The League of Extraordinary Statisticians: The Inaugural

The League of Extraordinary Statisticians is a weekly forum bringing together top analytical minds in the hockey world to answer a variety of questions that straddle the line between stats analysis and something you might hear floating around section 304.  They have agreed to answer these questions in a few paragraphs or less, and with minimal formulae.  Because this is a forum, we'd encourage you to use the comments section to answer the question yourselves, or to discuss or debate the answers given.  

The hockey stats community has been rapidly growing over the last decade, representing a wave of thought (partially sabermetrics-inspired, partially descendant from Corsi, Neilson, and others) spreading primarily from Canada.  The internet and standardized game-tracking measures have bolstered this development, and now the topic of hockey stats analysis refers to a widespread body of experts and articles.

This community is far less chaotic than it appears, though, as hockey statisticians build and debate with one another, creating an academic process worthy of any university.  It's easy to lose sight of the fact that they are hockey fans, just as likely to holler and drink beer at a hockey game as anyone else.  In that spirit, I wanted to create a forum around a combination of the high-powered minds of the stats community and questions a regular fan might ask.  Graciously, many of the people I asked agreed to answer some if not all of the questions.

As this is the inaugural week of the once-a-week series, I want to point out a couple of things.  For one, this is definitely not the entire of the hockey stats community; there were a number of people who were too busy or I didn't have an easy way of contacting.  Secondly (and as mentioned above), this is a forum, so feel free to contribute your takes on the questions or responses in the comments section.  There is strength in the wealth of informed ideas.

Finally, I just want to thank all the contributors to the forum.  As with most of us, they have lives beyond their work as statisticians, and their contributions here are a testament to their hard work and love of hockey.  Be sure to check out their sites linked next to their names.

This week's question: If you had to pick one NHL rule that you would institute, change, or remove, what would you choose?  Why?

If I had to change a single rule in the NHL right now, it would be the two-three point system for games. This system has wrought havoc with the integrity of games, encouraging teams to play forties and killing much third-period play. I would switch to a full three-point system with 3 points for a regulation win, 2 for an OT/SO win and 1 for an OT/SO loss.

If I could experiment more drastically, I would try going all 4-on-4 for some period of time. NHL players seem to have outgrown the standard-sized rinks, and since converting all NHL rinks to international ice surfaces is out of the question, the other way to create more space would be to remove a player. This would change strategy dramatically and create more open space, but scoring would not shoot through the roof: scoring at 4-on-4 in the last few years has remained below 3 goals/60 minutes.

– Tom Awad, Hockey Prospectus

I would remove shootouts and points for losing if you get to overtime. Why? Mostly as a hockey purist I want to take out some of the more outrageous examples of hockey standings being decided for non-hockey reasons.

Shootouts might be spectacle, but they don’t do a good job of determining which team wins a game any more than a hardest shot competition would. It merely adds a mostly random element to the standings. Teams can gain or lose significant ground in the standings due to shootout success – which appears to be largely random. Teams have not been able to sustain shootout success from one year to the next. Few players show significant success in shootouts for a sustained period of years. It is not clear that it is much of a consistent skill and it is overvalued. A player who scores three goals on three shootout attempts is worth almost two wins to his team that season and that is far too significant a value placed upon essentially a fluke.

Points for losing games that are tied in regulation serve to keep the standings closer than they should be, as a loss is not always two points less than a win. This allows weak teams to remain in striking range of a playoff berth and with a little luck sometimes make the playoffs.

As much as possible I want to see the best rise to the top. I want to see great games for first place and in the playoffs. Weak teams having a better chance to succeed because of shootouts and points for losing get in the way of this happening.

– Greg Ballentine, The Puck Stops Here blog at Kukla’s Korner

I’d change the icing rule so that it’s not a race to the puck/boards, but a race to some other point on the ice, like it’s been proposed (say the faceoff circle in the defensive zone…if the icing team beats the race to the circle, then it’s a race to the puck). I’d also maintain the rule even in shorthanded situations. If the league really thinks the shorthanded team is too much at a disadvantage without the special non-icing rule for them, then change it to a 90-second penalty or 110-second penalty to maintain the icing rule.

And I’d go with four-on-four for the whole game. For PP, I could go either way (5-4, or 4-3). Two-man PP would have to be 5-3.

– Tom Tango, tangotiger.net and author of The Book: Playing the Percentages in Baseball

The shootout. A hockey game should not be decided by a skills competition. Why not have a push-up contest, or a race to the red-line and back? Or have two goons slug it out at centre ice? It’s all so arbitrary, and has no statistical correlation to team skill.

Runner-up: get rid of head shots. I’d also get rid of no-touch icing, if I get a third choice.

In any event, I think NHL rules are fine the way they are, they just need to enforce them correctly and evenly rather than change the rules depending on who is winning, how much time is left, and how small the player that was hit is.

– Rob Vollman, Hockey Prospectus

Not a ton of love for the shootout, some suggestions for icing, and at least two calls for the wide-open, 4-on-4 hockey.  What say you, readers?  Like these ideas?  Have some of your own?

We have a lot of different, interesting respondents and responses in the coming weeks; stay tuned.