How playing poker can help you make decisions

Host: Nick Howe

Welcome back to the Nature Podcast. This week, a lesson in decision-making from poker.

Host: Shamini Bundell

And when peer review goes sour. I’m Shamini Bundell.

Host: Nick Howe

And I’m Nick Howe.

[Jingle]

Interviewer: Nick Howe

First up, how do you make a decision – go with your gut, toss a coin or simply roll the dice? Well, chances are that you do not think about your options probabilistically, weighing up the likelihoods of each outcome. Nothing personal, just, generally, humans are quite bad at thinking in this way. But there’s perhaps a game that could help you with that – poker. When it comes to playing the hand your dealt – making a decision in life – mathematician Jon Von Neumann thought that poker was perfect practice.

Interviewee: Maria Konnikova

He thought that poker was the perfect game to mirror strategic human decision-making because, like life, it’s a game of incomplete information.

Interviewer: Nick Howe

This is Maria Konnikova, a writer and former academic psychologist who’s written a book about poker and decision-making called The Biggest Bluff. After reading Von Neumann’s musings on poker and how it had helped him develop a mathematical model for human decision-making – also known as game theory – Maria wondered whether playing poker could help her understand uncertainty and make better decisions. For her, this was something particularly pressing as she’d been through a period of personal strife, with bereavement, job losses and health issues all affecting her and her family.

Interviewee: Maria Konnikova

And just all of these things happening within just weeks of each other made me realise that we just really overestimate often how much control we have over things, and I thought, ‘I want to figure this out. I want to try to dive into this further.’

Interviewer: Nick Howe

This wasn’t necessarily the first time Maria had been thinking about decision-making. In a previous life, she had researched decision-making at grad school. But with the extra push from her personal life, she decided to dive into this further again to better understand how to make decisions in an uncertain world. But this time, she took an unusual step. She took a year out to play professional poker, even though she didn’t know how many cards were in a deck.

Interviewee: Maria Konnikova

I knew nothing about poker. I mean zero.

Interviewer: Nick Howe

Despite this, after a lot of practice with professional players, Maria got into the pro poker scene. But at her first big tournaments, she did not do so well.

Interviewee: Maria Konnikova

I think you understated it somewhat. I mean I lost a lot of money at the beginning. I mean it was a very steep learning curve and the bottom of it was not pleasant.

Interviewer: Nick Howe

But after these setbacks, and with some psychology know-how and a dose of luck, within months, she was very successful. In fact, she won over US$84,000, beating out hundreds of competitors, at a prestigious poker tournament. But importantly, she found that poker really helped her think probabilistically. At every hand in poker, you have to estimate the odds of each person holding certain cards and weigh that up against your own hand. All the while, everyone is trying to deceive you. For Maria, hundreds of rounds of this intense probabilistic thinking helped her make better decisions. Whether it’s as simple as the decision to have breakfast or not or your future career choices, Maria thinks that if you want to learn to decide effectively, poker could be a useful tool.

Interviewee: Maria Konnikova

In fact, it’s not just a useful tool, it’s just genuinely the only tool I’ve found in my years of psychology training that teaches you to think probabilistically correctly. We learn much better from experience, so from actually doing something from physically experiencing it ourselves than we do from description, so when someone tells us something or says, ‘Oh, the probability of this is x or y.’ We’re bad at kind of reading that and internalising what it means, but if you force us to kind of go through something, we learn that way. Poker actually forces you to learn correctly. You’re learning through experience. You’re sampling, but you’re doing so over hundreds and thousands and tens of thousands of trials.

Interviewer: Nick Howe

This experience is mirrored by other poker plays too.

Interviewer: Liv Boeree

Poker is all about, ‘I think you’re bluffing in this situation, given everything I’ve seen, 80% of the time or 30% of the time, or whatever it is.’

Interviewer: Nick Howe

This is Liv Boeree. An astrophysics graduate turned professional poker play.

Interviewer: Liv Boeree

Because you’re making these kinds of decisions so frequently, where you’re having to kind of stick a probability on your predictions, you start to really build up a good database of when you’re correct and when you weren’t.

Interviewer: Nick Howe

According to Liv, this kind of repeated reinforcement of probabilities and their consequence can help people make better decisions in real life.

