As you’ve heard by now, the Mariners are close to a deal with Felix Hernandez on a 5-year extension. The deal would run from 2015-19, his ages 29-33, and is expected to make him the highest-paid pitcher ever, with an average salary of at least $25 million per season. (Initial reports of a $27.1 million AAV have been denied, but if I were a betting man….)

King Felix was already under contract for 2013-14 at about $20 million per.

I confess: I don’t really understand negotiating a top-of-the-pay-scale, 5-year extension for a pitcher two years before his free agency, when a mound of data suggests he’ll be declining by the time the contract kicks in, and quite possibly sooner. What does the club gain by doing this now?

Since the deal comes so far before his free agency, it’s worth guessing at what he might do not only during the deal, but also in the two years before it begins — two years that represent lost knowledge in Seattle’s decision-making process. Let’s see what we can learn from recent stars at a similar age.

Grounds for comparison: Age 23-26

To best capture Felix’s current status in the game, I’ll use his last four seasons, age 23-26. That span takes in his three best years, and all of his Cy Young Award votes and All-Star nods.

For ages 23-26, Hernandez totaled 20.4 rWAR and 24.1 fWAR.

• Note: I prefer to use rWAR because the Play Index is more robust. And trying to cover both methods in full would make even this minor study clunky. So most of the following treats only in rWAR. But we should bear in mind that Felix scores significantly better on the fWAR scale.

For retired pitchers since 1980 — approximating the universal adoption of the 5-man rotation — here are the
top 20 in rWAR from age 23-26:

This group averaged 19.6 rWAR from age 23-26. Hernandez would rank 8th in this group by rWAR.
(Using fWAR and the same search criteria, Hernandez’s 24.2 fWAR would rank 2nd compared to the top 20. That scale runs from 33.3 (Clemens) to 16.4 (Rijo) and averages 20.7 fWAR. Pedro, Saberhagen and Maddux all trail Felix by less than 1 fWAR for age 23-26.)

Age 27-28, as they might impact his market value for the new contract

Let’s take that control group and compare their average season from age 23-26 to age 27-28 — i.e., the years for which Felix was already signed:

Comparing their averages for 23-26 and 27-28:

• As a group, their average rWAR declined 29%, from 4.9 to 3.5 rWAR.
• 11 of the 20 declined by at least 47% of their base average rWAR —  including 3 of the 7 who topped Felix’s age 23-26 rWAR (Saberhagen, Stieb and Gubicza).
• Four significantly increased their average rWAR (Pedro, Rijo, Maddux and Clemens), ranging from 21% to 75%.
• One increased by 6%.
• The other 15 all declined by 16% to 96%.

By doing this deal now, Seattle buys out the risk that Felix takes his game and his market value to another level, à la Pedro, Clemens and Maddux. But I think those three were already on a different plane than Felix. At age 25-26, Pedro averaged 7.8 rWAR, Clemens and Maddux 6.0, Hernandez 4.0.

And how much higher could his market value go, anyway? The extension already represents the highest pitcher salary ever. Suppose he were to have a great year in 2013; would he have been able to command more than $30 million per year? I can’t see it.

The contract years: Age 29-33

Now let’s check that group’s performance at age 29-33, the years covered by Felix’s extension:

The breakdown:

• Maddux improved, and Pedro, Key and Mussina held their ground.
• The other 16 all declined by at least 30%.
• 13 out of 20 fell by at least 50%, and the group as a whole fell by 54%.
• 4 pitchers averaged 5+ rWAR/year (Maddux, Pedro, Clemens and Mussina). No one else topped 3.8 rWAR/year.

It’s tempting to think, “Well, Maddux, Pedro and Clemens were the superstars of this group, and they remained superstars.” But that’s a hindsight analysis. What about Saberhagen, Valenzuela and Stieb? All totaled at least 30 rWAR through age 26; Sabes had two CYAs, Fernando one, and even the underrated Stieb was a 4-time All-Star. At age 26 specifically, Stieb, Key, Soto, Radke and Fernandez all topped 6 rWAR; those five totaled one 6-WAR season from age 29 onward (Key, age 32).

