Not every player who is fast is also a good leadoff man. Willie Randolph fell into this trap last season when he gave Jose Reyes leadoff duties and stuck with him throughout the season despite the fact that Reyes led the major leagues in outs by a fair margin. Randolph’s initial 2006 batting order suggests he is incapable of learning from experience.

Traditionalists will still tell you that the batting order is about setting up innings. Your leadoff hitter should be fast in order to steal bases. The no. 2 hitter should be a guy with the ability to handle the bat so the manager can hit and run. The no.3 hitter is the team’s best hitter, the cleanup hitter a pure slugger, and so on down to the no. 8 hitter, the team’s worst before the pitcher.

This turns out to be an overly simplistic way of thinking about things. The real function of the batting order is to distribute playing time. We normally think of playing time as something the manager divvies up among his players in terms of games played, but games are too big a unit of measure; there are levels of playing time even within games. If Xavier Nady starts in right field and then is replaced by Endy Chavez in the seventh inning, both are credited with one game played. Chavez, though, might get one at-bat or none, while Nady gets three. All games are not created equal.

Similarly, a player’s position in the batting order determines how many times he comes to the plate during the course of the season. The leadoff hitter will bat more often than any other player on the team. The no. 2 hitter will bat more often than anyone but the leadoff man, and so on. The issue of whether the fastest player on the team will be in a position to steal bases is unimportant compared to the need to get, say, Cliff Floyd, as many plate appearances as possible. Last year, Floyd hit a home run once every 18.4 plate appearances. If he homers at a similar rate in 2006,moving him up from the sixth spot to the fifth would get the Mets an extra home run. A few stolen bases from Reyes don’t necessarily mean anything. An additional home run is one more possible game-winner.

A complementary purpose of using the batting order to get more plate appearances for the best hitters is to use it to minimize the exposure of the worst hitters. Last season, Reyes made 536 outs – meaning he personally consumed 20 games worth of outs for the Mets while creating only about 80 runs. In terms of fuel efficiency, Reyes is the baseball equivalent of a stretch Hummer, but it’s not all his fault. Randolph took a player who is still learning to hit and used the batting order to put him in a position to make outs more often than anyone else in the major leagues.

It’s not simply that Reyes is a weak leadoff hitter, he’s a weak hitter, period. Among players with more than 250 plate appearances last season, Reyes ranked 267th in on-base percentage. You can’t steal a base without getting to first on your own, much less score a run. There were nearly 300 players in baseball better at this than Reyes.

This cost the Mets games last year. A better leadoff man would almost certainly not have erased the seven game gap between the Mets and the Braves, or the six-game gap between the Mets and the Astros, but things would have been closer. Even Carlos Beltran, who was disappointing in his first year with the Mets, would have created an additional 10 runs if given as many plate appearances as Reyes received. Generally speaking, a team picks up a win for every 10 runs of offense that it adds.

This year that one win might actually matter. The Mets, Braves, and Phillies all have a fair shot at the NL East title. It behooves the Mets not to throw away runs because of a lack of strategic clarity.

There is nothing revolutionary about this evaluation of Reyes. You don’t need to be Billy Beane or a “Moneyball” believer to understand that Randolph is hurting the Mets with stubbornness; a .300 on-base percentage is a .300 on-base percentage. Toby Hall had a higher OBA than Reyes last year. So did Junior Spivey, Michael Tucker, and Terrence Long – none of whom have jobs right now.

Around baseball, teams are learning not to overvalue this kind of player. Speed kills. Only Randolph and the Mets are clinging to antiquated ideals (the Reds, leading off Tony Womack, don’t qualify as having ideals).

Randolph was a terrific player, ironically one with great leadoff skills. Few of his contemporaries posted a higher on-base average than his own career .373 rate. Given a full season (which he was rarely healthy enough to play), he was good for 90 walks. His career high was 119 in 1980, when the Yankees won 103 games. Randolph scored 99 runs that season, the same number Reyes did in 2005, while using 91 fewer outs.

It is fascinating how many stars of Randolph’s generation – Joe Morgan, Don Baylor, Dusty Baker, Randolph himself – missed the point of their own careers. It’s going to cost the Mets this year if Randolph doesn’t wise up.

Mr. Goldman writes the Pinstriped Bible for www.yesnetwork.com and is the author of “Forging Genius,” a biography of Casey Stengel.