May 7, 2007
The Clemens Signing
With a dramatic seventh-inning announcement, Roger Clemens made himself a returning hero, a difference-maker, and a whole big pile of cash.
Clemens, who was on his way into retirement the last time he was being cheered wildly by a Yankee Stadium crowd, agreed to a one-year, $28-million contract to don pinstripes for the remainder of 2007. The deal is the biggest in baseball history by average annual value, and includes the highest single-season salary ever paid to a major-league player. Clemens won't get quite that much money, as he won't join the Yankees for another month, but two-thirds of that figure, about $18.5 million, makes for quite the incentive to return. The cost is greater to the Yankees; throw in the luxury tax hit on this addition to the payroll, and they're committing about $26 million.
It's not for "four months," however, as has been reported--the Yankees don't make a signing like this without expecting to get an extra month from the player. Signing Clemens addresses the Yankees' biggest problem to date, a starting rotation that has left it behind the eight-ball-or eight-spot-in far too many games. It provides an upgrade from Kei Igawa, who looks overmatched, to a legitimate #2 starter, while also preventing Clemens from providing a comparable upgrade to the Red Sox, who were spurned in the bidding for their former ace.
What kind of impact are we looking at here? Remember that due to injuries, the Yankees have gotten a lot of starts from pitchers who weren't part of their master plan in February. Chien-Ming Wang and Mike Mussina are back, Andy Pettitte never really left. Darrell Rasner pitched well enough yesterday, especially on the heels of Igawa's stinkbomb Friday, to argue for a spot ahead of the Japanese left-hander. Philip Hughes is expected to return a bit after Clemens does. So let's say that Clemens will be replacing Igawa, while Hughes takes Rasner's spot.
Nate Silver's PECOTA system projects Clemens, conveniently, for 21 starts, maybe one or two shy of what he'll make for the Yankees. The system has Clemens posting a 3.34 ERA in a neutral environment, worth 28.3 runs over replacement, and 3.6 wins above replacement. None of that seems out of line; it would make Clemens one of the ten best starters in the AL on a per-inning basis. I can see an argument that the projection is slightly optimistic because Clemens' most recent work was in the NL, and he'd be pitching for an AL East team, but remember that both PECOTA and the Translations system are accounting for that.
PECOTA projected Igawa to post a translated 4.28 ERA in 180 innings, worth 22.9 points of VORP and 4.1 WARP. However, we have additional information on Igawa, his first 30 or so major-league innings, which indicate that those figures are high. Igawa has allowed eight home runs in 30 2/3 innings, leading to a 7.63 ERA. His translated ERA is 7.45, with negative VORP (-11.5) and WARP (-0.7) totals. If you weight the projection and the performance to date, allowing the latter to make up 1/6th of the line, you should get a more accurate projection.
Let's compare the two lines: Clemens' projection, and Igawa's adjusted projection for the two-thirds of the year for which he'll be replaced:
IP H BB SO HR ERA VORP WARP Clemens 127.2 109 39 118 12 3.34 28.3 3.6 Igawa 103.1 111 35 75 14 4.51 11.5 2.2
Is that upgrade-17 runs, a win-and-a-half-worth $26 million? That's hard to believe.
Now, this analysis assumes that Kei Igawa's first 30 2/3 innings in the major leagues are not an accurate reflection of his skill level, the same way that we would discount one month of play for any player with a track record. If you're of a mind to suggest that Igawa is, in fact, a 7.00+ ERA pitcher, then the gap between Clemens and him grows, although at that point you need to compare Clemens to Rasner and perhaps Tyler Clippard and other options, because Igawa wouldn't hold a job at that performance level. It may be that Clemens is two or three wins greater than the Yankees' replacement options in his absence; he also provides insurance against further health problems suffered by the remaining starters.
The specifics of this can be debated. What can't be debated is this: the impact of any starting pitcher, over two-thirds of a season, is limited by opportunity. The Yankees haven't signed vintage Roger Clemens, or even the version that won a Cy Young Award in 2001 while pitching them to the World Series. This version will only make 21-23 starts. It will most likely average six innings a start, perhaps less. It relies more on command and keeping the ball in the park-Clemens' translated walk and home-run rates as as Astro were his lowest since his Blue Jay years-than on whiffing batters, as his translated strikeout rates while in Houston were career lows.
Adding Roger Clemens makes the Yankees a better team, and improves their chance to catch the Red Sox in the AL East, or secure the wild card if they can't do so. They're not as bad as they've looked in the season's first five weeks, and the Red Sox are probably not quite as good as they've looked. (More on this later in the week.) However, the impact of Clemens is smaller than is perceived, because of three factors:
Here's the best argument for the Yankees spending $26 million on Clemens: he was probably going to the Red Sox if they didn't. So if you're the Yankees, you haven't just made a two-win, maybe three-win, improvement; you've prevented the Red Sox from doing about the same, given Clemens' edge over Julian Tavarez or a rehabbing Jon Lester. Taking that into consideration, and the signing is worth four to six wins, which makes it both economically sensible and gives it a greater potential to impact what should be a very good race in the AL East.
The real winners here are the fans, who get a couple dozen more chances to see the greatest pitcher of our lifetimes. Who says the Yankees are bad for baseball?