Imagine you’re a major league general manager, lying in bed at night dreaming of a pitching staff that will bring you an elusive World Championship. Perhaps you fantasize about signing an ace starting pitcher like Randy Johnson or Pedro Martinez to win that one big game in October. Maybe you’re working on a smaller budget, and you think of signing two or three mid-market starters, the sort of guys who win 12 games a year and fill out the back of your staff. Or you might want to focus on building a dominant bullpen to shut the door in close games down the stretch. So many pitchers, so many choices: How to get the most bang for your buck?

There’s no single correct answer to that question. But in 2004 at least, every playoff team shared one trait, no matter how good or bad their aces were – they got great performances from pitchers who weren’t among their top two starters.

For the purposes of this article, each pitching staff in the game last year was broken down into three groups – the top two starting pitchers, the bottom three starting pitchers, and all other pitchers on the staff. Each group was rated according to Value Over Replacement Player, a Baseball Prospectus metric that calculates the total number of runs prevented by a pitcher above what a freely available player (say, Scott Erickson) would produce. Breaking down the numbers this way shows you that there are several routes to success when it comes to your starters.

Let’s look at the rotation that led the Red Sox to their first title since 1918:

You really can’t ask for much more than two aces like Curt Schilling and Pedro Martinez at the top of your rotation. The Boston duo was the second best pair of top starters in baseball last year (only Johan Santana and Brad Radke of the Minnesota Twins posted a better combined VORP). But the back of the rotation wasn’t much to brag about in Beantown. Derek Lowe, Tim Wakefield, and Bronson Arroyo were only 15th-best when compared to other teams’ bottom three starters, with most of the fault resting on a brutal season by Lowe.

The Houston Astros were another team that used this studs-and-scrubs approach to the rotation. The front of the Astros’ rotation, Roger Clemens and Roy Oswalt, was the fourth-best in the game. But their back three starters – Wade Miller, Pete Munro, and Tim Redding – were even weaker than Boston’s, only good for 21st place.

Luckily for both the Sox and ‘Stros each team featured a dominant closer – Boston’s Keith Foulke ranked seventh among major league relievers with a 35.9 VORP, while Houston’s Brad Lidge led the N.L. at 39.0. The Sox also got excellent performances from Alan Embree and a handful of other relievers, and their bullpen ultimately played a decisive role in the team’s dramatic ALCS win over the Yankees.

Meanwhile, the National League champs in St. Louis featured a very different sort of staff:

Jason Marquis and Chris Carpenter‘s combined VORP was only the 12th-best in baseball for a team’s top two starting pitchers, trailing a bunch of clubs whose only action in October was bass fishing. But those bottom three starters were much stronger: Matt Morris, Woody Williams, and Jeff Suppan were the fourth-best in the game in combined VORP from the bottom three rotation spots.

The Atlanta Braves relied on a similar kind of depth to extend their streak of division titles in the NL East. The Braves’ top two starters (Jaret Wright and Russ Ortiz) were only 19th-best, but the bottom three (John Thomson, Mike Hampton, and Paul Byrd) were fifth-best among all teams. Like the Cardinals, the Braves might not have overwhelmed you at the top of their rotation, but there wasn’t a #4 or #5 starter who killed them, either.

The N.L. West-winning Dodgers used a third route to pitching success. Los Angeles got average performance across the board from their starting rotation: The top two starters (Odalis Perez and Jeff Weaver) were the 13th best in baseball, while the bottom three starters (Kazuhisa Ishii, Jose Lima, and Hideo Nomo) were just 20th. But the remainder of the Dodgers’ staff, led by ace closer Eric Gagne (28.2 VORP), was the most effective in baseball, racking up a VORP nearly 20 points higher than any other team’s spot starters and bullpen.

Right down the freeway in Orange County, the Angels succeeded with a similar emphasis on bullpen and swingmen. Their top two starters were 16th in VORP and their bottom three ranked eighth, but the rest of the staff was second only to the Dodgers; closer Francisco Rodriguez ranked as the majors’ fourth most valuable reliever, with a 37.6 VORP.

The Yankees were an interesting case in 2004.The team’s two most effective starters were Orlando Hernandez and Jon Lieber, whose collective 54.9 VORP ranked them near the bottom of the majors. Meanwhile the bottom three starters – Kevin Brown, Javier Vazquez, and Mike Mussina – were stronger, making the top 10.

The remainder of the Yankees’ staff was the eighth-best in the game, led by the dominating duo of Mariano Rivera and Tom Gordon, elite relievers who both easily topped the overall production of any of the staff’s starters, a rare feat given the difference in innings pitched. Gordon’s 39.6 VORP for the season was the highest of any reliever in baseball, while Rivera’s 37.9 was third. Of all the teams that played in the postseason last year, the Yankees had the least effective pitching staff as measured by VORP, but the strength of that staff was at the back end.

So, does all this mean that teams should ignore the front of their rotations, and instead focus their attention on building a killer bullpen? Of course not.The best bullpen in the world can’t do a thing if your starters are getting torched.

But given the general availability of good bullpen arms and swingmen at a lower cost than front-of-the-rotation starters, it might make sense to turn the traditional method of building a staff on its head, and instead try and work from the back of the game to the front, rather than the other way around. For teams flailing around without much of a plan and little money, it certainly couldn’t hurt.

This article originally appeared in The New York Sun newspaper. Baseball Prospectus contributes two articles a week to the Sun throughout the season. You can read those and other articles at

Thank you for reading

This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.

Subscribe now
You need to be logged in to comment. Login or Subscribe