| L/X | Player | Pos | Team | Opp | $ | Proj | $/pt | Own% | DK FPPG | L10 | Opp FPA | L 5v5 | Max exp % |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Loading player pool⦠| |||||||||||||
5V5 Hockey's NHL DFS Lineup Optimizer is a free tool for building daily fantasy hockey lineups on DraftKings and FanDuel. It uses a mixed-integer linear programming solver to find the highest-projected lineup that fits a site's salary cap and position slots, rather than just ranking players by value.
Under the hood, the optimizer treats lineup building as a math problem: maximize total projected points while respecting the salary cap and every position slot. Each rule described below adds a constraint to that solve, from "3 players minimum from this team" to "no defensemen in UTIL." The solver then searches the full player pool and returns the highest-scoring combination of players it can find under every rule you've turned on, and repeats the process up to 150 times per build. It stops once a lineup is provably within a couple of percent of the best possible score rather than exhausting the search to prove the last fraction of a point, which keeps a 150-lineup build to a sensible running time on a full slate. That margin sits well inside the uncertainty already in the projections themselves, and well inside the Randomness setting below.
A stack template tells the solver to concentrate multiple skaters on the same team or line instead of spreading salary evenly across the slate. The panel lists eight template names: 3-3, 3-3-1, 4-3, 4-3-1, 4-2-2, 5-3, 3-2-2-1 and 2-2-2-2.
Each number in a template's name is a floor for its own tier, and every tier is anchored by a different team: a 4-3 always draws from two distinct teams, at least four skaters from one and at least three skaters from a second, never doubled up on a single team. A tier's floor is a minimum rather than an exact count, so a team that ends up with five skaters in the final lineup can still anchor a "4" tier. The 3-3-1 and 4-3-1 templates each add a forced bring-back on top of the same tier floors as 3-3 and 4-3: at least one skater from the opponent of each team that anchors the largest tier. That bring-back is what makes them distinct lineup shapes rather than duplicates of 3-3 and 4-3. 3-2-2-1 carries the same forced bring-back on top of a 3/2/2 tier floor.
A "Lines, every combo" option cycles through every valid line pairing across a build instead of using a sized template. Pick Lines narrows which specific lines count toward whichever template is active, and only becomes available once a template other than None is selected; with no template chosen, the line picker is disabled rather than silently doing nothing.
Line affinity and PP affinity are separate toggles that sit next to the stack template. Line affinity requires at least two rostered skaters sharing a forward line (FL1 or FL2) on the same team; PP affinity requires at least two on the same team's top power-play unit (PP1). Bring-back requires at least one skater from the stacked team's opponent on top of whatever template you've chosen; it has no team to build the bring-back from when no template is selected or when Lines, every combo is active, so that control is greyed out in both of those states.
Click any player in the pool to lock them into every lineup in the build or exclude them from the pool entirely. Locking a player forces every lineup to include them, and also exempts that player from the ALL and positional exposure sliders below, since a lock is already a specific instruction and a slider is only a bulk default. Excluding a player removes them from consideration no matter how the rest of your rules are set. A per-player exposure cap set in the pool table's Max exp % column still applies even to a locked player, since that cap is a specific instruction too.
When you generate more than one lineup, an exposure slider caps the share of your build that any single player can appear in, from 1% up to 100%. That cap is enforced in whole lineups: a 20% cap on a 20-lineup build allows a player into exactly 4 of them, and a cap too tight to reach even one lineup out of the total still allows one, so the slider can narrow a player's exposure but never shut them out completely. Separate sliders repeat that same cap by position (G, D, C and W), so you can hold your goalie pool tight while leaving forwards looser, or the other way around. Whichever cap is tighter, the overall slider or the matching position slider, applies to a given player, unless you set a per-player cap in the pool table's Max exp % column, which overrides both sliders for that one player. Locked players are exempt from the ALL and positional sliders entirely, but a per-player cap on a locked player is still honored.
A set of rule toggles sits below stacking and locks:
Randomness adds noise to every player's projection before each solve, from 0% to 50%. At 0%, every lineup in a build chases the same projections and differs only because of your exposure and uniqueness rules. Raise it, and each lineup in the batch works off a slightly different set of numbers, which is how a 150-lineup Mass Multi build ends up with real variety in who gets rostered instead of 150 near-copies of a single best-value lineup.
Start from a Cash, Single GPP or Mass Multi preset, which set reasonable defaults for all of the above, or build every rule from scratch. You can also upload your own projections by CSV if you'd rather rank the pool yourself.
Cash games (50/50s, double-ups, head-to-heads) pay out for beating a fixed target or a single opponent, so the job is building the one lineup with the best floor, not chasing a ceiling. Tournaments, also called GPPs or guaranteed prize pools, pay out a small fraction of a large field, so a tournament lineup needs to look different from the field, even at some cost to its floor.
