Just yesterday I published a set of charts showing scoring chances relative over time for the forwards, and now I have a similar set for defenders. These charts show Fenwick For % Relative to a player’s team, but show the change over time per salary band.
Why Fenwick? Well, by removing blocked shots from the analysis, we can get a better idea of how good the defender is at shot suppression. If they’re blocking shots (as shown by Corsi) then they’re letting the opposing team get into a shooting lane with the puck, and that’s less than ideal.
The most important thing to know when reading these charts is that a higher Fenwick For % Relative is better, just like a higher Scoring Chance For % Relative is better for forwards. And if your team isn’t positive, you’ll want them to at least be higher than the league average.
Here are the NHL Averages:
Other than the <$0.6m and $4-5m band you can see a pretty clear trend that more recent signings have had better FF% Rel than in years past. I would suspect the reason that the $4-5m band took a hit with several UFAs moving into higher price ranges, however it could be the opposite – UFAs from the $2-4m bands moving up a level undeservedly due to cap inflation.
Again, I didn’t include Zone Starts in these graphs, partially because of this article challenging their usefulness on an aggregate level. (The larger reason is because it looked gross, and vanity couldn’t allow me to put out ugly charts.)
While there may be a trend on the NHL level, I’m seeing a lot of variance on a team level. This could be due to the small amount of data on a team level – even though the minimum was only 500+ minutes per season (all situations), most years you won’t get more than 7 defenders per team, which means around 14 salary data points per timeframe.
The other theory is that it’s due in large part to the very reactionary nature of GM & coach hiring and firing – a lot of teams have coach or GM (or both) turnover every few years, which makes it hard to establish a pattern of behavior over time.
As usual, I caution you against using these as a bar of what a defender “should” get paid on a contract, though it can definitely point you in the right direction when looking for comparables. Also, relative metrics don’t always translate from team to team – someone great on the Leafs, who have an overall very low Fenwick For %, could be completely negative on the Blackhawks, who historically have a high Fenwick For %.
Still, some of unaswered questions here:
1) WHEN WILL YOU LOOK AT UFA/RFA CONTRACTS?
2) How is all of this affected by quality of teammates? What about Quality of Competition?
Great questions guys. I’ll get back to you on that.
Anyway, here are all 30 NHL teams, in alphabetical order, presented without comment.
Arizona (includes time spent as Phoenix)
New York Islanders
New York Rangers
Winnipeg (includes time spent as Atlanta)