Can Rocket rescue the Yankees? Simulate it
Roger Clemens is due to make his first Yankees start of 2007 against the Pirates on Saturday. The Bombers, winners of four straight games, are still just 28-31, 10 games back of the Red Sox in the AL East and 5½ games behind the Tigers in the wild-card race.
Imagine SportsThe simulations were done using the Diamond Mind Baseball software, which was developed by renowned baseball statistics expert Tom Tippett. He started developing Diamond Mind more than 20 years ago and has been refining it ever since. Diamond Mind is now widely regarded as the most sophisticated and realistic baseball simulation software and was named PC Magazine's "Editor's Choice" for pc-based baseball software.
Each year, the Diamond Mind team uses the software to predict the upcoming MLB pennant races and successfully picked five of the six division winners in 2006. Diamond Mind is now owned by Imagine Sports, a Silicon Valley-based Internet company developing multiplayer online simulation games, including Total Baseball, an online baseball management game utilizing Diamond Mind that it launched last year.
ESPN.com asked us to project how the Yankees would fare over the remainder of 2007, with and without Clemens. That is a trickier task than it sounds because, although we're pretty good at simulating and projecting the results of games played by real players and teams, past and present, against each other, we can't literally see into the future.
So, for example, we might think there's a pretty good possibility that Mark Teixeira will be playing first base for the Yankees, and Eric Gagne will be closing for the Tigers by July 1, but we can't assume such things for purposes of our simulations. Kenny Rogers may come back strong for the Tigers, or Jason Giambi for the Yankees, but all we know is that right now, and for the indefinite future, they're out of action and not contributing.
So what we did was take the 30 major league teams as they stood at the close of play after June 4, and played out the rest of the AL schedule (from June 5 onward) 200 times. In 100 of those seasons, the Yankees had Clemens in their rotation; in 100, they didn't.
The first question our simulations answered wasn't whether Clemens could save the Yankees' season, but whether that season needed saving. And indeed it did. At the close of play on June 4, the Yankees were 24-31, 12½ games behind the Red Sox in the AL East and 7½ games behind the Tigers for the wild card. Without Clemens, over 100 simulated seasons, they averaged 60.5 wins and 46.5 losses the rest of the way, for an average aggregate record of 84.5 wins and 77.5 losses, finishing on average about 12 games behind the division-winning Red Sox and five games behind the Tigers for the wild card.
Adding Clemens to the rotation, the Yankees averaged 65 wins and 42 losses the rest of the way, an improvement of 4.5 wins. That wasn't enough to catch the Red Sox, who still finished about six games in front, but brought them to within less than one win of the Tigers on average for the wild card.
The Clemens Effect
After 200 simulations, Imagine Sports has some numbers on how much Roger Clemens will help the Yankees.
- Average Clemens record:
9-5, 4.24 ERA
- Yankees record with Clemens:
- Yankees record without Clemens:
- Finish with Clemens:
(Tied with Tigers for wild card)
- Finish without Clemens:
(5 games behind Tigers for wild card)
Clemens averaged 21 starts (which would lift him to fifth all-time in number of games started, passing Tommy John and Steve Carlton) in our simulated seasons. He posted an average record of 9-5 (which would give him 357 career wins) with a 4.24 ERA.
Strange things can happen in a baseball season, even more so over the last four months. Simulating and averaging the results of 100 seasons will filter out the more extreme and remarkable results. In 1978, the Yankees overcame a 14-game deficit on July 19 to tie the Red Sox for the division lead, and then beat them in a one-game playoff on Bucky Dent's home run over Fenway's Green Monster. And, in our 100 simulated seasons with Clemens, they managed to overtake the Red Sox and win the division 12 times. (They also tied twice, which would have necessitated a one-game playoff in each case, since their records were worse than the wild-card winner in each of those seasons.)
The most dramatic turnabout was in Season 76, in which the Yankees roared home to finish with a 100-62 record, a whopping 12 games ahead of the Red Sox at 88-74, with Clemens going 12-2 to lead the way. On the other hand, they managed to edge the Red Sox by one game in Season 7, despite the Rocket going just 5-4, while his best effort (16-1, 2.67) still wasn't enough for the Bombers to overtake Boston in Season 65.
Getting back to the overall averaged results, with the 4.5-win improvement contributed by Clemens, the Yankees finished 2007 with an average of 89 wins, which put them in a virtual tie for the wild card with the Tigers, who averaged 89.6 wins for the season. We played a one-game playoff between the Yankees and Tigers 1,000 times, with Clemens starting for the Yankees and Jeremy Bonderman for the Tigers, and the Yankees won 572 times.
One way of looking at these results would be like this: Each game our simulations say Clemens will add to their final regular-season win total will cost the Yankees about $4 million. Since, however, they don't reach the postseason without his contribution, he may very well be worth it.
MORE MLB HEADLINES
- Ortiz: Yankees lost 'face of ballclub' in Cano
- Sources: Yankees open to trading Gardner
- Yanks announce deals for Ellsbury, Kuroda
- Dodgers finalize Wilson deal, eye pen help