Radiolab - Are We Coins?

Episode Date: June 30, 2009

After we released our show about Stochasticity, we received a lot of comments about the idea humans can be just as predictable as coins. In that show, Jonah Lehrer was telling us about a study on the 82-83 76ers, and he was saying that even when a basketball player is supposedly hot – really on a streak – he is no more likely to make his next shot that any other time. Basketball players are slaves to their averages. Well, it turns out this isn't the whole story.

Wait, you're listening. Okay. All right. All right. You're listening to Radio Lab. Radio Lab. Shorts. From W. N. Y.

C. C? Yes. And NPR. Hi, I'm Robert Crulwich. This is Radio Lab, the podcast. Jad is not here with me right now for a really, really nice reason. He just had a baby.

He and his wife. So he's sort of blest out at his house. And so that just leaves me here. But that's just the beginning of my woes. Not only am I alone, but I have to deal with all of you. I mean you, you listeners who keep writing in little niggly things that you think we kind of got wrong here and there like the last show. We did the statistics show.

And, well, it just happened that we had all our microphones on and everybody was poised. and then all of a sudden, yeah, this happens. Hello? Hey, Jed, this is Steve Strogatz calling. Oh, hey, Steve. How's it going? Well, you know what? I was just listening, and I have to tell you, this isn't quite right what you're saying about streaks.

Wait, what do you mean? I should back in here and explain to you what's going on. We said in our show that if you love sports and you watch an athlete or a team have a brilliant, miraculous, off-the-chart streak of some kind. the 82-83 Philadelphia 76ers who had an extraordinary time, in a great season. So during the playoff, the 76ers were all incredibly hot. Jonah Lera walked in and told us, if you heard the show, that they really aren't as miraculous as they seem,

and in fact, what they did was statistically predictable and not miraculous at all. That can't be right. Some of these percentages are pretty damning. Statistics professors counted every single shot, not just the excitement. just the exciting ones, but every shot the athlete took on that exciting night to 34% and found that not just some, but all great athletes in the end, pretty much perform more or less inside their lifetime averages. So if you hit the basket 60% of the time or 50% of the time,

yeah, you may have an exciting moment, but when you count it all through, you did what you normally do. So I leaned over to Jed, it was looking a little bit crestfallen. And I said, don't. The fact is, Jed, you are, and Kobe Bryant even, is more like a coin than any of us had dared to imagine. No. Kobe has a pattern. In his case, it's what, 60-40?

On any given night with Kobe, you think, oh, this is, he's spectacular. What all he's doing is he's just having another night of his very 60-40 life. And that's just the way it plays out. Even on a shot-by-shot basis, you're saying. Yeah. Each shot seems to be kind of a random event. Exactly. Are you willing to concede that statistically this is a...

Not yet. It's so counterintuitive. I still as a basketball fan, I was just watching a game the other night saying, pass it to Kobe assholes because he's clearly hot. The only exception to this, whole, whole literature of streakiness is... Hockey.

No, it's... The sport. No one cares about. Is it Joe DiMaggio's hitting streak where he hits for 42 games in a row? I've got in the book somewhere. Actually, Joe DiMaggio's hit his streak was 56 games, 56. Yes. So Joe DiMaggio is just about the only outlier you can find in professional sports.

He's the only real hero. Yep. Well, at least I got Joe. Where have you gone? Joe DiMaggio? Our nation turns its lonely eyes to you. Oh. Hey.

So that's where we left. Joe DiMaggio proves that really at the end of the day we are not coins, that some people can be truly great, like something from the Iliad or the Odyssey, you know, and do something 56 times in a row, and that could just not happen. That's what we told you. That's what we told you,

but that's not what Steve Strogett's things happened. Well, as a mathematician, I like to run the numbers on these kinds of things, and I once checked with a student of mine, Sam Arbusman, whether it really was that unlikely for someone in the history of baseball to have a streak as long as Joe DiMaggio's. And it turns out it's not really that improbable. What do you mean? It's not that you ran the... Explain, explain what you mean. Well, we did simulations of the entire history of baseball in the computer 10,000 times.

And here's what Steve did. He took every player in baseball history, including the ones who did very, very well, like your Babe Ruth's and your Thai cops, and the ones who did, you know, very poorly. With each player hitting according to their batting average in a given season, and you keep doing it 10,000 times and compile all the statistics. And then if you look at what's the longest streak in the history of baseball in one of those simulations, well, one out of every six times, you'll see a streak as long as Joe DiMaggio's streak or longer.

