NHL Fantasy Draft Value: Goalie and Skater Valuation Models
NHL fantasy drafts operate on a split-position economy that no other major sport replicates quite so starkly: goalies and skaters are valued through fundamentally different lenses, using different statistics, different risk profiles, and different scarcity math. Getting either category wrong — overinvesting in a tandem goalie situation or undervaluing a power-play specialist — compounds across a full season. The models below apply to standard rotisserie and head-to-head formats across the major fantasy hockey platforms, including Yahoo, ESPN, and CBS Sports.
Definition and scope
Draft value in NHL fantasy is the expected statistical return a player delivers relative to the draft capital spent to acquire them. For the broader framework of how that concept applies across sports, Draft Value Analytics covers the foundational architecture. In hockey, the calculation forks at the position level in a way that matters immediately at the draft table.
Skater value is built around category volume: goals, assists, points, plus/minus, power-play points, shots on goal, hits, and blocks, depending on league settings. Goalie value is built around rate statistics and workload: wins, save percentage, goals-against average, and shutouts. These two valuation systems don't translate to a single currency naturally — a 0.920 save percentage and 38 wins don't convert to "points" the way a 35-goal season does. Draft models that paper over this distinction tend to produce rosters that are either goalie-heavy and skater-thin, or loaded with top-6 forwards and counting on streaming to fill the crease.
The scope here covers standard 12-team rotisserie and head-to-head formats with 8-10 categories. Dynasty and keeper-league adjustments, where aging curves for skaters diverge dramatically from goalie durability windows, represent a separate modeling layer.
How it works
A sound NHL valuation model assigns each player a z-score — a measure of how far above or below the replacement-level baseline they sit in each scoring category — and sums those scores into a single draft value number. The methodology traces directly to the Value Over Replacement Player framework, where replacement level is defined as the production available from the last player rostered at each position.
For skaters, the process unfolds in five steps:
- Apply positional scarcity metrics to identify where the talent pool thins fastest — historically, elite point-per-game centers disappear before comparable wingers.
For goalies, the z-score calculation runs against a goalie-only population — typically the top 24 starters in a 12-team league — because mixing goalie rate stats into a skater pool distorts both pools. A goalie's workload variable (starts projected over 82 games) scales their rate-stat value: a 0.920 save percentage across 60 starts is worth materially more than 0.920 across 35 starts. Most projection systems, including those published by Dom Luszczyszyn at The Athletic, treat starter probability as the primary multiplier on goalie value.
Common scenarios
The tandem goalie trap is the most common draft-day miscalculation in NHL fantasy. When two goalies split a crease at roughly 55/45 or 60/40, both carry depressed win totals and capped starts. Drafting the "top" goalie in a genuine tandem situation — Carolina's Pyotr Kochetkov and Frederik Andersen, for example, in seasons where both were healthy — at the same cost as a true 60-plus-start starter misprices risk by a significant margin.
Power-play specialist inflation runs in the opposite direction. Skaters who generate a disproportionate share of their value on the power play — say, a defenseman producing 40 of his 55 points with the man advantage — face sharper year-to-year variance than even-strength contributors. Custom scoring value adjustments matter here: in leagues that track power-play points as a separate category, that same defenseman's value jumps considerably.
Late-round goalie streaming represents a deliberate strategic posture. Some managers draft one confirmed starter in rounds 5–7, then stream backup-turned-starter situations across the season. This approach trades goalie certainty for extra picks on high-upside skaters — a reasonable exchange in leagues where the waiver wire turns over actively.
Decision boundaries
The model produces clearer guidance at the extremes than in the middle — which is true of most valuation frameworks in NHL fantasy draft value contexts.
At the top of the draft, the first 3 rounds are almost entirely skaters. McDavid, Draisaitl, and the two or three other true elite centers set the floor for the first 10 picks in most drafts. No goalie belongs in round 1; the expected value gap between the best goalie and a replacement-level starter is smaller than the gap between a top-3 center and a mid-tier center. That's the position scarcity argument applied specifically.
At the middle rounds (roughly picks 50–120 in a 12-team, 10-round draft), the decision boundary is whether to take a first goalie or continue stacking skaters. The defensible window for a first goalie is rounds 5–7. Earlier than that, the surplus value drafting math rarely justifies it. Later than round 8, starter certainty becomes genuinely uncertain for most available options.
At the back end, the correct posture is almost always skater-biased. Late-round goalie speculation on emerging starters is a reasonable gamble — late-round value targets covers that calculus in depth — but only after the skater roster is structurally sound across all categories.