clock menu more-arrow no yes mobile

Filed under:

Using stat tracking to explain the Canadiens' offense in Game One

Even-strength scoring-chances during game 1 of the opening-round series between the Montreal Canadiens and the Ottawa Senators were 11 apiece. Montreal produced 2 scoring-chances, while allowing 5 chances against in the first period. Chances were 7-2 for Montreal in the second, and 4-2 for Ottawa in the third.

Eric Bolte-USA TODAY Sports

Those people who know of me may be familiar with my previous work. I developed a system where every puck-possession event, and every change of possession is tracked. Events tracked include passes, dekes, shots, stick-checks, blocked passes, blocked shots, blueline holds, dump-ins, dump-outs, redline carries, controlled defensive-zone exits, controlled offensive-zone entries, and others. Each event is tracked as either successful or failed; allowing us to compare players equally based on their individual impact on puck-possession during a game.

My recent hiring at Sportlogiq has provided the opportunity for my previous work to move forward to the point where we are not only tracking every puck-possession event and change of possession, but also allowing for time-stamping, video-tagging, and event coordinates. It is the next step in analytics, providing us with data on the who? What? Where? When? and How? of every event that occurred during a game. Not only can this automated process track the event, we can see the result of every event by breaking down the plays in that series. Pulling the curtain back on how shot attempts are being created; and more importantly, how scoring-chances are being created. Expanding both fans' and teams' access to data above and beyond shot attempts.

Even-strength scoring-chances during game 1 of the opening-round series between the Montreal Canadiens and the Ottawa Senators were 11 apiece. Montreal produced 2 scoring-chances, while allowing 5 chances against in the first period. Chances were 7-2 for Montreal in the second, and 4-2 for Ottawa in the third.

The first period saw Montreal attempt only 6 controlled entries compared to 13 dump or chip-ins. They were 8 for 15 on their forechecking attempts, with 40 offensive-zone loose-puck recoveries. Montreal was 0 for 2 on their attempts to pass the puck to a teammate positioned in the slot, and 5 for 8 on their attempts to hold the puck in at the blueline.

*the XY viewer allows for video playback when any event bubble is clicked

Five of Montreal's chances in the game were produced off of controlled entries into the offensive-zone; with 4 of those occurring in the second period. The Habs attempted 11 controlled entries in that period, while dumping the puck in only 7 times. They were 8 for 10 on their attempts to remove possession from Ottawa in the offensive-zone by forechecking, and recovered 37 offensive-zone loose-pucks. They were also 3 for 7 on attempted passes to a teammate positioned in the slot, and 2 for 2 on their attempts to hold the puck in at the offensive blueline.

Holding onto a 1-goal lead in the third, Montreal attempted only 4 controlled entries, while dumping the puck in 14 times. They were however reasonably aggressive on the forecheck, going 10 for 20 on their attempts to remove possession from the Senators in the offensive-zone. They recovered only 24 loose-pucks, were successful with 2 of 5 attempted passes to the slot, and 2 of 2 blueline holds.

Understandably, Montreal increased their dump-outs late in the game, going from 14 successful dump-outs on 16 attempts in the first, to 11 for 16 attempts in the second, and 23 for 31 attempts in the third.

How does Sportlogiq provide this data?

Sportlogiq has developed a computer vision powered hockey analytics platform that provides detailed game analysis using nothing more than standard game footage. Similar player tracking systems require multiple cameras, or chips on the players themselves, but the sportlogiq system can generate data using videos downloaded from YouTube, or recorded at a game using a single camera.

Footage is analyzed by a patented computer vision engine (CV Engine).