Welcome to my first attempt predicting the summer 2027 NHL free agent class, none of which have signed yet because eligibility begins July 1st. Last summer featured Connor McDavid, who signed for substantially less than market value and this year features more generational talent, especially on the blueline with Quinn Hughes and Cale Makar due new deals. If either player decides to test the open market and take the highest bid, either could reset the price ceiling on defense, but Makar will probably stay in Colorado at a team-friendly price to continue competing for Stanley Cups.
If the Hockeynomics Factory Sportsbook were taking over/under wagers for these term/AAV and last June you bet all the overs, hypothetically, you would have broken my bank, driving HnF Inc into bankruptcy. That wasn’t because the “expected free agent value” (aka: EFA) modelling was flawed, it’s because players who receive/accept early extensions tend to be for more money on average (both nominally and relative to stat lines), generally required when asking players to forgo open market bidding (my predictions didn’t properly account for extension price). You get Blake-Stankoven young talent types who are overpaid long-term before producing a deserving stat line, but team understands their trajectory and willing to take that risk (which can backfire with Kotkaniemi types).
Whether it is older players aging well or young guns on upward trajectories, it tends to be players in those two upper echelons who get offered extensions. Meanwhile, players aged 25 to 32 are very unlikely to accept pay cuts before hitting the open market unless sufficient term is attached. At the top of my mind this attempt at early predictions, was actually contemplating extension price, often above what I believe the player is worth. Maybe the best way to do this is treating every 2026 and 2027 FA as pending, then building a model converting pending to extension, but I’m thinking this after the predictions are already finished and about to be published.
There is more discussion of my EFA process in the Final 2026 Predictions post, even more in the upcoming player valuation post, so I’ll avoid repetition here. EFA is 94% correlated to my prediction (lower correlation at higher income levels) and is the primary driver, but my process is listing every pending free agent against the recent comparables, then scroll up and down at different ages manually assigning values merging comps and modelling, first pass doing 2026 and 2027 together. Subsequent tweaking stacked 2027 alone and extension parameters pondered separately.
My idea of doing early and final together and building an extension model would be easy at different age and asset types to see what’s the bonus on top of normal EFA for signing before midseason (when my midseason predictions are revised and posted). A majority of these players will not be extended before midseason, especially at lower income levels, a thought crossing my mind several times while painstakingly plotting and re-tweaking for accuracy.
Remember, my primary goal is getting each and every one of these exactly correct, not preaching what each player SHOULD be worth, as many forecasters like Dom at the Athletic enjoy. My modelling is attempting to predict what the market will pay, and me too. It’s certainly not an efficient market, but I’ll assume on average it tends to be right. This list especially has several players with expected salaries above what I believe they’re really worth because they’re unlikely extending for a pay cut and/or teams tend to overpay for that asset class.
