Skip to content

Tue, Jan 5, 2027

Scores

xG
Expected goals: the model’s chance that each shot attempt scores, added up.
xG-Elo
A team rating built on xG margin, not results.

Win chance is the model’s projection, not a betting recommendation. Its accuracy is modest (held-out AUC 0.590, on a scale where 0.5 is a coin flip and 1.0 is perfect).

What the model sees · 2025-26 · 32 clubs · rank gap in spots

The model’s rating against the standings

The Boston Bruins (BOS) are 8th in points. The model ranks them 24th.

Rank gap in spots, 2025-26.

Drawn: the 9 clubs 7+ spots apart, of 32.

How the gap is measured — rank in points minus rank by xG-Elo; positive means the model ranks a club higher than its points do

Each bar is one club: its rank by xG-Elo, the model’s rating built from the chances a club makes and allows, against its rank in the standings. #1 is the top of each ranking, so the Boston Bruins, #8 in points and #24 by xG-Elo, have a gap of 8 − 24 = −16.

Green bars are clubs the model ranks higher than their points; red bars, lower. The words carry the direction too, so the colour is never the only sign.

2026-27 full-season pace

League leaders

Jacob Melanson leads with 419.9 projected hits from a limited sample of 37 games, ahead of Kiefer Sherwood at 402.7.

Limited sample means under 41 games, half a season. The pace stretches that short record across a full season, so it rests on limited evidence and can move a long way as games are added.

  1. 1Jacob MelansonSEA · R419.9 projected hits37-game sampleLimited sample · 37 games
  2. 2Kiefer SherwoodSJS · L402.7 projected hits267-game sample
  3. 3Yakov TreninMIN · C343.6 projected hits420-game sample
  4. 4Marc GatcombVGK · C330.2 projected hits89-game sample
  5. 5Jeremy LauzonVGK · D323.0 projected hits330-game sample
How the leaders are counted — a rate carried across every game of the club’s 2026-27 regular season, not a forecast of games he will play

A projection is the player’s recent rate per game multiplied by his club’s full schedule, assuming he plays every game. The sample is the games the rate rests on, and a dotted bar marks a projection where a solid one marks a count.

Full projections board

Injuries, lineups, transactions and rumours · newest first

News

Headlines from other outlets, shown as published, each linking to the full report.

  1. Lineup

    Penguins Assign Forward To Wilkes-Barre/Scranton (external site)

    The Hockey News - Pittsburgh Penguins via Yahoo Sports ·

    In this story: Rutger McGroarty, Pittsburgh Penguins

  2. Lineup

    Blackhawks Vs Mammoth: Projected Lineup, How To Watch, & More Ahead Of Game 2 (external site)

    The Hockey News - Chicago Blackhawks via Yahoo Sports ·

    In this story: Chicago Blackhawks, Utah Mammoth

All NHL news

What the model sees — not what happened

The model expects extreme finishers to move back toward average

Goals relative to each player’s position: 2025-26 beside the 2026-27 projection.

What happened, 2025-26Projected for 2026-27, not a result
  • Mackie Samoskevich−8.6 goals relative to his position in 2025-26, projected −1.6 in 2026-27
  • Tomas Hertl−8.8 goals relative to his position in 2025-26, projected −1.6 in 2026-27
  • Andrew Copp−8.9 goals relative to his position in 2025-26, projected −1.6 in 2026-27
  • Timo Meier−10.0 goals relative to his position in 2025-26, projected −1.8 in 2026-27
  • Anders Lee−11.5 goals relative to his position in 2025-26, projected −2.1 in 2026-27
  • Cole Caufield+13.2 goals relative to his position in 2025-26, projected +2.4 in 2026-27
  • Macklin Celebrini+13.2 goals relative to his position in 2025-26, projected +2.4 in 2026-27
  • Filip Forsberg+13.8 goals relative to his position in 2025-26, projected +2.5 in 2026-27
  • Morgan Geekie+14.5 goals relative to his position in 2025-26, projected +2.6 in 2026-27
  • Anthony Mantha+15.5 goals relative to his position in 2025-26, projected +2.8 in 2026-27

Both ends use the position’s rate rather than raw goals minus xG, and the projection assumes the same shot volume. A gap that persists may be finishing skill or shots the model under-rates. How this is measured