Kobe Theory: Adventures in Distorted Probability

[Doctor MJ]

With all the talk about Kobe's clutchness recently, I've written an article on the subject. I see it as a counterpart to what I described in "Chamberlain Theory". Where "Chamberlain" talks about the silliness of overly confident statisticians, "Kobe" talks about the silliness of overly confident human beings who don't put enough stock in statistics. Here's a snippet:

The root problem though is that human beings aren’t good at working with gradually accumulated large numbers without technological aids of some sort. People thus have a distorted sense of the probability that Kobe’s going to make that tough shot, because they don’t accurately remember how often he’s made similar shots in the past, compared to how often he’s missed similar shots.

And this is also why it absolutely makes sense for someone like Dwyer to say, “No, I don’t care if someone else is a millionaire because of his basketball knowledge, he’s wrong”. Because the flaw here is not one that is basketball-based at all. It’s just a human problem, and so we should expect that anyone analyzing the situation without the use technological tools is going to have problems with this. When NBA GM’s join the general chorus and exult Kobe his clutch abilities then, if they are being honest, they are unwittingly telling us that they have not properly used the tools that all humans need in order to properly analyze the situation.

http://asubstituteforwar.wordpress.com/ ... obability/

[ElGee]

Anyone who thinks humans do a good job judging probability or baseline rates of large-scale events
(a) hasn't studied humans or
(b) hasn't studied probability

I've always wondered if we took realgmers, casual fans and NBA execs/HOFers and had them watch 100 games -- in order to see each team a few times -- without seeing the scoreboard ever, and asked them to rank, the top 30 offenses and top 30 defenses and estimate points/per possession
(a) how far off they would be?
(b) if any of the 3 groups would be noticeably different in their accuracy?

It's a fun exercise - try and watch a game without looking at the scoreboard (or keeping score in your head) and see if you can guess the score at any given time.

[mysticbb]

Very good read about it, Doctor MJ.

@Elgee

Your idea reminds of me thinking sometimes by watching the game highlights that somehow the wrong team won. Well, sometimes the announcer and the highlights are giving the impression that the loser actually played better than the winner. It is a good example of a selective memory, imho. While the makers of the highlight videos are doing it on purpose, the human brain is somehow also filtering out the events which really led to the victory of the other team. Players, which produced some highlights, are more forgiven for their lapses on defense or even offense at the end. The same thing works here, game winners are huge highlights and the people are not remembering the misses very well.
They are also lack the complete informations, because they only watch a limited amount of games. Thus the player who produced the most highlights (in that case Bryant) is thought to be the best during those situations.

[Doctor MJ]

Excellent points guys, and I appreciate the kind words.

I sometimes just marvel that the difference between teams is often so small in basketball. In football, a great team regularly doubles the scoring of a bad team over the course of a season, so one can understand viewers feeling like they see one team clearly out performing the other. Literally, one team is more than 100% more likely than another to score at any given time.

In basketball the difference is that difference often below 10%. If two did the equally good for 9 possessions, and then one took the edge on the 10th, that would represent the difference - except that that ease of reading one happened doesn't happen because there is so much more randomness involved. I don't really buy that any human can grasp that difference simply by watching.

This helps explain why there are so many "bad habits" in basketball. What can a scout do but talk about what capabilities a player shows? How is a player supposed to realize he's behaving improperly when the difference between "good" and "bad" behavior is so small?

I'd like to see the project ElGee mentions.

I agree with mystic that highlights are killers. If you really follow the narrative the host attaches to plays, it becomes comical. A single slam dunk represents a momentum shift of epic proportions, and a stupid shot that happens to go in cements a legend.

[ElGee]

Excellent points guys, and I appreciate the kind words.

I sometimes just marvel that the difference between teams is often so small in basketball. In football, a great team regularly doubles the scoring of a bad team over the course of a season, so one can understand viewers feeling like they see one team clearly out performing the other. Literally, one team is more than 100% more likely than another to score at any given time.

