Introducing Possession-Independent Hitting and Shot-Blocking

You hear the word "grit" thrown around in the hockey world a lot, and while the term is subjective, there are measurable aspects of it that we can use to weigh the credibility of the claim that you need it to be successful. Two of the main factors that contribute to a player being seen as "gritty" is his ability to throw hits and to block shots. The NHL tracks these two statistics so that fans, media, and management can find out which players and teams throw the most hits and block the most shots.

There are two major problems with the tracking of these metrics, however.

1. There is significant scorer bias in the NHL, in the sense that what certain scorers consider a shot block or a hit varies by building, and therefore a team that plays in a building with a liberal shot-block counter is bound to rack up more blocked shots than another team without necessarily being more gritty.

2. A team can only record a hit or a blocked shot when it does not have the puck; that's just common sense. The problem is that some teams possess the puck far more than others, so the opportunity to record hits and block shots is unbalanced.

The solution to the first problem is to restrict our dataset to only road data. That way, teams face roughly the same distribution of scorers, and most of the bias is washed out. The way to solve the second problem is to make these statistics independent how much a team has the puck, or how many shots they attempt. Time of possession hasn't been tracked in the NHL in a decade, but possession metrics - namely Corsi - have been shown to be a good proxy in that regard. Corsi also allows us to come up with a shot blocking efficiency statistic. The two new real time stats are as follows:

Possession-Independent Hitting (PIH) - The number of hits a team records divided by the opposition's share of shot attempts in all situations.

Possession-Independent Shot Blocking (PISB) - The percentage of shot attempts against a team blocks in all situations.

Ideally, 5v5 would be separated from odd-man situations, but the hitting and shot-blocking data the NHL provides isn't that extensive so this will have to do. First of all, I'll list the team rankings of the two statistics from 2011-2012 and the shortened 2012-2013 regular seasons (Note: columns are sortable).

2011-2012:

Some of the lessons we learn here aren't particularly surprising. The New York Rangers under John Tortorella blocked a lot of shots. The Los Angeles Kings were a very physical team. Maybe surprising to some - although not to those of us who have looked into size and physicality in the past - is that Chicago is last in the league in PIH by a sizeable margin. They beat teams with skill, not physicality.

2012-2013:

Familiar faces appear at the tops and bottoms of these rankings. In fact, there is a significant correlation (R squared = 0.67) between Possession-Independent Hitting in 2011-2012 and 2012-2013. The same can be said, to a lesser extent, for PISB (R squared = 0.37).

So what do all of these numbers mean in the grand context of the NHL? There isn't a massive correlation between hitting efficiency/shot blocking efficiency and points or wins, but unlike with conventional real-time stats, the correlation is positive. That means that teams that are better at hitting and blocking shots - rather than those who hit or block more shots overall - tend to win slightly more games. This applies both in the regular season and playoffs.

The important conclusion to take from these findings, however, shouldn't be that hitting and shot blocking are ends unto themselves, but that they're a way to recover the puck, to prevent scoring chances, and to tilt the ice in your team's favor. Teams that manage to hit a lot when they don't have the puck likely recover the puck effectively using that technique. Same goes for shot blocking. There are, however, other ways to recover the puck, such as good stick-work, solid positioning, and the ability to read the game at a high level and make good decisions.