Projecting Arbitration Extensions
Expanding on yesterday's work
Projecting Extensions
Yesterday, I talked at length about how arbitration extensions compare to free agent deals in terms of value. Along the way, I came up with a relatively simple way to project pre-arb extensions that works pretty well.
While the median free agent contract since 2000 has had negative returns, we can use the long-run median value of $200,000/year and say that in general, one should expect the typical free agent deal to be simply a neutral return of $0/year. That is, people are paid what they’re worth, and past salaries can be used as a basis for what future players should get in an extension. Remember - the issue with free agency is not necessarily the salary, but the term. Nobody has a problem with Trea Turner at $28 million per year, but rather the 11 year length. Most exorbitant contracts that come to mind aren’t part of the database yet, so there’s no need to give them special consideration.
Therefore, we can use rough base cases (this is harder than it seems, with the caliber of players we’re talking about) to find their expected earnings based on the minimum, arbitration, and FA years covering the extension. Given we can establish a worthwhile return at about $5 million per year for an Arb extension (~42% of all extensions reach that threshold), we can remove $5 million per season to compensate the team for the added risk by guaranteeing years that normally would be team control. While a value of around $3 million would be considered the 50th percentile return based on the data, I’m trying to keep it to clean, simple numbers, and I’m not sure that would be quite enough value for a team to justify all of the extra risk taken on. Also keep in mind that while I’ll try to be fair about the comparisons, I’ll air towards the side of the team when possible, and focus only on hitters, which is where most of the value is. Of course, market inflation is also taken into consideration.
Here’s some test cases to check how it does with past extensions:
Yordan Alvarez
3 years of arbitration - $45 million total (Vlad Guerrero as base)
X years of FA - $32.5 million (Corey Seager, Rafael Devers as base)
6-year extension = 45 + (32.5 * 3) - (5 * 6) = $112.5 million
Real 6-year contract: $115 million
Vlad Guerrero is a similar bat-first player, albeit without the consistency of Alvarez. Seager and Devers are great left-handed bats that are more valuable defensively and on the bases than Alvarez, but slightly less so at the plate. We have to be fast and loose with the base cases - they’re more comparing quality of player than exact skillset and position.
Bobby Witt
1 year at the minimum - $1 million
3 years of arbitration - $55 million total (Juan Soto as base)
X years of FA - $40 million per year (Corey Seager, Aaron Judge as base)
11-year extension = 1 + 55 + (40 * 7) - (5 * 11) = $281 million
Real 11-year guarantee = $288 million
7-year guarantee = $141 million
Real 7-year guarantee = $148 million
Witt is not the type of player that generally hits FA, so I’m stuck with Soto, Seager, and Judge as comps. In doing so, I’m basically saying he would have likely set all the records if he hit the market.
Austin Riley
3 years of arbitration - $60 million total (Nolan Arenado, Rafael Devers as base)
X years of FA - $35 million per year (Nolan Arenado, Rafael Devers as base)
10-year extension = 60 + (35 * 7) - (5 * 10) = $255 million
Real 10-year extension = $212 million
The Braves are well-known for getting players to sign team-friendly contracts, and that certainly appears to be the case here for Riley. Arenado was a touch better when he signed his big deal, but that was 5 years ago at $32.5. Devers recently got $32 per year, and I believe Riley is a decisively better player.
Matt Olson
1 year of arbitration - $20 million (Pete Alonso as base)
X years of FA - $27.5 million per year (Freddie Freeman, Paul Goldschmidt as base)
8-year extension = 20 + (27.5 * 7) - (5*8) = $172.5 million
Real 8-year extension = $168 million
Olson comes in right on the nose, although it’s hard for me to pinpoint the AAV he’d get in free agency. It felt like Freeman got less than what he deserved. Maybe $30 million is more reasonable for Olson’s FA years? In that case, it would be another Braves steal.
Will Smith
2 years of arbitration - $25 million total (JT Realmuto as base)
X years of FA - $25 million (JT Realmuto as base)
10-year extension = 25 + (25 * 8) - (5 * 10) = $175 million
Real 10-year extension = $140 million
Will Smith seems to have taken a team-friendly deal here. Based on my little model, he took $20 million per free agent year, despite posting 5 fWAR/650 PA the past 3 seasons; this year, he’s slightly down to around 3.5 fWAR/650. One could also argue that the Dodgers tacked on extra years for luxury tax purposes, with an 8-year contract being pretty much in line with what I’d expect. We can say with some confidence, then, that the model works for catchers as well.
Based on these tests, it seems to fit pretty well - the largest error per year from the deals I tested is around $3.5 million. When projecting, one would generally like to create a margin of error, which usually is expressed in a confidence interval. While I won’t manually calculate it each time here, keep in mind that we can expect it to be off by up to $5 million per year in either direction. Based on our test cases, we’d expect it to over-project for team-friendly deals more so than under-project, but our test sample isn’t particularly robust.
