Some 4 year guy isn’t worth 15 million dollars because no 4 year man is ever going to hit the FA market. His dollar value should be tied to what 4 year arb eligible guys are currently getting not what some 8 year veteran is getting when he shops his services around to 20 different teams.

Much more of the money is being paid for near-term performance than longer-term, even though the annual breakdown of salary doesn’t look that way. Long-term contracts, eveluated over their life, would, I suspect, show a much higher price per win than the apparent value.

Think about this in terms of roto-ball. In the initial bidding process when all fantasy players are bidding on baseball players for the most part the bids are restrained and conservative because people have to fill the roster with many players. But then after the season starts everybody is allocated a free agent payroll and when the bidding commences on free agents you get really whacky numbers when compared to the initial draft. You suddenly get a small handful of guys who might contribute a little to your team getting paid huge % of your free agent payroll. This happens because there is a small supply of these free agent players and an even smaller supply of them are of any value to a team and the bidders no longer need to fill the roster with 20 or 30 players. They are only on the lookout for one or two players and have a large budget to use to get them.

In otherwords by using free agents dollars for all players you are using a relatively closed off supply of players to set the price for all players. Mark Texiera doesn’t set the price for all players because Mark isn’t competing against all those other players for free agent dollars. He is competing against other free agents. In the past we have seen weak free agent crops still get top dollar simply because they were all that was available and the teams had the money. If all the players were available or a greater supply then those players would not have signed for so much.

I would think a better option would be to compare 0-3 year service time players salaries to one another, look into the arb year players to see how they should be broken down, do free agent dollars, and then use those numbers for players based on where they are at in terms of service time.

I was just looking back at this post and comparing it to work I have done. FWIW, I came up with a figure of $5.1 million paid per free agent WSAB in 2007, a full million higher than you did. It could be that your lower replacement level accounts for some of the difference, but I think the margin for error here is huge and I just want to say again that the potential for misinterpretation of this sensitive material is huge.

It’s the only thing Fangraphs has done that I would disagree with.