Earlier this week, we learned from "one plugged-in agent" via one plugged-in reporter that Aroldis Chapman "is looking for $100 million." The day might come when we should talk about whether Aroldis Chapman is worth $100 million, but more urgently, we should really talk about that rumor construction: "Looking for."
Is there any power in these aspirational clauses?
To figure that out, we dove deep into the past almost-decade of hot stove rumors, collecting every rumor we could find of a free agent seeking, looking for or asking for anything of quantifiable value. We searched MLB Trade Rumors archives, along with the Twitter archives of Buster Olney, Ken Rosenthal and Jon Heyman. (We would have continued with other individual writers, but we found that most of these reporters' tweets had already been aggregated by MLBTR and showed up in our initial searches.)
The result was a spreadsheet with 78 lines of demands, some every bit as juicy at the time as Chapman's nine-figure request is now, some every bit as guffaw-inducing as Gomez' five-year ask. As a body of literature, these 78 rumors cover the full range of hedgy hot stove linguistics: the hyperbole, the suggestion, the tease, the peek at an agenda. As a collection of numbers, though, they reveal some actual truth.
The smallest rumor in our data was the shortstop Adam Everett seeking $3 million in the winter before the 2010 season. (Or, perhaps, Brian Shouse and Chris Capuano seeking two years.) The largest was Robinson Cano, seeking $305 million in his free agency before the 2014 season. (To simplify, we'll refer to every offseason by the year of the season following it. So you are living in the 2017 offseason right now.)
In some cases, players sought different things throughout the course of their free agency; for our spreadsheet, we went with whatever demand was the most. In some cases, these demands were expressed as a range, e.g., Mike Morse is "looking for $7m-$8m." For our spreadsheet, we logged it at the lowest figure in the range. So: highest demand on record, but lowest end of defined ranges.
Alongside our "seeking" column was a "got" column, and alongside that we judged whether the player got what he sought. Each outcome could be rated one of four ways:
Yeah: The player got more or less exactly what he was asking for. Shin-Soo Choo was seeking a contract "worth more than Jayson Werth's $126 million deal." ("Well that's hilarious.") Choo got seven years and $130 million.
Eh: The player didn't get his target, but the request and the result were in the same region. It looks, in other words, like the spread of a normal negotiation. Wei-Yin Chen was seeking five years, $100 million. ("He's not worth half that!") He ended up getting five years, $80 million.
LOL: The player's demands were either so high that they look silly in retrospect, or the player was so stubborn that he ended up getting frozen out of the market. Jason Varitek's agent "suggested Varitek should get a deal comparable to Jorge Posada's four-year, $52.4MM contract." He would sign for one year and $5 million.
We actually started with 80 players, but Joakim Soria's request (a no-trade clause, he didn't get it) didn't lend itself to the rest of this analysis, and Hisashi Iwakuma's demands from the A's were anomalous because Iwakuma had a posting fee attached to him. He was willing to go back to Japan and hit free agency without a posting fee the following year.
Of the 78 free agents remaining:
+: 4 players
This isn't as clean as we'd hoped. There are enough LOLs -- more than a third of the outcomes -- that we have to take seriously the possibility that we shouldn't take the player's demands (or his agent's demands, or the second-hand telephoning of his supposed demands) seriously. Most day-to-day human interaction depends on our faith that people aren't outright lying to us, that the waiter isn't copying down our credit card number so he can steal our identity, that the cop is a cop, that I really did put together this spreadsheet that I keep claiming to be working off of, that our mom really does love us. If a third of these players are seeking something totally unrealistic, it causes something close to a crisis of confidence in the whole thing.
On the other hand, the majority of these asks -- nearly two-thirds -- are perfectly reasonable, either as legitimate targets or as honest positions in a negotiation process.
So what do we do with an individual goal such as Chapman's? We could decide it lacks good faith, but as noted in the examples above, even in the rare case when the player exceeds his demands, there will be people who find the target laughable. We are, as a population, horrible at putting player demands in perspective. We're always a few years too late to adjust our brains to the actual market.
We could take into account the agent involved and decide whether he's an honest broker of rumored demands, but to kill that idea: all four of the examples I gave above shared the same agent, Scott Boras. You can't throw out the Variteks without losing the Choos.
So we prefer to keep everybody together and trust that the large cohort would quiet that noise. We took each player's goal and calculated what percentage of the goal he ended up getting paid. Some players were asking for years, so we looked at how many years he got. Some were asking for money, so we looked at how much money he got. (Usually we used average annual value. For three players -- Fielder, Ervin Santana and Ricky Nolasco -- only total contract dollars sought were available in our rumors, so we used total contract dollars.) Some demands were expressed as both -- as when Jhonny Peralta was looking for four years and $56 million, or Hisanori Takahashi was seeking three years and $12 million -- so we calculated both outcomes (years and AAV) separately.
Here's what we found: the median outcome for a player who was looking for X years was: 87.5 percent of X. A player who asked for three years got, on average, 2.625 years. (The median result was actually Chris Davis, who asked for eight years and got seven.)
And the median outcome for a player who was looking for X dollars was: 87.5 percent of X. Exactly the same! A player who asked for $18 million per year got, on average, $15.75 million.
That puts Aroldis Chapman in line to get $87.5 million, and I wouldn't be surprised at all. It will still come as a shock to the system, but most precedent-setting contracts do, until the precedent is set and we quickly get used to them.
The concept of seeking a specific contract is, of course, total nonsense. This isn't an eBay sale. There's no "buy it now" button, and if a player who seeks $100 million gets an offer for $100 million, he's going to start seeking one that's worth $105 million.
But the linguistics of the offseason aren't arbitrary. A player "seeks" a certain contract for a lot of reasons: he wants to anchor the negotiations at a certain range, or he wants to encourage comparisons between himself and another player. His contract will ultimately reflect the value that the sport places on him. The contract he seeks reflects the value that he places on himself. There's meaning in it. There's also, broadly speaking, useful information to the consumer of the rumor.
A few leftover details that might be relevant:
Why median instead of mean?
We went with median instead of mean because a few extremely unrealistic players dragged down the average. That might or might not have been the right decision, so here are the mean outcomes: For years, the average player got 76 percent of what he was seeking; for dollars, he got 79 percent.
Does the date of the rumor matter?
It does! Rumors from November, October and September turn out to be closer to the actual signed contracts than rumors from December and onward. This goes counter to my hypothesis that agents would start high but settle into more realistic territory once actual negotiations have happened. My new hypothesis is that once December comes along, the data starts to gather the unsigned players who are simply unrealistic about themselves, or whose stubbornness will end up suppressing their salaries. It might also be that the later the offseason gets, the more likely an outrageous request will be talked about and reported. (Also, the difference is small.)
Do I have a favorite?
I do have a favorite! In the offseason before the 2009 season, Adam Dunn asked for four years and $56 million. That was apparently so absurd that even a rival agent mocked it, saying he wouldn't get more than $5 million per year. Dunn ended up settling for two years and $20 million, and over the next two years he did almost exactly what he had done before his free agency:
2007-2008: 80 HR, .919 OPS, 206 RBIs, 2.5 WAR
2009-2010: 76 HR, .910 OPS, 208 RBIs, 2.0 WAR
So he hits free agency again and asks for ... four years and $60 million. The best explanation here is that two years earlier, Dunn had an honest view of himself and his place in the market. The world rejected it, but Dunn meant what he said and he held to it. Given a second chance to describe himself to the world, he chose the same words, the same story. What Dunn sought, he sought with total honesty.
The second time, he got paid four years and $56 million.