THE BOOK--Playing The Percentages In Baseball

Wednesday, December 01, 2010

What IS a FA worth?

http://baseballanalysts.com/archives/2010/11/whats_a_free_ag.php#comments

Rich Lederer has two points based on some limited data:  One, relief pitchers as a whole are paid a lot more than everyone else, per win.  I assume he means as a general rule and not just a one or two year fluc, which could be the case.  Two, excluding relief pitchers, FA are paid a lot less than 4 or 4.5mm per win.  Sort of a third point is that even WITH relief pitchers, FA are not paid as much as we think (4-5mm per win).  Obviously it depends on the year we are talking about, but Rich seems to think that whatever recent year we choose, we (sabermetric wisdom) are too high.

Tango mentioned that you have to separate one year and multi-year deals. I am not sure why, unless you want to give a figure for each and you think they are substantially different.  If you are just telling us the $$ per win, then you want to look at all deals.  If you think that there IS a substantial difference, then you should tell us what the 1-year and the multi-year numbers are, lest we incorrectly apply the one-size fits all number to all deals (in which case, we will always be making a mistake whether the deal is one year or multi).  Whenever I read about the 4-5mm numbers, I don’t recall that it is restricted to multi-year deals. Maybe it is and we just forget to repeat that.

Tango also talks about the difference between performance and true talent. Of course the $$ per win is based upon true talent wins. But, if we want to compute $$ per win retrospectively, like by looking at all the FA performances in 2010 and compare that to their 2010 salaries (adjusting for any backloading of course), then performance does equal true talent since all combined FA performance in 2010 is an unbiased estimate of their true talent.

I don’t see how there can be much controversy.  It is relatively easy to compare FA dollars versus FA performance in any given number of years.

I don’t know about Rich’s relief pitcher contention. If it is true that relief pitchers get “overpaid” by a lot, then, like with the 1 and multi-year deals, it would behoove us to have two $$ per win numbers.  Then again, we might need more than that, as different positions probably get paid different $$ per win - not just relief pitchers.

Then again, just because a relief pitcher gets paid $5 mil per win and the rest get paid $3 mil per win (made up numbers) does not mean that if a team signs a relief pitcher to a 5mm per win contract, that it is “correct.” It is actually incorrect if the team could have gotten more value in someone else at a different position for the same amount of total money.  So the overall $$ per win is an important numbers since all positions are to a large extent fungible (obviously some teams in some situations can upgrade at one position and not another so are forced to pay the going rate at that position that they CAN upgrade, even if the going rate is an inflated number, like Rich suggests it is with RP).

I am interested in what you guys think of this article and its theses…