The honest value board: who actually drove value in 2025-26
The whole chart, rather than one player's Evolving-Hockey bar at a time: every Shark, split into the four jobs (even-strength offense, even-strength defense, power play, penalty kill), in one picture. Here it is.
One methodology note, because it changes how you read this. Evolving-Hockey's full GAR app is subscriber-only, so this board is built from MoneyPuck's components: individual expected goals for offense, on-ice expected goals against for defense, and the on-ice rates on each special-teams unit. It is a raw view (process, not results): it measures chances created and allowed, not who finished or who Askarov bailed out. Where it can be cross-checked against the Evolving-Hockey GAR numbers from the Fear the Fin reviews, it agrees, with one instructive exception I flag at the end.
The forwards#
| Forward | EV offense (ixG/60) | EV defense (xGA/60, lower=better) | PP (min) | PK (min) |
|---|---|---|---|---|
| Chernyshov | 0.91 | 2.80 | 31 | 0 |
| Graf | 0.84 | 2.51 | 38 | 196 |
| Sherwood | 0.80 | 2.58 | 123 | 105 |
| Gaudette | 0.80 | 2.43 | 62 | 1 |
| Smith | 0.74 | 2.72 | 215 | 4 |
| Misa | 0.73 | 2.75 | 52 | 0 |
| Celebrini | 0.72 | 2.63 | 285 | 14 |
| Eklund | 0.72 | 2.63 | 211 | 31 |
| Toffoli | 0.65 | 2.49 | 219 | 0 |
| Goodrow | 0.59 | 2.38 | 4 | 155 |
| Ostapchuk | 0.52 | 2.20 | 2 | 67 |
| Dellandrea | 0.68 | 2.98 | 6 | 118 |
| Wennberg | 0.35 | 2.35 | 252 | 164 |
The defense#
| Defenseman | EV offense (ixG/60) | EV defense (xGA/60, lower=better) | PP (min) | PK (min) |
|---|---|---|---|---|
| Dickinson | 0.22 | 2.65 | 22 | 64 |
| Orlov | 0.13 | 2.56 | 197 | 62 |
| Desharnais | 0.08 | 2.15 | 2 | 166 |
| Mukhamadullin | 0.14 | 2.48 | 3 | 69 |
Departures who carried real load (for context, all UFA or moved): Ferraro killed a team-high 268 penalty-kill minutes, Klingberg ran 156 power-play minutes, and Leddy, Skinner, and Reaves ate depth shifts. Cagnoni (3 NHL games) is below the sample cutoff.
What the four components actually say#
EV offense: the kids carry it, the blue line doesn't. Every forward who generates real individual chances is young (Chernyshov 0.91, Graf 0.84, Smith, Misa, Celebrini, Eklund all around 0.72-0.74). And every Sharks defenseman is under 0.22, with Dickinson the lone exception, which is the entire statistical argument for why this team needs a puck-moving, shot-generating defenseman (Reid, Carlson) more than it needs anything else.
EV defense: Desharnais is the back end's quiet MVP. Lowest expected-goals-against on the team (2.15), and it is not close among the defensemen. The checkers (Ostapchuk 2.20, Wennberg 2.35, Goodrow 2.38) defend; the young scorers leak (Chernyshov 2.80, Smith 2.72) and Dellandrea (2.98) was the worst on the roster by a wide margin.
Power play: the chances were there, the finish wasn't. The unit ran 7-to-9 expected goals per 60 with Celebrini, Toffoli, Smith, Eklund, and Wennberg, which is good chance creation, and it still ranked 17th. That gap is conversion and a point shot, not chance generation, which is exactly the hole a quarterback like Carlson is meant to plug.
Penalty kill: leaky, and about to lose its workhorse. Most killers sat at 7.5-to-9.5 expected-goals-against per 60 (the 28th-ranked unit), with Desharnais (6.52) and, quietly, Graf (6.52 over 196 minutes) the only real suppressors. The problem: the man who ate the most kill minutes, Ferraro, walks in free agency. The 28th-ranked PK is losing its number one, which makes re-signing Desharnais and finding kill minutes a bigger deal than it looks.
Does it agree with Evolving-Hockey's GAR? Yes, except once.#
Fear the Fin's Evolving-Hockey reads had Desharnais as the best Shark D by GAR (defense-driven), Dickinson as offense-first, and Klingberg as offense-positive with defense the flaw. This board lands on all three independently: Desharnais best EV defense plus best PK, Dickinson the only D who creates offense, Klingberg high chances created and high chances allowed.
The one divergence, and the lesson in it: Orlov.
By raw on-ice numbers he looks like a top driver, his team controlled expected goals well with him out there. But Evolving-Hockey's adjusted GAR rated him one of the least valuable Sharks defensemen, and he finished minus-28. Both are true. His individual EV offense is near zero (0.13), his on-ice number is flattered by deployment and partners, and the adjusted model strips that away. The takeaway for reading any value board: a good raw on-ice rate is not the same as driving value. Orlov's real 2025-26 was elite power play, fine defense, no even-strength offense, and a bounce-back bet for next year.
The bottom line, no narrative#
- Drove value: the young forward core on EV offense (Chernyshov, Graf, Smith, Misa, Celebrini, Eklund), Desharnais on EV defense and the kill, and Graf as the two-way utility knife (real offense plus a top PK role).
- Specialists, not all-around: Wennberg (thin EV offense at 0.35, but heavy PP and PK) and Orlov (elite PP, no EV offense) earn their keep on special teams, not five-on-five.
- Drags: Dellandrea (worst EV defense on the team), and the fourth line as a unit generated little while not always defending enough to justify it.
- Structural read: the forwards create, the defense does not, and the special teams lose two of their highest-usage pieces (Ferraro's kill, Klingberg's power play) to free agency. That is the value case, in four numbers a player, for the whole offseason plan.
Built from MoneyPuck 2025-26 component data (individual xG, on-ice xG for/against, situational TOI), 20-plus-game skaters. This is a raw, process-based view (chances), not a results or finishing view, and not Evolving-Hockey's adjusted GAR (subscriber-walled), though it cross-checks cleanly against the Evolving-Hockey reads in the Fear the Fin player reviews. EV = five-on-five.
