xJawn NHL expected goals

← Players · ← 2026-27 season

TOR Chris Tanev

D · career · 16 seasons · value in goals above league average per 60 minutes of ice

GAA/60 by season

How to read this chart

One point per season: GAA/60, the accounting sum of that season's 5v5 value — chances created (Off xG-RAPM) + chances suppressed (Def xG-RAPM) + finishing above expectation — in goals above league average per 60 minutes of 5v5, power-play and penalty-kill ice. The players table shows the same accounting as a season total, which keeps ice time inside the number; this page divides it out, because comparing a player to himself across seasons should not be driven by how long each season was. The chart opens on the 3-season weighted average: each point is the ice-time-weighted average of the player's GAA/60 over that season and the two before it. A single season of RAPM is noisy (year-over-year correlation ≈ 0.5, the defensive half worse), so the smoothed line is the honest read — it answers "is this level real?" rather than "what did one 1,000-minute sample say?". The Single season toggle shows the raw per-season number instead, swings and all.

The grey bands are the league that same season, in the same view: the brighter fill is the middle half of skaters (25th–75th percentile), the faint fill the 5th–95th, the dashed line the league median — all over skaters with at least 300 minutes of 5v5 ice time (window-summed in the 3-season view). Where a point sits inside the bands is its percentile; the label above it makes it exact. Hollow points fall under the floor — they draw (the line is the career) but get no rank.

The vertical axis is GAA/60 — a linear goals-above-league-average scale, so heights add up. It opens on just the orange net line over the bands. Click Off / Def / Fin in the legend to stack that piece in: chances created and finishing (when positive) fill up from the dotted zero line, chances against fill down when a player is bleeding them, and the net line is their sum. Toggle each one independently — the line says how good, the stack says where the value came from.

Season by season

SeasonTeamTOIOffDefFinGAA/60GAA (goals)Percentile
2025-26TOR164 -0.00+0.13-0.00 +0.24+0.7 —
2024-25TOR1,438 -0.02+0.34+0.00 +0.35+9.6 95th
2023-24CGY1,604 +0.05+0.19+0.00 +0.17+5.3 82nd
2022-23CGY1,062 +0.15+0.28-0.00 +0.42+8.8 96th
2021-22CGY1,486 -0.05+0.06+0.01 -0.05-1.3 49th
2020-21CGY974 +0.01+0.18+0.01 +0.20+3.8 90th
2019-20VAN1,271 -0.06-0.05+0.01 -0.20-5.1 11th
2018-19VAN866 -0.12-0.00+0.01 -0.17-3.0 12th
2017-18VAN656 -0.04+0.04+0.00 -0.03-0.4 50th
2016-17VAN870 -0.09+0.09+0.01 +0.01+0.1 59th
2015-16VAN1,155 -0.06+0.11+0.01 +0.07+1.6 74th
2014-15VAN1,280 -0.01+0.06+0.01 +0.12+3.1 84th
2013-14VAN1,049 +0.05+0.05+0.01 +0.16+3.3 80th
2012-13VAN567 -0.03+0.08+0.01 +0.13+1.3 77th
2011-12VAN446 +0.03+0.10+0.01 +0.20+1.6 94th
2010-11VAN448 -0.01+0.09+0.02 +0.14+1.1 86th

2026-27 is not rated yet. The curve and the table end at 2025-26, the last season the value artifact covers; this season's counting stats are on the season card and the season joins the curve once it has enough games to refit.

The table is always the raw single-season record — the ledger the 3-season line smooths. TOI is 5v5 minutes. GAA (goals) is the rate × that season's ice time — the season's total value in goals. Percentile is against that season's qualified skaters, single season.

GAA/60 comes from value artifact value__f4dc1877 (accounting layer, docs/player-value.md): single-season xG-RAPM plus the modeled finishing rating over the player's own 5v5 ice time; the 3-season view is the trailing ice-time-weighted average of those. It is a separate surface from the box-score xG model badged in the footer.