These being the Mets, they had to be swept by the Phillies in degrading fashion last week, cutting their division lead to two games, before they could sweep the hated Braves in Atlanta and increase their lead to five games. They would not be the Mets if they did not cause suffering.

These being the Mets, they also got to enjoy a near-brilliant pitching performance from young Mike Pelfrey on Saturday and a highly encouraging one from Pedro Martinez on Monday, which raised the central question that the team’s brain trust needs to focus on right now: What will this rotation look like in October? Naturally enough, even an excess of pitching can cause problems in Flushing.

From a remove, though, the biggest problem isn’t the excess of pitching, but rather a set of conflicting imperatives, and while it would be nice to say that these could be easily reconciled if only the Mets would commit to winning, that’s not actually true. There are some hard choices to be made, and I don’t envy the men who will have to make them.

The Mets’ decision making process, like that of all teams, follows basic rules and priorities. The first and most important is that you put your best pitchers in the playoff rotation. The next is that veterans with playoff experience get deference. Once the veterans are taken care of, hot pitchers get priority, and after them pitchers with good stuff. Health, ability to adapt to the bullpen, and the way certain pitchers match up against certain lineups or work in certain ballparks all have to be accounted for as well, but these aren’t first order concerns.

For most teams, these priorities aren’t issues. For the Mets, they are. If you fed all of them into a computer and told it to set up a rotation, its circuits would start smoking. The first rule, in the Mets’ case, isn’t particularly helpful, because they don’t have any bad starters, nor do they have any who are a clear cut above the rest. This makes things a jumble. You can make pretty convincing cases for and against five Mets starters, and it’s just possible that a light went on for Pelfrey on Saturday, which would make things all the more perplexing, at least for fans and writers who will agitate for him to get a start in certain circumstances. (That won’t happen, but I’d like to suggest to Pelfrey that he mug rancid reliever Guillermo Mota, hide him in a meat locker somewhere, steal his uniform, and start wearing it to Shea. This will get his filthy sinker into key spots in important games.)

Martinez, because of who he is and what he means to the team in an abstract, intangible sense, will start so long as he’s healthy. It’s an easy call.

Orlando Hernandez may be the trickiest case. He’s been perhaps the Mets’ most reliable pitcher and his exploits under pressure are legendary. On the other hand, he is, as usual, injured, and he’s not just the only starter with relevant bullpen experience, having helped the White Sox win a title two years ago while pitching in relief, but one particularly well-suited to relief, given his traditional weakness against left-handed hitters.

Tom Glavine may be the next most difficult case. The all-time leader in postseason starts, Glavine is enormously respected in baseball, and it’s hard to imagine him being benched in October, especially as he’s never pitched an inning in relief. On the other hand, he’s been the Mets’ fourth best starter this year, may be the fifth best with Martinez’s return, and while he’s been terrific over the last five weeks, running up a 2.90 ERA, he’s also getting by on nothing but guile, having struck out 20 while walking 17 in 49.2 innings.

John Maine is something like the opposite of Glavine right now. Before his last start, he hadn’t pitched six innings in a game since July 24; despite that, there’s a good case to be made that he has been the Mets’ most effective starter, and in crushing the Braves with seven great innings last Friday, he looked like an ace.

Finally, there’s Oliver Perez. On pure stuff, he’s the Mets’ best starter, and the only one capable of truly dominating another team. Like Maine, though, he’s not a veteran, and going into last night’s game he had a 5.19 ERA in his last five starts. He also has a reputation for, depending on how you’d like to put it, either giving highly variable performances or being a head case.

In the end, this really looks like a choice between Perez and Maine, whatever the merits of the cases. Martinez and Glavine are simply going to start; whatever you might say about where Glavine really is in the pecking order, baseball respects age. El Duque is only slightly less a lock, assuming he’s healthy; his bullpen experience will be canceled out by his October pedigree and his consistency this year. And this means that Perez and Maine, both of whom were just brilliant in last year’s playoffs, will be competing for a rotation spot down the stretch. The math just won’t work any other way, unless someone gets hurt or one of the veterans quietly tells management to do the right thing, seniority be damned.

Is this right? Probably not. Perez and Maine have earned the right to start in the playoffs, by how good they’ve been this year and how good they were last October, when the Mets turned to them in desperation and were rewarded with the sight of two players they’d picked up off the scrap heap blossoming into quasi-aces. But this is Mets baseball, and it wouldn’t be if something bad or obviously wrong weren’t happening. The bright side? They usually find their way right in the end. The suffering just comes along the way.

tmarchman@nysun.com