Now we can look at possible future extensions.
Gunnar Henderson
1 year at the minimum - $1 million
3 years of arbitration - $55 million total (Juan Soto as base)
X years of FA - $45 million per year (Corey Seager, Aaron Judge as base)
10-year contract = 1 + 55 + (45 * 6) - (5 * 10) = $276 million
12-year contract = 1 + 55 + (45 * 8) - (5 * 12) = $356 million
My prediction before the year was 12/$300 million, with the general vicinity of $275-325 million to be worth consideration. That would have been right in line then, when I gave him an FA valuation of around $40 million per year. Now, it’s $45 million, after an even better year.
This prediction can explode, though. If he plays at this level through the end of next year, and uses up his last minimum year, he will not only be older when the contract ends, with stronger base candidates, they will have to enter scary territory in buying FA years:
3 years of arbitration - $65 million total (Juan Soto as base, I guess)
X years of FA - $50 million per year (Aaron Judge as base, I guess?? No real comparison)
10-year contract = 65 + (50 * 7) - (5 * 10) = $365 million
12-year contract = 65 + (50 * 9) - (5 * 12) = $450 million
I’m not sure if one can even say such a large commitment would be worth it. I’d offer 12/$325 million tomorrow to beat Witt’s deal convincingly and beg him to accept it (he likely wouldn’t without similar opt out concessions).
Anthony Volpe
1 years at the minimum - $1 million
3 years of arbitration - $30 million total (Dansby Swanson, Carlos Correa, Trea Turner as base)
X years of FA - $30 million per year (Dansby Swanson, Trea Turner as base)
10-year extension = 1 + 30 + (30 * 6) - (5 * 10) = $156 million
My feeling was around $150 million, with an acceptable range down to maybe $120 million. Remember that Swanson was an average offensive hitter for his entire career up to FA with great defense and got $25 million a year. $30 million a year is very reasonable along the same lines when projecting a 23-year old 10 years into the future.
Elly de la Cruz
1 year at the minimum - $1 million
4 years of arbitration - $70 million total (Nolan Arenado as base)
X years of FA - $40 million per year (Nolan Arenado as base)
10-year extension = 1 + 70 + (40 * 5) - (5 * 10) = $221 million
12-year extension = 1 + 70 + (40 * 7) - (5 * 12) = $291 million
I had no real predictions here. He’s a two-way threat like Arenado, but trades a little bit of hitting value for a lot of baserunning value. He’s also 22, by the way. Perhaps he should get $45 million per year consideration like Witt and Henderson, but I’m not quite comfortable with that yet. It took some back-and-forth to get him to $40.
Adley Rutschman
3 years of arbitration - $35 million total (JT Realmuto, Will Smith as base)
X years of FA - $25 million (JT Realmuto as base)
8-year extension = 35 + (25 * 5) - (5 * 8) = $120 million
It seems low, but catchers generally do not get paid much on the Free Agent market. Rutschman will be 30 when he reaches it, hurting his case even further. By age 34, he will probably go the way of Mauer and Posey; either a first baseman, or largely irrelevant to the team’s setup. Posey’s massive extension was out of line with one should expect (i.e., Bogaerts in recent years).
CJ Abrams
4 years of arbitration - $55 million total (Trea Turner as base)
X years of FA - $30 million (Trea Turner as base)
10-year extension = 55 + (30 * 6) - (5 * 10) = $185 million
12-year extension = 55 + (30 * 8) - (5 * 12) = $235 million
This offer may look high now, but in a few years it may be a steal if Abrams continues on this trajectory. Abrams is comparable to the bat-and-baserunning mold of Turner - just don't ask about his glove.
Riley Greene
3 years of arbitration - $25 million total (Brandon Nimmo, George Springer as base)
X years of FA - $30 million (Brandon Nimmo, George Springer as base)
10-year extension = 25 + (30 * 7) - (5 * 10) = $185 million
I think Greene may still be underrated here, honestly. He looks far better than Abrams currently.
Jarren Duran
4 years of arbitration - $60 million total (George Springer as base)
X years of FA - $30 million (George Springer as base)
10-year extension = 60 + (30 * 6) - (5 * 10) = $190 million
Duran is the real deal, some questionable hot mic decisions aside. His baserunning value is seriously underrated.
Jackson Merrill
2 years of minimum - $2 million
3 years of arbitration - $40 million total (George Springer as base)
X years of FA - $32.5 million (George Springer as base)
10-year extension = 2 + 40 + (32.5 * 5) - (5 * 10) = $154.5 million
12-year extension = 2 + 40 + (32.5 * 7) - (5 * 12) = $209.5 million
It’s hard to project Merrill this far out, and 2 years of the minimum remaining lowers the asking price. I’ve given him credit as a significantly better George Springer, which seems more than fair.
Let me know what you think of these projections. I think they worked out quite well. While there is subjectivity in the base cases, it can create a solid starting point.
Sources
Spotrac for salary data