The Benefits of the 4-Forward Powerplay

On Friday night I wondered aloud why coaches don’t play a 3F/1D setup during overtime, given the unpredictable nature of the shootout and the single point they’d already guaranteed themselves. And so naturally, as I often do, I started to look into an entirely different (although in this instance at least somewhat related) problem: why do coaches almost always play with a 3 forward/2 defenseman approach when on the powerplay? To me, this seemed like it might be a case of the old wisdom taking precedence over analytical thought. After all, in baseball bunting with a runner on first still remains a popular play, while in football coaches are still hesitant to going for it on 4^th down, even though both of these strategies have been shown analytically to be poor decisions in most cases. Perhaps the decision to ice a standard powerplay lineup is another example of coaches playing too conservative when the numbers suggest a non-conventional approach is superior.

And with that, I started to dive into the data to look whether playing 4 forwards on the powerplay really did provide an advantage over the standard 3 forward approach. To do this, I started out by looking at the shooting percentage of teams playing under both approaches from the 2010-2011 season through to the 2012-2013 season. I broke the stats down into 5v3 and 5v4 situations to eliminate any potential contamination from the 2-man advantage situations. While this analysis doesn’t look at whether or not the 4 forward approach generates more or less offensive chances, given the number of shots we’re looking at it should give us a decent sense of whether the 4 forward approach is generating higher quality chances and/or chances taken by higher percentage shooters.

The table above outlines the overall shooting percentage and Fenwick shooting percentage for both 3 and 4 forward approaches in 5v3 and 5v4 situations. The usual caveats about the accuracy of the NHL’s play-by-play data apply here, but the results do point towards the 4 forward approach providing a statistically-significant advantage, at least in the 5v4 case (unfortunately, given the limited sample size we can’t say anything conclusive about the 5v3 case).

But what about the defensive side of things? The obvious counter-argument to the 4 forwards approach is that it decreases the defensive ability of the unit and puts the team at risk of giving up a short-handed goal. And if we look at the data presented in the table below, we see that that does seem to be true: short-handed teams shoot better when they’re playing 4 forwards rather than 3 (note that once again, the 5v3 numbers aren’t significant).

The question that we have to ask then is whether the offensive benefit provided by using the 4 forward approach outweighs the defensive cost. Ideally, we’d look at the Fenwick For/Against rates for both 4 forward and 3 forward situations but I’m too lazy to write a query to pull that data out but that TOI and Shot/Fenwick data isn’t straightforward to put together. We can, however, put together an estimate of how many additional goals a 4 forward approach is worth to an average team by looking at the average Shots For/Against of a team on the power-play and comparing whether the increase in expected shooting percentage outweighs the decrease in expected save percentage.

I put together the table above by taking the average 5v4 Shots For and Against from 2007-2012 from stats.hockeyanalysis.com and comparing the expected shooting/save percentages in 4 forward and 3 forward setups based on the data above. As you can see, the benefit of playing with 4 forwards clearly exceeds the risk: teams that played an all 4F approach would be expected to score an additional 3.3 PP goals over the course of an 82 game season, which works out to roughly half a win. This, of course, also ignores any effect on the shot rate that the 4F approach would have. It’s entirely possible that 4 forwards produce more shots per powerplay minute than the standard approach further increasing the value of this approach. In either case it does seem like one of those cases where the aggressive approach should payoff: regardless of the defensive risk that teams are taken by putting a forward on the point, the offensive benefit appears to outweigh it, and in the long run, teams that take this approach should see a non-negligible bump in their powerplay percentage.