March 14, 2013
Greater Than Plus Minus
NHL Realignment, Part 4

by Brian Macdonald

In Part 1, Part 2, and Part 3 of our look at the NHL realignment work that we did in this paper, we estimated team and league travel without knowing the schedule, presented some of our best 6-division realignment solutions, 4-division solutions, and solutions for some hypothetical franchise moves and expansion to 32 teams. Since the writing of those articles, the NHL and NHLPA have settled on and approved a new plan. As promised, we'll analyze this new plan.

But first, let's review. Here was the NHL's proposal from last year:

We estimated that this configuration would have caused the league an extra 100,000 miles of travel per season. Part of this increase would have been due the choice of conferences. It was not optimal from the perspective of minimizing distance to have TB and FLA with the current Northeast Division teams.

But one thing that we did not focus on enough last time was the effect of the change in the schedule structure that accompanied this proposal. Each team would have played 5-6 games against teams within their conference, and 2 games against everyone else. This schedule would have been a bit more balanced than the current schedule. The minimum number of times you would play a team each season would be two instead of one. A more balanced schedule means more travel for the league overall.

This schedule structure would have been good for some Central Division teams, but would have been bad for most teams, especially the West Coast and Florida teams. The West Coast teams would play a road game at every east coast team each year, instead of visiting the east coast teams once every two years.

The most recent proposal is similar to the one above, except that COL, DET, CBJ change divisions:

Let's compare travel under this proposal with travel under the old proposal.

Reminder: Black dots are the estimated miles for the NHL's old proposal. The green arrows pointing down indicate that teams would have better travel (fewer miles) under the new proposal. Red indicates that teams would have worse travel (more miles) under the new proposal.

It is a little worse for most teams, which is partly because the schedule proposed along with the new realignment was even more balanced. Teams play 4-5 games within their division instead of 5-6 games, 3 games instead of 2 against non-division teams in their conference, and 2 games against teams out of their conference.

It might be surprising that there is not a bigger improvement for CLB and DET, but they were actually pretty well off last time as far as distance. Instead of being far from DAL and WPG, they'll be far from TB and FLA, so that's almost a wash. The main benefit for them will be the Eastern time zone games.

TB and FLA are still with the Northeast division teams, which is not optimal for them, for the other teams in their division, or for the league overall. So let's look at the "best" solution, according to distance, subject to a couple of constraints. Let's suppose the NHL is set on having 14 teams in the West and 16 in the East, and CBJ and DET are the two teams that move East. Let's also assume that the NY teams have to stay together, and PHI and PIT have to stay together, but that PHI and PIT do not have to be with the NY teams. Here's what we get:

League travel is better, but only by about 3,000 miles. Since the West stays the same, let's focus on the change in travel for the 16 teams in the East:

As we would expect, TB and FLA would have better travel, but it is only better by about 700 miles. So there's not a huge difference. This might be a little surprising, but we have to remember the new schedule is more balanced. Under the new proposal, Eastern teams would play an average of 4.3 games per season against teams in their division, and 3 games against the teams in the other Eastern division. So in our best solution, TB and FLA would play the Northeast Division teams 3 times a year instead of 4.3 times a year. That's not a huge difference.

Optimizing the divisions has less of an effect on travel with this more balanced schedule than it would with an unbalanced schedule. Consider the extreme case where the schedule is completely balanced, and a team plays every other team the same number of games. The choice of divisions wouldn't matter at all.

At first glance, TB and FLA get the short end of the stick on this one. But this analysis suggests that it's really not that bad…it just looks bad.

Of course, in all of this analysis, we have to keep in mind that distance is far from (no pun intended) being the only concern. Changes in travel time and travel costs are both important factors. Changes in TV revenue when teams are in the same time zone rather than different time zones is another. The NHL has presumably determined that the benefits of their plan outweight the increase in travel miles. Given our analysis of the old proposal, the upcoming realignment, the best proposal, and the small differences in travel among them, that conclusion seems pretty reasonable.

If we knew that the next two expansion teams would be in Seattle and Kansas City, or Houston or some other western conference city, we would understand this realignment plan even more. But what happens if the next new teams include Quebec, or southern Ontario, or both? Plus, in the mean time, there is that oft-discussed issue with an uneven number of teams in each conference. It should be interesting to see how it all unfolds.

Brian Macdonald is an author of Hockey Prospectus. You can contact Brian by clicking here or click here to see Brian's other articles.