Collin Graf
2025-26· Data as of July 27, 2026
Status derived from CBA rules; edge cases (Group 6 UFA, 10.2(c), 35+ contracts) may be misclassified. Years remaining are not yet tracked.
Surplus value ~$2.1M/yr
Pending restricted free agent (RFA) His contract expires this season, but as a restricted free agent his current team keeps his signing rights and can match rival offers. This is a rating, priced in dollars — what the model rates his on-ice value at, not a forecast of what he'll sign for. See the projected signing price on the Free-Agent board.
Realizable today Entry-level rights let the team realize about +$304K of this as signable/tradeable surplus; the rest is locked-in cost control — real value the team enjoys but can't directly cash out while he's signed this cheap.
Track record Judged as-of — with only the model and data available each season — this deal has returned +$1.5M over 2 seasons through 2024-25. See how past calls held up →
+0.02 cap-years of surplus ≈ +$1.5M
Judged as-of — each season priced with only the model and data available then — Collin Graf has delivered 0.02 cap-years more value than he's been paid across our data window. One cap-year = a full season's salary-cap ceiling.
Career still active — this total will keep moving.
Low-confidence total — thin or short seasons drive a large share of the swing, so read it with caution.
- Frank Nazar +$156K
- William Eklund +$741K
- Noah Ostlund +$185K
Why this verdict
Why this verdict
Where this contract sits vs. the market
Each dot is a real free-agent signing (2015-2026): 1082 comparable forward deals. The line is the going rate at each talent level, the market convention we price against. Real signings scatter around it; comparable players don’t all sign for the identical number.
The page’s headline verdict, +$304K/yr, refines this by adjusting the market rate for age, contract term, and restricted-free-agent status. The plot compares to the raw market curve; the verdict adjusts it.
This player is on an entry-level contract. Part of the gap below the curve is CBA cost control (teams pay restricted players below open-market value by rule), not evidence of a market inefficiency.
Structurally similar active deals
This contract sits in the Young / bridge region of the contract map. These are the active contracts whose shape is nearest to it, closest in spend, term, age, and role (not how they aged):
- Frank Nazar $950K +$156K
- William Eklund $863K +$741K
- Noah Ostlund $887K +$185K
- Stephen Halliday $950K +$251K
- Fraser Minten $817K +$305K
- Justin Sourdif $825K +$520K
Comparable by the shape of the deal, not by team or playing style. Archetypes are broad regions on a continuum; a deal near a boundary is genuinely in-between.
The range reflects pricing-model error, not measurement — fair value is a modeled estimate. The sample badge reflects games played, not dollar accuracy.
Outlook
Outlook
Approaching his peak
Production index for forwards. The dot marks Collin at age 23.
Empirical production-by-age index (delta-method WAR fit, 2010-11+). Illustrative of typical aging; the dollar model prices age with a separate, softer decline.
How this deal's surplus has accumulated — each elapsed season judged as-of,
scored with only the model and data available that season (no hindsight) and priced at the
player's age that year. This retrospective runs through 2024-25; the
current-season verdict is shown above. The waterfall builds the running total left to right; the subtotal below is a tally to date, not a forecast. Single-season value is noisy, and early-year
deficits partly reflect a lower salary cap — the same AAV was a smaller share of
the cap then — while the model tends to underprice an elite player's early-career
season, so read the early seasons with extra caution.
Partial record: part of this deal falls outside our as-of coverage, so only its 2023-24–2024-25 seasons are graded here.
Full stats
Full stats
On-ice expected-goals impact per 60 — RAPM-adjusted (teammate- and competition-adjusted) where career 5v5 ice time clears the reliability floor, a raw on-off xG differential otherwise. Not wins. The RAPM figure is career-pooled (one value across seasons); the on-off fallback shows the most recent season.
Counted outcomes — totals and rates that carry shooting and goaltending luck.
| Season | GP | G | A | P | TOI |
|---|---|---|---|---|---|
| 2025-26 | 81 | 21 | 25 | 46 | 16.4 |
| 2024-25 | 33 | 5 | 6 | 11 | 15.9 |
| 2023-24 | 7 | 0 | 2 | 2 | 16.6 |
On-ice shot shares and modeled expected goals — these stabilize faster than goal totals, but xG is an estimate, not a count.
| Season | CF% | xG | xGF |
|---|---|---|---|
| 2025-26 | 44.3 | 20.0 | 55.3 |
| 2024-25 | 44.2 | 4.2 | 18.5 |
| 2023-24 | 34.9 | 0.7 | 3.0 |
- CF%
- Corsi-for percentage — share of shot attempts while on the ice
- xG
- Individual expected goals from the player's own shot attempts
- xGF
- On-ice expected goals — team's xG while this player is on the ice