Interviewer: Liv Boeree

An example I often give is like I needed to park somewhere. I was running late for a meeting and I couldn’t find any parking spaces anywhere, and I knew I was only going to be in the meeting for half an hour. So, I thought, ‘Do I risk it and park on the double yellow line? Let’s do the expected value calculation. I think a traffic warden will come along maybe 10-20% of the time. The parking ticket it is £100. So, on the expectation, I’m going to lose about £15 here, would I pay £15 to not be late for this meeting? Yes, I would. Okay, fine, I’ll take the gamble.’

Interviewer: Nick Howe

So, whilst a fine may be £100, there’s only a small chance you’ll be hit with it, so realistically, if you take the risk, you’re only betting a proportion of that fine. A hard thing to get your head around, but for both Maria and Liv, poker can help, as it’s essentially like playing a sped-up simulation of life.

Interviewer: Liv Boeree

Poker is definitely a great analogue for life because it emulates the messiness of life. Whereas a game like chess, the best player always wins almost every single time. Whereas in poker, if I sat down against a complete beginner and we only pay 100 hands, I’ll probably only win like 51-52% of the time. It’s only if I played someone for like 10,000 hands or more that that edge actually starts to even get above like 90% likelihood that I’ll win. So, because of this sort of decoupling between results and actual quality of strategy, that is very much the case in life because life is very messy. There’s a lot of randomness. There’s a lot of unpredictable variables. Poker gives you an opportunity in a kind of more like – sounds silly – but low stakes, controlled environment to mess around with these confusing things that can make decision-making hard.

Interviewer: Nick Howe

Now, poker isn’t the only game to have randomness and messiness in it, but for Maria, it’s got enough elements of skill and luck to hit a sweet spot to emulate life. With enough practice, Maria thinks this life-emulator will allow people to make better decisions in an uncertain world and, like life, if you go all in, poker shows you the consequences.

Interviewee: Maria Konnikova

It’s funny, I talked to a man named Frank Lantz who designed games for a very long time, and he told me that poker is actually a really horrible game design from a modern standpoint because people think it’s rigged all the time because the probabilities don’t function normally, just it’s random and not controlled. And he said that in all modern games that you play on the computer, the designers actually screw with the random number generator because too many people complain when it’s truly random. What he loves about poker is that it doesn’t pander to you. It actually says, ‘Yeah, you think it’s rigged? Well, here you go. That’s how the deck played out.’ It doesn’t care that you’re supposed to win. There’s no such thing as ‘supposed to’. Probability does not have a memory. So, when you’re playing most games on the computer, they are actually rigged, so just know that because otherwise people complain and think, ‘Oh, well, this doesn’t make sense. I can’t believe I’m losing. This is a terrible game.’ Poker doesn’t do that. Poker just says, ‘Screw you. This is how the cards work.’

Interviewer: Nick Howe

That was Maria Konnikova. Her new book about her poker journey and understanding decision-making is called The Biggest Bluff: How I Learned to Pay Attention, Master Myself, and Win. You also heard from Liv Boeree who’s written a review of the book in this week’s Nature. We’ll pop a link to that in the show notes.

Host: Shamini Bundell

Later on, we’ll be having a discussion about how to keep things civil in the peer review process. I’m looking at you, reviewer two. Right now, though, it’s time for the Research Highlights, read this week by Dan Fox.

[Jingle]

Dan Fox

A newly developed synthetic skin can sweat droplets of fluid when stimulated by radio waves. The fluid is stored in numerous micrometre-sized pores surrounded by liquid crystal molecules which can move like a liquid, but when exposed to an electric field align themselves neatly like the atoms of a crystal. When the radio waves are turned on, the liquid crystal molecules twist, wringing liquid out of the pores in the process. The synthetic skin even reabsorbs excess fluid when the radio waves are turned off. Systems like this that secrete fluids on demand could be used to keep surfaces clean or to directly apply medicines, and the team behind the work have already demonstrated that the skin can release the painkiller ibuprofen. Read that research in full at Matter.

[Jingle]

Dan Fox

Mars may be the red planet, but its atmosphere glows green. Forty years ago, researchers predicted that Mars’ atmosphere should emit a green light, similar to the Northern and Southern lights here on Earth. But detecting this glow has proven tricky. Now, a team have spotted the elusive green hue using the European Space Agency’s Trace Gas Orbiter spacecraft to scan the planet’s edge against the dark background of space. The glow is given off by oxygen, which forms when radiation from the Sun breaks apart carbon dioxide in the Martian atmosphere. The researchers were able to measure the glow’s intensity in both visible and ultraviolet wavelengths, and calculate a ratio between the two. They suggest that this ratio should hold true for other planets, allowing other teams to calibrate their instruments when studying emitted light in future planetary research. Shine a light on that research at Nature Astronomy.

