from Amazon
Friday, March 19, 2010
Arbitration and bias
It is practically impossible for a team to sign a bad multi-year contract for a player who is not eligible for free agency for the life of that contact.
How can I make that statement?
Generally speaking, a player who is eligible for arbitration for the first time will sign a one-year contract that is about 30 to 40 cents on the free agent dollar. That is, take two comparable players in talent, one is a free agent and one is eligible for arbitration for the first time, and the player still under team control will sign at over a 60% discount. Things get a bit better for the second-year arbitration-eligible players, as they sign at 50 - 60 cents of the free agent dollar. And for third-year arbitration-eligible players, they sign at 70 - 80 cents of the free agent dollar.
Suppose you have players like Justin Morneau and Ryan Howard that the arbitration process values much higher than the free agent market would value them. The teams would still be happy with the arbitration award, or the pre-arbitration settlement, because they are going to get a discount regardless. And, if they didn’t think they would get a discount, they can just release them without offering them arbitration. So, the one-year arbitration deals will keep teams from making egregious mistakes. Indeed, the arbitration awards follow a pretty clear pattern such that both sides have a pretty good idea of what each side will offer and what the arbitrator will award, because the entire process is comparables-player-driven. The outliers (Tim Lincecum, Howard, Mariano Rivera) is where there is great uncertainty.
Teams are in an easy position to offer multi-year deals for arbitration-eligible players because of this known pattern. Cole Hamels for example signed a 3 year, 20.5 million dollar contract last off-season as a first-time arbitration-eligible pitcher. If he was a free agent, he’d have signed for two to three times that amount. So, he signed at say 35 to 50 cents on the free agent dollar. Had he gone year-to-year, he would have averaged 50 to 60 cents on the free agent dollar. He gives up about 10 cents on the dollar for the security of a long-term deal.
Mark Reynolds was not arbitration-eligible this year. He actually missed the cutoff by two service days. So, the first year of his new deal is whatever the Diamondback would say it should be. Based on the standard practice, it would be for far less than one million dollars, probably close to 500,000$. The remaining portion of his contract buys out his first two years. First-year stars are mostly in the three to four million dollar range, while second-year stars are in the five to seven million dollar range.
Mark Reynolds signed a three year deal with this breakdown according to AP: $1 million signing bonus and a $500,000 salary this season, then $5 million in 2011 and $7.5 million in 2012, and a 500,000$ buyout in 2013. It looks like the Diamondbacks overpaid relative to what players of Reynolds’ quality would have gotten, not to mention that they got no discount for signing him to a multi-year deal. However, remember I said that there are certain types of players that are overvalued by the arbitration process, the Justin Morneau type of players, and Mark Reynolds would have presented a similar challenge next year: likely lots of home runs, likely lots of RBIs, but that overall, he’s a good but not great-player.
On the flip side is a player like Denard Span in which the arbitration process doesn’t see everything that that free agent market sees: good fielder, premium position, low power but overall a plus hitter. The free agent market would value Span in the same ballpark as Reynolds. But Denard Span, one-year behind the service clock compared to Reynolds, signed a five year, 16.5 million dollar contract. His arbitration years are being paid out at roughly three years and 15.5 million dollars, compared to the two years and 14.0 million dollars for Reynolds.
I presume that the Diamondbacks wanted to avert the Morneau-kind of situation with Reynolds. And so, they pay him a premium relative to similar-quality players, likely pay him comparably to similar-profiled players, and still get a huge discount relative to what is available on the free agent market.
Why doesn’t the arbitration process see more than this? Why do teams and players (collectively all teams and players, separately as all teams and all players, and individually as teams and players) go along with an arbitration system that really doesn’t correspond to true value?
Almost all teams use value metrics like WAR right? And any agent who isn’t engaging in malpractice has to know what his players’ true values to teams are right?
So what gives? Is there a set list of criteria that the arbiters must follow? If so, who set that criteria?
If it was set through collective bargaining, why not change it? Are the transaction costs too high (because owners and players would prefer different metrics? because teams like getting some players for way under value?)? Are there tensions between individual teams or types of players that keep the old metrics in place?
What actually would be the appropriate way to measure players?
Are our defensive metrics too variable right now to include them in arbitration arguments?
Is injury risk factored in? If so, how? If not, why not? Should it be? If so, how should it be?
Yeah, that is a long list of questions, but hey, crowdsourcing.