Analytics to evaluate defensive play

[WestCoastExpress94]

Aside from shot supressions, what advanced stats in particular are evaluated to determine if a player is good defensively?

Click to read more...

 

[solidprospect]

Well not a big one but dZS%. Defensive zone starts. If they're in the %65-75 range they're used defensively more often that other players. It would be a good starting stat to add onto others.

what about on ice save percentage?

 

[Shaffer]

I like comparing Corsi Against and Blocked Shots.

 

[NickH8]

It's hard because corsi and shot suppression and all that doesn't account for the quality of the shots. Is there a way to measure the mean distance and angle from the net of each shot/shot attempt to show how well a defenseman prevents those opportunities? Heat maps maybe?

 

[Germant]

I'd think SH/TOI would be key. Though not "analytics" per se, it's pretty indicative of defensive acumen most times.

 

[hatterson]

Well not a big one but dZS%. Defensive zone starts. If they're in the %65-75 range they're used defensively more often that other players. It would be a good starting stat to add onto others.

what about on ice save percentage?

I'd think SH/TOI would be key. Though not "analytics" per se, it's pretty indicative of defensive acumen most times.

One thing to note about dZS% and SH/TOI is that they're not measures of defensive performance/acumen, they're measures of player usage.

Now, generally speaking, coaches aren't idiots and defensive usage is heavily correlated with defensive quality, but there's several things that can confuse those measurements.

If the coach doesn't really care about matchups and purely rotates lines on a fixed schedule, you won't have anyone with a high or low dZS%. A "good" PKer will receive much more icetime on a team of scrubs than he will on a team of elite PKers.

If you're looking for an analytical measure of his defensive performance, rather than just usage, you'd need to look at things like shot/chance/goal suppression.

It also brings up the question of what type of defensive performance you're looking at. Theoretically the best defensive player is one whose team possesses the puck 100% of the time. It's impossible for the other team to score if they never touch the puck, but many may not consider a player like that when they're thinking of defensive skill.

 

[morehockeystats]

One thing to note about dZS% and SH/TOI is that they're not measures of defensive performance/acumen, they're measures of player usage.

Now, generally speaking, coaches aren't idiots and defensive usage is heavily correlated with defensive quality, but there's several things that can confuse those measurements.

If the coach doesn't really care about matchups and purely rotates lines on a fixed schedule, you won't have anyone with a high or low dZS%. A "good" PKer will receive much more icetime on a team of scrubs than he will on a team of elite PKers.

You might want to adjust it in some way to PIM. You can't kill penalties from the sin bin.