[Jingle]

Host: Shamini Bundell

Next up, disagreement, argument, critique – these are all things at the heart of scientific research and publishing. But there’s a fine line between being critical and just being rude. This week, reporter Anand Jagatia takes a look at what happens when things turn sour during peer review.

Interviewer: Anand Jagatia

There’s a longstanding joke in academia about the dreaded ‘Reviewer 2’. Reviewer 2 represents the absolute worst of the peer-review system. They don’t read the paper properly, they’re overly critical or often downright rude, and always have a huge list of unreasonable demands. There are Facebook pages and Twitter handles devoted to complaining about Reviewer 2, as well as countless memes – my personal favourite being, ‘Go back to the shadow, Reviewer 2.’ But even though these pages make light of the problem, there is a serious issue underlining them.

Interviewee: Mina Razzak

I’ll start by saying, the majority of our interactions with researchers are perfectly civil, very pleasant, very collaborative and constructive.

Interviewer: Anand Jagatia

This is Mina Razzak, the editorial director at Nature Reviews.

Interviewee: Mina Razzak

But I’ve handled, on my team, a few papers where some of the language and the discourse was just fraught with unhelpful language and unfair criticisms, and that was in the peer-review process from the referees towards the author, but also between the author and the editor.

Interviewer: Anand Jagatia

For Mina, this culminated in an incident with some authors after her team had taken the difficult decision to reject their review paper.

Interviewee: Mina Razzak

The angry email that came back from the authors was terrible. One author came back to us with a lot of vitriol about how we let them down and that they did a lot of work, and then another of the authors replied all with us still on the email chain, and essentially it just turned into the authors complaining about us and saying that we were the patron saints of the blind and we didn’t know what we’re doing. It was unnecessary. It was difficult to take. We thought it was unfair as well. And it doesn’t actually help science, so that experience and then the few others around that got me thinking about this, and that’s when I started talking to my colleagues to ask them, ‘Is this something that you’ve experienced? Is this happening more? What can we do about it?’

Interviewer: Anand Jagatia

After speaking with colleagues, Mina found that most editors had stories of similar experiences at some point in their careers, either with rude reviewers of with abrasive authors. But it isn’t something that’s talked about very much.

Interviewee: Mina Razzak

Part of me feels that we spend a lot of time as editors thinking about papers and how to promote science and get these papers across the finish line and into the public domain. We are champions of papers and we think a lot about author service, so when an author complains and in that kind of way, it feels like a failure. So, there may be some element of shame or ‘I didn’t do my job right’ and ‘I’ve made a mistake’ and I think that’s at least one of the reasons why we haven’t talked about this as an industry. So, what we ended up doing was we decided to do a little bit of an informal, small survey amongst our teams, and I guess the results kind of back up what I’ve been saying. The vast majority of our interactions are positive, but about a fifth of editors at Nature said that they had experienced some sort of unfriendly or hostile, sometimes abusive, language in the course of making a decision.

Interviewer: Anand Jagatia

This survey suggests that this kind of behaviour isn’t exactly uncommon. I reached out to Nature editor Karl Ziemelis to get his take.

Interviewee: Karl Ziemelis

We have to be mindful that reviewers are very busy. They’re usually doing the task of peer review on the side from their regular work. Some would say it’s a rewardless and thankless task as well. But it is part of the scientific discussion and really it should be kept at a civil and at a friendly level. But at the same time, it should be critical and challenging. That’s how science operates. But criticism isn’t necessarily hostile. It can be challenging without being rude. It’s one thing to draw attention to that problem and explain why that is a problem. It is another thing to say that the authors do not know what they’re talking about. That’s when you’re moving into personal hostility more than anything else.

Interviewer: Anand Jagatia

So, what’s the best way to deal with reviewers or authors who cross the line during the peer-review process?

