Bold opening: The real story here isn’t who signed where—it's how our predictions held up under real market heat and what that means for future forecasting. And this is where the conversation gets interesting...
As I type, spring training is underway on a second screen, a reminder that another baseball season is almost here. By now, 48 of the Top 50 free agents from the winter have landed with teams, leaving Zack Littell and Lucas Giolito as the lone holdouts. That sets the stage for my annual assessment of contract predictions, focusing on both my work and the crowd’s performance.
I’m a big believer in reviewing my own forecasts to improve, and I also enjoy evaluating the crowd’s crowdsourced bets simply because it’s entertaining to see how sharp collective judgment can be. Our crowdsourced predictions have consistently impressed—arguably outperforming many industry experts—so sharing them adds extra satisfaction for me.
Scope and method: I categorized the forecasts into three groups—hitters, starting pitchers, and relievers—and also looked at the entire Top 50 without position breaks. My chief metric follows a formula I discussed earlier this winter, and I also examined average annual value (AAV), total guarantee, and contract length to gauge accuracy. I compared predictions to actual deals to measure how well we anticipated the overall market and how precisely we forecasted individual contracts. This approach helps answer two questions: which group nailed the market as a whole, and who predicted each individual free agent’s value most accurately.
I reused the review structure from last year, compiling predictions from myself, the crowd, and several colleagues across the industry. Each Top 50 contract was matched against the corresponding prediction for that player. If a writer didn’t forecast a contract for a given player, that player was excluded from that writer’s analysis. What follows is a broad analysis of our results, and in the appendix you’ll find a complete record of all measured categories plus notable performances from non-FanGraphs prognosticators. Quick note: in this article, a positive number signifies an overestimate, while a negative number signals an underestimate.
First, the predictions that ran hot.
Too Hot
You Aren’t a FanGraphs Member
It appears you’re not logged in or aren’t a FanGraphs Member yet. We’re not mad, just disappointed.
We get it—you want to read this article. But before you dive back in, here are a few compelling reasons to consider becoming a Member:
- Ad-free viewing so you aren’t interrupted by ads
- Unlimited articles; non-members get 10 free reads per month, but members never hit a limit
- Dark mode and Classic mode options
- Custom player page dashboards to arrange the cards you care about
- One-click data exports for your personal projects
- The option to remove homepage photos (some readers asked for it, and we listened)
- Expanded Steamer projections, including handedness, percentile, and context-neutral projections for Members
- FanGraphs Walk-Off, a personalized year-end review showing how you used FanGraphs relative to others
- A weekly members-only mailbag column
- Support for FanGraphs and our staff, helping us continue to improve the site
We hope you’ll consider a Membership—as a gift or for yourself. And yes, this has been a long sales pitch, so we’ve pared back other ads in this article to keep things smooth.
My Biggest Overestimations in 2026 Contract Predictions
Player – Estimate vs Actual – Miss
- Munetaka Murakami — 7 years, $154M vs 2 years, $34M — $90M
- Kyle Tucker — 10 years, $370M vs 4 years, $228.4M — $37M
- Tatsuya Imai — 5 years, $100M vs 3 years, $54M — $26M
- Eugenio Suárez — 2 years, $50M vs 1 year, $15M — $25M
- Josh Naylor — 4 years, $100M vs 5 years, $92.5M — $24M
A key takeaway: I misread the market for NPB stars this winter. There are two possibilities. One is that Murakami and Imai have some unique factors my style misses; but that’s unlikely since other prognosticators were stumped too. A structural pattern may be at play: beyond the very top, the posting-fee system encourages shorter, smaller deals, allowing players to reach free agency sooner and limiting fees back to the posting team. Apart from that anomaly, the misses fall in the expected range for star hitters, “pillow” contracts, and my usual enthusiasm for Josh Naylor.
Biggest Crowdsourced Overestimations in 2026 Contract Predictions
Player – Estimate vs Actual – Miss
- Munetaka Murakami — 6 years, $132M vs 2 years, $34M — $71M
- Eugenio Suárez — 3 years, $60M vs 1 year, $15M — $28M
- Brandon Woodruff — 3 years, $66M vs 1 year, $22M — $20M
- Gleyber Torres — 4 years, $72M vs 1 year, $22M — $19M
- Zac Gallen — 2 years, $44M vs 1 year, $19M — $15M
While Murakami’s deal captivated the crowd, overall the crowd’s misses skewed toward pillow contracts and signings that didn’t materialize as expected. This pattern aligns with a broader observation: the crowd’s estimates tended to be lower overall than the pros’ estimates.
For the record, I also overestimated the overall market, roughly to the same extent as the crowd underestimated it. The one systematic miss was the NPB contracts, while other misses were scattered and didn’t reveal a clear bias.
Too Cold
My Biggest Underestimations in 2026 Contract Predictions
Player – Estimate vs Actual – Miss
- Michael King — 1 year, $22M vs 3 years, $75M — -$29M
- Cody Bellinger — 5 years, $140M vs 5 years, $162.5M — -$23M
- Devin Williams — 2 years, $24M vs 3 years, $51M — -$19M
- Pete Alonso — 4 years, $120M vs 5 years, $155M — -$15M
- Alex Bregman — 4 years, $140M vs 5 years, $175M — -$12M
Ouch—King’s QO expectation backfired completely, with him landing three times as much. Outside that, the underestimations reflect a market with slightly higher demand for productive players—minor tweaks in dollars or years, but not a fundamental misread. The misses span hitters, starters, and relievers, with no single archetype dominating.
