|NLL face offs: "Yeah, it's kinda like this."|
The key to winning a game is scoring goals when you have possession of the ball and stopping your opponents from scoring when they are in control. That's pretty much common sense to every sports fan and athlete out there so let's go a little deeper. Not only do you want to score when you have possession, but you want to maximize how many possessions your team has.
How do you increase your possessions in lacrosse? One of the ways is through winning face offs. If you win the face off then you have control of the ball but a loss in the circle could lead to a goal against.
Mike came to me the other day wanting to know just how much effect face off percentage had on a team's winning percentage or goal scoring totals. If you know me well (or even just a very little), you know that I can't say no to someone when they give me an opportunity to combine my love of mathematics and sports. I compiled the team stats that NLL.com had for every team in the last three years and calculated the correlation between different stats.
Before I get into the results, let me explain to you what correlation is. Correlation shows the relation between a two variables (a dependent and an independent). Correlation is represented by the correlation coefficient (r) and can show us how much effect face offs have on anything. r is measured on a scale from -1 (negative relation) to +1 (positive relation). If data has an r value of -1, this means that as the independent variable increases then the dependent decreases and vice versa. An example of this would be as someone exercises more, their body fat decreases. An r value of +1 means that both variables increase or decrease at the same time. For example, as someone eats more food their weight goes up and as they diet their weight goes down. If r is around 0, there is no relation between the two variables.
Back to lacrosse. Here are the correlations between FO% / FO% ratio (FO% divided by opponent's FO%) and multiple stats.
For a correlation to be considered "strong", the absolute value of r should be at 0.6 or higher. As you can see, all of the r values for both FO% and FO% ratio are down under 0.2 (except for shots ratio which measures how much a team is out shooting their opposition). This leads me to the conclusion that winning face offs aren't nearly as important as I once thought.
So, why aren't they so important if they give a team extra possessions? An easy answer would be to cite turnovers. In the last three years, there have been approximately the same number of turnovers per team per game as there have been face offs per game. This means that the ball changes hands faster than teams can score goals. In addition to turnovers, how often do we see teams win the face off and then sit back until they get the right personnel onto the floor? Quite a bit. If teams were to pressure more off of the face off, I think that winning the FO might lead to more goals.
I'm not saying that winning face offs aren't important. They just aren't as important as I originally assumed. If I had access to overtime only stats, I would be willing to bet money on the team that wins the face off scores the goal the most often. In other words, face off wins have more leverage in overtime than they do in the first 60 minutes.