#15 Hot Hand Fallacy: Why we love seeing hot streaks

You’ve just made your last three shots, won your last three hands or closed your last three deals. What do you think will happen next? In this episode, Mel and Dan explore the fallacy - and reality - of ‘catching fire’ in games, business and life.


Mel: 00:14 Hi, and welcome to Bad Decisions.

Dan: 00:21 The podcast that helps us understand why we choose what we choose.

Mel: 00:24 Why we think what we think.

Dan: 00:25 And how to exploit this stuff for fun and commercial gain.

Mel: 00:27 I'm Dr. Mel Weinberg, I'm a performance psychologist.

Dan: 00:30 And I'm Dan Monheit, co-founder of Hardhat, a creative agency built for today.

Mel: 00:34 You did it.

Dan: 00:34 I did it, now you do it Kops. [music playing]

Mel: 00:41 So Dan Monheit for this episode, I want to welcome you to Mel's Casino!

Dan: 00:44 Hey hey, it's great to be here.

Mel: 00:47 So here's the situation alright, we're flipping coins.

Dan: 00:50 This is a decidedly low-budget casino.

Mel: 00:53 This is a coin toss game, and I have just flipped the coin three times, and it has resulted in three heads. Your job, for a million dollars is to tell me whether you think that the next coin will be a head or a tail.

Dan: 01:10 So we've had heads, heads, heads?

Mel: 01:11 Yup.

Dan: 01:12 Due for a tail, I'm going with tails.

Mel: 01:13 You think you're due for a tails.

Dan: 01:14 Due for a tails.

Mel: 01:15 Willing to bet your life savings on it.

Dan: 01:19 I thought I was betting a million dollars that wasn't mine.

Mel: 01:22 Well we can go either.

Dan: 01:22 Either way.

Mel: 01:22 So you ready-

Dan: 01:26 Let's keep the show moving. I'm going with tails.

Mel: 01:26 Heads, heads, heads, tails. Alright. Okay so now I wanna switch things up a little-bit, and I want you to go in your crazy imagination, to the NBA playoffs.

Dan: 01:36 Okay, I'm there.

Mel: 01:37 You are, Dan Monheit. You are a fifty percent shooter from the field.

Dan: 01:42 Wait, I'm playing in the playoffs?

Mel: 01:43 Yeah.

Dan: 01:44 Oh, this is good. Okay yes.

Mel: 01:45 This is easy for you to imagine?

Dan: 01:46 Oh yeah.

Mel: 01:46 You're right there.

Dan: 01:47 Yes.

Mel: 01:47 Okay. You're a fifty percent shooter from the field.

Dan: 01:50 Yes.

Mel: 01:50 You've just hit your last three shots.

Dan: 01:50 That's good. Yes.

Mel: 01:52 Yes, it's NBA playoffs.

Dan: 01:52 So normally when I think about playing basketball I'm on the bench, but this is good. I'm in the game. I'm hitting shots, I've hit three in a row.

Mel: 02:02 Very realistic.

Dan: 02:03 Okay, so I've hit three in a row, feeling good.

Mel: 02:05 Alright, Coach Mel calls a timeout. Gonna draw up a play. We need a shot to win. You've just hit your last three. Should I put the ball in your hands?

Dan: 02:14 Gimme the ball, "get out of the way."

Mel: 02:17 Dan Monheit for the buzzer beater to win?

Dan: 02:17 Absolutely.

Mel: 02:18 You will back yourself every time?

Dan: 02:20 Oh I've just hit three in a row.

Mel: 02:21 Yeah, but you see, when we just flipped three heads in a row, you were pretty convinced that the fourth one was gonna be a tails.

Dan: 02:28 Because we're due for tails.

Mel: 02:29 We're due for tails, and this time?

Dan: 02:30 Yeah, well it's different.

Mel: 02:32 Okay, I can tell, why?

Dan: 02:33 I'm not a coin.

Mel: 02:34 Right, who are you?

Dan: 02:35 I've hit three in a row.

Mel: 02:35 You're super star bench player Dan Monheit right here.

Dan: 02:39 I'm no longer bench player.

Mel: 02:40 Sorry.

Dan: 02:41 Yes, this is my coming out party.

Mel: 02:42 So, what we're talking about here, what we're talking about today, what we've tried to just illustrate, is the idea of the Hot Hand Fallacy.

