NHL power play opportunities for Friday, Oct 2

How the forecast works

1. Expected PP opportunities = league rate × the team's rate of drawing penalties × the opponent's rate of taking them, with a small home-ice edge. Each rate blends this season's games with last season's rate (pulled halfway or more back to the league average), weighted like 80 games, so a hot October does not swing the forecast.
2. Expected PP time = expected opportunities × the team's seconds per power play (about 99 seconds league-wide; shrunk toward the league).
3. Each player's share of team PP time comes from the last 10 games with any power play (recent games count more), from NHL shift data, pulled 15% toward a typical share because power-play hot streaks cool off. PP1 and PP2 are the five-man units with the most power-play time together in the last five games.
4. Player expected PP minutes = share × team expected PP time. Only players who dressed in one of the team's last three games are listed.
5. Referee crews are shown when the NHL posts them, usually close to puck drop. They are not in the forecast unless the crew effect beats the no-crew model on seasons it was not fitted on (see the referee page).

Power-play chances are mostly luck game to game: the forecasts spread only about ±0.35 opportunities around the league average, and a single game swings by ±1.5. We tested the model the honest way: for each season from 2020-21 to 2025-26 we refit it on earlier seasons only and forecast every game. Against a league-average-plus-home-ice baseline it was slightly better per game in 5 of those 6 seasons and no better in 2025-26, where it matched the baseline. Over 10-game stretches it tracked each team's actual opportunities better in 6 of 6 seasons (correlation 0.36 → 0.45 on average). The range beside each forecast held about 9 in 10 games in those seasons. Player PP minutes were tested on 2024-25, a season the player constants were not tuned on: they were off by 62 seconds per game on average (a season-average guess: 65), and players projected for 3–4 minutes averaged 3:12 (the rare 4+ minute projections ran about half a minute high). Most of the miss is how many power plays the team actually got, which no one knows before the game. Use it for weekly PPP planning and close start/sit calls, not as a lock.

Data: NHL team power-play and penalty reports since 2017-18, NHL shift charts processed by our on-ice engine, and NHL game-day officials.

FAQ

Related

• NHL lines todayEvery team's lines and PP1 / PP2 units from shift data
• NHL refereesPenalty and power-play tendencies of every referee
• Stack FinderBest line and PP1 stacks tonight
• Schedule EdgeRest, travel and back-to-back edges by game
• Who BenefitsWho takes an injured player's PP1 spot