Another sports-history experiment is over: Our NBA Tuesday revisions of the American Basketball Association’s MVP and Rookie of the Year awards. What we learned is that the voters from the era shot 66.7 percent from the sidelines on picking the right winners for the two awards.
In 9 seasons, from 1968 to 1976, we confirmed two thirds of both the MVP and ROTY winners, taking under consideration many factors laid out here in our original NBA Tuesday column. This is pretty impressive and somewhat demonstrates that modern basketball sabermetrics are balanced effectively enough to include traditional evaluative methods.
Let’s look more closely at the years where the original voters were off on the awards:
- 1969 MVP: Indiana Pacers center Mel Daniels was, at best, the fourth-most valuable player in the league.
- 1971 MVP: Daniels was, again, not the best or most valuable player in the league.
- 1971 ROTY: Co-winner Charlie Scott, Virginia Squires shooting guard, was nowhere near as good or valuable as co-winner Dan Issel, Kentucky Colonels center.
- 1973 MVP: Carolina Cougars power forward Billy Cunningham may have had a great season, but he was not the most valuable.
- 1973 ROTY: New York Nets point guard Brian Taylor did not play a full season, and this gave him much less value overall.
- 1975 ROTY: Spirit of St. Louis power forward Marvin Barnes had the best rookie season, but he did not have the most value among rookies.
We went heavy on the sabermetrics for these retroactive analyses, and the basketball metrics are great for balancing the aforementioned modern and traditional measurements. For example, Win Shares illuminate value while Player Efficiency Rating emphasizes individual quality.
We’ve been merging both for as long as we have had the ability to do so, and it creates a great blend. However, the voters in the past didn’t have those same tools, although they did a pretty decent job overall of assessing value in their voting processes.
Compared to the voters for the American Football League, for example, the ABA voters did the job significantly better. Of course, the AFL voters went just 1-for-10 in their MVP picks, so almost any group of voters can do better.
For comparison’s sake, however, the NBA voters—probably many of the same people, in truth—picked 12 of 18 MVP and ROTY winners correct, by our estimate, in the same time frame (1968-1976).
That is consistency, and again, it shows that modern sabermetrics for basketball are very well blended between modern and traditional statistical analysis. The same cannot be said for football (yet), obviously.