Interviewee: Karl Ziemelis

I think that there probably are at least two things that can be done here. First is on the part of the editor. If they see such behaviour then I think working with the reviewer to edit the comments, to remove the potentially inflammatory remarks can only help because at the end of the day, you want the discussion between the reviewers and the authors to be a civil discussion. You want it to concentrate on the science. What you don’t want it to do is to descend into a war of words. I would also say there’s a responsibility on the part of the authors as well. Seeing some comments which either rightly or wrongly they’re interpreting as unduly hostile is not to respond in kind. Take the high ground. In addition to that, I think just making it as clear as possible to the reviewers that really, we want to keep this civil, and most reviewers fully understand that.

Interviewer: Anand Jagatia

One suggestion along these lines would be to create something like a code of conduct or a set of ethical guidelines which authors and reviewers are expected to stick to. But what about another approach of making peer-review reports more transparent? Nature policy is that reviewers do have the option of signing their reports if they feel comfortable doing so. So, would deanonymising or publishing those encourage people to play nice?

Interviewee: Mina Razzak

I think that the ideas behind open peer review are valid and they’re definitely worth exploring. I think that one of the criticisms or one of the concerns is that you may find very weak or diluted comments to avoid almost the opposite, to avoid offending and to avoid causing conflict. So, I think it’s something that we should study. I would like to see more on it, understand the behaviour of researchers. I think one of the things that editors can do is we should call that kind of behaviour out.

Interviewer: Anand Jagatia

Something that Mina and Karl both agree on is the importance of an open discussion about issues like these so that people feel like they can speak out if they have bad experiences, and to help foster a culture of respect that ultimately makes science a nicer, more inclusive place to work.

Interviewee: Karl Ziemelis

I think it’s absolutely important that such issues are discussed because only by doing so will we potentially raise awareness of what is and isn’t acceptable behaviour. Some may say it’s common sense but for some people, presumably, it isn’t common sense. But also, as individuals, to challenge ourselves. It’s all very well critiquing and challenging others but just taking a step back and looking at how we behave ourselves as individuals and saying, ‘Is this really appropriate?’

Host: Shamini Bundell

That was Nature editor Karl Ziemelis. You also heard from Mina Razzak, editor at Nature Reviews. Nature has put out an editorial on civility during peer review, and we’ll put a link to that in the show notes.

Host: Nick Howe

Finally on the show, it’s again time to look at some other non-corona science news highlighted in the Nature Briefing – that’s Nature’s daily pick of science news and stories. Shamini, what have you found this time?

Host: Shamini Bundell

So, there have been a couple of articles, both in Nature, this week about what researchers are doing to try and combat racism in policing. So, the first article is about how researchers are trying to and have been trying to collect data and use science to get some evidence-based policies to stop racism in the police force and, in particular, obviously, situations which end up with people dead or in injured.

Host: Nick Howe

And so, what does this data reveal? What sort of techniques can police use?

Host: Shamini Bundell

Well, actually, the main point of the story is kind of that there is a lack of this data and studies at the moment. So, particularly in recent years, so particularly since 2014 when Michael Brown was shot in Ferguson and Eric Garner was killed in New York, people have been sort of doing a lot more research, and there is data about the racial differences in if you are black versus white in America, you are more likely to be killed by police. If you get shot by police and you’re black, you are twice as likely to be unarmed. So, there’s definitely data showing the scope of the problem, but what there isn’t so much of yet is data that could suggest how to improve it, and one of the reasons for that is that different police departments providing data, that’s sort of voluntary. So, so far, it’s actually been difficult to get enough data and get a broad range of data.

Host: Nick Howe

Okay, so there may be a dearth of some of this data, but from the data we have, what are the results showing?

Host: Shamini Bundell

Well, so, one way to get around the lack of data is to use sort of sophisticated algorithms and models. So, one of the things people are doing is looking at sort of correlative factors. What situations end up with guns being fired or with excessive force being used? So, it might be looking at the race of the police officers called to a scene, and that seems to be correlated with if it’s a black or white neighbourhood. White officers in a black neighbourhood – those situations are more likely to end up in the officers firing their guns. And the other thing that was mentioned a lot was using body cameras. So, there’s some sort of conflicting evidence on this, but it seems like a consistent use of body cameras is something that can help. But the important conclusion is that we need to make sure that whatever policies are made, as with everything, that these are evidence-based policies.

Host: Nick Howe

Right, so I guess more data needed, but we have some insight into what we’ve got so far. And what was the second story? What did that talk about?