Dan: 02:50 This is not a fallacy. I actually had a hot hand. I hit three in a row, and I'm gonna hit the fourth.

Mel: 02:55 Remember this was your imagination, let's come back, let's bring it back into the podcast.

Dan: 02:58 Alright.

Mel: 02:59 The Hot Hand Fallacy, is the belief that we're going to be more successful in future attempts, if we've been successful in past attempts.

Dan: 03:06 I think NBA Jam has a lot to answer for, in this fallacy.

Mel: 03:10 Queue the audio.

Dan: 03:12 “He’s on fire. Razzle-Dazzle. It’s good!”

NBA Jam Audio: 03:12 Boom-Shaka-Lacka! Kaboom!

Dan: 03:18 So I tell you what Dr. Mel. I mean, I'm not a doctor but I don't know, I have some thoughts about this.

Mel: 03:23 Yeah, you're allowed.

Dan: 03:24 I'm allowed to have thoughts? With the fact that it's a fallacy, I totally get it for your first example, when we're tossing coins, or doing other completely random games, so maybe roulette or, something else like that. Where anybody that's done, you know, in maths, understands each event is completely independent and has nothing to do with the previous event and we kinda construct this story that, we're on fire. And I think, actually, casinos kinda know this and tap into it where at a roulette wheel they'll actually put up, what the last twenty spins or the last forty spins have been, to sort of help people construct a story.

Mel: 03:57 Yeah, that's not to actually give you information. It's actually to mess with your mind.

Dan: 03:59 Yeah, got it. So, you know that's a thing that we do, but, isn't sport different? In sport it's not completely random like if I have just hit three shots in a row, there's probably some things working in my favour. So like, my confidence is probably feeling better, my teammates might be looking for me, and maybe something's working against me as well. Like maybe the defence, trying to close out.

Mel: 04:19 100%.

Dan: 04:20 Or maybe taking stupid shots, cause, my confidence is up?

Mel: 04:22 Yeah well, your shot difficulty changes right?

Dan: 04:25 Yeah.

Mel: 04:25 With each one. What you're saying is that there's two different scenarios here. Right? One when we're talking about casino type stuff, mostly down to chance. Right? So the odds are fifty-fifty in those scenarios no matter what. Whereas, in the heat of a game, yeah there's so many other things that could influence whether or not you're going to make the shot.

Dan: 04:41 Exactly so with the casino stuff I'm sure there was research, but we don't really need research to know if that's a thing. We know it's completely random and we all know that we make up stories to tell ourselves that it's not cause we're all special snowflakes and we need to construct a narrative where the world makes sense around us.

Mel: 04:54 Well see that last bit, I'm not sure about being special snowflakes but we certainly do need to construct narratives how the world operate. Because we need to have meaning in our world it gives us a feeling of familiarity, predictability, it would be really uncomfortable for us to live in a world where we had no idea what was gonna happen next. So we tend to make patterns out of random things.

Dan: 05:13 Yeah, like how people see Jesus on toasted cheese sandwiches.

Mel: 05:16 Yeah.

Dan: 05:17 Yeah.

Mel: 05:18 We see faces. We see them in the clouds.

Dan: 05:20 Yes, and we're wired to do that, you know, even as a baby one of the first things you get to do is recognise what a face looks like, so, it makes sense we see patterns and repeated things everywhere we look.

Mel: 05:30 So, next time you see Jesus in the back of a tree, you're completely normal.

Dan: 05:34 Exactly, but with sport I guess it's, it's kinda different. Right? Like is there a fallacy or really is there a hot hand?

Mel: 05:41 Well so now we're going to enter ourselves into an argument that has been going for, about the last 35 years. Alright it started with a piece of research in 1985 by Gilovich, Vallone, and Tversky

Dan: 05:57 Did you say Tversky?

Mel: 05:59 Yeah.

Dan: 05:59 Because if there is a hot hand in research, this guy has it. He's basically on a Bad Decisions hot streak. He's been mentioned in every single episode.

Mel: 06:06 But here's the thing. When you get to be as good as Tversky was, you get to do really cool things with your research.

Dan: 06:11 Such as?

Mel: 06:12 Well, they spent their time, this is so totally different to my experience in Academia. Well it's a little-bit, cause I did watch a lot of basketball when I was in Academia, but they basically just got to look at and analyse the statistics of basketball games. So in 1985 they were looking at the shooting records of the Philadelphia 76ers.

