We have seen a number of articles and social media activity speculating about the NBA’s desire to have Miami advance to the NBA finals. It’s a nervous time for the NBA because the other 3 teams in the conference finals are from “small” markets. In some ways, the success of small market teams is a welcome outcome as all professional leagues tend to be nervous about large market dominance resulting in competitive imbalance, but the overwhelming short-term concern is obviously about how this situation will impact the television ratings for the final.
We have seen speculation about the impact of having small market teams such as Indiana, San Antonio and Memphis in the finals, but not a great deal of analysis. To fill this gap, we have developed several statistical models that forecast TV ratings as a function of the characteristics of the two teams who are participating. As a starting point we collected data on market population, winning percentage, home attendance, pricing, road attendance, and the number of All-star game starters and reserves for each team participating in the NBA finals over the last several years. In this case, we have only a limited number of data points, so the key to the analysis is in identifying which of the variables are the best predictors.
We tried a great many combinations of the previously listed variables and found that the two best predictors were the sum of the two participants’ home box office revenues and the number of All-star game starters participating in the series. A model with these two variables yielded an R-squared value of 0.53, and both explanatory variables had t-stats with p values of less than .05.
Our speculation is that combined home revenue captures the market size and fan intensity of the two teams. This metric seems to be much more effective than population simply because not all large market teams are equivalent draws. For example, in LA, the Lakers are a more powerful brand than the Clippers, and in New York, the Knicks have dominated the Nets (let’s say the New Jersey Nets to avoid any additional angst from the Brooklyn contingent).
We also found that All-star starters was the right metric rather than total All-stars. In hindsight, this is also an intuitive finding. The NBA has long been known as a “Star” driven league. In fact, if you look back in history, the Michael Jordan era had amazingly high ratings compared to the last decade. Based on the data, it appears that finals ratings are driven by the number of extremely high profile players.
In the tables below, we report actual ratings for the last six finals and our model’s predictions for the possible NBA Finals matchups. As expected, the most promising matchup is Miami versus San Antonio. What are really notable are the predicted ratings for the least promising matchup. We predict that an Indiana – Memphis matchup would result in an epic failure in terms of ratings.
As a reality check for our prediction, consider the most recent finals matchup of small market teams. In 2007, San Antonio defeated Cleveland, and the finals achieved a 6.2 rating. While this number is much higher than our prediction, the San Antonio and Cleveland series had a significant advantage relative to an Indiana-Memphis matchup. The difference was that the Cleveland and San Antonio series featured LeBron James and a Tim Duncan still close to his prime. These types of stars would be sorely lacking in a Memphis – Indiana series. This is, however, not a criticism of the Grizzlies or the Pacers, but more an indictment of how the NBA markets itself. The NBA’s practice of emphasizing a few marquis players means that ratings will suffer when teams without these high brand-equity players make the finals.
The other problem for the NBA is that fans also understand the league’s dilemma. This means that a meaningful percentage of fans believe that the NBA clearly prefers a series that features Miami. This is a significant problem if fans believe that marketing considerations influence outcomes.
Mike Lewis & Manish Tripathi, Emory University 2013.