Host: Shamini Bundell

So, the second story is similar. It’s about a group of mathematicians in the United States, and they have written a letter basically saying that they and their mathematician colleagues should stop collaborating with police because of these issues of racism in law enforcement agencies – this is again in the US – and in particular, the kind of things that the mathematicians might be working on is things like predictive policing algorithms.

Host: Nick Howe

So, predictive policing algorithms, are these like in Minority Report where they’re trying to

predict where crimes are going to happen? What’s that about?

Host: Shamini Bundell

Yeah, that’s basically it. So, for example, in recent years, there have been algorithms developed to look at huge amounts of data and help police reduce crime by perhaps suggesting where a crime is most likely to occur and then you can put more resources into that particular area. But there’s sort of been debates going on already about how useful that is and about whether there could be actual biases in the data that these algorithms are already being fed. So, for example, if you’re looking at reports of drug crimes, police respond to a lot more drug crimes in black communities even though the rate of drug use is estimated to be the same in black and white communities. So, if the algorithm gets fed data that includes that then it might form a sort of self-fulfilling prophecy or a feedback loop where there’s more crime and therefore the police focus on that and therefore report more crime.

Host: Nick Howe

Oh right, so it’s almost doubling down on where crimes are. But I wonder, though, if mathematicians don’t get involved in this, is there a concern that these things will just be done anyway?

Host: Shamini Bundell

Yeah, so it’s tricky because you don’t want to say let’s not get science involved, but the worry that these mathematicians have is that their work is sort of lending this veneer of respectability and ‘Ah yes, it’s scientific’, and that’s kind of burying the actual racism that is at the core of many problems, which is what they don’t want to be a part of.

Host: Nick Howe

Are there any solutions for this?

Host: Shamini Bundell

Yeah, so several researchers quoted in this piece are concerned that both mathematicians and people from other fields do need to come together to make sure that these kinds of problems are countered and to make sure that researchers can engage with communities and can benefit in a positive way.

Host: Nick Howe

Well, hopefully researchers are able to come together to find some sort of solution to this, but for my story this week, scientists may have found a hint of dark matter.

Host: Shamini Bundell

Oh, have I heard this story before? Does this just happen every so often?

Host: Nick Howe

It definitely seems to be something that crops up every now and again, that scientists say they’ve found a bit of dark matter, and I did emphasise the ‘may’ there a little bit because it’s definitely not confirmed. But basically, there’s this huge experiment which has this giant vat of xenon, essentially, and this is a noble gas so it doesn’t react with much, and what you’re able to do because of that is monitor it to see if there’s any weird ripples in it to see if anything’s interacting with it, and that could be an indication of things like dark matter.

Host: Shamini Bundell

So, is there dark matter floating through our xenon tanks or is this some kind of faint signal of some distant dark matter?

Host: Nick Howe

So, what this might be is a hypothetical dark matter particle called an axion, so this has never been proven to exist but this could be good evidence for it. However, the researchers do also say there are two other possibilities. These ripples could be caused by a novel property of neutrinos, which is something that physicists would also get super excited about because it’s never been seen before or, and this is probably the scenario which is most likely, it could just be contamination.

Host: Shamini Bundell

I can’t believe you’ve let me get this excited and then said oh, but it’s probably just contamination. But they must think that there is a chance that this could be the real thing, right?

Host: Nick Howe

Yeah, I think they’re just being really cautious, as scientists often are, and they’re just waiting for confirmation. So, what they’re doing is they’re getting an even bigger xenon tank to monitor this to see if this is replicated, if they have these excess ripples again, which would be indicative of maybe this particle of dark matter or a novel property of neutrinos.

Host: Shamini Bundell

That is physicists’ answer to everything – build a bigger detector.

Host: Nick Howe

I mean when is that ever not a solution? Well, thanks again for chatting to me, Shamini, and listeners, we’ll put links to everything we discussed in the show notes. And if you’re interested in more but instead as an email delivered daily, then make sure you check out the Nature Briefing. We’ll put a link to that in the show notes as well.

Host: Shamini Bundell

That’s all for this week. If you want to get in touch with us then you can reach us on Twitter – we’re @NaturePodcast – or send us an email – we’re podcast@nature.com. I’m Shamini Bundell.

Host: Nick Howe

And I’m Nick Howe. See you next time.