Some managers are more into mysticism than the airiest new age guru. They put more stock in yesterday’s successes than they do in today’s disasters, make decisions based on intangibles like experience, and cherrypick the evidence for ideas that support their prejudices and reject those that don’t.

No, this isn’t another lecture about government. It’s a discussion of playoff pitching possibilities. The Mets and the Yankees each have a starting pitcher who has — by traditional measures—been successful. Each team will probably try to give that pitcher a prominent role in the playoffs.

The Yankees’ Randy Johnson and the Mets’ Steve Trachsel were big winners this year: Johnson, who will miss his last start with a “locked-up” back, finished with 17 wins; Trachsel currently has 15. Trachsel will likely be the Mets’ Game 3 starter in the division series. Manager Willie Randolph has hinted that the veteran might be slotted ahead of the inexperienced John Maine.

Never mind that without Maine the Mets might not have run off with the division. Since the All-Star break, Maine has pitched 68 innings with a 3.31 ERA. An apparent “throw-in” on the Kris Benson deal, Maine gave the Mets a consistent starting pitcher while Martinez and Tom Glavine dealt with injuries.

Trachsel has a 5.40 ERA during the same period. His season has been a gift from the Mets’ offense (the National League average ERA is 4.90). For half the season he’s been a half-game worse than the typical pitcher and yet he’s added seven wins because the Mets have supported him with nearly seven runs of offense a game.

In a sense, what Trachsel has done is give the Mets exactly what they’ve needed. If the team is more than likely to score seven runs, then a guy who allows five or six isn’t so bad. The problem, of course, is that offense isn’t predictable, especially since the Mets will be facing the best pitching in baseball during the playoffs. The National League playoff picture is unsettled, but if the season ended today, the Mets would play Los Angeles in the first round. The Dodgers don’t have great pitching, and a potential Game 3 matchup of Trachsel against Brad Penny — a Cy Young contender at the All-Star break who’s pitched terribly since — could turn into a dangerous slugfest.

The only argument for starting Trachsel over Maine is his win total and his (shaky) track record, neither of which say anything about his ability to actually pitch. A win isn’t so much an evaluative statistic as a description of a certain kind of event — the pitcher lasted five innings and left with a lead. It doesn’t say that the lead was 1–0 or 14–13.

Randy Johnson has received the best run support in baseball this year, 7.51 runs per game. With run support like that, Minnesota’s Johan Santana would go 30–0, but Johnson has lost 11 games and won 17. His ERA is 5.00; the league’s is 4.94.

Right now the Yankees hope that Johnson will be healthy and effective by the Game 3 of the first round so they don’t have to start Jaret Wright or Cory Lidle in that spot. Yet it’s uncertain whether Johnson is a better option than those very mediocre pitchers since Johnson has himself been below average. Whether the root cause is his bad back or simply his age, Johnson has been hammered during about half this past season.

The Yankees do have better reasons for crossing their fingers and starting Johnson than the Mets do Trachsel. Trachsel is that deadly combination of pitch-to-contact stuff with a tendency to give up fly balls. This means not only that hitters can put balls in play against Trachsel, but that better hitters can hit them out of the park. Worse, Trachsel walks four batters per nine innings, meaning there’s a good chance there will be runners on base when the home runs come. All of this is expressed in his 1.01 strikeout–walk ratio, one of the worst in the game.

Conversely, if you looked at Johnson’s raw numbers without knowing his ERA — 205 innings, 194 hits, 60 walks, 172 strikeouts — you might guess that he was far more effective than he actually was. This is the tragedy of Johnson, and of aging in general. At 43, there are still signs of the old Cy Young winner in his frame, but he can’t tap that ability consistently from start to start or even inning to inning.

The possibility exists that Johnson will be able to dial up a semblance of his old self for six or seven innings. It’s happened a few times this season, but there have been no vintage Johnson starts this year, nothing dominating, just some very good starts scattered amongst a lot of poundings and average outings.

If Johnson was just some guy named Bob instead of a future Hall of Famer, the Yankees might not care if he started at all. He’d be just another random, unreliable pitcher. And for all purposes, that’s what he is — his experience counts for nothing if his back hurts, and if he no longer has his stuff.

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