It's tough to be an NHL referee – though the days of being physically attacked by fans and players are history, there's no one else on the ice with no performance upside: either they make the right call (and possibly get booed) or they make the wrong call (and probably get booed). They will never be able to simultaneously please coaches, players and fans.
But just because everyone on the ice and in the rink has a vested interest in having calls go their way, it doesn't mean that referee performance is identical. Overall, referees call 6% more penalties against the visiting team, which might reflect home-ice advantage (ie – visiting teams are more likely to give up chances, and thus need to take more penalties) or it may reflect a human bias in officiating. On top of that, there are additional biases: some refs call a tighter game; some are more likely to penalize the visitors.
I looked at every referee who worked 500+ regular season games since the 1995-96 season and adjusted the penalties they called each game based on the season and the teams that were playing. The results are below:
| Games | Extra PIMS | Away | Home | Homer Bias | |
| Don Koharski | 935 | -1.00 | -0.26 | -0.74 | -0.35 |
| Bill McCreary | 921 | 0.10 | 0.92 | -0.82 | 0.88 |
| Paul Devorski | 921 | -1.27 | -0.57 | -0.70 | -0.74 |
| Dave Jackson | 916 | 1.51 | 0.92 | 0.59 | -0.33 |
| Dan Marouelli | 901 | -0.49 | 0.02 | -0.51 | -0.36 |
| Kerry Fraser | 899 | -0.55 | -0.05 | -0.49 | -0.35 |
| Rob Shick | 840 | 1.31 | 1.58 | -0.28 | 1.11 |
| Michael McGeough | 829 | -0.84 | -0.20 | -0.64 | -0.31 |
| Dennis LaRue | 805 | -1.11 | -0.18 | -0.94 | 0.04 |
| Dan O’Halloran | 727 | -0.10 | 0.10 | -0.20 | -0.43 |
| Brad Watson | 683 | 0.49 | 0.39 | 0.10 | -0.66 |
| Greg Kimmerly | 656 | -0.57 | -0.04 | -0.53 | -0.41 |
| Mike Leggo | 646 | -1.38 | 0.09 | -1.47 | 0.70 |
| Stephen Walkom | 631 | -2.24 | -0.64 | -1.60 | 0.22 |
| Terry Gregson | 620 | -0.16 | 0.27 | -0.42 | -0.14 |
| Tim Peel | 584 | 0.00 | 0.30 | -0.29 | -0.22 |
| Kevin Pollock | 578 | -0.01 | 0.33 | -0.34 | -0.10 |
| Brad Meier | 577 | 0.45 | 1.10 | -0.66 | 0.83 |
| Marc Joannette | 572 | 0.09 | 0.48 | -0.39 | -0.04 |
| Rob Martell | 545 | -1.13 | -0.23 | -0.90 | 0.00 |
| Mark Faucette | 542 | 1.93 | 2.01 | -0.08 | 1.39 |
| Mike Hasenfratz | 535 | 1.28 | 1.49 | -0.21 | 0.77 |
| Stephane Auger | 518 | 1.60 | 1.43 | 0.17 | 0.39 |
| Kelly Sutherland | 514 | 0.20 | 0.60 | -0.40 | 0.14 |
| Paul Stewart | 508 | -2.62 | -1.06 | -1.56 | -0.35 |
| Average | +0.40 | -0.43 |
I don't know the reputations of individual referees – I assume they're all hated by someone – but Marc Faucette, Rob Shick and Bill McCreary appear to tilt the most strongly towards the home team. Overall, penalty totals tilt 0.83 minutes against the away team, and no one favors the visitors anywhere near as much.
It's important to keep in mind what the impact of this 'homerism' is: if a referee penalizes a visiting team one additional penalty minute per game in a seven-game playoff series (which strikes me as fairly extreme), that's worth 0.7 goals, which tilts the series about 3% to the home team. Of course, if those additional penalty minutes are all called in overtime of Game 7, the impact is much much larger.
Update: Japer's Rink (or as we call it at SBNation, Jasper's Rink) has a different look at the same data.
Further update: Gooseman did a great job of analyzing this same issue last season.
