League parity


This 6-article series presents some crazy and interesting findings from all NBA games over the last 14 seasons (from 2002/2003 to 2015/2016).

Today we answer question #2:

Over the past 14 NBA seasons, has league parity increased or decreased?

Parity is a key part of the general health of any sports league, as some fans may lose interest if some teams totally dominate their opponents, while others lose on a consistent basis. Unfortunately, there is strong evidence of a significant decrease in parity in the NBA over the last 14 seasons.

In order to investigate this question, we have used two different approaches.

Approach #1: A league with parity will tend to have more teams whose record is around .500, and few teams with very high or very small win percentages. Following this line of thought, we introduce a measure of parity:

Measure of parity = ( |Team 1 win % - 0.5| + |Team 2 win % - 0.5| + ... + |Team 30 win % - 0.5| ) / 30

Let’s clarify things:

  • For each team, we are first calculating the distance between its win percentage and .500.
  • Then, we are taking the “absolute value” of the number obtained, where the “absolute value” simply means making it positive. In other words, the absolute value of 3 is 3, while the absolute value of -3 is 3. In mathematical notation: |3| = 3 and |-3| = 3.
  • Finally, the measure of parity is the average over all 30 teams.

IMPORTANT NOTE: a small measure of parity means more parity (lots of teams around .500), while a large measure of parity means less parity.

This measure is calculated for each of the past 14 NBA seasons. Let’s see the evolution over the years:

The horizontal line was drawn at 0.126, which is the average measure of parity across all 14 years. There is very clear indication of an upward trend, as the first 5 seasons yielded below-average measure of parity, while the last 9 seasons were all above-average. Remembering that high values of parity measure correspond to less parity, we conclude to a clear decrease in NBA parity.

Approach #2: For each team, we calculate the percentage of times it came back to win a game when trailing after quarter 1, quarter 2 and quarter 3. We use them to assess parity by calculating the average comeback percentages across all 30 teams for each season. If parity has indeed decreased over the years (as suggested by the previous analysis), we are going to observe less comebacks occurring recently.

Let’s start with the case of comebacks following a deficit after the first quarter:

The horizontal line corresponds to the average quarter #1 comeback percentages across all 14 seasons. One could argue that there is some downtrend, although not totally clear (omitting 2003 would make it more obvious). However, the last three years have been clearly below-average regarding comeback % once trailing after quarter #1.

Let’s move on to the same graph as above, but this time focusing on the proportion of times a team overcame a halftime deficit (i.e. through 2 quarters instead of 1):

The plot above leaves absolutely no doubt about league parity decreasing in recent years. As a matter of fact, the last 5 seasons all produced below-average comeback percentages, with last year being significantly lower than usual.

We finish off by presenting the corresponding graph related to 4th quarter comebacks (i.e. coming back to win a game after trailing through 3 quarters):

The downward trend is not completely clear, much like the quarter 1 plot shown earlier. If not for last season, we would have had 5 straight below-average comeback percentages. It’s fascinating that last year had much less second half comebacks, but still a fair share of fourth quarter comebacks.

In summary, the two approaches considered in this study point in the same direction: the NBA league parity has been clearly in a decreasing mode over the last few seasons.


Thanks a lot for reading! Here is what’s coming up in next articles:

Over the past 14 NBA seasons,

  • …is there evidence of game momentum? (you will learn that, unlike popular belief, there is no such thing as “having the momentum” during a game. In fact, it seems like when a team goes on a roll throughout a quarter, it tends to strike back for the rest of the game. Like a wake-up call to their opponents!)






Source: https://www.basketball-live-betting.com/