basketball is a markov process … and that’s why i hate it
I hate watching pro-basketball because it’s a Markov process. This bugs me a lot.
I first noticed this as a kid. Didn’t matter what happened the first 46 minutes. If Larry Bird was had the ball, there’s a good chance he’d win. Then came Jordan. No matter the score at the 2:00 mark, he could pull off a win. As time passed, I noticed how deeply Markov pro-basketball was. For example, last year, my father in law took me to a Pistons/Lakers game last year. Lakers did bad. Heck, they did bad all season and then went on to win the championship. Other sports aren’t like that. If Brazil goes to 0-2 in the first half, they aren’t winning, even if every player was a Romario/Pele love child. You don’t see many Superbowl champs with an 8-8 record. History matters in most sports, but not in basketball.
There’s a deeper lesson here. People watch sports for two reasons. First, we like to see excellence. We want to see Usain Bolt become the fastest human being in the world. Awesome. Second, there’s a narrative we buy into. Dynasties, underdogs, new faces. The Markov property undermines the whole narrative because the first act has no bearing on the outcome. That’s why, ultimately, police procedurals are often boring, if satisfying in the short term. Yes, the intros are fun, but who cares if the bad guys are caught every single time in pretty much the same way? Why should I invest in regular season NBA games if they aren’t going to affect the final outcome?
orgtheory.net 
The Lakers won the Western Conference last year. While they likely coasted a bit at times, I don’t think it’s fair to say that they “did bad.” Their regular season performance earned them home-court advantage and helped them overcome difficult playoff series vs. Phoenix and Oklahoma City, and may have made the difference in the Finals, which went the full seven games against Boston. In this case, I’d argue that the regular season clearly was consequential towards, if not also affected, the eventual outcome.
Further, if Brazil trails 0-2 at halftime, they’re coming back occasionally. There are relatively few unequivocal propositions one can make about the future in sports, and anywhere in life when people are involved, in my opinion. The ordinary and banal set the stage for the rare and dramatic. It’s why the media in my native Canada can market overwrought sentiments of tragedy, after the national junior hockey team blew a 3-0 3rd period lead to the Russians. Somebody looked up the records, and history suggested that in theory, such a lead should be safe over 99% of the time. Still, black swans tend to have disproportionate influences on the world.
Further, the process of gaining an advantage can be pyrrhic, underpinning an eventual loss or equilibrium change in the competition later on (e.g., revealing of strategies, overconfidence, player injuries, suboptimal risk-aversion in protecting a lead). I’d suggest that differing groups, teams, organizations, etc. have various mixes and equilbria between memorylessness and imprinting from the past. Some games may be more ‘Markovian’ than others, but basketball never struck me as extremely memoryless.
LikeLike
ksiler
January 14, 2011 at 1:26 am
What do you mean by Markov?
In the usual sense, it means that the state of next minute only depends on the state of this minutes plus some new development. But the past can also affect future by affecting the state of the nature now. (In linear model, Markov process can be represented by an AR model, where the AR coefficient can be close to 1. I am not sure if the case of unit root is still Markov.)
LikeLike
passerby
January 14, 2011 at 2:00 am
Lousy soccer example: http://en.wikipedia.org/wiki/2009_FIFA_Confederations_Cup_Final
LikeLike
Simon
January 14, 2011 at 2:10 am
i’m guessing you must really hate quidditch
LikeLike
gabrielrossman
January 14, 2011 at 3:25 am
harvard quidditch — http://www.youtube.com/watch?v=-YmJob5rbNM
LikeLike
teppo
January 14, 2011 at 3:44 am
I think part of the problem with NBA basketball is there are too many possessions per game. The difference between the best and worst offense in the NBA this year is only 12.2 points scored per 100 possessions, and most teams get close to 100 possessions per game. It’s a large enough sample that regardless of how teams start, we will probably see regression to the mean by the 4th quarter.
All other sports have rarer scoring (soccer, hockey) or fewer possessions (football, baseball, even college basketball), making regression to the mean less likely during the course of a single game.
LikeLike
Noah
January 14, 2011 at 4:56 am
Thanks, Noah. I think that’s the issue. I had under appreciated the crazy number of possessions.
LikeLike
fabiorojas
January 14, 2011 at 4:57 am
The beauty of basketball, in my opinion, is that it has the fluidity and athleticism of soccer while incorporating the most sophisticated instrument under God’s green earth, the human hand.
More to the point of the thread, it’s the only sport that truly has runs–as in a bucket followed by a steal or a turnover and another bucket, followed by a steal or a turnover and yet another bucket. Football has more dramatic plays, hands down, but the building of momentum in basketball is unparalleled.
LikeLike
Bob
January 14, 2011 at 7:11 am
Quidditch is the main reason I’m lukewarm on the whole Harry Potter phenomenon. Whoever came up with those lousy rules is clearly not a sports fan.
LikeLike
brayden king
January 14, 2011 at 7:35 pm
I’ll bet there’s a Northwestern Quidditch team that you could join. And sure enough (a quick google search reveals) there’s not just a team but a NW intramural Quidditch league: http://www.dailynorthwestern.com/campus/intramural-quidditch-takes-flight-1.2001328
LikeLike
teppo
January 14, 2011 at 9:06 pm
I hate that “sport.”
LikeLike
brayden king
January 14, 2011 at 11:55 pm
[…] The NBA as a Markov Process. […]
LikeLike
Links for January 17, 2011 « Marc F. Bellemare
January 18, 2011 at 11:05 am
You’re completely correct but it’s no reason to hate the game. Think of it as an excuse to not watch anything other than the last 10 minutes of any basketball game. In fact, since making the same observation couple years ago, I’ve basically taken to only watching the last 10 minutes of the championship games. It’s saves me tons of time and I get to see all the excitement that actually matters.
LikeLike
Benjamin Mako Hill
February 6, 2011 at 12:16 am