Roster construction has become en vogue again, as it is wont to do, on the Internets. Just a few things about roster construction from the last month:

• In a recent broadcast, Gary Cohen and Keith Hernandez discussed the wisdom of carrying 13 pitchers.
• Christina Kahrl of Baseball Prospectus verbally assaults the Brewers for carrying fourteen pitchers.
• Jeff Sackmann of The Hardball Times comes to the Brewers’ defense and cites an article he wrote about the practically-expanded rosters of the modern game.
• Paul DePodesta of the San Diego Padres discusses the emergence of the 12-man pitching staff.

Being that it’s now in the discourse, I feel like it’s worth discussing. The Met decisions, however, have pretty much rendered the discussion moot. On Tuesday, the Mets added Wise and Vargas and shipped out Figueroa and Sosa.

I feel, then, that it would be good to take stock of the Mets’ current roster construction. Are there redundancies? Where does everyone fit in? What do the Mets lack?

Down the line, the Mets are going to have the issue of carrying 13 pitchers return. It begs a couple of questions:

1. How often do the Mets stress their bullpen? In other words, how long do Mets’ starters, on average, stay in the game?

2. Is the Met bullpen overworked at this point in time? How does one define “overworked?” What is a fair workload?

Some of these questions, as fans, we just can’t answer. We simply don’t know how long different Met relievers need to bounce back from a 15-pitch outing, or a 30-pitch outing, or a 40-pitch outing. Every pitcher is different, certainly. Still, the rest of these questions can be answered simply by looking at the data.

We’ll start by looking at average length of start.

Interestingly, it is San Diego who has the least need for the expanded bullpen (even with DePodesta’s defense). The much-maligned Phillies’ rotation gets pretty deep into games, relatively speaking. And the Met rotation, which I would have guessed is at the bottom of the league, is actually the closest team to the league average (which is 5.73 IP/GS). So in an average game, the bullpen is counted on to get… 10 outs or so. Extra innings, of course, make this a fairly inexact science, but extra innings are the exception, rather than the rule.

My point here is that the Mets, as an average starting-length team, have no pressing need for extra bullpen.

Here’s some data. I’ve combined the percentage of lefties and righties that each pitcher has faced, the total number of batters they’ve faced, the average length of outing, and their “leverage” score from Baseball Prospectus.

Billy Wagner: Closer. Randolph usually opts to save him on the road for the save situation in extra inning games. Wagner usually gets the ninth in tie games at home.

Aaron Heilman: Effective against both righties and lefties normally, struggled in the early going this year but seems to have settled down (before last night, anyway). At this point, he’s the picture of an average middle reliever.

Duaner Sanchez: Rapidly regaining use in the later innings. More effective against righties.

Pedro Feliciano: Scarely used in the early going. Used for tough lefties, but he’s not kept from righties.

Scott Schoeneweis: Closest thing in the ‘pen to an OOGY (one-out guy). Much more effective against lefties than righties, but Randolph prefers to use Scho for more than one batter at a time.

Joe Smith: Gets high-leverage innings these days but still hasn’t quite solved lefties—they’re hitting .455/.538/.545 against him.

Matt Wise: ?

Sosa found his way into too many high-leverage situations, but the truth is the Mets haven’t had all that much mop-up work to go around. Sosa seems to have been reassigned, in a way, into lower leverage situations of late (he came into Monday night’s game with the Mets down by three). Out of necessity, the Mets designated Sosa for assignment in the Tuesday afternoon massacre.

The only criticism I could levy at this point is that Wagner simply has not been used enough, but few managers would actually use the closer outside of save situations, and Wagner, with his propensity to tire in August and September, is best saved. Really, it’s hard to criticize the implementation of resources. No one seems particularly overworked, though I would try to dial back Heilman and crank up Feliciano’s innings a bit.

Also worth noting: the Mets don’t really have a long reliever to speak of at this point. Sosa was capable of the role but never lasted long enough in an outing to be a “long reliever.” (For reference, Aaron Sele, a true long man in 2007, averaged 1.58 innings per outing.) While long relievers don’t do much to win the game in which they are pitching, they can be very useful for other games. A good long-reliever gives you the freedom to save a pitcher without his best stuff from laboring and getting hammered, and allows you to save your middle relievers and closer for the high-leverage innings. Over the past few years, the Mets have had their fair share of good long relievers. Darren Oliver almost won them a pennant. Pat Mahomes was quite useful. I am a little uncomfortable with the Mets not having a long-man in that group. I would almost (key is almost) prefer to keep Claudio Vargas or Nelson Figueroa on the roster ahead of Matt Wise when the time comes—none of these bullpen pitchers can actually extend for a few innings.

Indeed, I find Minaya’s track record to be far more spurious than Randolph’s. Minaya has exhibited a dangerous propensity to pay a premium for fungible, lightning-in-a-bottle-style talent. He did it with Mota last year and Sosa this year. He found those guys on the scrap heap initially, and could have cut ties and found another in a similar fashion. But I digress.

The hitters present their own unique challenges. In addition to the eight position players, Castillo needs a player to spell him (Damion Easley), Moises Alou needs a defensive replacement (Endy Chavez), and they need a guy who can fill in around the infield (also Easley), and a right-handed pinch hitter (also Easley, who is actually quite valuable on this team), and a guy who can fill in at first base for Delgado (Marlon Anderson), and a backup catcher (Ramon Castro).

The rest of the bench is now reserved for Fernando Tatis, thus bringing the grand total up to 13 hitters. I like carrying 13 hitters more than 13 pitchers, certainly, but I have to ask: why Tatis?

Tatis continues to be a third baseman, with few exceptions. He’s played third in 27 games down in New Orleans. He’s played a couple of games at second, one game at shortstop, five in left field, and three as the designated hitter. Defensively, Tatis’ skills don’t really fit in on the Mets unless David Wright is out for an extended period of time.

To his credit, though, Tatis has raked in May, hitting .409/.667/1.076 in 36 at-bats. Hot and cold streaks are difficult to discern, but if hot streaks exist, Tatis is certainly in the midst of one. Perhaps Minaya and Co. are hoping to capitalize on Tatis’ hot bat. Sadly, the Mets lack any better options in Triple-A at this point. Looking over that roster, only Brady Clark makes sense. In Double-A, the two young first basemen, Nick Evans and Mike Carp, have hit quite well, and one might find his way onto the big league club if Delgado struggles to get things going for a couple of months.

I really want another bat who brings more to the table. I like Jacque Jones at this point; if the Tigers grant him his outright release, he would be a major bargain and can do a lot of damage against righties. For the short term, Tatis might be okay. Still, it is important that the Mets not be seduced by a couple of early homers. We know what Tatis can do at this stage of his career. Even Chip Ambres, who filled in admirably last year, would be a nice fit. Sadly, he’s in Triple-A Portland in the San Diego organization at this point, doing what he does.

I like to talk about free talent and how good the Mets are at finding it, but, for now, the Mets are pretty limited on the “injury replacement” front. It’s a long season, of course, and rosters are fluid throughout. Periodically taking stock of it, though, is one of those essential tasks. At this point, I think the Mets have some work to do.