NHL Goaltending Composite Efficacy
Research Report: Comparative Goaltending Efficacy (2021-22 Stanley Cup Playoffs)
NHL Goaltending Composite Efficacy — 2021–22 Stanley Cup Playoffs. Five-pillar composite ranking with the complete qualifying table.
Inclusion Criteria
Restricted to goaltenders with a minimum of 5 Games Played during the 2021-22 Stanley Cup Playoffs.
Primary Findings: The Postseason’s Elite
The following table represents all qualifying goaltenders for the 2021-22 Stanley Cup Playoffs, ranked by their average standing across all five pillars.
| Rank | Player | Team | GP | Composite Score |
|---|---|---|---|---|
| 1 | Pavel Francouz | COL | 7 | 3.0 |
| 2 | Jordan Binnington | STL | 6 | 4.2 |
| 3 | Antti Raanta | CAR | 13 | 6.0 |
| 4 | Andrei Vasilevskiy | TBL | 23 | 6.0 |
| 5 | Darcy Kuemper | COL | 16 | 6.2 |
| 6 | Mike Smith | EDM | 16 | 6.4 |
| 7 | Jake Oettinger | DAL | 7 | 6.5 |
| 8 | Igor Shesterkin | NYR | 20 | 8.0 |
| 9 | Jeremy Swayman | BOS | 5 | 8.1 |
| 10 | Jonathan Quick | LAK | 7 | 10.2 |
| 11 | Jack Campbell | TOR | 7 | 10.7 |
| 12 | Jacob Markstrom | CGY | 12 | 11.7 |
| 13 | Louis Domingue | PIT | 6 | 12.0 |
| 14 | Sergei Bobrovsky | FLA | 10 | 12.5 |
| 15 | Ville Husso | STL | 7 | 13.6 |
| 16 | Marc-Andre Fleury | MIN | 5 | 13.7 |
| 17 | Ilya Samsonov | WSH | 5 | 14.2 |
Detailed Logical Analysis
1. The Cup-Winning Tandem
Colorado, the Stanley Cup champion, produces two of the postseason's top five Composite finishes — Pavel Francouz at #1 (3.0) and Darcy Kuemper at #5 (6.2). Francouz's profile is built on a perfect 6-0 record across 7 games (Win Rate #1, Loss Suppression #1) plus a top-tier Shutout Rate (#1). Kuemper's profile is more balanced (#3 WinRate, #5 Svs/GA, #3 LossRate). The two profiles reflect the season-long Colorado tandem signal at scale: when both goaltenders are deployed during a championship run, the team-system component of the Composite produces parallel elite results.
2. Vasilevskiy’s Workload Achievement
Andrei Vasilevskiy carries the highest postseason workload — 23 GP — and finishes #4 with a Composite of 6.0, statistically tied with Raanta. His profile (#4 WinRate, #5 Svs/GA, #6 LossRate, #5 GA/W) holds up across nearly four full series. In a small-sample playoff field where 5–7 game samples can produce extreme rankings (Francouz, Binnington), Vasilevskiy's Composite represents the era's first clear case of high-volume, deep-run efficiency.
3. The Sample-Size Reality
Of the seventeen qualifying goaltenders, ten played fewer than 10 games. Francouz (#1, 7 GP) and Binnington (#2, 6 GP) post the postseason's lowest Composites on samples too small to draw multi-series conclusions from. The methodology applies the same model to playoff data as regular-season data, but the playoff sample sizes change the reliability calculus — single rounds produce extreme rankings, multi-round runs produce more durable ones. The 5-GP inclusion threshold preserves the data; interpretation requires the GP column.
Conclusions
The 2021-22 postseason produces three structural findings the Composite captures: a Cup-winning dual-goaltender system (Colorado), a workload-overcoming run by the era's premier playoff workhorse (Vasilevskiy), and a structural reminder that playoff samples behave differently from regular-season ones. The methodology's value in the postseason is not predictive of outcome but descriptive of efficiency once the games have been played.
Thesis: The Playoff Composite as Descriptive, Not Predictive
The Cup-Tandem Confirmation
Colorado's #1 and #5 placements during a Cup-winning playoff run mirror the same team-tandem signal the regular-season Composite identified. The postseason data does not contradict the team-system finding from the year's regular-season analysis — it reinforces it. The playoff Composite functions as a checksum on the regular-season conclusions.
The Volume-Validation Apex
Vasilevskiy's 23-GP, top-five Composite is the playoff equivalent of the workload-tax counterexample. Where the regular season tends to penalize 60+ GP volume goaltenders (the workload tax), the playoff Composite rewards goaltenders who maintain efficiency across a full run. The volume signal inverts in the postseason because deep playoff GP is itself a marker of effectiveness.
The Sample-Size Caveat
The playoff sample size is the methodology's natural limit. Five-game samples produce extreme Composites (Francouz #1, Binnington #2) that the model cannot distinguish from sustained excellence. The 5-GP threshold preserves the data but the interpretation must remain literal: the model reports what happened in the games played, not what the goaltender's true level was.
Final Conclusion
The 2021-22 postseason produces a Composite that confirms a Cup tandem (Colorado), surfaces a volume-validation outlier (Vasilevskiy), and exposes the playoff sample-size limit. Read alongside the regular-season report, the postseason analysis is the descriptive complement — what the goaltenders produced once the games actually mattered.