The Cash preset reflects that: it builds a single lineup on a 3-3 stack, keeps the no-skaters-vs-goalie rule on, and resets randomness and exposure to their defaults. The Single GPP preset moves to a tighter 4-3-1 stack and adds 15% randomness across 20 lineups, trading some floor for lineup-to-lineup variety. The Mass Multi preset goes further still, building up to 150 lineups on a 3-3-1 stack with 25% randomness and exposure capped at 35%, so no single player dominates the whole build.
Treat any of the three presets as a starting point, not a fixed rule. Raise exposure caps and randomness as you move from a single cash lineup toward a large multi-entry build, and lower them again around a specific stack or player you want in every lineup you submit. For a deeper walkthrough of tournament lineup construction, read how to build winning NHL GPP lineups.
Stacking is the central idea in NHL DFS because scoring is correlated within a shift. When a line or a power-play unit scores, more than one skater on the ice is usually involved directly, drawing a goal, an assist or a secondary assist on the same play. Rostering several players from the same line or power-play unit means a single strong shift pays out across multiple roster spots at once, instead of landing on only one of your players.
Stacking is also a deliberate tradeoff between floor and ceiling. Concentrating skaters on one line or team raises your ceiling if that line has a big night, since several of your players score off the same shift, but it concentrates risk the same way: a quiet night for the stacked line affects every lineup spot built around it at once, instead of spreading that risk across the whole roster. For what each stack template actually guarantees, see Stacking Templates above.
There's no single "best lineup" on a given night; the best lineup is the one that fits your contest type and your read on the slate. The method behind finding it stays the same regardless of what that lineup turns out to be:
Working through goalies and lines before rules keeps the optimizer's math running on the best information available instead of stale projections.
An NHL DFS lineup optimizer is a tool that uses math, specifically mixed-integer linear programming, to build the highest-projected daily fantasy hockey lineup that fits a salary cap. This optimizer applies your stacking, exposure and matchup rules on top of that math, so every lineup it builds is one you'd actually consider playing.
Yes. This optimizer is free to use for both DraftKings and FanDuel, with no paywall on building, exporting or saving lineups. Set your rules, generate up to 150 lineups per build, and export them as a CSV ready to upload to your DFS site.
Yes. This optimizer supports both DraftKings and FanDuel, and switching sites reloads that site's salary cap and player pool. DraftKings uses a $50,000 cap with a flex UTIL slot; FanDuel uses a $55,000 cap with no UTIL slot. DraftKings builds include an optional rule to keep defensemen out of UTIL, a rule that FanDuel builds don't show since FanDuel has no UTIL slot to restrict.
Pick a stack template in the settings panel, such as 3-3 (at least three skaters from each of two distinct teams) or 4-3 (at least four skaters from one team and at least three skaters from a separate team), or choose Lines, every combo to test every valid line pairing across a build. The 3-3-1, 4-3-1 and 3-2-2-1 templates add a forced bring-back on top of those floors: at least one skater from the opponent of each team that anchors the largest tier. Pick Lines narrows which specific lines count toward whichever template you've chosen, and becomes active once you've selected any template other than None. Line affinity and PP affinity are separate toggles that require two rostered skaters sharing a forward line or a team's top power-play unit, without pinning you to a team-based template.
How many lineups to enter depends on the contest. Cash games and head-to-heads lean on a single carefully built lineup, since the goal is a reliable floor rather than upside. Tournaments reward more entries, since additional lineups let you cover multiple stacks and spread exposure across more players. This optimizer's Cash, Single GPP and Mass Multi presets default to 1, 20 and 150 lineups, matching those three contest types, up to a maximum of 150 lineups in any single build.
Line stacking means rostering multiple forwards from the same line, or the same power-play unit, in one lineup. Goals are correlated within a shift, so when a line scores, more than one forward on it is often directly involved in the play. Stacking is a bet that a strong shift pays out across several roster spots at once instead of landing on just one of your players.
If your settings conflict, for example a stack size larger than your max-per-team limit, or a minimum salary too close to the cap, the optimizer shows a Build Infeasible message instead of a lineup. It lists which settings are in conflict so you can loosen the one that's blocking the build instead of guessing.
This optimizer draws from the day's slate data: DraftKings salaries come from that day's posted salary file, FanDuel salaries come from your uploaded CSV, and point projections are built from the same underlying projection data as the DFS Player Projections tool, recalculated using each site's own scoring rules rather than pulled from a separate projection system. You can override any individual player's projection by hand, or upload your own full projections CSV if you'd rather rank the slate yourself.