Quick Thoughts: 4-Forward Power Plays and Protecting A Lead

One important point about the 4 forward power play is that it increases the overall rate of scoring, both for and against.

Usually this is fine, because on the whole your goal differential is significantly better with 4F (~6.1 GD60 w/ 4F vs 4.9GD60 w/ 3F since 09-10)... https://t.co/uywIvBtvj3

— Matt Cane (@Cane_Matt) December 4, 2017

But when you start to get later in the game, the math changes a bit - the value of an extra goal isn't as great as the cost of giving up a goal when you're leading late.

— Matt Cane (@Cane_Matt) December 4, 2017

Because of this there's a cutoff point if you're leading when it makes more sense to use 3 forwards than 4 forwards on the power play.

— Matt Cane (@Cane_Matt) December 4, 2017

We can figure out what this point is by looking at the probability of a power play/shorthanded goal with 4F and 3F respectively, and multiplying it by the change in expected points when a team scores/allows a goal at a given point in the game (the leverage)

— Matt Cane (@Cane_Matt) December 4, 2017

This chart shows at which point teams should switch from a 4F to 3F setup based on the score (+1 or +2) and venue (home/away). i.e. if a home team is up 1 and gets a power play with < 8.2 minutes left, they should use 3 forwards rather than 4 to minimize the risk of a SH goal. pic.twitter.com/m3gqafnxPa

— Matt Cane (@Cane_Matt) December 4, 2017

(Note that there may be some weird things going on here since I used the empirical data rather than Micah's smoothed leverage data)

— Matt Cane (@Cane_Matt) December 4, 2017

Tl;dr:
-If you're up by 1, you shouldn't use 4 forwards when there's less than 6-8 minutes left in the 3rd
-If you're up by 2, you shouldn't use 4 forwards when there's less than 9-11 minutes left in the 3rd.

— Matt Cane (@Cane_Matt) December 4, 2017