Free Agent Predictions Retrospective, Part Two
Yesterday, I looked into how the predictions at FanGraphs, both crowdsourced and those produced by me, matched the contracts awarded to free agents this offseason. Today, I’d like to dive into a few cases where I made mistakes in individual player predictions, as well as a few I think I did well on. In each case, I’ll try to come up with some takeaways for predicting contracts in the future.
My prediction: 3 years, $141 million
Crowd prediction: 3 years, $120 million
Actual contract: 5 years, $185 million
My lesson here: Don’t predict an unprecedented contract if you’re aiming for accuracy. It made sense to me that Jacob deGrom would sign a deal that outstripped any before him when it comes to average annual value. He’s the best pitcher in the game when healthy, and that was enough for me. Why wouldn’t he have the biggest contract?
That’s a silly way of thinking about it in retrospect. He merited a huge contract, and he got one, but why in the world would someone with his injury history want a short-term deal? In all honesty, it doesn’t need to be more complicated than that. Predicting something outside of the ordinary is fine, but extraordinary predictions should require extraordinary confidence, not merely “I think this would be neat.”
It’s one thing to predict this kind of deal for, say, Justin Verlander (I predicted two years and $80 million, the crowd came in at two and $70 million, and he got two years and $86.67 million). There’s a close comp in Max Scherzer, and Verlander is old enough that a long-term deal was never likely. High-dollar contracts for pitchers are within the realm of possibility, and indeed, deGrom’s $37 million per year average annual value is still solid on that front. But if I want to get better at predictions, I’ll need to stop trying to shoot the moon. This one was particularly frustrating because my estimate would have been quite good if I’d gone with a five-year contract; I just wanted a cool thing too much.
My prediction: 2 years, $18 million
Crowd prediction: 3 years, $39 million
Actual contract: 2 years, $33 million
My lesson here: Don’t let a personal evaluation of a hitter override the broader market view. I’m a big Josh Bell fan personally; I was a member of his Nationals book club and really enjoyed his discussion of the mental game of baseball. Heck, we literally read The Mental Game of Baseball in that book club, and he did a great job walking through how he thinks about it. The evaluation I’m talking about here is Bell the hitter, though, and I just couldn’t wrap my head around him as a difference-maker offensively.
From a projections standpoint, that’s roughly right; he looks like an average contributor overall, right around 2 WAR per 600 plate appearances. Those kinds of players haven’t been valued very highly in free agency in recent years, and I let that general view guide my evaluation. I shouldn’t have, though, because Bell’s pedigree is better than that. He was an All-Star in 2022 and his potential remains sky high; when he’s on, he’s a power hitter with good patience. He had a 143 wRC+ in his stint with the Nationals last year, for example.
If a team knew they were getting that 143 wRC+, he would have gotten more than he did. But I just look at Bell’s statistics and Brandon Drury’s and don’t see much difference. It’s pretty clear that the industry as a whole doesn’t share my view – Bell finished a lot higher on various lists of top free agents than he did on mine, and teams generally treated him like more of a name than was my instinct. I would have preferred to go after one of the many plug-and-play bats – Trey Mancini, Drury, Matt Carpenter, Wil Myers, or even Carlos Santana.
This probably sounds bull-headed of me, but I still think I’m right. The lesson here, though, is that just because you have a different view on the market than everyone else doesn’t mean that you should project a contract that matches your own view. I personally would not have extended Bell the deal that the Guardians did if I were their GM. But I’m not. Mike Chernoff is, and Chris Antonetti is their President of Baseball Operations. My job is to predict how those people think, not to play armchair GM, and that’s a good lesson to learn.
My prediction: 1 year, $9 million
Crowd prediction: 3 years, $30 million
Actual contract: 3 years, $40 million
This was an easy one: Remember to dig through the bulk bin of “pitchers who had decent seasons in relief” and see if any of them are actually mislabeled starters. Zach Eflin fell into the reliever bucket on my initial sort of players by position, and if he were going to be a reliever, he wouldn’t have gotten this contract. But he’s actually an oft-injured starter who was pitching in relief partially to stretch back out from injury, and the Rays valued him as such. I don’t have much to say about this one other than that it’s a good reminder that pitchers can flip back and forth between roles without their long-term outlook changing.
My prediction: 3 years, $30 million
Crowd prediction: 2 years, $15 million
Actual contract: 3 years, $33 million
The lesson here: Bet on good relievers with good peripherals, even if they have a hiccup. I thought Taylor Rogers was the second-best reliever coming onto the market this year behind Edwin Díaz before the year started, and I continued to think that even as he posted a 4.76 ERA. The stats are just good, and for multiple years. I assumed that would be a valuable commodity even if he had an ERA above 5.00 in his time in Milwaukee.
As it turns out, I was pretty much spot on his actual contract, though given that I undershot in most places I was maybe a hair optimistic about Rogers. I’m calling it a win, though, because I thought he’d be in demand and then he turned out to be in demand. Relievers with multi-year track records are a rare commodity these days. Teams still seem interested in paying a competitive rate for them, particularly when the alternative is betting on whoever popped in September.
My prediction: 7 years, $217 million
Crowd prediction: 6 years, $168 million
Actual contract: 11 years, $280 million
I’m actually not sure whether to call this a win, a loss, or a lesson to improve upon. I think it’s a bit of all three. I was very high on Xander Bogaerts‘ market heading into the offseason. Maybe I trust projections too much, but they have him as nearly a match for Trea Turner and Carlos Correa in 2023, and he’s the same age as Turner. He’s also produced at an All-Star level throughout his career, and has been remarkably durable to boot. On a blind résumé basis, he should have merited a contract that looks a lot like Turner’s. I debited him two years compared to Turner’s projection because his defense feels less projectable, but the crowd and most prognosticators docked him by meaningfully more than that.
What do you know – Bogaerts got a contract that looks like Turner-light. Sure, I didn’t have Turner’s market right, and I missed the lengthening of contracts that happened this year, but it didn’t make sense to me that someone who has been as good and as consistent as Bogaerts would have to settle for a second-tier contract. He didn’t, and again, I’m not particularly surprised by that outcome. In the future, I’m going to continue to trust projections and accumulated statistics. I’m also going to continue mostly ignoring concerns that someone will have to switch positions, or that their particular skill set will age poorly. Being good for a long time is a great predictor of continuing to be good, as far as I’m concerned.
For me, this was a fun and enlightening free agency period. There were a ton of big ticket free agents, including the most dominant pitcher of our generation and a man who just hit 62 home runs. Last year’s market was dour and marked by the looming presence of the lockout; the year before was derailed by COVID. It’s fun to make predictions when teams feel confident with their finances and good players are changing teams. I hope to improve upon these predictions next year, and given the takeaways from this article, I think I’ll be able to.