Do we really need 82 games?
Somewhere around the time it became known that after the labor stoppage the NBA was, in fact, going to have a 2011-12 season -- and that it would feature only 66 games per team instead of the usual 82 -- the thought dawned in various circles of fans and the media:
"Why do we always play 82 games, anyway?"
ESPN's Kevin Arnovitz asked that very question last October, eventually endorsing a 44-game slate instead, and J.A. Adande advocated shortening the schedule to 76 games.
The usual arguments were in play here: reducing the number of back-to-back contests and "schedule losses" (games teams essentially punt at the end of long trips) would significantly raise the quality of basketball being played.
A major byproduct of shortening the schedule would be pulling star-studded teams like the Lakers and Heat closer to the pack, creating a more exciting regular-season product in the process.
There's plenty of scientific evidence to back up those sentiments. In 2010, researcher Daniel Myers found that teams coming off at least one day of rest played more than two points better per 100 possessions than they did on the final leg of a back-to-back. Given that 49.7 percent of games last season featured at least one team playing its second game in two nights, more than any year since 1999 (when it happened in 54.1 percent of games), it's no coincidence that my timeline-adjustment research this summer found the average 2011-12 NBA team to be slightly worse than it was in 2010-11, despite the league's overall talent being on an upward trajectory since 2005.
While 66 games sounds like a vacation compared to 82, last year's compressed schedule packed an average of 8.1 games into each day of the regular season, which was actually more than the 7.3 per day the league played in 2010-11. And even the league's ordinary 82-game schedule sees at least one team facing a back-to-back in 40 percent of all games.
So how many games should the NBA play?
The ideal schedule, from a rest perspective, wouldn't just shave games off of the docket, it would spread the remaining games over the league's typical five-and-a-half-month timespan, rather than the mere four months we saw last year. It would give teams more rest days and reduce the grind-it-out state the current 82-game regular season often devolves into.
Too many games
There's hard evidence and research to suggest that an 82-game schedule is flat-out unnecessary. Comparing the length of a league's schedule to the spread of its teams' won-loss records, statisticians can estimate the amount of certainty that any team's observed record accurately represents its true underlying skill level. And when we look at the big three North American sports under that microscope, it's clear that we learn much more, far sooner, about the relative true abilities of basketball teams from their win-loss records than we do in either baseball or football.
For the NBA over the 2005-11 period, it took roughly 12 games, or 15 percent of the league's 82-game schedule, to glean as much knowledge about the relative true abilities of its teams as we learn in 88 Major League Baseball games (54 percent of the schedule) and 10 NFL games (64 percent of the schedule).
Both MLB and the NFL have very similar actual season lengths, relative to the number of games required to arrive at a given level of certainty between observed records and true talent. The NBA, however, is a major outlier in this regard. In order to achieve the same level of certainty as we see in NBA records after 82 games, NFL teams would have to play a 71-game season, while MLB teams would require a 610-game schedule. And you thought the baseball season drags on too long now!
Basketball's design
Why do we see such a high level of certainty so quickly in basketball? It seems to come down to the nature of the sport. For starters, basketball has by far the smallest number of players on the court/field at any given time, a reduction in moving parts that allows the cream of the crop to truly shine.
In that same vein, there are also no rule-based limits on player usage in basketball. While a great baseball slugger can only bat once every time through the order, a basketball team can funnel as much as 35-40 percent of its possessions through a superstar like Kobe Bryant or LeBron James. And although an ace starting pitcher has more influence in a given game than a basketball alpha dog or NFL quarterback, he only pitches once every five games. For game-in, game-out control by a single player, it's tough to beat the NBA's stars.
Furthermore, the structure of each game naturally promotes differences in variance. The typical NBA game contains about 92 possessions per team; by contrast, MLB teams play nine innings per game (a single inning being roughly the equivalent of a possession in basketball) and NFL teams average 11-12 drives per game.
By the Law of Large Numbers alone, there are far more opportunities for the better team to exert its talent advantage in basketball than in either baseball or football. Likewise, because scoring in those latter two sports is rarer than in basketball, they can be modeled with more built-in random chance, which contributes to situations where the less talented team wins.
Problems with short schedule
Whatever the underlying reasons for the greater certainty in basketball, the numbers speak loudly about the structural benefits of a shorter NBA schedule. It would be a tough sell, though, to ask the owners to slice their home dates in half (or more) or ask the players to take a possible reduction in pay.
The league's biggest question would be whether it could offset the reduction in games by increasing revenues elsewhere, perhaps generating more fan interest by evening the playing field between big-market and small-market teams.
After all, one of the commissioner's stated goals of last year's lockout was to promote parity. In the wake of the infamous canceled Chris Paul trade -- and the Lakers' subsequent acquisition of Dwight Howard -- it appears that such an aim cannot readily be achieved via collective bargaining, but the numbers above suggest that a shortened schedule would be an easy way to instantly make the league more balanced, even in today's top-heavy, "Big Three"-laden environment.
A major byproduct of shortening the schedule would be pulling star-studded teams like the Lakers and Heat closer to the pack, creating a more exciting regular-season product in the process.
Cutting down the schedule won't fix every problem the NBA currently faces, but it would go a long way toward improving the quality of the game, whether by ensuring more rested players or eliminating extra contests that, mathematically, are completely unnecessary when compared to the schedules of professional baseball and football.
In other words, the next evolution of the game might be to realize that less can indeed be more.
Neil Paine is a writer for Basketball-Reference.com. Follow him on Twitter: @Neil_Paine
Filed under: NBA, Classic Posts