Hoops Hypothesis

Hypothesis: The outcome of any pick-up basketball game depends more strongly on the match-up between the two worst players on each team than the match-up between the two best players on each team.

Argument: If the talent differential between the worst players is sufficiently large, then on defense, the better of the two is essentially free to double-team the other team’s best player on every single possession, while the reverse is not true. Thus, the team with the single worst player is effectively playing four-on-five against the team with the second-worst player, and it’s rare for the best player to be better than the second-best player by enough to overcome that numerical disadvantage.

Conclusion: This means that, when it comes time to pick teams, it’s more important to avoid getting the worst player on your team than to actively seek to get the best player on your team.

Discuss.

8 comments

  1. Thanks You! I forwarded this to my two guards (son and daughter). One’s in college, the other on a 6th grade feeder team, they both play sports, and into ChemE and NASA. I think they’ll get it and appreciate it too!

  2. Im Danish, so soccer for me. But i would disagree. Then there would be no reason for the big spread at pay day. But that’s in the pro world. After hours ball, i dont know basket to see if it’s true. But in soccer, hands down, it is better to have the best man.

  3. Mads Keller, if you are looking at how much players are worth, you must take into account the commercial value they represent. It’s not only about Beckham being better than Roberto Carlos (or vice versa), but also about Beckham being a… better brand. Yeah, it’s sad how soccer is commercialized these days.

  4. You just verified every non-athletic fifth-grader’s recease nightmare. Well done sir

  5. I must have played a different sort of pick-up game than you. In my experience (typically with a lot of enthusiastic but not
    particularly polished players) it’s the talent differential at
    the top one or two players. And if one team has a point guardish outside guy and a good athletic tall inside guy and they know how to play together, that’s pretty much it. Although I’m assuming 5 on 5 here, 3 on 3, or 2 on 2, number 3 or two is more important. But numbers 4 and 5 mostly do hustle stuff (rebounds, loose balls, garbage buckets). YMMV and all that.

  6. I wasn’t intending to crush anyone’s self-esteem– this is just an observation from the last couple of weeks of lunchtime games.

    I suppose the sample may be skewed by the fact that this is a regular game, and the players are always a subset of a limited number of people. We all know each other, and know each other’s game, so everybody in the gym knows which guys can be left completely alone on defense.

    I suspect that the same effect should hold for more random pick-up situations, though it might be masked by the lack of clear knowledge about who you can leave alone, and who can score on their own.

  7. I’d agree that the dropoff in talent in your typical pickup game doesn’t happen between the two best players, but I’m not sure it happens at the two worst players. The generalization would be that there’s a 50% chance of having an odd number of competent players on the court and the team with more competent players pretty much always wins.

Comments are closed.