New tracking project: Tracking hits that lead to possession

Whether you believe in them or not, advanced stats are becoming a staple in the hockey community. These stats are being used in conjunction with scouting to evaluate players, teams, and every facet in hockey. However unlike sports like baseball, hockey is still very “raw” in regards to the information provided in the stats. That gap though is being closed by amazing hockey minds such as Ryan Stimson and his passing project, Micah Blake McCurdy, Jen Lute Costella and many others. Before I begin, I do want to say thank you to them because there work has not only inspired me but also many others, their information is truly invaluable.

As a fan, I have noticed the resistance against advanced stats for a handful of reasons. One reason I’ve seen is because advanced stats may go against notions that some have already created through the eye test. Theoretically the resistance does make sense, people naturally cling on to what they see and remember.

Another reason that I have seen and it is recently being seen in a larger quantity is the concept of grittiness and intangibles. Kevin Hayes is seemingly a perfect conduit for the little argument between fans of opposing views. Some call Kevin Hayes out as lazy because of what they see (which in my opinion is practically blasphemous, large skaters like him who don’t need to peddle are beautifully effortless skaters. That doesn’t mean he is lazy).

To back up their argument some then say Kevin Hayes does not go to the “dirty areas”. To those who believe that, I respect their opinion but my question becomes; What is your proof? You don’t have a library of every play Hayes decided maybe not to run someone through the boards. You only have your word, which simply isn’t good enough in evaluating players since no facts are backing it up.

Don’t get me wrong I do think physical play does have a role in hockey. Growing up playing the sport, it was easy to realize how a good clean hit on the forecheck and grinding the defensemen along the boards can benefit the team. But looking back at it now, it wasn’t the hit that ended up qualifying it as successful. It was getting the puck back on our sticks. In the end hitting the player with the puck in the defensive zone or offensive zone was strictly to get the puck back and go on our own offense, yet we don’t quantify that.

The ability to get possession off a meaningful hit should have more value than simply firing yourself into the boards after the puck is already off the stick of a player. That is the purpose of this project: To track these types of plays, the hits that end up with your team gaining possession.

Myself and @Mactrucked –who is my first tracker– have been able to track three Ranger games with a few more on the way. I’ve bee able to provide a visualization for some of the things that we have seen in the games against New Jersey (2/23) Dallas (2/27), and Columbus (2/29). I refuse to make any decisions on player evaluations just yet, there is simply not even close to enough data. This is an introductory post to hopefully show what information may soon be available.

The bar graph above  shows the Rangers who have attempted hits and the amount of hits that actually led to puck possession over those three games. These are all during even strength.

This graph is similar to the first graph, however instead of just the raw numbers this shows the percentage of the hits that led to puck possession. While once again this is simply too small of a sample size to make any executive decisions, hopefully over time once more games have been tracked we will be able to see which players are most efficient at creating turnovers with hits. Kevin Hayes for example, although he only had five hit attempts over the three games, but four of them led to puck possession.

Going head to head between teams is fun too, what is even more fun is bringing up the terrible Rangers-Devils game. After tracking the game the Rangers who were vastly out possessed looked like they outhit the devils. But when looking at the percentages even though the NYR outhit NJ New Jersey’s hits were more efficient in stealing possession.

Over the course of tracking games I hope to have a few things completed for further study:

1. A database that shows the “meaningful” hits that lead to puck possession.
2. See the impact of score effects on the frequency of both attempts and hits that lead to puck possession.
3. Allow the stat to be seen as a rate/60 for the specific players.
4. Correlate the stat to goals and already used stats such as Corsi to see the connection.

If you are interested in helping to try to track games that would be much appreciated, school is rough around midterms. Email me at [email protected].