ESPN.com - Simulations predict Yankees, Red Sox victories

ESPN.com: MLB Playoffs 2007

Tuesday, October 2, 2007
Simulations predict Yankees, Red Sox victories

By Charles Wolfson and Luke Kraemer, Imagine Sports
Special to ESPN.com

The 2007 regular season was a topsy-turvy one. Not one division winner from 2006 came out on top again in 2007, and only the Yankees made it back to the postseason. The Cubs went from worst to first, battling the upstart Brewers until the final weekend; the Phillies and Rockies treated their fans to spectacular late-September runs; and New Yorkers experienced the agony and the ecstasy of the Mets' historic collapse and the Yankees' tremendous comeback.

Imagine Sports

The simulations were done using the Diamond Mind Baseball software, which was developed by renowned baseball statistics expert Tom Tippett. 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.

Diamond Mind is now owned by Imagine Sports, a Silicon Valley-based Internet company developing multiplayer online simulation games, including Diamond Mind Online, an online baseball management game licensed and promoted by MLB Advanced Media.

We've employed our Diamond Mind simulation software to project the results of the upcoming League Divisional Series. Before revealing our projected results for the ALDS, here is a brief explanation of how we did it:

• We updated our projections and ratings for each player based on his 2007 regular-season performance.

• Although playoff rosters were not yet finalized, we used our best judgment as of Sunday night about whom the teams would select, the starting rotations, the batting orders, and the bullpen and bench roles.

• We made judgments about players carrying an injury into the postseason. Beyond deciding, for purposes of setting these roles, whether or not a player would play, we did not attempt to simulate the degree to which such an injury might hamper the player, except to the extent that it already was reflected in his regular-season performance.

• Then we played each series 1,000 times.

As unpredictable as the outcome of the regular season can be, luck plays an even greater role in a short series. Nevertheless, the results of our simulations project the Yankees and Red Sox as heavy favorites to face off again in the League Championship Series.

The Yankees won 572 of the 1,000 series against the Indians, with the most frequent outcome being the Yankees in four games:

YANKEES VS. INDIANS
Winner/Games	NYY in 3	NYY in 4	NYY in 5	CLE in 5	CLE in 4	CLE in 3	Overall
NYY-CLE	147	248	177	210	133	85	572-428

The key difference was the inability of the Indians' right-handed starters, particularly Jake Westbrook (5.06 ERA) and Paul Byrd (6.75 ERA), to handle the Yankees' bats. Alex Rodriguez was the hitting star for the Yankees, shrugging off his past misfortune in the postseason.

The Red Sox were even more dominant against the Angels, led by David Ortiz (1.044 OPS) and Mike Lowell (.915 OPS), winning 591 series with a very large plus-1.7 average series run differential:

ANGELS VS. RED SOX
Winner/Games	BOS in 3	BOS in 4	BOS in 5	LAA in 5	LAA in 4	LAA in 3	Overall
BOS-LAA	156	245	190	165	153	91	591-409

The magnitude of Boston's plus-1.7 average run differential can be better appreciated if it is extrapolated over a full season. On a per-game basis, the Red Sox averaged 4.36 runs over the 1,000 series and the Angels 3.94 per game, which is equivalent to 706 and 639 runs, respectively, over a 162-game season. A 706-639 run differential equates to a record of 89-73, a .550 winning percentage, using Bill James' Pythagorean formula for projecting team wins based on runs scored. Sure enough, Boston's aggregate record against the Angels over the 1,000 simulated series was .549 (2,256-1,852).

Of course, one big caveat is that anything can happen in a short series, especially a best-of-five series. Sometimes, a single play can seem to turn a series, such as Derek Jeter's defensive gem to cut down Jeremy Giambi at home plate late in Game 3 of the 2001 ALDS. And any player can be a hero in a short series -- there are many examples of that in postseason lore, the most recent being David Eckstein's MVP performance in last year's World Series. Here are some of the more amazing individual performances from our simulation runs:

Alex Rodriguez

Third Base
New York Yankees

Profile

2007 Season Stats
GM	HR	RBI	R	OBP	AVG
158	54	156	143	.422	.314

• Rodriguez belted six home runs in the ALDS -- twice, in the 328th and 849th runs of the Yankees-Indians series.

• The Indians' C.C. Sabathia tossed two complete-game shutouts in the 16th simulation run.

• Curt Schilling matched that feat in run No. 892 of the Red Sox-Angels series.

• Chone Figgins ran wild with seven stolen bases in three of the simulation runs against Boston.

Another big caveat is who wins Game 1. If the Angels or Indians win that crucial first game, the odds swing dramatically.

When we ran our simulation again assuming that the Indians and Angels took their respective Game 1s, they became the favorites, with the more dramatic swing being in the Red Sox-Angels series. Our simulation runs show the Red Sox with a 59 percent chance of winning that series. But when the Angels win Game 1, the odds swing to 67 percent that the Angels will win. It appears that Josh Beckett is the difference -- beating him in Game 1 leaves the Red Sox quite vulnerable. The Angels also gain homefield advantage with a Game 1 victory, something the Indians already have entering their series with the Yankees.

After all is said and done, however, our Diamond Mind simulation says the winners will be the Red Sox and the Yankees, both in four games, setting up another epic battle between these rivals … but all bets are off if either team loses Game 1.