Dan: 06:30 Jordan's rookie season.

Mel: 06:31 And what they were trying to do was look at whether shooters were more likely to make a shot, depending on the result of their previous shot.

Dan: 06:39 So they were trying to determine, if I just made a shot, am I more likely or less likely than average, to hit the following shot.

Mel: 06:45 Yep, and here's the thing, when they asked both the players, and the fans watching the game, if they were more likely to make the next shot after just hitting, the previous one, everybody was like, "Yeah, definitely more likely to."

Dan: 06:54 Yeah.

Mel: 06:55 Actually what the research found, was that there was no evidence for the influence or the suggested idea of a hot hand. The hot hand was a fallacy, and the hot hand didn't actually exist.

Dan: 07:09 Well that's upsetting.

Mel: 07:09 Isn't it?

Dan: 07:09 That's like, really upsetting.

Mel: 07:10 Sorry imaginary Dan in the NBA playoffs, apparently you are not a thing.

Dan: 07:14 I bet also, just like as a guy watching sport, you kinda like to think that this is happening right?

Mel: 07:19 Well yeah, it makes it entertaining, it makes it fun of course. The crowd goes nuts when somebody's hit three in a row.

Dan: 07:23 Exactly. So it may lead to the question I mean, you said this is a contentious issue, we’ve got a paper in 1985 that says no, his is absolutely not a thing ... surely it can’t end there?

Mel: 07:32 Well so then there was a whole-bunch of other studies that came out to replicate it, right? And then all of a sudden you got all these coaches going, "Na uh," and elite sports people going, "Na uh. I don't believe this. I don't care what your statistics show. I know that if I'm feeling hot, God damn it I am feeling hot!" Right? Now for 30 years this research stood, right? There was no such thing as a hot hand, in basketball or in sport.

Dan: 07:55 Well maybe not according to academics and statisticians.

Mel: 07:57 Right. According to everybody else and so a couple of researchers came out and they're like, "Not uh, this has got to be a thing and, we are gonna show you that it is a thing."

Dan: 08:04 "We're gonna keep researching this until we get it right."

Mel: 08:07 "And, we are going to keep watching basketball games and calling it research for as long as we possibly can."

Dan: 08:10 I see where you're going here.

Mel: 08:13 And so, in 2014 some researchers from Harvard, came out with a study.

Dan: 08:18 Well they're smart. They're smart right?

Mel: 08:18 Right, I mean they're from Harvard right? They came out with a study where they looked at 83,000 shots in the NBA.

Dan: 08:22 Tough gig.

Mel: 08:23 Yup, and they incorporated all these factors that weren't included in the earlier studies. So they were looking at factors that contribute to shot difficulty, like the things we were talking about like is another defender there, how far are you from the ring, etc. etc. What point in time is it in the game? And what they found was that there was, actually some evidence for the idea of a hot hand.

Dan: 08:41 Hey, see?

Mel: 08:43 If you look long enough. You'll find something.

Dan: 08:44 Exactly, you will find it’s in the data somewhere.

Mel: 08:47 So, currently, apparently, as it stands, the hot hand is not a fallacy. The hot hand is actually an effect.

Dan: 08:53 Right like what are we talking here. How much more likely am I to hit my 4th shot if I just hit the last 3.

Mel: 08:58 So here's the thing the effect sizes were marginal. I'm talking like an extra 1 or 2 percent. But look, if you think about in the contents of sport, like I always say in sports sometimes it's the smallest differences that have the biggest impact.

Dan: 09:12 I guess thinking back about what we were saying earlier about our desire to want to create stories around this stuff. Maybe the reality is you only have 1 or 2 percent difference in likelihood of hitting a shot but when you add on to that bit of extra swagger from the player, the crowd going off and cheering the person’s name because they just hit 3 in a row, the commentator starting to talk up that they have seen not seen a shooting streak like this at this count since December 1987 and this is one of the greatest performances likely to see. All sudden the story fits, even if the reality is questionable.

Mel: 09:44 Right the perception is way more fun.

Dan: 09:46 Yeah, and we want it right? We want the story. It's a much enjoyable experience to watch a game where there is actually momentum and hot streaks. Then when it’s just a bunch of random shot attempts.

