Did home scorer bias affect Roy and Hasek?

Updated: May 31, 2012 at 7:28 pm by Cam Charron

As Eric noted yesterday, Martin Brodeur’s true talent may not be captured by his save percentage, since the home scorer in New Jersey probably did not count shots as effectively. Brodeur faced much fewer shots against in Jersey and had a lower save percentage. But his goals against was also much lower as well, leading me to believe that the Devils ought to have played better at home, but on the surface, Brodeur didn’t.

It’s not just shots that Jersey doesn’t record properly. Simply eyeballing the team’s RTSS data on NHL.com can make one a little suspicious. The Devils had 741 recorded hits at home and 936 on the road. They blocked 322 shots at home and 606 on the road. They missed 331 shots at home and 431 on the road. The only thing that looks remotely consistent are the giveaway : takeaway numbers.

But how did this affect some of the other great goaltenders from the era? Dominik Hasek and Patrick Roy specifically. Since all three goaltenders played a good chunk of their prime years between the lockouts of 1994-95 and 2004-05, it’s best to compare only that era, since Brodeur played some time post-lockout during a period in which save percentages slowed.

The Splits

Here are the top 10 goaltenders by save percentage:

Hasek 0.926 2.20
Roy 0.916 2.35
Brodeur 0.912 2.15
Burke 0.912 2.63
Hebert 0.911 2.76
Belfour 0.910 2.30
Kolzig 0.910 2.53
Khabibulin 0.910 2.61
Joseph 0.908 2.60
Osgood 0.908 2.41

From this look, Hasek is superior, Roy is up there, and Brodeur a little lagging towards the rest of the field among goaltenders who played at least 400 games.

But what are the home-road splits between the three top goaltenders of the era? How did shot recording bias affect each one?

  Home SA/60 Away SA/60
Roy 27.2 29.1
Hasek 29.3 29.9
Brodeur 23.2 25.7

These numbers aren’t really sorted any particular way, side-note. While Roy probably played in front of a better defence than Hasek did, Brodeur’s New Jersey Devils were four shots better on the road. I have no idea what to make of this, but it doesn’t look like Brodeur had to do too much work as a goaltender.

Save percentages? Well, the major take-away is that Martin Brodeur had a higher road save% in this era than Patrick Roy did:

  Home SV% Away SV%
Roy 0.918 0.914
Hasek 0.928 0.923
Brodeur 0.909 0.915

Despite facing fewer shots at home, Roy and Hasek’s save percentages were higher at home than on the road. That makes them consistent with their GAA totals:

  Home GAA Away GAA
Roy 2.23 2.49
Hasek 2.11 2.30
Brodeur 2.10 2.19

For what conclusion to draw from this data I don’t really know, but it looks like Roy’s home scorer either undercounted shots or the Colorado Avalanche were better at defence at home. Either way, smaller shot totals at home didn’t affect their save percentages, only Brodeur sees that effect.

Perhaps when I have a bit of time, I’ll look to see how the other goaltenders from that era fared, but I think it’s legitimate to say the following things:

A – New Jersey had a real good defence between lockouts.

B – Teams play better at home than on the road.

C – Goalies do too.

D – Dominik Hasek was probably the best goalie of the era, but Roy and Brodeur are closer than I initially thought.