Tendency index, league-wide trends, and ranked dataset findings in one place.
10 findings ranked by effect size and sample size across 2016-10-12 – 2026-04-16. Descriptive historical tendencies, not betting advice.
Loading findings…
10 patterns from 123 officials. Strong-confidence findings appear first; thin samples at the bottom.
Why it matters: Two different cuts of the same data: each ref's personal over rate asks whether they beat the 6.3 combined goals in most of their games; the league-wide share counts every game individually. A ref can lean over while many of their games still go under. This uses a fixed benchmark, not sportsbook pricing.
60 of 62 officials (97%) finish under the 6.3 combined goals line in a majority of their own games (personal over rate below 50%). Across all 12,282 games, 36.4% cleared the benchmark and 63.6% went under.
Marc Joannette averages 14.8 PIM per game (-2.3 vs league) across 442 games, the largest whistle delta among 50+ game refs.
Why it matters: Year-over-year context from 10 seasons in data/baselines.json (12,282 total games). League-wide baselines, not ref-specific.
From 2024-25 to 2025-26, combined goals per game held steady (+0.2 goals) while penalty minutes rose (+1.7 PIM). Two-minute minors averaged 8.3 per team (+0.4 minors YoY). Overtime rate: 24.8% (+4.1 pp vs prior season).
27.7% of Mike Leggo's 65 games finish under 6.3 combined goals, 22.3 pts from a neutral 50% baseline.
Why it matters: Minor penalty rate is descriptive game-management pattern from NHL logs. Not a live foul-trouble or power-play edge.
Mike Leggo averages 7.2 two-minute minors per team per game (-1.2 vs league) across 65 games.
Dave Jackson reaches overtime or shootout 13.8% of the time (9 of 65 games), 11.0 pts below the 24.8% league rate.
Mike Leggo's 65 games average 5.2 combined goals (27.7% over 6.3), one of the largest scoring deltas in the pool.
Why it matters: Matrix splits compare ref×team win rate to each team's overall record in this dataset. Standout cells require 8+ games; descriptive history only.
Ottawa Senators are 10-2 (83.3%) in 12 games with Brad Watson, vs a team sample baseline of 44.3%.
Why it matters: Matrix splits compare ref×team win rate to each team's overall record in this dataset. Standout cells require 8+ games; descriptive history only.
Colorado Avalanche are 1-8 (11.1%) in 9 games with Tom Kowal, vs a team sample baseline of 56.9%.
Highest: Cody Beach on Tampa Bay Lightning (8.9 avg). Lowest: Cody Beach on Los Angeles Kings (4.2 avg).