How competitive was the WNBA in 2012?

With the end of the 2012 WNBA regular season, it is time to compare this season against the 15 completed seasons that have come before it.

We do this using a method called a Noll-Scully Measure, which compares the final standings in the season to the final standings in an imaginary league where all teams are of equal strength. In this imaginary league, not all teams will finish .500, for the same reason that when you flip a coin 10 times it doesn't always end up as five heads and five tails. There will still be "winners" and "losers" in this equal strength league. Without going into the math, an equal strength league gets a ranking of 1.00, and in Noll-Scully terms, the closer the ranking is to 1.00 the more competitive the league is - you can't get more competitive than every team being of equal strength.

Here are the combined Noll-Scully ranks for various leagues from 1996 to 2010, weighting across year.

Noll-Scully Measures for 1996-2010

NFL: 1.58
NHL: 1.68
WNBA: 1.74 (more on this later)
MLB: 1.84
NBA: 2.77

Noll-Scully ranks the NFL as the most competitive of the major leagues and the NBA as the least competitive. Why this is the case is academic. It could be that a league's Noll-Scully Measure could be affected by the following:

1) Role of luck during individual games.

2) How strongly the gain or loss of an individual player affects a team.

3) How long players remain in the league.

We now look at the 2012 WNBA season. The 2012 WNBA Noll-Scully Measure was 2.49, the second highest Noll-Scully ever measured for a WNBA season. One complaint about the 2012 season was that there didn't seem to be a lot of drama, and Noll-Scully agrees - there were good teams and there were bad teams and everyone seemed to know what they were just a few games into the season.

Noll-Scully Ranks Per WNBA Season

1997 1.40
1998 2.60
1999 1.62
2000 2.22
2001 1.96
2002 1.64
2003 1.65
2004 1.26
2005 2.00
2006 2.12
2007 1.53
2008 1.84
2009 1.13
2010 1.98
2011 2.30
2012 2.49

The least competitive season in WNBA history was 1998. You had a 27-3 Houston team on its way to its second WNBA championship and a miserable 3-27 expansion team in Washington. In general, expansion tends to boost a Noll-Scully value and contraction tends to reduce it. Expansion guarantees that some of the teams will be very bad, at least as far as the WNBA season is concerned, and dilutes the overall talent pool. Contraction distributes at least some modest talent across the league and raises the benchmark for league entry. However, contraction is a very bad way for a league to lower its Noll-Scully Measure.

If you look at the last four seasons, the WNBA has gotten less and less competitive over time. Once again, the reasons why are academic. The relocation of Detroit to Tulsa distributed all of the talent that Detroit had out of Tulsa (and in some cases, overseas). Washington's decision to combine the coach and general manager job after 2010 merely resulted in the Mystics losing both, and Washington has played very much like an expansion team the last two years. In the other direction, capable management and a collection of high draft picks has turned Minnesota into a powerhouse.

Is this lack of competition good for the WNBA? It depends on where you want your league to be. The NFL seems to do just fine with its low Noll-Scully Measure; no one complains that it's too hard to form a dynasty. Furthermore, the NBA seems to do just fine with its high Noll-Scully Measure, no one complains that superteams hurting the league's popularity (or if they do, the non-Miamis seem to be making enough money not to care).

Now that we have a way to look at the competitiveness of a league, we also have a way to determine how good a team is compared to other teams in other years of the league. We simply take the z-value of every team in the league, and compare how many standard deviations its winning percentage is from a team with a winning percentage of .500. The standard deviation of winning percentage will change from year to year, but all teams are compared against the same imaginary .500 team.

Unfortunately, neither the Minnesota Lynx of 2011 or 2012 rank among the WNBA's best all-time teams.

Last year's Lynx was a team 1.489 standard deviations above .500. This year's team was 1.377 deviations above .500. In a non-competitive season, you have to make a greater impact, and 27-7 would be more impressive if there were a lot more teams below Minnesota that could compete. How impressive can a 27-7 record be, goes the argument, when you get to feed on the likes of Phoenix, Tulsa, and Washington?

However, the 2012 Washington Mystics made the roll of dishonor in finishing as the #14 worst team of all time by this measure. They were actually worse than the team the year before.

Note the 1.74 Noll-Scully for the WNBA we mentioned at the beginning. That value only covers 1996-2010. If you use the same weighting system, and extend the 1996-2010 frame an extra couple of years, the Noll-Scully for the WNBA would rise to 1.89.

My personal opinion is that steps need to be taken in the WNBA to make teams more competitive. Given the weak state of some franchises, the league is not stable enough to survive the "superteams" phenomenon of the NBA. Washington shows that fans are not going to attend games from a team that can't prove itself competitive out of a sense of loyalty, unlike teams from the other major leagues.

As for the Lynx of 2011-12 not making the list of the Top 15 WNBA Teams of all time, I suspect that they won't be so distraught. "Let us have our two championship rings," they would say, "and you can keep your list just as you please."