Draft helper

These are historical comparison tiers, not projected points or a guarantee of future performance. Pool scoring, line changes, injuries and team strength can change draft value.

Skaters: 50% adjusted points/game, 30% total points, 15% shots and 5% power-play ice time. Rates use a 20-game prior of 0.50 points/game for forwards and 0.30 for defensemen. Goalies: 40% wins, 30% adjusted save percentage and 30% starts; save percentage uses a 20-game prior of .900.

Metrics become percentiles within the selected season roster and position. Missing metrics are removed for the entire position and remaining weights are normalized. Ranking requires 20 skater games or 15 goalie appearances. Ties use games played, then NHL player ID. The index is relative, not a probability.

Tier rank cutoffs: forwards 6 / 24 / 60; defensemen 4 / 16 / 40; goalies 4 / 12 / 24. Remaining qualified players are Tier 4. Search and shortlist filters do not change ranks.

Rising-star row signals: last-season rookies or players age 24 or younger on October 1 of the draft year can receive evidence notes. Skaters need 20 GP, at least 0.40 P/GP (F) or 0.30 (D), and either a 0.15 P/GP increase over a prior 20-game sample or 2:00 PP time/game. Goalies need 10-35 appearances and .910 SV%. These notes do not predict a promotion or starting job.

Last-season rookie badges come from NHL rookie-filtered reports. New faces have a small sample or no appearances in the loaded prior season; that is not proof of rookie status. Unknown age, usage and history remain unknown. New faces are unranked, not prospect projections.

New faces default to NHL points for skaters or wins for goalies, then games played. Missing values stay last in either direction. This is early NHL production, not a prospect ranking; players with no NHL sample cannot be compared from these stats alone.

Incoming prospects are ordered by their actual NHL draft selection, not a fantasy projection. The listed team drafted their rights; this does not confirm a signing, NHL roster spot, or pool eligibility. Amateur club and league are draft-time details. Prospects without a verified NHL player ID are kept separate from the shortlist.

Preseason watch uses verified final NHL GameCenter box scores because the NHL preseason club-totals feed may remain empty. It shows cumulative appearances and results for players who dressed. Split squads, changing opponents, experimental lines and small samples mean it cannot confirm an opening-night roster, line combination, power-play role or pool eligibility.

TOI and PP time are per game. All figures cover the displayed prior regular season. NHL trades update the selected roster team, not these historical totals.

NHL skater reports · NHL goalie reports · NHL usage data · NHL draft selections

Projection covers every player on the loaded current-season roster. Skaters show projected NHL goals, assists and points; goalies show projected wins because ordinary goalie value is not represented by player points.

Every player uses the same local model; there are no player-specific or article-derived overrides. Established NHL players use the two prior NHL regular seasons, availability, position, age, scoring or win rate, ice time, starts, power-play usage and a very small preseason weight.

For a recruit without NHL history, the model selects the most recent substantial non-NHL regular-season sample from the official player history, converts its scoring rate with a conservative league-equivalency factor, then applies expected NHL games and estimated deployment. Junior, NCAA, AHL, KHL and European leagues use separate factors; an unknown league uses a deliberately low fallback.

Verified top-10 NHL draft position provides only a small, capped roster-opportunity adjustment. A productive 10-to-19-game NHL audition carries more weight than feeder-league production, and its short late-season call-up is not treated like games missed through injury.

A rostered prospect with little NHL history follows a preseason recruit path after a verified final preseason appearance. Preseason can add only a few expected games and at most three game-equivalents to the scoring rate; it cannot confirm a roster spot or outweigh substantial NHL or feeder-league history.

Thin-sample projections stay pending until the official player history cache is available; the page does not publish a misleading generic low estimate while enrichment is queued.

Expected games and projected TOI are shown beside each estimate. Expected GP blends the player's average recent participation with the estimated NHL role. Non-NHL games are normalized for that league's usual schedule before comparison with an 82-game NHL season, and young recruits receive an additional uncertainty adjustment. These are draft clues, not confirmed lineups. Preseason participation can slightly move a thin-sample model but never overrides established history.

Likely ranges are intentionally wider for rookies, recruits, unknown leagues and players with little history. The projection remains an approximation, not a guarantee.

Local range calibration uses a one-season walk-forward test: the older regular season predicts the displayed historical season. F: 396 players, median absolute error 6.7; D: 200 players, median absolute error 5.2; G: 56 players, median absolute error 2.8.

Projection is independent of custom pool scoring, injuries after the refresh, and pool eligibility. Use it as a fun draft input alongside the historical tiers, current news and your own judgment.

Your personal list of players to revisit before the draft. Star a player to add them, then open Shortlist. It does not draft or reserve a player and is not shared with your pool.

The list is saved in this browser for the selected season, not across devices. PDF export temporarily sends the selected public player IDs to generate the document; the server does not save your list.

CSV includes all matching players for the current position. PDF groups all positions into at most three pages, using the current view, search and sort. Unsupported sort columns use that position's default order. Each page shows its included and total counts; use CSV for the complete list.

PDF badges use prior-season statistics: RS follows the rising-star row-signal rules; R marks last-season rookies. With 20+ GP, P requires 0.45 assists/GP for forwards or 0.35 for defense; S requires 0.30 goals/GP or 0.15; PP requires 2:00 PP time/GP. L1/L2 estimate forward usage from 18/15 minutes TOI per game, not confirmed lines. ST marks goalies with 40+ starts. Missing data does not earn a badge.