You and I are in the wrong line of work: The only job worth having is representing free agent pitchers.

Imagine you were charged with selling a service or asset that, even if it proved as valuable as expected, would be priced above market rate. Further imagine that the odds of it performing as expected were on the order of one in five, and further imagine that just for putting your service or asset up for bid, you were assured that a half-dozen or more wealthy people would be clamoring for it. I do believe that anyone capable of opening their wallets or counting loose change can see the possibilities and pleasures inherent in such a job.

This year’s free agent starter signings will on the whole, prove disastrous as they always do. The reasons are obvious. By the time a player reaches free agency, he’s in his prime. Because of the mechanisms of the free agency market, a pitcher is usually paid in accordance with best-case scenarios, and a pitcher is far, far likelier to decline than he is to improve or even to sustain his performance. Pitchers are also more prone to injury and spectacular ineffectiveness than position players. For these reasons, when you look at any class of free agent pitchers, they are, taken together, spectacularly overpriced. Take the class of 2004 — pitchers like Pedro Martinez, Derek Lowe, and Jon Lieber have been pretty effective, and worth the money they’ve been paid. On the other hand, you have pitchers like Matt Clement, Russ Ortiz (7.33 ERA in 178 innings), and of course Carl Pavano. You’d almost be better off at a three-card monte stand than you would be picking up free agent pitchers.

Of course, picking free agent pitchers isn’t a pure gamble. Teams have plenty of information available, ranging from advanced statistical information to which everyone has access to confidential medical reports and the opinions of scouts who can tell whether a pitcher’s delivery and body frame are more or less likely to allow him to hold up over time. Still, when ostensibly intelligent people use all that information to arrive at the conclusion that Pavano, having pitched as many as 140 innings worth of league average baseball in his career exactly once, is worth $40 million, something is obviously a bit haywire.

Really, there are two qualities worth paying for in a free agent pitcher: Reliability and upside. Reliability is worth paying a lot for. The team that signs Tom Glavine knows more or less exactly what it’s going to get: 200 innings and an ERA between 3.50 and 4.00. Upside is worth paying for, but only if you’re not paying a lot for it. If your team doctor and pitching coach are absolutely convinced that a new training method they’ve devised can keep Mark Mulder healthy and effective as a starter, it’s worth signing him, but not at a price that means you’ll have to rely on him as a rotation anchor. This last part is the problem with most signings: It’s one thing to take a calculated risk that Pavano has matured into a valuable no. 2 starter, quite another to pay him like one. All of this depends of course on circumstances; the Mets had the money and were in the position to take a gamble that Martinez would stay healthy. The Twins would have been psychotic to take the same risk.

Looking at things this way, it shouldn’t be too hard to tell which are the better pitchers on the market this year. The top name is probably Barry Zito. There really is a lot not to like about Zito. He walks a lot of hitters, for one; he’s also quite reliant on his defense, consistently putting up ERAs much better than those suggested by his underlying statistics. He’s been worked quite hard in his career, he doesn’t have a great fastball, he’s going to get paid like an ace, and he has floppy hair.

That’s all quite true, but none of it much matters. His relatively high walk rate and apparent reliance on his defense are a bit troublesome, but a pitcher with a big, sweeping breaking ball like Zito’s can get away with a pitching strategy based on pitching around hitters and inducing bad contact. Far more important, though, Zito is incredibly consistent. Every year he goes out and pitches 220 innings of quality about equal to those Glavine pitched this year. It’s easy to gloss over that innings total and focus on his ERA that isn’t quite elite, but 200 innings a year is the benchmark for durability today, and Zito exceeds it by 10% every year. In any given year there will be between 10 and 20 more valuable pitchers, but very few who are a good bet to pitch at that level for the next three or four years, and none who is on the market right now. Zito might be more like Glavine than Greg Maddux, but that’s hardly a reason to damn him.

The next-best bet is Daisuke Matsuzaka, ace of the Seibu Lions and Scott Boras client. His Japanese numbers are so good they’re easily discounted — 741.1 innings over the last four years with a 2.51 ERA, a 768/188 K/BB ratio, and a record of success under pressure in championship games. Under the posting system by which American teams gain access to Japanese players, the rights just to negotiate with him will cost probably upwards of $20 million, and then it will take an ace-level contract to sign him.

Is an unknown worth all that money? I’d guess yes, though it’s an awfully big risk. We know how good the Japanese leagues are — better than AAA, worse than the majors — and also why. There’s not as much power there. Statistically, Matsuzaka looks like a 200-inning, 3.25 ERA or so starter; at his age, that’s exceptionally valuable. The problem is that you can’t do a linear, one-to-one translation of his statistics and call it a day.

Matsuzaka is eerily reminiscent of a young David Cone, from his low slingshot delivery to nearly patternless pitching based around a wide variety of breaking balls and a fastball that reaches the mid-90s but sits a bit lower than that; should his fastball prove a bit short for the majors, though, and thus more vulnerable to the longball, he might just be the world’s priciest junkballer. Despite that, he has youth and a proven track record of excellence. For a team with enough money to sign him and not collapse if he proves a bust, he’s worth the risk.

And honestly, that’s about it for the market this year, excepting St. Louis’ Jeff Suppan, essentially a right-handed Zito without the youth or long-term track record, and Maddux, at this point in his career a reliable, innings-eating no. 4 starter. Every other pitcher on the market is exactly of that class that’s most likely to prove disastrous, whether because of age and a history of recent injury (Jason Schmidt, Mike Mussina, Andy Pettitte) or because of not being very good (most everyone else). If the right teams pay out for Zito, Matsuzaka, Maddux, Suppan, and Glavine and put them in the right roles, they’ll probably be very happy, and if the right teams sign Schmidt, Mussina, and Pettitte understanding that they can’t count on them to be aces, they might well be happy, too. Everybody else will just be rolling the dice.