Really? So that means out of 10,000 totally imaginary baseball histories in 1,66666 cases, somebody produces a 56 game hitting streak or longer. People used to think it was like a million to one shot or something like that. It's not that. Oh, man. So you're telling me that Joe DiMaggio, too, is like some version of a coin. He's like a six-sided die, basically. And if you roll a one or a two or three or whatever,

that's the same chance as this streak happening. Yeah, that's right. The numbers are about the same as dice. If you win when you get a six, you know, you're not going to get a six every time, but once in a while you will get a six. That's about as often as you would see a streak as long as Joe DiMaggio.

So you're telling you that even the last guy standing, the last guy, the last hero, you're trying to take that from me now? Well, I mean, you could take Solowell. in this, that, I mean, it shouldn't, by all right, have been Joe DiMaggio, but it was. Huh. I'm not sure how to make sense of that. He is statistically predictable yet, yet high performing nonetheless? Yeah, he was, we know he was a tremendous player. He had a very high batting average. Obviously,

he was fantastic, but he's not usually the one who has the streak. It's not usually Joe DiMaggio. It's someone like Ty Cobb or Napoleon LeJoe or someone like that. I mean, DiMaggio is great, and there are a lot of great players in the history of baseball, but it's kind of surprising that he's the one who holds the record. Statistically, it shouldn't have been him. Also, his era, you know, he hit that streak was 1941. In our simulations, that was one of the most improbable times for the streak to have been set. So in a way, you could say DiMaggio really truly was remarkable,

because by all odds, it shouldn't have been him, and it shouldn't have been when he actually did it. So the fact that it was him and that he did it in 1941, you know, maybe there's something really special about the guy. Well, my consolation, my little piece of real estate here is shrinking, but I'll take it. Okay, I'll take that consolation. Well, let me ask you this. What has been, I can't imagine that baseball fans have been very pleased with your findings. That's right. Yes, we got slapped around quite a bit when our findings were first produced.

Was it just by mathematicians or did you get baseball fanatics? Oh, absolutely, it's baseball fanatics. The mathematicians are totally gullible like myself. Because for one thing, a lot of us don't know much about sports. You know, these are the kids that didn't really play baseball. So you're talking to a crowd of weaners to some extent. There were a few other things that have come out since then. You know, when you have to do these simulations, like you think in terms of flipping a coin

or doing what a statistician thinks of as independent events. That is, one game doesn't affect the results of the next game. That's also known to not be correct, and that was recently shown by a young baseball expert named Trent McConnor, who did a careful statistical check on is one game independent of the other games before it, and found that, no, they're really not. And this is something all baseball fans will believe,

that what's happening in a streak is, that players are doing something to keep the streak going. They're changing their style. They're changing, you know, which pitches they'll swing at, which ones they will let go, and they're trying to keep the streak alive in some way such that, in fact, they end up having longer streaks than they would have otherwise than their performance would have suggested statistically. And you're saying that that is in fact true, or that's just a sort of... That's real. That's real. That was only documented in the past few months by Trent McConnor. And so our assumption in our computer experiment that they were coins wrong.

Yay! Yeah, okay, that's good for you, right? That's good, because if you're saying that the psychology of the hitter can sway the length of the streak, well, then, yeah, the whole coin analogy is not quite right anymore. It's not quite right. That's the truth. And now, can I start crying a little bit? I mean, you were the crybaby before.

Now it has to be me, because this makes my life really bad. Now I don't know how to do the calculation. Really? That's so interesting. Yeah, this has become a much harder problem now. If the games are not independent, one to the next, all our statistical thinking becomes much, much more difficult. And the real truth as of this phone call is that we don't know what the odds are of Joe DiMaggio's streak. And no one does. It hasn't been figured out yet. Now, if we were to take the realization you just put forward that psychology matters, does it matter in hot hands?

as a whole in the same way it matters in Joe Amagea's specific case? It's not something I could really say in terms of evidence. My guess is it probably does matter, although if there is a psychological effect, it's been very hard to find it. But in the case of baseball with hitting streaks, it's really been clearly documented now. In other words, the players are aware of who they're playing on a given day, and if they've got a streak going, they're doing something to make the streak longer than it would have been, that were just in randomized game order.

And teasing out the role of statistics, which is sort of always there and it's rigid and it's hard to escape from. It's not everything. It's a big part of the story, but there is still room for psychology and for human voibbles. Oh, man, you made me so happy right now. You took me from the depths of depression to absolute elation. This is fantastic. So the reason Jed is so happy, not that you really need to know, because we know. We are not coins.

