Puckerings archive: Unbiased Standings (31 Jul 2001)

What follows is a post from my old hockey analysis site puckerings.com (later hockeythink.com). It is reproduced here for posterity; bear in mind this writing is over a decade old and I may not even agree with it myself anymore. This post was originally published on July 31, 2001 and was updated on November 13, 2002.

---

Unbiased Standings
Copyright Iain Fyffe, 2002

---

As we all know, unbalanced schedules (that is, not playing the same number of games all other teams) create bias in the final standings for any given year. Teams that play within weaker divisions may be overrated, since they play more games against weaker teams, and teams playing in stronger divisions may be underrated.

Using the 1995/96 NHL season, I calculated an unbiased ranking of teams. That is, for each team I calculated the winning percentage versus every other team, and computed the simple average of these percentages. This eliminates the distortion created by playing different numbers of games against certain teams. I then converted these percentages into a number of points for 82 games, which produced a new overall ranking. Here are the team-by-team results:

Team	Old Pts	New Pts	Old Rank	New Rank
Anaheim*	78	82	t-17	13
Boston	91	89	8	t-9
Buffalo*	73	70	20	20
Calgary	79	81	t-15	t-14
Chicago	94	100	6	5
Colorado	104	102	2	3
Dallas*	66	63	t-22	23
Detroit	131	128	1	1
Edmonton*	68	68	t-22	23
Florida	92	89	7	t-9
Hartford*	77	84	19	12
Los Angeles*	66	65	t-22	22
Montreal	90	90	9	8
New Jersey*	86	81	12	t-14
NY Islanders*	54	56	24	24
NY Rangers	96	97	5	6
Ottawa*	41	46	26	26
Philadelphia	103	107	3	2
Pittsburgh	102	101	4	4
St.Louis	80	79	t-13	16
San Jose*	47	47	25	25
Tampa Bay	88	87	11	11
Toronto	80	74	t-13	18
Vancouver	79	76	t-15	17
Washington	89	91	10	7
Winnipeg	78	72	t-17	19

For most teams, the difference is minor. However, some are worth noting.

Hartford wins the Divisional Whipping-Boy Award, for being the most victimized by the unbalanced schedule. They move up 7 points when the bias is removed, and move up 7 places in the overall standings. Quite a difference, eh? Imagine Hartford in the playoffs!

Toronto wins (?) the Abuser of the Year Award, for taking advantage of the system. They lose 6 points, and 5 places in the standings when everything is equalized. They shouldn’t have even been in the playoffs, with their standing of 18th overall.

The playoff picture would have been quite different. Teams marked with a star in the above table missed the playoffs in reality. Ideally, the top 16 teams would make the playoffs, rather than the mishmash of divisional winners and conference-based determinations.

The playoff teams that shouldn’t have made it are Toronto, Vancouver, and Winnipeg. Speaking as a Canadian hockey fan, maybe this system isn’t so bad after all.

The non-playoff teams that should have made it are Anaheim, Hartford, and New Jersey. It’s a travesty that the Devils, ranked 12th in the league overall in reality, were denied the chance to defend their championship.

I’m not arguing that unbalanced scheduling should be abandoned. The realities of travel costs prohibit that. However, it is an interesting exercise to see who benefits and who suffers from the arrangement. Just food for thought.