Passing and Zone Entries are so last year.

When Corey Sznajder decided to track microstats for the upcoming season and began incorporating my passing concepts into his work on last season’s playoffs, I wondered if we really needed to track this season. Instead, Corey and I chatted a bit and decided the best use of everyone’s time would be if myself and the other passing project volunteers continued to work on last season*, with the hope that we can build a solid sample by the time Corey finishes the 2016 – 2017 season. Having two (nearly) full seasons of data would be excellent to have.

I’m going to be tracking microstats for the upcoming NHL season & I will make the data publicly available. https://t.co/tMrM7BaJTB

— Corey Sznajder (@ShutdownLine) August 20, 2016

However, this also gave another idea to explore something we really haven’t done a lot of: forechecking.

This began as an email chain to fellow Hockey-Graphs members Charlie O’Connor and Prashanth Iyer. Twenty emails, several templates, and a couple of test games later, we decided on the parameters for how we would go about quantifying forechecking and its impact on the game. Kevin Winstanley, who tracked passing data on half of the Devils games season last year, and Alex Novet, another HG writer, joined us just as the season got started. Anyone interested in tracking should reach out to Prashanth.

What We Will Be Tracking

First off, we had to define exactly what a forecheck is, since hockey is such a chaotic game at times. We settled on the below definition:

A forecheck is defined as occurring when the team without the puck has two players in the offensive zone, one of which must be below the faceoff dots.

This gives us a simple framework with which we can begin quantifying forechecking. As long as the team without the puck meets this requirement, we can record the formation they are in, along with how aggressive (defined as how many forecheckers are below the faceoff dots) they are.

The next thing we had to address were all those situations in which a team doesn’t set up in an easily-identifiable formation, but the team is still pressuring the opposition. We decided on the following categories, provided each situation meets our definition of a forecheck:

• Formation: (1-2-2, 2-1-2, 2-3, etc)
• Rebound: When a shot is taken (any shot) that results in a 50/50 puck battle
• Faceoff: For OZ draws
• Turnover: When the offensive team loses the puck and immediately applies pressure to get it back

This allows us to isolate when the team is in a forechecking formation, which formation that is, how aggressive it is, or how aggressive they are in a different phase (rebound, faceoff, turnover). We’ll also record which forwards were on the ice that were in on the forecheck.

The Flip Side

Of course, breakouts are very much the other half of forechecking. Which breakouts are most effective against specific forechecks? How well does an aggressive forecheck break up zone exits? Force turnovers? Which team’s defensemen are most active in pinching to keep play alive? How risky is that? By also recording information on each breakout, we can start to answer these questions.

To start, we agreed we would define the breakout plays as described here, but these are your common plays: “Up”, “Over”, “Wheel”, “Rim”, and “Reverse.” We also will record when a team sets up in a controlled breakout, when the goalie plays the puck, and also will record situations when the forecheck prevented a breakout from starting.

We’re also recording the lane, with respect to the team breaking out, of the destination of the breakout play, how much support (teammates below the faceoff dots) they had, and the result (controlled/uncontrolled exit, turnover, etc).

Period, time, and score state will be recorded so we can see how tactics change based on time and game state.

What We Hope to Do

Analytics in general are really about distilling various pieces of information down into an answer to a single question: “What Will Help My Team Win?” Exploiting market inefficiencies on the trade market, in free agency, or even in the draft are all things the analytics community does reasonably well. However, apart from work on zone entries and passing plays, there’s still a lot of work to be done on quantifying and recommending specific tactics and strategy. The next great market inefficiency to exploit is in how teams play the game.

Ideally, defense should prioritize getting the puck back: smart, aggressive pressure should be a better path to take than conservative approaches. This stems from coaching philosophy as well. It shouldn’t be: “My opponent has the puck and what is the best way to defend as they come up the ice?” but should be, “My opponent has the puck and what is the best way to pressure them to get the puck back.” I believe more aggressive tactics should result in better shot-suppression numbers, something I’ve looked at recently and something that Charlie looked at at the end of last season.

Lastly, each phase of the game should flow into the next for a well-structured team. With enough data to analyze forechecking and breakouts, evidence can be used to tactically optimize a team’s strategy. Teams should really work to optimize their tactics in three phases: what they do when they have the puck, what they do when don’t have the puck, and how they transition between the two. There would likely be slight changes based on the zone phase occurs in as well, but I believe this is how coaching can be improved as well.

We will keep you all  updated on our progress.

*Analysis on the passing data will continue as we accumulate more and more games. This does not mean it will cease to be looked at. If you’re still tracking games from last season, I’m still working on different ways to analyze and use the data.*