Whatever you think your comfort level is, take it down a notch. Signing a free-agent pitcher to a lengthy deal is a risky proposition, with all sorts of causes for angst.
We say this based on precedent and come armed with the numbers to illustrate our point.
ESPN Stats & Information has compiled a list of all of the free-agent signings since the 1990-91 offseason.
In that span, there have been 52 instances in which a pitcher signed a contract of four years or more.
How many of those have been success stories? It depends on how you want to define success.
We're going to use two stats to measure it -- starts and adjusted ERA+.
Starts is self-explanatory. You figure if the pitcher is healthy and pitching well, he'll have a bunch. An injured or ineffective pitcher won't.
Adjusted ERA+ is a stat tracked by Baseball-Reference.com that measures how much better/worse a pitcher's ERA is than his peers that season, making slight tweaks for the ballpark in which he primarily pitched.
If a pitcher has an adjusted ERA+ of 120, it means he was 20 percent better than the rest of the league. That's a good rating. (For reference, CC Sabathia's 3.18 ERA matched to a 134 ERA+ in 2010; Andy Pettitte's 3.28 ERA rated a 130.)
You figure if a team gives a pitcher a four-plus year deal, it has high expectations for that pitcher's durability and skill, perhaps hoping for 30 starts a season and an adjusted ERA+ of 120 or better over the life of the contract.
How many times out of the 52 did the pitcher hit that criteria?
Now there are some flaws here. The shortened seasons of 1994 and 1995 impacted some pitchers' ability to average 30 starts (it didn't with Maddux). And perhaps our standards are too high. So let's tweak the criteria slightly.
Of those 52 contracts, how many times did the pitcher average 20 starts a season, pitching 10 percent better than the league average?
Fourteen -- about one in four.
Let's lower the standard even further. How many times did a pitcher average 20 starts a season, and pitch to a league average or better ERA+?
Twenty-two. That's 42 percent.
That means that there are a lot of bad results from a lot of formerly good pitchers.
Now in fairness, some of those contracts were a bit of a stretch to begin with. Should we judge Lee's potential against the deals signed by Jeff Suppan (4 years, $42 million in 2007), Chan Ho Park (5 for $65 million in 2001) and Jaime Navarro (4 for $20 million in 1996)? These are pitchers who were slightly above average who took advantage of favorable markets.
But there are some comparable cases.
Ex-Yankees prospect Doug Drabek was 30, two years removed from a Cy Young and coming off a three-year stretch in which he had a 124 ERA+ when he signed with the Astros for four years and $19.5 million in the 1992-93 offseason.
Over the life of that contract, Drabek averaged 30 starts a year, but the results were average -- a 4.00 ERA, which at the time converted to a 98 ERA+.
The Astros didn't do that badly when you consider some deals of more recent vintage.
Mike Hampton averaged 18 starts a year and a 96 ERA+ over his eight-year deal, which actually doesn't look as bad as it did after Hampton struggled through his initial stint in Colorado. It took trading him to the Braves and overcoming injuries to make him a formidable pitcher again.
Bartolo Colon got a four-year deal from the Angels in the 2003-04 offseason, and won 21 games and the AL Cy Young in the second year of that contract. Problem was, in the last two years, he made 28 starts with a 5.90 ERA and 77 ERA+. That's not worth $30, let alone the $30 million that Colon got.
The most recent bad-deal poster child is Giants starter Barry Zito, four seasons into a seven-year, $126 million contract with a 4.45 ERA and 97 ERA+ to show for it (on the positive side, he's averaged 33 starts per year).
But hey, the Giants managed to live that mistake down, winning a World Series title in spite of it (Zito wasn't even on the postseason roster). It just goes to show that one signing is just that -- one out of 25. It's how much value you get out of everybody else that counts just as much.