“All managers are losers,” Ted Williams once said. “They are the most expendable pieces of furniture on the face of the Earth.”

Of all baseball’s traditions, the worst may be giving the manager an ultimatum or deadline. Usually, this happens when a talented team such as the Mets plays badly, with the fans in revolt and the players disregarding the skipper. With the fans and the team lost, and the season in danger of slipping away, the general manager or the owner lets it be known that if the team doesn’t win an important series, or show improvement by a certain date, the manager will be out on his ear.

This is ridiculous both because by the time it comes it’s too late to do any good, and because it presupposes that being fired is a punishment, which it shouldn’t be. Barring embarrassing or criminal conduct, the only reason a manager should be fired is that he’s going to do badly. If you think he’s likely to do damage, waiting until he’s already done it makes no sense.

The question that needs to be asked about Willie Randolph, then, is whether he’s likely to manage well or not. There are some good reasons why he should keep his job, not least that the Mets have won more than they’ve lost and are just a half game out of first place. The point those of us who think he should be fired are making, though, isn’t that he’s so awful that he should be turned into a burnt offering to appease the angry baseball gods, but that under his leadership the Mets are likely to keep muddling along, not doing too badly but not doing especially well, either. The team is capable of something more.

Part of the reason that this is a difficult argument to make is that there is no generally accepted way to judge a manager other than by wins, which is the point Williams was making. Dusty Baker was a great manager in San Francisco when he had Barry Bonds, and in Chicago when Carlos Zambrano, Mark Prior, and Kerry Wood were all healthy; when he no longer had the best player of all time or a trio of aces, he was ripe to be fitted for a dunce cap. Along this line of reasoning, Randolph has done fine; the Mets may have looked flat in April, but then Moises Alou and Pedro Martinez were hurt, and after all he’s won more than he’s lost three years running. It’s far too early to say he should lose his job, some would say, especially coming off a series win this weekend over Arizona, whom I claimed would flay the Mets.

With the notable exception of 2006, though, the Mets haven’t done especially well under Randolph. In 2005, they won seven fewer games than they should have, given how many runs they scored and allowed; last year, they won two more, but kicked away a seven-game lead with 17 left to play, and this year, under pressure, they opened the season playing apathetic, unfocused baseball. I’ve come to think this isn’t coincidence, and that Randolph’s style is largely to blame.

It’s not quite as hard as you may think to pin this down; style isn’t an abstraction, but a function of the discretionary moves a manager makes. Baseball Info Solutions tracks this information, and annually publishes it in the Bill James Handbook.

Randolph is one of nine men to manage three full seasons in the National League between 2005 and 2007. The other eight — Bruce Bochy, Bobby Cox, Clint Hurdle, Tony La Russa, Charlie Manuel, Bob Melvin, Jim Tracy, and Ned Yost — are his direct peers, established managers who have been running teams in the same league at the same time. A review of what he’s done during that time, and a comparison to what they’ve done, shows the holes in his approach.

During Randolph’s tenure, he’s run his pitching staff more or less the same way the other eight managers in question have run theirs. Randolph’s teams have averaged 21 starts of 110 or more pitches per year, for instance, the same as his rivals; he’s used relievers on consecutive days an average of 105 times, against 111 for the comparison group. The one area where he really stands out is in the total number of relievers used. He’s averaged 455 per year, against 478 for his peers; only Yost uses fewer relievers. One common complaint about Randolph is that he doesn’t spread the bullpen’s workload enough, and there seems to be something to that.

When it comes to using his position players, Randolph is far more of an outlier. He prefers a set lineup, for one thing. Over three years, he averaged 103 lineups per year; only Cox and Manuel used fewer. He also really does call an exceptionally active running game, as he well should, given his players. He averaged 207 stolen base attempts per year, 57 more than anyone else, and ranked fourth in setting runners in motion with the pitch, which he did 95 times per year. He also ranked second in annual sacrifice attempts, with 95.

These are matters of preference, really; one suspects that a team with several really exceptional base runners and power hitters wouldn’t need to sacrifice so much, but it’s not that big of a problem. What is a problem is Randolph’s inability or unwillingness to run a game.

Over the three years under review, Randolph ranked dead last in the use of pinch hitters and pinch runners. The other managers averaged 273 pinch hitters and 28 pinch runners per year, Randolph 246 and 13. He was also third from last in defensive substitutions, though not far off the average, with 34 per year. This, combined with his preference for a set lineup, paints a picture of a manager whose style is just badly suited to run this roster.

Think about the lineups Randolph has had since 2005. They’ve been built around great young players who take the field more or less every day, such as Jose Reyes, David Wright, and Carlos Beltran, and older, injury-prone players who can’t field and often have platoon issues, such as Carlos Delgado, Moises Alou, Mike Piazza, and Luis Castillo. These teams have also featured useful, often athletic reserves such as Lastings Milledge, Endy Chavez, Ramon Castro, and Ruben Gotay. Given this, the Mets should logically rank near the top of the league in pinch hitting, pinch running, and defensive substitutions, as the manager shuffles caddies onto the field and into the lineup to take advantage of his speed and depth, especially in critical game situations, and keep the more brittle regulars fresh. Instead, they’re right at the bottom in these areas.

Randolph is sometimes painted as a man out of his depth, which I’ve always found absurd and patronizing. What is clear, though, and what the numbers back up, is that he really does have a classic American League style. This makes sense; he played nearly his whole career in the junior circuit, and spent nearly a decade on the bench with Joe Torre. The Mets, though, aren’t the Joe Torre Yankees — full of iron men and first-rate role players. They’re a team with a lot of older players who need to be kept fresh and swapped out tactically, and a lot of players with narrow talents who need to be used just right to be effective.

It’s one thing for a manager’s style not to fit the needs of his team; it’s another for it to run directly counter to them. There are other reasons the Mets should seriously consider getting rid of Randolph, from the curious lack of development of some very talented younger players to the team’s real lapses of concentration to his dubious abilities as a judge of talent. But the most important is that to expect Randolph to bring this specific team to their full potential is to expect him to become something other than what he’s showed he is. It’s not fair to the team, it’s not fair to the fans, and, in the end, it might not even be fair to him.

tmarchman@nysun.com