NHL DFS Strategy: Line Stacking, Power Plays, and Goalie Selection

NHL DFS is a line-correlation game. Goals in hockey are shared events between linemates, top lines play together for 40+ minutes a game, and the joint distribution of a stacked line is heavier-tailed than any uncorrelated construction. Add the goalie decision — win probability times shot volume times opposing scoring floor — and you've covered 80% of what decides NHL slates.

Hockey DFS runs quieter than NBA or NFL in public conversation but rewards the same discipline: correlated construction, ownership leverage, and honest variance pricing. What's distinctive is the scoring structure. DraftKings adds power-play and shorthanded multipliers that concentrate value on specific ice-time roles; FanDuel takes a simpler scoring approach that flattens the ceiling. Same players, different optimal builds across the two sites — as with every DFS sport, per-site config matters.

Line stacking: the correlation math

A top-line center projected for 3.5 fantasy points and his winger projected for 3.2 fantasy points do not sum to 6.7 expected in the joint distribution. When the line scores a goal, both players benefit. When the line generates shots, both accumulate shot bonuses. Correlated Monte Carlo prices the joint distribution honestly — a stacked top line's 90th-percentile combined outcome exceeds the sum-of- independent-percentiles by ~20% at typical line correlations.

A properly-modeled top-line 3-stack (C + LW + RW) has a 90th-percentile ceiling roughly 25% higher than three uncorrelated skaters at the same mean projections. That gap is why line stacking, not player picks, drives NHL GPP wins.

The specific pattern that wins: full first line (C + LW + RW) plus a first-pair defenseman who logs meaningful power-play time. Four skaters from one team, all correlated via shared ice time. See our stacking-strategy piece for the general theory.

Power-play + shorthanded multipliers on DraftKings

DraftKings adds +0.5 for power-play goals and assists, +2.0 for shorthanded goals and assists. Those bonuses aren't cosmetic — they materially reshape optimal builds. First-unit power-play skaters see roughly 40% of their goal events on the PP; that's +0.5 on 40% of scoring events, a ceiling bump the field routinely underweights.

The practical takeaway: a top-line winger who runs PP1 is a different player than the same winger's twin on PP2. Simulating the difference requires per-unit ice-time projections — the DFS Simulator NHL config uses those inputs so the ceiling column reflects the multiplier math correctly. On FanDuel, the multiplier system is absent; builds shift toward middle-six skaters and away from PP1-only leverage plays.

Goalie selection: three inputs, in order

1. Win probability. The win is worth 6 points on DK; nothing else on the goalie's stat line approaches that value. Vegas moneylines are the leading indicator; a -180 favorite carries a ~64% win probability.
2. Shots-against volume. Each save is worth 0.7 points. A goalie facing 35 shots scores 24.5 save points; a goalie facing 22 shots scores 15.4. That gap (9.1 points) equals a full skater's expected floor.
3. Opposing team scoring floor. Goalie scoring is penalized -1.0 per goal against. A goalie with a high win probability facing a soft opponent AND a low-scoring one has the shutout-plus-win ceiling that wins GPPs.

Cash-game goalies optimize on inputs (1) and (2). GPP goalies flip to input (3) — take a slight underdog whose defensive matchup suppresses shots enough that a low-scoring win plus a shutout becomes a leverage play.

DraftKings vs FanDuel NHL construction

The same NHL slate produces materially different optimal lineups on DK vs FD. DK's power-play multipliers pull builds toward PP1-heavy stacks; FD's cleaner scoring rewards balanced middle-six production. DK caps skater stacks differently than FD, which changes what a "top line + D" stack looks like structurally.

See NHL on DraftKings and NHL on FanDuel for the site-specific scoring configs and slate-type coverage.

Portfolio construction for NHL multi-entry

1. Pick 3-4 stackable top lines across the slate (Vegas team totals, matchup quality, first-line minutes).
2. For each stack, build 2-3 variants: different D-man pairings, different goalies (never pair a stack with the opposing goalie — that's anti-correlated), different punt skaters.
3. Cap exposure per top-line stack at 25% of your portfolio; per goalie at 30%.
4. Sort by ceiling for GPPs; by floor for cash. NHL variance runs high — the delta between ceiling and floor is usually 25-30 points on a full lineup.

Related

• DFS stacking strategy — the general correlation math
• Ownership and leverage — how to fade chalk lines without over-fading
• DFS bankroll management — NHL variance is wide; bankroll discipline matters
• NHL DFS simulator