Biggest Crowdsourced Underestimations in 2026 Contract Predictions
Player – Estimate vs Actual – Miss
- Kyle Tucker — 8 years, $280M vs 4 years, $228.4M — -$40M
- Pete Alonso — 4 years, $107M vs 5 years, $155M — -$31M
- Dylan Cease — 5 years, $130M vs 7 years, $187.25M — -$28M
- Cody Bellinger — 5 years, $135M vs 5 years, $162.5M — -$20M
- Tyler Rogers — 1 year, $8M vs 3 years, $37M — -$20M
Kyle Tucker’s case illustrates a market that was highly polarizing, with the Dodgers ultimately rewarding him with a long, lucrative deal. Outside that, the misses mirror my own: expectations were a little low on several players, with a notable structural bias not tied to a specific player type. The misses spread across hitters, starters, and relievers. Overall, this wasn’t a wild departure from the norm, just a reminder that free agency can surprise in the specifics even when the general direction is clear.
Just Right
My Closest Predictions in 2026 Free Agency
Player – Estimate vs Actual – Miss
- Ranger Suárez — 5 years, $130M vs 5 years, $130M — $0
- Ryan O’Hearn — 2 years, $30M vs 2 years, $29M — $1M
- Brad Keller — 2 years, $24M vs 2 years, $22M — $2M
- Dylan Cease — 5 years, $155M vs 7 years, $187.25M — $2M
- Robert Suárez — 3 years, $48M vs 3 years, $45M — $3M
I excluded one-year deals here, because they tend to be treated separately in retrospective analyses. Overall, my grip on the top tier of the pitching market was solid, and I also had a decent read on several relievers. The winter’s contracts generally had clear market signals, especially among the top pitchers, which helped refine my estimates for those deals.
Closest Crowdsourced Predictions in 2026 Free Agency
Player – Estimate vs Actual – Miss
- Edwin Díaz — 4 years, $84M vs 3 years, $69M — $0M
- Josh Naylor — 4 years, $80M vs 5 years, $92.5M — $1M
- Shota Imanaga — 2 years, $38M vs 1 year, $22M — $1M
- Kazuma Okamoto — 3 years, $48M vs 4 years, $60M — -$1M
- Tatsuya Imai — 4 years, $64M vs 3 years, $54M — -$2M
The crowd’s performance on the Japanese players was especially impressive, with Imai nearly pinpointing his deal within a few million dollars. Excluding the NPB cases, the top performers among the crowds showed there’s no shortage of sharp, data-driven thinking within the group.
Overall Market and Final Thoughts
The pitcher market heated up early and cooled off later, while hitters, led by Murakami, never truly took off. Across five prognosticator groups (myself, the crowd, and three external analysts), the crowd and I were among the few not to overshoot the market. In a nutshell, most others projected higher totals than what actually transpired, while a few stood out for accuracy in total market size.
I adjusted a future-year multiplier to better compare contracts of different lengths, finding robustness across a range from 0.5 to 0.75 (two-thirds, as I used previously, appears to be a solid baseline). Across that span, the crowd, my predictions, and ESPN’s Kiley McDaniel were all strong contenders in different aspects.
What drove errors among the top predictors varied. If you remove the three NPB cases, my top-10 predictions were nearly spot-on, within about a third of a million per player. Excluding those also improved Kiley’s results, suggesting a touch of groupthink may have affected him—something I’ll watch for next year. The crowd, meanwhile, tended to undervalue the top nine players, implying a potential rethinking about the cost of a win in free agency. If you drop the top five players, crowd predictions become nearly perfect, though the top slices of the market are precisely what many teams find most compelling.
All told, readers like you did an exceptional job forecasting contract outcomes. The top tier came in hotter than expected, but the overall results were excellent. If a team-side analyst asked for a solid baseline, I’d recommend leaning on FanGraphs’ crowd for the broader market and using my top-tier estimates for marquee players—while probably avoiding forecasts for players from foreign professional leagues.
Other commendations go to Kiley for close alignment on total market size after adjusting for length, and MLB Trade Rumors for strong relief-pitcher predictions. In short, this was a season where nerdy spreadsheets and disciplined analysis paid off, delivering a remarkable snapshot of contract dynamics.
Appendix
The article includes tables with error metrics across multiple scoring methods, comparing my predictions, the crowd’s, and top industry peers. When two prediction sets tie on a metric, both are shown. The year-adjustment method from my earlier write-up is applied in both average and absolute terms. Highlights include:
- Overall market misses by group (Starter, Reliever, Hitter) and by model (Ben, Crowd, Kiley/McDaniel)
- Absolute misses by group and model
- AAV forecasting errors and total guarantee forecasting errors
- A non-monetary view of predicted vs actual contract years
Bottom line: the collaboration between my analysis and the crowd’s insights produced one of the strongest forecasting performances in years. If I had to pick a path forward, I’d blend the top-end precision from my own estimates with the crowd’s broader-market perspective for the rest of the field, while steering clear of forecasting foreign-league signings without additional context. And you, reader—what’s your take? Do you think the market overcorrected in either direction, or did the season’s actual results strike a fair balance? Share your thoughts in the comments.
