Comments on: Wasting Money on K-Rod

By: Thomas

Thomas — Mon, 15 Dec 2008 22:11:13 +0000

90% of the time K-Rod pitches it will matter and be a save situation. They blew more games after the 7th inning last year than just about any team. It is irrelevant to compare the money spent on him, to spend it on Lowe. So what if Lowe gets you a lead, then leaves in the 8th only to have Billy Wagner blow it. If he was on the team last year, they would have made the playoffs. Would you pay someone $13 million a year if they were your missing link to the playoffs? I think they made the right choice.

By: tangotiger

tangotiger — Sun, 14 Dec 2008 20:28:17 +0000

The first published implementation of WPA was by the Mills Brothers in 1970, and named Player Win Average. You can click on my name for an excerpt. Doug carried on the tradition in fine style. Whereas Mills’ used division, Doug used subtraction.

As for replacement level, it is, at its heart, simply the production of the player at which he would earn the minimum salary. That level of production is roughly 2.25 wins per 700 PA, or .01 wins per IP, below the league average. I agree with JC that you don’t need replacement level, and you can get there with just using league average.

By: JC

JC — Sun, 14 Dec 2008 13:10:43 +0000

WPA is a fine stat; in fact, it was developed by my good friend Doug Drinen. For the purpose of generating marginal revenue product estimates, the problem with measures that value the exact change in the state of the game is that they give credit/blame to players for things they are not responsible for. It is quite useful to know the potential WPA when deciding between a mop-up reliever and a fireballing out-machine. I don't find the replacement player benchmark to be useful, and I think it creates more confusion while adding nothing over the average benchmark.

By: birtelcom

birtelcom — Sun, 14 Dec 2008 05:19:20 +0000

Win Probability Added (WPA) over at Fangraphs has Lidge as the second most “valuable” pitcher in the majors in 2008 (only Lee is ahead of him, Sabathia is just behind him) if you use WPA as the basis for “value”. K-Rod is just behind Brandon Webb in WPA. I take it the consensus then is that WPA is not really a useful proxy for value in comparing high IP pitchers such as Sabathia or Webb to low IP pitchers such as Lidge or K-Rod. I take it the big problem with WPA in this respect is that it is a measurement against average rather than against replacement, which means it fails to value innings eaters who add win probability above what a replacement would do (but not necessarily above what an average player would do) in large numbers of innings a season.

By: TangoTiger

TangoTiger — Fri, 12 Dec 2008 20:52:15 +0000

In my system, it works out that the Leverage Index to use is roughly halfway between how he should be used and 1. So, a closer, with an expected LI of 2.0, will have an effective leverage, for salary purposes, of 1.5.

By: JC

JC — Thu, 11 Dec 2008 12:17:04 +0000

I don’t know how the Mets plan to use Putz.

By: JC

JC — Thu, 11 Dec 2008 12:13:38 +0000

A.J. Burnett's value: $13 million per season.

By: Brian

Brian — Thu, 11 Dec 2008 08:03:12 +0000

JC,

This was John Dewan’s stat of the week from Sept. 5:
http://actasports.com/sow.php?id=182

I found it interesting that K-Rod only had one “tough” save at the time while Thigpen had eight of 10 in 1990. Plus, it is more common in baseball today for closers to earn saves coming in when the first batter is neither the tying or winning run.

Should K-Rod be the only closer or will Putz get some matchups in the ninth? In Seattle, Putz was dominant against David Ortiz (0/7, 5 Ks) and Manny (2/11, 7 Ks). Yet, I do find it funny that both closers come from the AL West.

By: TomL

TomL — Thu, 11 Dec 2008 05:12:43 +0000

JC,
Sorry if you’ve covered this, but what do you have AJ Burnett valued at?

By: Jesse G.

Jesse G. — Thu, 11 Dec 2008 02:19:50 +0000

I have a question. I agree that closers are dramatically overvalued even if you buy the importance of leverage. This is because the marginal improvement of a top closer over even an average to below average “closer” (this could really be almost any pitcher with one great pitch or two or more above average pitches). However, I do wonder if, given current usage patterns, excellent closers do have more value to top teams than we would think based purely upon performance and innings pitched/appearances because at that level of competition (trying to win a division and especially in a playoff setting) the small increases in efficiency and the ability to increase usage beyond what is typical (think Mariano Rivera routinely pitching more than a full inning in the playoffs) are what makes the difference for them? I realize this concept assumes that a lot of the inefficient strategic norms and usage patterns are a given but I just wonder if the data might bear this out (or if it could)?
Anyway, I guess I am just interested in your thoughts of whether at the top level (Yankees/Red Sox, Mets/Phillies, etc.) the importance of small improvements at the margins add some value to players like Rivera in New York.