In basketball the difference is that difference often below 10%. If two did the equally good for 9 possessions, and then one took the edge on the 10th, that would represent the difference - except that that ease of reading one happened doesn't happen because there is so much more randomness involved. I don't really buy that any human can grasp that difference simply by watching.

This helps explain why there are so many "bad habits" in basketball. What can a scout do but talk about what capabilities a player shows? How is a player supposed to realize he's behaving improperly when the difference between "good" and "bad" behavior is so small?

I'd like to see the project ElGee mentions.

I agree with mystic that highlights are killers. If you really follow the narrative the host attaches to plays, it becomes comical. A single slam dunk represents a momentum shift of epic proportions, and a stupid shot that happens to go in cements a legend.

I'd love to run that experiment. I'd need some volunteers, and I'm not entirely sure the NBA people I know would be willing to do it.

Anyway, here's my tangential post on this topic: http://elgee35.wordpress.com/2011/02/01 ... ch-debate/

I've added some 4th Q numbers from the FInals and noted a change in Bryant's play (and as a result, team performance) in that period, which may also be what happens at the very end of these games in which LA trails. Of course, people (will) just ignore that information and call him King...

[GreenHat]

I've always wondered if we took realgmers, casual fans and NBA execs/HOFers and had them watch 100 games -- in order to see each team a few times -- without seeing the scoreboard ever, and asked them to rank, the top 30 offenses and top 30 defenses and estimate points/per possession
(a) how far off they would be?
(b) if any of the 3 groups would be noticeably different in their accuracy?

It's a fun exercise - try and watch a game without looking at the scoreboard (or keeping score in your head) and see if you can guess the score at any given time.

Every group would be way off. The differences are too small, a couple of points per 100 possessions for someone to notice with great accuracy other than the extremes (ie Boston is really good at D and Toronto is really bad at it, Lakers are really good at offense and the Cavs are horrible at it). I would say at best knowledgeable people who were good with numbers could get 3 or 4 of the top and bottom five teams but then they would get a lot wrong in the middle 20. Also ranking 30 of anything is very hard for humans. Think about how long it would take and how inaccurate it would be if someone gave you 30 names and you had to alphabetize them in your head. In that example the ranking is explicit and its still hard for someone to rank 30 things.

Even though everyone would be way off it would be amazing to see how off people would be and which things specifically people are the furthest off from.

If instead of the generic "realgmers" you used the statistically inclined on Realgm I would take them to outperform even the NBA Execs/HoFers

[Doctor MJ]

Wrote a bit of a follow up piece after an LA Times article by Bill Plaschke.

http://asubstituteforwar.wordpress.com/ ... ned-plant/

What set me off:

The Lakers are near the top of the league in rebounding but are only 15th in the league in field goal percentage in the fourth quarter of games they trail. Kobe needs help closing, and Anthony gives him that help. The Lakers' offense needs a second option outside, and Anthony can take that shot. The Lakers don't shoot as well as their biggest rivals, and Anthony would fix that.

As per usual I talk about a variety of things, but the "Drowned Plant" refers to what happens when an amateur gardener kills a plant by overwatering ("Still not looking good, better give it more water").

Appreciate your feedback,
Cheers

[Gongxi]

And this is also why it absolutely makes sense for someone like Dwyer to say, “No, I don’t care if someone else is a millionaire because of his basketball knowledge, he’s wrong”. Because the flaw here is not one that is basketball-based at all. It’s just a human problem, and so we should expect that anyone analyzing the situation without the use technological tools is going to have problems with this. When NBA GM’s join the general chorus and exult Kobe his clutch abilities then, if they are being honest, they are unwittingly telling us that they have not properly used the tools that all humans need in order to properly analyze the situation.

I know I'm late to the party but: Exactly. So much misunderstanding in not just basketball, but all sports, and not just sports, but business and politics come down to these types of cognitive mistakes that aren't even related to the topic at hand.

Mlodinow's The Drunkard's Walk most applicably applies to this phenomenon as it relates to statistics, but others touch on it these things too: Kathryn Schultz, Carol Tavris, Robert Burton, and to an even lesser extent Gladwell and the Freakanomics guys.

It's a serious problem. Think of the children!