemens signing will mean 27,410 more fans at Minute Maid Park this season. With an average price of $26.66 a ticket, that means that the Astros will see about $750,000 extra. But we also have to account for concessions, parking, and merchandising sales. Let's take a stab in the dark and say that doubles the amount of money the 'Stros will make with Clemens in uniform. That puts the attendance effect at $1.5 million, for a total of $12.5 million in value being provided by Clemens.
Summing it All Up
This is not meant to be a to-the-decimal calculation. If you told me that Clemens was actually worth $12 million or $13 million or any reasonable number, I'd nod my head and agree. What's important is that despite his enormous contract, Rog-ah does not seem to be overpaid. The Astros did their math about right, and ended up with a pretty good deal.
Another interesting thing to me was how little Clemens affected Houston's playoff chances and attendance. His main value is derived from being good, and man is he ever. Clemens may very well be the greatest when it's all said and done, and if he is, he'll have cemented his legacy closer to home. Both when he went to closer to home in Toronto, and now, only miles away from where he lives in Houston.
Rog-ah will always be a controversial pitcher, but he's undeniably one of the best out there, even at 43. So no matter how exorbitant his contract, he's certainly worth it.
References and Resources
Some might feel that impact on playoff odds and attendance is already included in the marginal wins to dollars calculations. This is incorrect. Imagine an average pitcher signed to an average team. He changes neither their playoff odds nor their attendance, yet we still do the marginal dollars calculation for him. And when we use some kind of a replacement level, those marginal dollars will be multiplied by a coefficient equal to average minus replacement.
Thus, Clemens' impact on Houston's attendance and chances of making the playoffs must be measured separately; it is not included the in dollars per marginal win calculations.