Mel: 09:57 Right, so let's break this down in to a couple of reasons why we might fall for this. Okay, why we might think the Hot Hand Fallacy is a thing? In addition to it just being more fun to believe in. I will give you a couple other reasons why we tend to think this way. The first is the fundamental attribution era. Which is the idea when good things happen we attribute them to internal characteristics of ourselves. Right? So, when you've just hit 3 shots in a row you're thinking, "I'm hot, I'm on fire." The flip side of that when bad things happen we don't think that has anything to do with us. Right we’ll put things down like chance or bad luck.

Dan: 10:30 Exactly and so if we think about that example. Like in the casino where, you know if we are losing, oh you know, oh we get 3 heads in a row we think. We think we're due for tales..

Mel: 10:39 Yeahs, it's not because there's something wrong with us. It was bad luck.

Dan: 10:42 Exactly.

Mel: 10:43 We are due for a win.

Dan: 10:44 Due for a win. Totally.

Mel: 10:44 We're a good person. Good people deserve wins in life.

Dan: 10:47 Exactly, whereas if I just won 3 hands in a row. I'm probably due for a 4th.

Mel: 10:51 Of course. Because it was you.

Dan: 10:52 It was a skill. It was all skill.

Mel: 10:54 So, the next thing that comes in to place the idea of the law of small numbers. Basically, it's the idea when we know that something exist in the population. We assume it's gonna also apply in the small tiny little sub section of the populations. When we have small numbers. So, for example if we flip a coin 100 times we expect that we get 50 heads and 50 tails. Right? The law of small numbers said if I would of flip a coin 4 times. You would of expect to get what.

Dan: 11:22 2 heads and 2 tails.

Mel: 11:23 2 heads and 2 tails. Right, there’s just the same likelihood that we would have 4 heads in a row as if we would have head, head, tails, tails. That sequence “head head head head” is exactly the same likelihood happening as “head head tails tails” in a small sample.

Dan: 11:44 Right. So what you're saying about this law of small numbers, we failed to see the sample is merely part of a much much bigger experience. Much bigger population and we think that's the whole universe.

Mel: 11:50 Correct. If you go back to your hot streak in basketball. Right, you've just hit 3 out of 3 of your previous shots right, even though I know that you missed the last 20 before that and you're not really a great shooter. I'm looking at the little streak and I'm only looking at that and I'm neglecting all the information prior to that. I'm neglecting your overall shooting percentage which really, let's be honest isn't that great.

Dan: 12:12 High 40's probably.

Mel: 12:13 And I'm just looking at the small numbers. That little small instance that I’ve got there making my judgement  based on that.

Dan: 12:19 So you're not assessing my skill as a shooter over my lifetime. You're saying in the last 3 minutes this guy is a 100 percent shooter.

Mel: 12:27 Yep. Whenever we restrict the available range of data, we are going to make errors in estimation.

Dan: 12:32 Sure

Mel: 12:33 The last thing that comes into play is something that we have spoken about in previous podcast episode, which is the idea of availability basis. Which is if I want to think the likelihood of the next thing happening is, my brain is going to search for the most recent available information and “oh you hit your last 3 shots? Well of course you're going to make your next one.”

Dan: 12:48 Yeah and probably made shots more memorable than missed shots unless I missed a really critical shot or missed a layup or something like that. Chances are made shot that are memorable than missed shots. So that playing into that as well. So I guess we have all those things working in our favour or against our favour. It does seem pretty clear that we are excellent constructing stories where we want to. Maybe it's further evidence of that.

It seems funny that we only make the stories where we want them to. So we talk about hot streaks for shooting but we don't talk about hot streaks for other statistics like nobody gets on a steal hot streak or a block hot streak or a successful inbound pass hot streak.  

Mel: 13:24 And look that is one of the key criticisms of this sort of research when it comes to sport, that we only look at sort of one metric, one indicator of success and we make all of that judgements about future success based on that. Where as you say there's a whole-lot of other things that could come into play that we just don't pay attention to.  

Dan: 13:38 So the question for you Doctor Mel. The Hot Hand Fallacy is it a thing or is it not a thing?

Mel: 13:44 Look I think as long as we believe it's a thing. It's a thing.

Dan: 13:48 Right.

Mel: 13:48 Which is the point of the whole thing.

Dan: 13:50 So, Doctor Mel, the truth fairy. It is a thing or is it not a thing.

Mel: 13:52 Do you believe it's a thing?

Dan: 13:54 Well.

