Sheehan: Foulkelore (January 29, 2004)
Sheehan takes a look at my favorite reliever: Keith Foulke.
I disagree with removing Sutter, Gossage and Hernandez. His study was not trying to figure out anything about "usage patterns", but rather "maintaining effectiveness". I think it was a poor choice on Joe's part.
Anyway, he goes on to list the WARPs for age 26 to 30, and they averaged 5.0 wins over that time frame. This is of course their observed performance, and not their probable true talent. Doing a quick regression, I'll assume that the true talent level of these pitchers was 4.5 wins above replacement at age 28.
Taking a guess that a pitcher's true talent decreases by about 0.2 wins from age 28 to 34, I'd expect this group to be:
28: 4.5
29: 4.3
30: 4.1
31: 3.9
32: 3.7
33: 3.5
34: 3.3
This does not consider injuries.
So, from the age of 31 to 34, the "out of sample" years, we expect this group of pitchers to exhibit a performance consistent with a true talent of 14.4 wins above replacement (3.9+3.7+3.5+3.3).
This is what Joe presented:
WARP figures for only pitchers throwing at least 15 innings in a season
Age 30: 4.99
Age 31: 4.97
Age 32: 4.05
Age 33: 3.53
Age 34: 3.57
His age 31 to 34 pitchers totaled 16.1 wins. That was higher than what I would have expected. Of course, we do have selective sampling issues. Going to the previous chart Joe presented, and the total was 12.5 wins.
Bottom line is that Joe's study really reinforces what we think we know about the aging patterns of pitchers.
I do agree that you should give greater care to giving out long-term contracts to pitchers. But, the effectiveness / true talent level of our relievers is exactly what we think it will be in their 30s. The more interesting thing would be to look at injury rates.
(While insurance covers 70% of the first 3 years of a contract, you still have to pay out those premiums.)
With no research and off the top of my head, I would have paid Foulke: 8 million for 2004, then 7, then 6, then 5. 26 million for 4 years.
--posted by TangoTiger at 11:53 AM EDT
Posted 1:56 p.m.,
January 29, 2004
(#1) -
tangotiger
Hmmm... just thinking about it now. I said:
28: 4.5
29: 4.3
30: 4.1
31: 3.9
32: 3.7
33: 3.5
34: 3.3
If I add in the 26 and 27 lines, I would have had:
26: 4.5
27: 4.7
28: 4.5
29: 4.3
30: 4.1
31: 3.9
32: 3.7
33: 3.5
34: 3.3
The 26-30 group we expect was actually a total of 22.5. So, all the numbers should bump up by 0.1, so that we get:
26: 4.6
27: 4.8
28: 4.6
29: 4.4
30: 4.2
31: 4.0
32: 3.8
33: 3.6
34: 3.4
The 26-30 group comes in for a total of 22.6. That's the regressed performance (true talent) of our pitchers.
Their out-of-sample true talents would be 14.8 wins.
Not much change, but just wanted to clear it up.
Posted 5:19 p.m.,
January 29, 2004
(#2) -
bob mong(e-mail)
(homepage)
Thanks for your great, as usual, look at mini-studies. I read this article yesterday and was immediately annoyed: Aren't the results (I thought to myself) he presented exactly consistent with every baseball player's (or at least every pitcher's) aging pattern? He didn't show any evidence that relievers were worse than anybody else.
So anyway, I pulled similar data for starting pitchers since 1980, and got basically the same results.
So, here's the data, in case anybody is interested, for starting pitchers with the most WARP between ages 26 and 30 (only for pitchers whose age-26 season occurred in 1980 or later):
First, here are the fifteen pitchers.
Greg Maddux
Pedro Martinez
Roger Clemens
Orel Hershiser
Frank Viola
Mike Mussina
Mark Langston
Teddy Higuera
Jack McDowell
Tom Glavine
Chuck Finley
Dave Stieb
Randy Johnson
Jack Morris
David Cone
No real surprises? Greg Maddux was worth 53.9 WARP3 from age 26 to age 30, David Cone was worth 32.7.
Here are the age-by-age averages:
26: 7.7
27: 8.5
28: 8.0
29: 7.5
30: 7.9
31: 6.7
32: 6.2
33: 5.5
34: 5.3
Total from 26-30: 39.6
Average from 26-30: 7.9
Posted 6:54 p.m.,
January 29, 2004
(#3) -
tangotiger
Great stuff, Bob!
Posted 12:24 a.m.,
January 30, 2004
(#4) -
MGL
I've discussed this with Tango before, but I'm not convinced that there is a strong cause/effect relationship between age and pitching performance, except towards the end of a pticher's career. Research I did a long time ago suggested that major league experience was the "cause" for the observed aging patterns and not aage itself. I'd like for someone to do a quick age "delta thing," but control for ML experience, or vice versa. Remember that the two reasons why batters get better and then worse with age is that: 1)first they get bigger and stronger but slower and then all their physical skills decline AND 2) experience (I assume that's why their walk rates go up alomst forever). For pitchers, until they get so old that they start to have physical problems, I see no reason why age should have much of an impact on pitching performance....