A Better Way To Measure Pitchers’ Performance
This article is from the archive of The New York Sun before the launch of its new website in 2022. The Sun has neither altered nor updated such articles but will seek to correct any errors, mis-categorizations or other problems introduced during transfer.

It’s a baseball truism that pitching wins championships, and the battle for October raged throughout the past winter. The Yankees replaced over half of their rotation; the Red Sox brought in three new starters; Pedro Martinez is a Met. In all, major-league teams have invested nearly a half a billion dollars on starting pitchers since October. With that in mind, it’s worth asking who has the best starting five in baseball.
Many conventional pitching statistics can be deceiving, none more so than win-loss records. A starter’s record is based on not only his own performance, but also that of his team’s offense, defense, and bullpen. There were 58 games last year where a pitcher threw at least eight innings, allowed one run or less, and did not receive a win. On the flip side, there were nine games in which a pitcher threw fewer than six innings, allowed at least five runs, and still eked out the win.
To correct this problem, Baseball Prospectus has developed a stat called “Expected Win-Loss,” an estimate of how many games a pitcher would have won and lost had he been given average run support, average bullpen support, and an average defense behind him. Expected Win-Loss removes the biases inherent in the pitcher decision system; instead, pitchers are credited with the average wins proportional to their performance.
For example, Jarrod Washburn was tagged with a loss for his September 9 effort against Texas, in which he pitched eight innings and allowed no earned runs. Expected Win-Loss, on the other hand, compares Washburn’s outing to all starters who allowed no earned runs over eight innings in 2004. If those pitchers’ teams won 95 out of 100 games, then Washburn is credited with 0.95 wins and 0.05 losses. Assigning fractional wins and losses based on league averages credits the pitcher for innings pitched and runs allowed, but does not unduly punish or reward him for run support, defense, or bullpen performance.
According to this metric, the Cubs top the list – hardly a surprise with a rotation anchored by three top arms in Kerry Wood, Mark Prior, and Carlos Zambrano. With the crafty Greg Maddux and the surprising Glendon Rusch bringing up the rear, this is a deep staff.
Prior spent nearly half of 2004 on the DL with an Achilles injury, but he pitched five great starts in September and has looked healthy thus far this spring. There are serious questions about Wood’s vague triceps problem that cost him two months last year, but the Cubs hope that both he and Prior will return to pitch substantial innings in 2005. If they do, look for the Cubs to dramatically increase their Expected Win-Loss totals.
The Yankees’ rebuilt rotation looks to be the best in the AL, but there are rumblings of concern in the Bronx. All three major acquisitions last winter – Randy Johnson, Jaret Wright, and Carl Pavano – are major question marks heading into 2005.
Johnson has compiled a record of performance that places him among the greatest southpaws in the history of the game, but his advanced age may finally catch up with him. Wright enjoyed a career renaissance in 2004 with Braves pitching guru Leo Mazzone, but before his 3.28 ERA last year, Wright had amassed a career mark of 5.68. Pavano struggled along with a 4.59 career ERA before his sudden breakthrough with a 3.00 ERA in 2004.
Baseball Prospectus’s Pecota projection system sees Wright with a 4.57 ERA this year and Pavano with a 4.51, a flag of caution that the Yankees may have in vested too much on too short of a performance record. With the likelihood of their regression to their previous career performances and Mike Mussina struggling to an Expected Win-Loss of 8.0-10.9 last year, the real bombers in 2005 may reside in opponents’ lineups.
If the Cubs are the new Big Three, the Houston Astros are the new Boston Braves. The old motto “Spahn, Sain, and pray for rain” is now “Oswalt, Clemens, and score at least 10.” Roger Clemens and Roy Oswalt combined for more Expected Wins than any two teammates in 2004, with Clemens taking home his record seventh Cy Young award in the process.
Don’t bet against them repeating that trick in 2005, but the rest of the Astros rotation – Andy Pettitte, Brandon Backe, and Peter Munro – combined for an Expected Win-Loss record of 12.2-13.5 in 2005.Pettitte has reportedly recovered from a flexor tendon injury that cost him most of 2004; his healthy return is the Astros’ best hope for returning to the postseason.
The Braves place fourth in Expected Winning Percentage from 2004, but their win total is one of the lowest in the league because former closer John Smoltz didn’t start any games. The Braves hope that he can regain his old form, though there are serious concerns about Smoltz’s stamina, especially considering his reconstructed elbow and the forearm tendonitis that bothered him in late 2003.Yet Smoltz was one of the league’s top starters for nearly a decade, so if any reliever can make the transition, he can. Alongside the newly signed Hudson and an underrated supporting cast of Mike Hampton, John Thomson, and Horatio Ramirez, Smoltz should make the Braves’ rotation the best in the pitching-rich N.L East.
Rounding out the top five is the Minnesota Twins, led by Johan Santana. The A.L. Cy Young winner notched the best Expected Win-Loss record in 2004 with a 16.0-8.4 mark. He’s followed closely by Brad Radke, re-signed by the Twins early in the off-season for an economical $18 million over two years. Radke placed fourth in the AL in ERA and fifth in innings pitched, giving the Twins a 1-2 punch unmatched in the junior circuit.
The only question in the Twins’ rotation is whether Joe Mays can recover from his 2003 Tommy John surgery to contribute. Full recovery from the now regular procedure can take up to two years, so the Twins will count on Carlos Silva, Kyle Lohse, and top prospect J.D. Durbin to pick up the slack.
Each of the top five rotations has significant risks attached, usually involving injuries or past performance issues. If trouble strikes, the Marlins – with veteran Al Leiter and a trio of top young pitchers in A.J. Burnett, Josh Beckett, and Dontrelle Willis – could vault to the top of the league. But in March, when everyone’s healthy, the Cubs are the leader in baseball’s arms race.
Mr. Click writes for Baseball Prospectus.For more state-of-the-art baseball content, visit www.baseballprospectus.com.