That's what we've just kind of proven, at least strongly suggested here. We are creatures of will. We are the authors of our own success because we have feelings, and the feelings lead to will. If you make the right move, if you say the right prayer, things can change, and you can change them. I mean, listen to Paul Simon. Paul, tell them what you know. And that's our lesson. with one unfortunate exception, which begins with the word but.

Do we really have control over our choices, over our destiny? I mean, do we really believe that? Well, I want you to join me here and play in a little game. Snick, Snick, Snick, Snuck. This is our reporter Gregory Warner and Paul Glimpshire, a neuroscientist at New York University. Okay, schnick, Snick, Snack, Snuck is like rock, paper, scissors, but there are four choices.

They are rock, paper, scissors, and well. When someone plays well, the scissors falls into the well, the rock falls into the well, but the paper covers the well. So always choose well? No, then I'm going to always choose paper in your doom. Ready? Oh, how you show it? The well is a little tube made with your hand. Okay, that kind of looks like a rock but open. Right. Ready? Schnick, schnick, snuck. Oh. Okay, you beat me. I played well because I thought there's no way it'd stick with paper again. Schenick, schnick, schnick. When you stand there and you go, ah, ah, paper. You do not actually at the instant of making that decision go,

oh, well, let's see. Last time, and the time before, and the time before, and the statistical likelihood of, and if I take the likelihood and derive the maximum, you know, you stand there, you go, ah, paper. I think that that, uh, experience is the randomness in your parietal cortex. Those cells, chuggling along until one of them randomly jumps up.

And that's the one you pick. To illustrate this point, Paul does a little experiment, which he showed Greg at his lab. He takes a monkey, puts the monkey in his seat, and gives the monkey a choice. Here, you can look at this target, it's red, or you can look at this one. It's green. This is a meaningless choice. Doesn't really matter. Red, green, it's all the same.

That's kind of the point. All the while, as the monkey's deciding where to look, Paul is listening to its brain. That's, in fact, the sound you're hearing, the sound of individual neurons inside the monkey's brain, chattering away, as it's about to make a decision. And this coming up is the sound of it actually deciding, right? There. Each time you hear that stuttering,

that's actually the monkey deciding to look at that target. It's him deciding to look straight down. When you actually hear the cell firing away, one of the things you may notice is it doesn't sound like a little clock going tick, tick, tick, tick, tick, tick, tick, tick, tick, tick, tick. It sounds more like... You can actually hear the randomness in the way it's behaving.

And we think that's not an accident. That, we believe, is the randomness that makes us unpredictable when we're making choices. One of the things I hear all the time is, often from scientists, is why would you think there's randomness in the brain? That seems crazy. You're a material scientist. Surely what you should believe is that everything in the brain

measurable and predictable. Well, I think the answer to that is you don't actually want in real life to always be predictable. People have to be unpredictable if they are to survive in the natural world. I actually think that the experience we have of actually choosing in these hard hang fire choices is not the experience of, well, he's going to and I'm going to and he's going to. You can feel this hanging moment. This pregnant pause. We both did paper. And the decision is constructed by the randomness in your brain.

But that's not how it feels to us. And in fact, that's not even how it feels to you. Because when you said, oh, I thought you were going to do well, I never thought you'd do paper a third time. You were kind of making... But I think we make up that story afterwards. You made up that story? Yeah, I think we all make up those stories afterwards. Like, I try and give meaning to my choices, just like you.

you know, gosh, why did I pick paper? I picked favor because I thought you were going to pick well. That's what I thought. But, you know, is there really compelling evidence that at the moment I had to pick I was thinking about that? Yeah, not so compelling. Well, you know better because you've looked in the brain. It doesn't look that way in the brain.

But randomness is kind of a nasty thing, right? It's sort of the last thing you can't measure. And what we're arguing is sometimes it's there on purpose. Now that's an aphanima to the traditional scientific method where what we do is try and make our measurements more and more and more and more accurate. And then the stuff we don't understand, the stuff that's unpredictable will go away.

What we're saying is it'll never go away. Special thanks to Gregory Warner, who I know is dreaming of making the American schnick-snack team. They're having their prelims in Waco, Texas coming up, and I don't know that he's going to make it, but all our best to Gregory Warner. Thanks for the peace. Who else I got? Michael. Jazz is not here. I can't remember anything. Who else do I have to thank?

Oh, Steve Strogatz. Yes, Steve Strogatz, for bringing us out of the darkness into the light. Thank you, Steve. And it's always nice to say something special about Alfred P. Sloan, who's got this foundation that helps pay for us, and to the National Science Foundation, and to the Corporation for Public Broadcasting. And Jad will be back, but, you know, oh, well.