I ranked Broda towards the top of my rankings on a preliminary basis since I don't know if I'll be able to post on Sunday. It's probably the ranking I'm the least certain of, but he's not crashing down anytime soon.
There's something to be said for getting consistently done in the playoffs, PLUS, there's a determined amount of credit that has to be given, and I always considered Ted Kennedy to be somewhat overrated in the grand scheme of things. And there's the longevity too. And I'm also weary of giving undue weight to coaching, since too many stupid shit is sometimes attributed to this factor, like Jean Perron having somehow a POSITIVE EFFECT on a netminder.
Luongo is firmly mid pack.
I'd suggest that the voters giving undue credit to coaching are probably the ones not really factoring in coaching at all. That said, I'm absolutely happy to have a conversation or take feedback about any of my data.
With respect to Jean Perron, let's look at the with-or-without you stats of the goalie with the second-most career NHL games played for that coach, shall we?
Brian Hayward, NHL Career:
| GP | SA | SV | SV% |
| With Perron | 76 | 1989 | 1782 | 0.896 |
| Without Perron | 281 | 7854 | 6871 | 0.875 |
It really doesn't strike me as far-fetched to suggest there was a save percentage advantage derived from that environment.
To be clear, nobody should take any save percentage numbers entirely at face value, including these coaching ratings. When you try to isolate pretty much any impact in hockey using team data, there are absolutely other factors that are going to get in the way. I am trying to estimate the save percentage impact of coaches by effectively looking at their career save percentages while behind the bench, but like any save percentage analysis it has the potential to be affected by any of the factors that impact save percentage generally. In addition, regular save percentage has the goalie themselves in net, while coaching impacts are more indirect.
Here are the three variables that need to be kept in mind at all times when evaluating anything from a coaching perspective, all of which I have referred to in my previous analyses using the coaching numbers:
1. Team Quality
This is I believe where your issue with Jean Perron falls. I'm assuming you think he was an idiot coach who had Chris Chelios and Larry Robinson and Guy Carbonneau and Bob Gainey and the rest of them carrying him to glory anyway. I actually don't have a problem with that perspective at all.
Here's the thing, though, if we're trying to evaluate the environment of a goalie who played for that coach
on that same team, then the quality of the team is effectively "baked in" to that rating. It becomes much more of a problem when a coach plays in different environments and has team quality disproportionately impact that. So it's super rough on Mario Gosselin playing for Perron in Quebec in 1988-89, who I'm sure was doing his absolute best to just to hang on to his .868, but for Roy and Hayward it's probably not too far off the truth (as the above stats for Hayward would suggest). Let's say Perron is a bad coach who hurts save percentages in general (I'll pluck a number out of the air and say 1.04), but his team is a great defensive team that hugely boosts save percentages (maybe 0.89 or so, in line with all-time great elite defensive results). Multiply them together, and you get 0.93, which is Perron's estimated save percentage impact based on the raw numbers, and we're kind of where we're aiming to be anyway.
My best example of how team quality can actually become a real problem would be something like Tom Barrasso's 1997-98 season coached by Kevin Constantine. I think Constantine was a very good defensive coach who was definitely helping his goalie's save percentages, yet his career save percentage is a mediocre 1.03 relative to league entirely because his career stats are spiked by coaching the awful expansion San Jose Sharks and getting destroyed by the rest of the league. Obviously that adjustment factor would be incorrect and shouldn't be used for an environment like the '97-98 Penguins.
2. Team Discipline (Separate from the Coach)
Team discipline really tends to follow coaches around, and in that case that level of team discipline (and potentially even special teams effectiveness) might be already included in a coaching rating. Occasionally it happens though that a team might deviate pretty significantly from the coach's prior norms.
A relevant example here is Carey Price's 2010-11 season under Jacques Martin. Martin's teams have historically been quite disciplined (0.96 relative to league average in terms of PP opportunities against), but the Habs that year led the league in power play chances against, which would have made Price's job more difficult than the coaching numbers might suggest.
3. Goalie Quality
What's the old canard, "Show me a good coach and I'll show you a good goalie?" Some coaches definitely were luckier than others with the talent they had between the pipes, and that impacted their team save percentages. Does a high save percentage mean a defensive system carrying a goalie, or a goalie carrying a defensive system? Could be either of them, and since the relationship is so interdependent this is one of the harder ones to tease out. It's easier for coaches who had lots of different goalies, but for the pairings where one coach was closely linked with one goalie (e.g. Fred Shero and Bernie Parent, or Punch Imlach and Johnny Bower) you really won't ever have an objective answer, it's just a matter of assessing the overall evidence. In those cases, I think the backup numbers are better evidence, as again I've argued in previous discussions.
Overall, I think evaluating coaching is very important to evaluating goalie results. People can decide how to incorporate it however they want, if you have a subjective method you think is better than have at it, but I think there is a lot of value in a consistent, objective system and I think it mostly provides meaningful feedback on various team environments which is worth the few times where it doesn't quite portray things accurately.
