Special Teams
Which power plays and penalty kills are real, and which are riding a hot finish or a hot goalie. The draws that start every unit, and what the playoffs do to special teams. Per-team detail lives on each team's Special Teams tab.
Power-play forensics - mirage or real?
A hot power play is either built on winning draws and creating chances (sustainable) or riding a hot finish that regresses. Winning the offensive-zone draw is a repeatable skill; finishing above expected is not. 2015-16 to present.
Biggest mirages - great PP%, carried by a hot finish
| Team-Season | PP% | Draw | Finish | Verdict |
|---|---|---|---|---|
 MTL2025-26 | 23% | 48% | +17 | Mirage |
 TBL2018-19 | 28% | 54% | +34 | Mirage |
 CHI2024-25 | 25% | 46% | +19 | Mirage |
 COL2017-18 | 22% | 49% | +12 | Mirage |
 NJD2015-16 | 20% | 53% | +14 | Mirage |
| CHI2015-16 | 23% | 55% | +15 | Mirage |
 STL2015-16 | 22% | 57% | +18 | Mirage |
| TBL2023-24 | 29% | 52% | +18 | Mirage |
 BUF2022-23 | 23% | 50% | +9 | Mirage |
 WSH2017-18 | 23% | 50% | +9 | Mirage |
 WPG2024-25 | 29% | 53% | +22 | Mirage |
| WSH2020-21 | 25% | 47% | +8 | Mirage |
Robbed - good process, cold finishing (regression due up)
| Team-Season | PP% | Draw | Finish | Verdict |
|---|---|---|---|---|
 NYI2024-25 | 13% | 64% | -26 | Robbed |
 PIT2023-24 | 15% | 60% | -23 | Robbed |
| COL2016-17 | 13% | 60% | -15 | Robbed |
 BOS2024-25 | 15% | 57% | -18 | Robbed |
| NYI2018-19 | 15% | 50% | -20 | Robbed |
 LAK2021-22 | 16% | 60% | -19 | Robbed |
| NYI2016-17 | 15% | 56% | -14 | Robbed |
 NYR2024-25 | 18% | 56% | -15 | Robbed |
 NSH2019-20 | 17% | 56% | -11 | Robbed |
 TOR2015-16 | 15% | 58% | -12 | Robbed |
| LAK2019-20 | 17% | 56% | -11 | Robbed |
 VGK2020-21 | 18% | 54% | -9 | Robbed |
Penalty-kill forensics - real or a hot goalie?
The mirror of the power play. A good kill is either built on winning D-zone draws and smothering chances (sustainable) or riding hot goaltending that regresses. On the kill the goalie matters even more than the shooters do on the power play. 2015-16 to present.
Biggest mirages - great PK%, carried by hot goaltending
| Team-Season | PK% | Draw | Goalie | Verdict |
|---|---|---|---|---|
 EDM2020-21 | 82% | 43% | +11 | Mirage |
| BUF2025-26 | 82% | 41% | +15 | Mirage |
 CBJ2016-17 | 83% | 42% | +11 | Mirage |
| NJD2015-16 | 83% | 41% | +10 | Mirage |
| EDM2019-20 | 84% | 41% | +14 | Mirage |
 ANA2016-17 | 85% | 54% | +20 | Mirage |
| TOR2025-26 | 81% | 49% | +18 | Mirage |
 SJS2021-22 | 85% | 43% | +17 | Mirage |
| STL2015-16 | 85% | 47% | +12 | Mirage |
| NYR2021-22 | 82% | 41% | +10 | Mirage |
| COL2017-18 | 83% | 39% | +7 | Mirage |
| SJS2017-18 | 85% | 48% | +14 | Mirage |
Robbed - good process, cold goaltending (regression due up)
| Team-Season | PK% | Draw | Goalie | Verdict |
|---|---|---|---|---|
 PHI2020-21 | 73% | 53% | -14 | Robbed |
| LAK2022-23 | 76% | 46% | -18 | Robbed |
 ARI2015-16 | 77% | 53% | -14 | Robbed |
 VAN2017-18 | 78% | 49% | -18 | Robbed |
| COL2016-17 | 77% | 52% | -10 | Robbed |
| LAK2018-19 | 77% | 48% | -13 | Robbed |
| WSH2018-19 | 79% | 47% | -13 | Robbed |
| TOR2019-20 | 78% | 44% | -10 | Robbed |
| BOS2024-25 | 76% | 46% | -11 | Robbed |
 OTT2023-24 | 75% | 54% | -5 | Robbed |
| EDM2017-18 | 77% | 46% | -12 | Robbed |
| OTT2016-17 | 80% | 48% | -8 | Robbed |
The faceoff men - who wins the special-teams draws
The draw is the one special-teams event a single skater controls, and it's a real, repeatable skill. These are the centers behind every team's draw bar - observed win% regressed to a true rate (toward the league mean, so a hot 50-draw sample doesn't top the board). Wingers and defensemen don't take the draw, so they aren't graded here.