Looking at seasons of 5+ rWAR:

• From age 23-26, this group of 20 pitchers totaled 38 seasons of 5+ rWAR; 19 had at least one, and 15 had two or more.
• From age 29-33 (five years compared to four), they totaled 20 seasons of 5+ rWAR, by 10 guys; only 5 had two or more.

Some of the most costly decisions involving those pitchers:

• St. Louis made a big trade for Mark Mulder, two years from free agency, having averaged 18 wins and 4.5 WAR from 23-26. He gave them one pretty good year (2.3 WAR) and then got hurt, but they re-signed him anyway. In all, they paid Mulder $25 million over 4 seasons, for totals of 22 wins, 5.04 ERA, and -0.4 rWAR. Dan Haren, the centerpiece of the trade, averaged 3.5 rWAR over 3 years with Oakland, then brought them a big payoff in trade.
• Alex Fernandez signed a big deal with Florida having averaged 4.4 rWAR from 23-26. They paid him $35 million over 5 years; he got hurt in the 2nd year and totaled 6.9 rWAR in 3 seasons. His salary ranked 2nd, 5th and 11th among all pitchers in the deal’s first 3 years.
• The Reds re-signed Jose Rijo after a fantastic age-28 season that brought his 23-28 average up to 5.2 rWAR. On a 4-year deal, Rijo’s salaries ranked 17th, 3rd, 2nd and 8th among all pitchers. He got hurt in the 2nd year, and totaled 3.6 rWAR during the contract.
• Age 20-26, Fernando Valenzuela averaged 4.2 rWAR. His salaries for age 27-29 ranked 2nd, 8th and 7th among all pitchers, but he totaled just 0.4 rWAR. After age 26, he never had another 3-WAR season.
• The Mets dealt for Bret Saberhagen after a strong comeback year that gave him a 21-27 average of 5.4 rWAR. After two injury-shortened but productive years (totaling 3.9 rWAR), they re-signed for 3 years at salaries that ranked 8th, 5th and 4th among all pitchers. He pitched well when able, but totaled just 330 IP and 7.3 rWAR over 3 years.
• Greg Swindell signed with Houston having averaged 4.0 rWAR from 23-27. His salary averaged 22nd among all pitchers during the 4-year deal; he totaled -1.2 rWAR.

What about active pitchers?

For convenience, I limited the main study to retired pitchers. But a look at the active rWAR leaders from age 23-26 is no more encouraging.

* Age 29-33 data presented only for those who have completed age 33.

Three of these 11 gained in value at 27-28 — but the other eight all dropped at least 20%. Does the 29-year-old Lincecum project to earn his $22 million this year, or the 32-year-old Zambrano his $19.3 mil? (As a setup reliever?) Think the ChiSox are glad they spent $48 million for Peavy’s 3-year 25-25, 4.07? Do any Tribe fans expect to see the Ubaldo of 2008-10 ever again? Would anyone take the contracts of Zito or Santana?

Similarity breeds contempt?

If all we’ve seen so far isn’t scary enough, consider the pitchers judged “Most similar” to Felix through age 26:

1. Larry Dierker (973)
2. Dennis Eckersley (955) *
3. Greg Maddux (954)
4. Frank Tanana (953)
5. Bret Saberhagen (942)
6. Joe Coleman (941)
7. Ken Holtzman (939)
8. Milt Pappas (933)
9. Mike Witt (930)
10. Catfish Hunter (927) *

Similarity Scores may be unscientific, but still — yikes!  From age 29-33, that group averaged 1.9 rWAR per year. Without Maddux, the average falls to 1.4.

“Don’t fence me in”

One last point: I think the M’s are missing a chance to exploit park-based illusions. They’re bringing in the fences at Safeco this year, so there’s a good chance that Hernandez’s unadjusted ERA will go up. It wouldn’t mean a thing to his real value, of course. But all other things being equal, isn’t it likely that his market value would decline just a little with a higher raw ERA?

____________________

For the sake of downtrodden Mariners fans, and for the beauty of his game, I hope that King Felix lives up to the expectations of this new deal. But, you know, a lot of my friends are actuaries, and that’s lent me a sober acceptance of risk patterns. If I were a betting man….