The complete table
17 qualifying goaltenders · every pillar value and rank
| # | Player | Team | GP | GS | W | L | SA | Svs | GA | Sv% | GAA | SO | Win Rate | Rk | SO Rate | Rk | Svs/GA | Rk | Loss Rate | Rk | GA/Win | Rk | Composite |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Pavel Francouz | COL | 7 | 4 | 6 | 0 | 171 | 155 | 16 | 0.906 | 2.81 | 1 | 0.857 | 1 | 0.250 | 1 | 9.69 | 10 | 0.000 | 1 | 2.67 | 2 | 3.0 |
| 2 | Jordan Binnington | STL | 6 | 6 | 4 | 1 | 176 | 167 | 9 | 0.949 | 1.72 | 0 | 0.667 | 2 | 0.000 | 14 | 18.56 | 2 | 0.167 | 2 | 2.25 | 1 | 4.2 |
| 3 | Antti Raanta | CAR | 13 | 13 | 6 | 5 | 322 | 297 | 25 | 0.922 | 2.26 | 1 | 0.462 | 9 | 0.077 | 8 | 11.88 | 4 | 0.385 | 5 | 4.17 | 4 | 6.0 |
| 4 | Andrei Vasilevskiy | TBL | 23 | 23 | 14 | 9 | 752 | 693 | 59 | 0.922 | 2.52 | 1 | 0.609 | 4 | 0.043 | 10 | 11.75 | 5 | 0.391 | 6 | 4.21 | 5 | 6.0 |
| 5 | Darcy Kuemper | COL | 16 | 16 | 10 | 4 | 386 | 348 | 38 | 0.902 | 2.57 | 1 | 0.625 | 3 | 0.062 | 9 | 9.16 | 13 | 0.250 | 3 | 3.80 | 3 | 6.2 |
| 6 | Mike Smith | EDM | 16 | 16 | 8 | 6 | 560 | 511 | 49 | 0.913 | 3.37 | 2 | 0.500 | 7 | 0.125 | 6 | 10.43 | 6 | 0.375 | 4 | 6.12 | 9 | 6.4 |
| 7 | Jake Oettinger | DAL | 7 | 7 | 3 | 4 | 285 | 272 | 13 | 0.954 | 1.81 | 1 | 0.429 | 11 | 0.143 | 4 | 20.92 | 1 | 0.571 | 10 | 4.33 | 6.5 | 6.5 |
| 8 | Igor Shesterkin | NYR | 20 | 20 | 10 | 9 | 720 | 669 | 51 | 0.929 | 2.59 | 0 | 0.500 | 7 | 0.000 | 14 | 13.12 | 3 | 0.450 | 8 | 5.10 | 8 | 8.0 |
| 9 | Jeremy Swayman | BOS | 5 | 5 | 3 | 2 | 146 | 133 | 13 | 0.911 | 2.63 | 0 | 0.600 | 5 | 0.000 | 14 | 10.23 | 8 | 0.400 | 7 | 4.33 | 6.5 | 8.1 |
| 10 | Jonathan Quick | LAK | 7 | 7 | 3 | 4 | 228 | 206 | 22 | 0.904 | 3.43 | 1 | 0.429 | 11 | 0.143 | 4 | 9.36 | 12 | 0.571 | 10 | 7.33 | 14 | 10.2 |
| 11 | Jack Campbell | TOR | 7 | 7 | 3 | 4 | 203 | 182 | 21 | 0.897 | 3.15 | 1 | 0.429 | 11 | 0.143 | 4 | 8.67 | 16 | 0.571 | 10 | 7.00 | 12.5 | 10.7 |
| 12 | Jacob Markstrom | CGY | 12 | 12 | 5 | 7 | 354 | 319 | 35 | 0.901 | 2.95 | 1 | 0.417 | 13 | 0.083 | 7 | 9.11 | 14 | 0.583 | 12 | 7.00 | 12.5 | 11.7 |
| 13 | Louis Domingue | PIT | 6 | 5 | 3 | 3 | 187 | 168 | 19 | 0.898 | 3.65 | 0 | 0.500 | 7 | 0.000 | 14 | 8.84 | 15 | 0.600 | 14 | 6.33 | 10 | 12.0 |
| 14 | Sergei Bobrovsky | FLA | 10 | 10 | 4 | 6 | 303 | 276 | 27 | 0.911 | 2.7 | 0 | 0.400 | 14.5 | 0.000 | 14 | 10.22 | 9 | 0.600 | 14 | 6.75 | 11 | 12.5 |
| 15 | Ville Husso | STL | 7 | 6 | 2 | 5 | 228 | 203 | 25 | 0.89 | 3.67 | 1 | 0.286 | 16 | 0.167 | 2 | 8.12 | 17 | 0.833 | 17 | 12.50 | 16 | 13.6 |
| 16 | Marc-Andre Fleury | MIN | 5 | 5 | 2 | 3 | 159 | 144 | 15 | 0.906 | 3.04 | 0 | 0.400 | 14.5 | 0.000 | 14 | 9.60 | 11 | 0.600 | 14 | 7.50 | 15 | 13.7 |
| 17 | Ilya Samsonov | WSH | 5 | 4 | 1 | 3 | 148 | 135 | 13 | 0.912 | 2.97 | 0 | 0.200 | 17 | 0.000 | 14 | 10.38 | 7 | 0.750 | 16 | 13.00 | 17 | 14.2 |
Pillars: Win Rate (W/GP) · Shutout Rate (SO/GS) · Save Efficiency Ratio (Svs/GA) · Loss Suppression (L/GS) · Victory Efficiency (GA/W). Composite = mean of the five pillar ranks; lower is better. Full definitions: methodology.