One for Bakes: Icing Percentage, a Good Indicator of Team Performance?

A statistical look into Icing the puck

Jamie Baker asked a question during yesterdays broadcast about icing. His impression was that the Sharks have been forcing teams to ice the puck from their defensive zone far more often this home stretch. Unable to change, this has created favorable match-ups which the Sharks have exploited.

I believe Jamie's theory is completely sound, and it would be great if the NHL tracked such data. Unfortunately, the NHL PBP files do not track where the puck was prior to to the icing, thus we don't know if the icing was a result of an attempted zone entry (not such a bad thing), or from being hemmed in your own zone (definitely a bad thing).

Regardless, I ran the numbers for Bakes

Icing Statistics

Games Icing against Opponents Icing Against sharks Icing%
Games 1-30 126 126 0.500
Games 31-37 31 34 0.477

Icing against opponents indicates opponents icing the puck, with the draw coming in the Sharks offensive zone. Icing Against Sharks indicates icing against the sharks, the puck coming into the Sharks defensive zone.

The Sharks have been icing the puck more frequently over this home stretch, (actually the data set includes the game 31 away against ANA, this is when the Sharks started using Pavs as a 3rd line center). My thoughts are this; the sharks have been controlling play to a better degree, probably have the puck more often, and as a result have more frequent icing calls due to attempted zone entries. The icing while hemmed in is probably less frequent, distorting he significance of the numbers above.

Essentially, I don't think this data is particularly useful, at least as presented. What Jamie needs is someone to go back and watch every icing play this season, tracking where the puck was prior to the icing. Then we could separate these different types of icing.