Now that the 2025/26 offseason is essentially over, we’re checking our work to see how well our model correlated to the league – in two markets. First up was our self-check on our model’s performance in the trade market. Now it’s time to look at how we did in the free agent market.

 

The numbers

• Number of signings: 120
• Average AAV delta between our model and the actual deal, in absolute value terms: $3.4M
• Percentage of signings that fell within our average delta or less: 73%

Last year, we came to the conclusion that the best metric for measuring market correlation was the AAV delta, in absolute value terms. It’s based on this analysis last year from Ben Clemens of Fangraphs, who came to a similar conclusion when measuring their site’s results. For example, the AAV on Josh Naylor’s deal was $18.5M. Our model projected an AAV of $19.7M, so it was off by $1.2M (whether it was high or low doesn’t matter for this exercise – but more on that below).

So on average, our model was off by $3.4M relative to the AAV of each deal, across all 120 signings.

Keep in mind that, unlike other popular outlets, we don’t predict the precise amount of years, as we think that’s a bit of a reach. In our free agent value calculator, you can toggle between years to see how the AAV changes, as the risk profile is affected by each additional year you add on.

 

Breaking it down a bit further

So we have the aggregate results, but let’s see how things shook out more specifically.

• Number of deals with an AAV delta of $2M or less: 57 (48%)
• Number of deals with an AAV delta between $2.1M and $4M: 31 (25%)
• Number of deals with an AAV delta of $4.1M or more: 33 (27%)

Looking at it this way, we can see that, in our model, about half of the deals were closer than the average delta; about a quarter were right around the average; and another quarter were misses – a bit too far off.

 

Very fair deals (48%)

(Those with an AAV delta of $2M or less between our model and the market) – more prominent names in bold:

Josh Naylor, Gleyber Torres*, Brandon Woodruff*, Ryan Helsley, Miguel Rojas, Mike Soroka, Steven Matz, Gregory Soto, Kyle Finnegan, Hoby Milner, Mark Leiter Jr., Colin Holderman, Lane Thomas, Christopher Morel, Danny Jansen, Merrill Kelly, Josh Bell, Adolis Garcia, Tyler Alexander, Caleb Thielbar, Jason Foley, Adrian Houser, Caleb Ferguson, Andrew Knizner, Alexis Diaz, Drew Pomeranz, Dustin May, Shawn Armstrong, Jacob Webb, Pete Fairbanks, Kirby Yates, Pierce Johnson, Ryne Stanek, Tyler Kinley, Andy Ibanez, Victor Caratini, Jakob Junis, Jorge Mateo, Luis Garcia, Taylor Rogers, Austin Hays, Luis Arraez, Carlos Santana, Isiah Kiner-Falefa, Framber Valdez, Jose Urquidy, Brent Suter, Scott Barlow, Paul Goldschmidt, MJ Melendez, Jonah Heim, Evan Phillips, John King, Paul Sewald, Jose Quintana, Max Scherzer, Starling Marte

*Accepted Qualifying Offer 

 

Close enough to fair deals (25%):

Deals that fell right around the average delta (between $2M and $4M) between our model and the market: 

Phil Maton, Trent Grisham*, Devin Williams, Edwin Diaz, Mike Yastrzemski, Robert Suarez, Jorge Polanco, Ha-Seong Kim, Chris Martin, Jordan Romano, Michael King, Rob Refsnyder, Ryan O’Hearn, Sean Newcomb, JJ Bleday, Alex Bregman, Willi Castro, Seranthony Dominguez, Miguel Andujar, Justin Verlander, Tomoyuki Sugano, Erick Fedde, Gary Sanchez, Kike Hernandez, Luis Rengifo, Zac Gallen, Shelby Miller, German Marquez, Ramon Urias, Griffin Canning

*Accepted Qualifying Offer 

 

Deals that fell too far off the average delta (27%):

These deals fell either too high or too low (more than $4M off the average AAV delta between our model and the market). Let’s break them down by high and low:

 

Overpays: 16

Raisel Iglesias, Emilio Pagan, Shota Imanaga, Kyle Schwarber, Pete Alonso, Tyler Rogers, Amed Rosario, Kenley Jansen, Luke Weaver, Brad Keller, Michael Lorenzen, Kyle Tucker, Bo Bichette, JT Realmuto, Cody Bellinger, Harrison Bader

 

Underpays: 17

Dylan Cease, Cedric Mullins, Paul Blackburn, Hunter Harvey, Zack Eflin, Tyler Mahle, Ranger Suarez, Yoan Moncada, Eugenio Suarez, Marcell Ozuna, Chris Paddack, Nick Martinez, Aaron Civale, Jordan Montgomery, Miles Mikolas, Chris Bassitt, Zack Littell

 

Now let’s look deeper into this last bucket for some possible factors:

 

Scarcity of impact players

We saw several overpays involving some of the biggest names of the offseason, including Tucker, Bichette, Schwarber, Alonso and Bellinger. Remarkably, Tucker's and Bichette’s deals pushed past previous limits in AAV terms (non-Ohtani) of around $43M.

In a way, that shouldn’t be surprising, as they’re all impact players. The reminder here is that teams will overspend on impact, beyond what is reasonable, due to an imbalance of supply and demand – there’s a scarcity factor there. In some ways this has always been a no-brainer, but modelers tend to skew more conservative than the market with this category.

 

Non-scarcity of lower-impact players

The inverse is also true: The amount of underpays we saw that were off relative to the fair-value threshold was about the same number. As you look at the names – like Moncada, Paddack, and Civale – the key difference is that most are not impact players. They tend to be good, not great, performers who were scooped up towards the end of the offseason, when prices dropped and bargains could be had. Smart teams know this dynamic, and wait out the market for these types of players to fall into their laps. 

 

Pitchers are also riskier, which dents their value

The other thing you’ll notice is that a majority of players in this underpay bucket are pitchers. You might be surprised to see names like Cease, Suarez, and Bassitt on this list, as they’re all quality arms. But pitchers are also riskier, so even though good quality starting pitching can be hard to find, the scarcity factor is mitigated somewhat by the injury risk, especially in longer-term deals like that of Cease. 

Meanwhile, pitchers like Blackburn, Eflin and Mahle are useful, but when you combine their injury risk with their replaceability factor, their price goes down.

 

Overall takeaways

So what can we learn from this? Mainly, that the free agent market is still inefficient – impact matters, supply and demand matter, and timing is a real factor. It’s still largely an auction-based model (the highest bidder usually wins), but not always. If you’re a team that really needs impact, don’t wait, and don’t hold back. Paying up will often get you the player you covet. But if you’re okay with spending a bit less on more pedestrian production, you’ll often do okay by waiting out the market.

As for us, the above trends will be priced a bit more into our model going forward. You’ll see slightly higher prices for impact scarcity, and slightly lower prices for more plentiful, average-ish arms. And don’t be surprised if the price estimates go down later in the offseason.

As a reminder, free agent estimates for the future can be accessed by all GM-level subscribers -- details here.