Mel: 13:54 If you think it, it is real.

Dan: 13:56 Okay, so the Hot Hand Fallacy is not a fallacy. The hot hand attribute is an attribute.

Mel: 14:01 According to the latest research there are some evidence for a hot hand effect.

Dan: 14:05 Alright, well I think we should go with that and I mean at the core of it, it’s because people like stories and a few consecutive shots made or few consecutive hands won at a poker table, or roulette wheel give us all the ingredients we need to construct a story about how wonderfully awesome we are.

Mel: 14:19 They also make good for Hollywood blockbuster movies. Right?

Dan: 14:22 This is true.

Mel: 14:22 About sports.

Dan: 14:23 Nobody's going to see a movie about a series of random shots that may or may not go in.

Mel: 14:28 So how do we take this and how do we apply this understanding of a Hot Hand Fallacy whether it exists or not. What does it mean in brand world?

Dan: 14:35 So if we look at that underlining ingredients for the Hot Hand Fallacy it’s people's desire to create a story and a story about somebody winning. And so, brands it makes sense to do whatever we can to look like we’re on a hot streak. Right. I guess often times with brands, and with agencies and other types of organisations we will look for big news and big wins to talk about our momentum moving forwards. But, probably a great lesson out of these is we should be looking for any sorts of wins we can talk about. Whatever that is individual performance reviews, or PR marketing we do out to the audience. The more made shots we can get out there the more we're contributing to the stories people are creating in their heads about our organisation or us as an individual being on a hot streak.

Mel: 15:20 Yeah, you're just trying to set the tone. Write the narrative.

Dan: 15:23 Exactly. So talking about this there are a couple of industries that come to mind straight away. Alright, maybe the umbrella over all of these is the awards industry, and how many awards are there that are set up just to give people good news to get themselves on hot streaks. So, I have not done the research maybe you would want to do the research but it would seem that there are more awards for cars available than there are cars, available in Australia. It seems that every car has won car of the year award for something. They've usually won it at least 4 or 5 times in a row. Total hot streak get on board with that.

I think superannuation is another one where you know, all of these different funds have been voted or recognised as a top something fund for the last x number of years and of course past performance is the best indicator for future performances, the fund is on a hot streak, make sure you get involved with that.

Mel: 16:10 See it’s funny that, right? Where past performance is the best indicator for future performance. It totally depends.

Dan: 16:15 Yeah who knows but I mean, actually it doesn't really matter because we like to believe that it is, and if you look at the ads, not picking on superannuation here but you will have the first 28 seconds of the ad talking about how amazing how their past performance been, and the last 2 seconds of the ad saying “past performance is not the best indicator for future performance” basically this whole ad was bullshit.

I should just say the ad is not bullshit. It's just the conclusions that we know people will draw from that are probably bullshit. The ads are true. Don't sue me.

Mel: 16:43 Alright so to bring it back and to really sum what we have been talking about. In circumstances where we don't have a whole lot of information or complete uncertainty. We have to understand our brain is going to look for patterns.

Dan: 16:55 For hot streaks.

Mel: 16:56 We're going to look for hot streaks. We're going to look for any information that can make us feel like we're in control of whatever is going to happen next but at the end of the day it's a story that our mind just made up.

Dan: 17:05 Right and clearly for completely random events, it’s 100 percent a story and for not completely randomly events, maybe sport, investment, buying and selling shares. It's like 98 percent a story. And 1-2 percent real.

Mel: 17:20 It's pretty much a story, our brain is making up stories.

Dan: 17:24 Alright it's all bullshit. It's a good way to end.

Dan: 17:26 No, actually you know what's a better way to end?

Mel: 17:28 What?

Dan: 17:29 How do people find you on the internet?

Mel: 17:30 Well they will either Google Melissa Weinberg or they can find me @DrMelW.

Dan: 17:36 Cool and if you finish talking to Mel and you would like to talk to me you can find me at @Danmonheit

Mel: 17:42 @Danmonheit #NBASuperstar.

Dan: 17:45 Yes. I think that is it for today. What are we going to talk about next time Mel? Are we going to keep the hot streak going?

Mel: 17:49 We are going to have to keep the hot streak going.

Dan: 17:51 Yeah I reckon we definitely talk about another bad decision making heuristic.

Mel: 17:54 It's a good chance. It's a good chance that will happen.

Dan: 17:57 Good chance. Seen you then.