Aside from the standard team stats like wins, losses, and goals, and the standard players stats like goals, assists, and points, faceoffs might be one of the most talked about stats in the game of hockey. You can’t go five minutes in a hockey game without the broadcasters talking about how the teams and players have done in faceoffs, and they always talk about how it helps a team get possession of the puck, and how it helps the team win.

But, is that actually true? Are faceoffs actually important in a hockey game, or are they just statistical noise disguised as a notable part of a hockey game? I decided to do a bit of a deep dive into how much they impact the game.

So, you may be wondering, how can you just simply find out the value of something in sports. Its actually a lot easier than you may think. You remember in math class when they made you figure out the equation of a line? That’s going to play a big role in figuring this out.

Now, this isn’t something I discovered on my own (I’m not that smart). I’m using a process I first discovered with current Carolina Hurricane’s assistant GM Eric Tulsky’s work from over 10 years ago. Basically, what figuring out the equation of the line tells us is how many points a team with a 0 faceoff differential should get, theoretically.

So, first I took each team’s faceoff differential and points from the 2018-19 to 2020-21 NHL seasons (it’s still a bit too early in this year’s season to include that data) and threw the data on a scatter plot in Excel. Once I got that, I just do a simple right click to add a trendline, and I have all the info I need.

So, our data’s trendline has an equation of 0.0227x + 91.159. What that tells us is a team with a 0 faceoff differential will finish with 91 or 92 points, and that every extra faceoff win is worth about 0.02 points. Since 0.02 equals to about 1/50, that means that 50 more faceoff wins than losses is worth a point, and 100 more wins than losses is worth a win.

Now, that doesn’t mean this is always the case. In fact, our R2 equation tells us how much our x-axis’ data correlates with our y-axis’ data. With an R2 of 0.0542, that tells us our data does not correlate at all, which if you know anything about data, you probably didn’t need the equation to tell you that with one look at the chart.

So, as far as faceoffs’ impact on winning a game, what we’ve learned from this is that faceoffs have almost no correlation with winning, and even if it did, a team would have to be absolutely dominant in the faceoff dot for this to actually have an impact.

Another point often made with faceoffs is how it helps a team’s possession numbers. In theory, it should, because it allows you to start the play with the puck. So, let’s take this process, and compare a team’s faceoff differential to their shot attempt differential.

So, there isn’t a lot to pull from the actual equation itself, but the R2 comes into play again here. The R2 of 0.083 tells us that once again, there is almost no correlation between faceoffs and a team’s possession ability. Good possession teams will find ways to get the puck on a play whether they win or lose the faceoff.

This isn’t to say that no faceoff is ever important. There are definitely faceoffs that have massive importance in a game, particularly late in a game when a team is looking to win to get the game tying goal, or to win to clear the puck out and secure the win. But, that’s part of the randomness that comes with faceoffs. For every really important faceoff that creates a goal or wins a game, there are 50-100 that have almost no impact aside from who starts the shift with the puck, and as we’ve learned, good teams will find ways to get the puck regardless of their faceoff ability.