Best on the power-play draw
| Center | Draws | Won | True |
|---|---|---|---|
| Dylan StromeWSH | 293 | 66% | 61% |
| Jordan StaalCAR | 156 | 69% | 61% |
| Vincent TrocheckNYR | 157 | 66% | 60% |
| Nathan MacKinnonCOL | 152 | 66% | 60% |
| Gabriel LandeskogCOL | 101 | 68% | 60% |
| Nico HischierNJD | 255 | 62% | 59% |
| Sidney CrosbyPIT | 246 | 62% | 59% |
| Bo HorvatNYI | 248 | 62% | 59% |
| Anthony CirelliTBL | 77 | 69% | 59% |
| J.T. MillerNYR | 123 | 65% | 59% |
Best on the penalty-kill draw
| Center | Draws | Won | True |
|---|---|---|---|
| Claude GirouxOTT | 165 | 70% | 55% |
| Jean-Gabriel PageauNYI | 221 | 59% | 52% |
| Scott LaughtonTOR | 153 | 62% | 52% |
| Aatu RätyVAN | 119 | 60% | 51% |
| Michael McCarronNSH | 208 | 53% | 50% |
| Jake EvansMTL | 220 | 55% | 50% |
| Mark KastelicBOS | 152 | 57% | 50% |
| Anze KopitarLAK | 137 | 54% | 49% |
| Robert ThomasSTL | 104 | 54% | 49% |
| Curtis LazarEDM | 71 | 56% | 49% |
Special teams in the playoffs
Special teams get harder in the playoffs - tougher opponents, tighter officiating. Across 181 playoff team-seasons the average power play drops 21.8% → 19.8% (-2.0 pts) and the penalty kill 80.8% → 78.5% (-2.3pts). The teams below compare each club's own regular season to its own playoff run - a single run is a small sample, so these are descriptive, not predictive.
Biggest playoff power-play collapses
| Team-Season | RS PP% | PO PP% | Δ |
|---|---|---|---|
| NJD2024-25 | 28% | 0%15opp | -28 |
| LAK2023-24 | 23% | 0%12opp | -23 |
 FLA2021-22 | 24% | 3%31opp | -21 |
| TOR2023-24 | 24% | 5%21opp | -19 |
| OTT2025-26 | 24% | 5%21opp | -19 |
| NYR2019-20 | 23% | 7%14opp | -16 |
| PIT2018-19 | 25% | 9%11opp | -15 |
 DET2015-16 | 19% | 4%25opp | -15 |
| PHI2015-16 | 19% | 4%24opp | -15 |
| TBL2024-25 | 26% | 11%18opp | -15 |
| STL2016-17 | 21% | 7%30opp | -15 |
 DAL2021-22 | 22% | 8%24opp | -14 |
Biggest playoff power-play surges
| Team-Season | RS PP% | PO PP% | Δ |
|---|---|---|---|
| WPG2022-23 | 19% | 42%12opp | +22 |
| LAK2024-25 | 18% | 40%20opp | +22 |
| COL2020-21 | 23% | 41%29opp | +19 |
| BOS2022-23 | 22% | 41%27opp | +19 |
 CGY2016-17 | 20% | 38%16opp | +17 |
| BOS2020-21 | 22% | 36%33opp | +14 |
| EDM2022-23 | 32% | 46%39opp | +14 |
| MTL2024-25 | 20% | 33%15opp | +13 |
| ANA2025-26 | 19% | 32%38opp | +13 |
| BOS2017-18 | 24% | 36%33opp | +13 |
| COL2023-24 | 24% | 37%30opp | +12 |
| DAL2025-26 | 29% | 40%25opp | +11 |
How to read this.A special-teams unit's result splits into what repeats and what doesn't. Winning the draw is a repeatable skill; finishing above expected (and on the kill, hot goaltending) historically regresses. Real = built on the sustainable parts; Mirage = results ahead of the process; Robbed= good process undone by cold finishing or goaltending, due to climb. Faceoff true win% is the observed rate regressed toward the league special-teams mean, so a hot 50-draw sample doesn't top the board. Team forensics cover 2015-16 to present (faceoff-zone data); regular season unless marked playoffs.