# LMS Popular Lecture Series 2010, Clutching at Random Straws, Matt Parker

Jun 06, 2021
that a king is something else, then the queen, can you ever flip through it and you're just bragging now, aren't you? Now she decided to shine not when she needed you, okay? Can you have it? Flip through and see if there isn't one card at a time. You just think you would have learned your lesson last time, right? Okay, can you take a look? Let me know if you found them next to each other for the next two. back, the next two and over here yeah, I'm going to say yeah, you did it right, three next to each other, give these guys a big round of applause if I say what are the chances if you shuffle a pack of cards in which the king ends up next to the queen, you'd think that would be a pretty unlikely event, but it just happened three times in a row and I was actually pretty lucky, in fact, side by side comes out at around 48% chance and I'm being vague for a reason that I will explain in a moment uh and next to everything, one aside it is more than 67.2% uh and I have heard arguments that it is around 74%, it is not as easy to work Do it as you might think, if you are boring, give it a try, it's a brilliant activity.
Start with the next one and you should get around 48% if you try to solve it.Logically, uh, and then if you master it, try a separate one. I've seen people argue that it should be around 74% in theory and I've seen people fake it and get out. about 67%, but both methods are more complicated than you expect, so please try it, no, now, some people don't, yes, you, with a pen, I conceived Marel, well, well, unless you are writing a note to the person next to you. You wouldn't believe what he's asking us to do now is um yeah maybe tonight or tomorrow night you're busy tonight maybe tomorrow night invite some friends over have a massive party yeah you've had a masquerade party , they are great.

## lms popular lecture series 2010 clutching at random straws matt parker...

It is not like this? So, we can talk later. Can? You get along very well with the group. Theory guy. I'll introduce them later, seriously. This is anyway. Whether or not they are at a math party, you can try it. and solve it and it's harder than you think, but it's certainly more likely than you expect. It's yet another case where our expectations are of a much lower probability than what actually happens again, partly because you underestimate the number of opportunities. It had to happen, there are a lot of PA pendings in a pack of cards and although it is a fairly unlikely event, there is a good chance that the same thing will happen with the birthday, there are so many pairings when we reach 30 people that even something quite improbable becomes very likely.
My all-time favorite example of this is this photo here. This is a photograph taken at Disneyland and Disneyland, it bothers me for several mathematical reasons, but one of them is: is there a single salt in the notion of scale, how can you have a giant version of a dwarf that is not part of that Does it make sense anyway? um the point the point in case here is this this character here with uh her arrow, not mine, is indicating uh Donna, this is a girl, a picture from when she was a child, now she's growing up and she was sitting with her fiancĂ© Alex and they were flipping through photos of their childhood because we all know that's a fascinating experience, uh and and they were looking at photos from their trip to Disneyland and Alex looked at, wait a minute, that guy in the background who looks like me dad and that's Alex, they didn't meet until decades after this photo was taken, but in their When they were kids, they were accidentally photographed together because they were at Disneyland on the same day and I think it's amazing, this ranks as one of my favorite photos of all time, I think it's absolutely phenomenal, and for the people involved, it's incredible.
These guys, I mean, what a thing to have, uh, and they checked that they had photos of Alex's trip and compared them, and his father was wearing exactly what he's wearing in the background of this photo, an incredibly unlikely event, but I would say that a despite being incredible. It's an unlikely case that's still to be expected just because of the sheer number of photographs we all appear in, if you think about how many times throughout your life, particularly now, you've been photographed with other

## random

people in the background and try to figure out how many

## random

people you've been photographed with, it's not that alarmingly unlikely, it's still incredibly unlikely, but not overwhelmingly so, and then if you think that applies to each of us, there are so many opportunities for one of us has been photographed with someone we later met later in life.
So, in my opinion, it's a bit like winning the lottery. It's amazing if it happens to you, but it shouldn't be a surprise. happens at all uh very improbable events will continue to happen given enough opportunities uh with enough opportunities they become inevitable and this is my second and a half point, there are things that we believe are really improbable, in fact, it is much more likely and many times we do not appreciate How many times has something had to happen. I don't want to downplay amazing things when they happen. These strange coincidences are amazing, but we shouldn't be surprised when they happen. happen, we should be more scared if they never happened because then we know that something strange is happening, but the fact that they happen periodically and if nothing strange ever happens throughout your life, you're a little unlucky if something surprising happens. that's your assignment, I guess you could say uh and and enjoy it, it's phenomenal, but that doesn't mean it's unexpected.
I think I've worked on that point enough. I'll work on it a little more later, please come see me if you need more work anyway um I actually wrote down important points that I wanted to get right here, this is kind of fun, uh, moving on to my third point, we want to see patterns like humans. , we thrive on patterns and I think that's great, I originally wanted to say that I'm not very good with biology and stuff like that, but you can see why evolving the ability to look for patterns and be hypersensitive to risk gathering is a very good tactic. of survival and the notion of ourselves that we detect. patterns that we round to the nearest pattern is what I'm trying to say here and it's great to have that and it's the reason we do math because we love detecting and working with patterns, that's why math exists in part, people you can come and argue with me later, uh, but I'll make you work, um, and I think that's great, but we forget how attuned we are to wanting to see patterns where there may be none, um, and one of my favorite examples of all the Times of this, uh, it is what it is. called back masking, which is where a band is recording an album and they're playing the songs and they think, guys, you know Satan is our Unholy Master, we should record some messages about him backwards and hide them in the song and this.
It was all the rage in the '70s and '80s and people used to I don't I don't want to I don't want to speculate who sits at home listening to album after album backwards hoping I'll find someone trying to brainwash the youth of uh, well, yesterday, um, but they're still going, so here's an example. This, uh, for those of you who aren't familiar, is uh, uh, Lady Gaga, her husband, Lord Gagar, is not in this photo. um and you guys should be ashamed for laughing at that I should make it very clear anyway uh uh she had a hit that had several hits and one of them was uh Paparazzi and if you're not familiar with that it says uh not like that I'm yours, it I'll do, baby, there's no other superstar, you know they'll all be you, dad, dad, okay, that's the song forward.
Now I'm going to play part of it backwards, see if you can detect what people think, this says, here we go. and suppose anyone wants to guess even a word they think might be in there Lucifer, all right I guess, babab, cute, you're just trying to outdo Trump on his AR syllable count, so you're saying maybe I, maybe time no one. I heard evil, save us the stars above above, we model it after the Arts of Lucifer, two points for you, uh, what I heard exactly, you heard exactly that where it might be better for you to go a different way out, are you a master a bit?
I'm kind of dabbling I'm not invited to school but I'm going I like you more and more anyway uh I'll play it again watch this and you'll hear this word for word are you ready? it's there you don't seem convinced no, but you can do it the second time. I'll mention it again. Now people are often not convinced how important the suggestion is. People think you might listen to it multiple times. You will start to notice it. or uh, it's a slightly different recording every time and I want to clarify all those notions. I want to show you that it is only the fact that you have been partial before you heard it and if you are partial, you will detect the words. if you're not you won't and to do that I'm going to use a different song this is uh Freddy um from Queen who was gay who knew him um uh you should be more embarrassed about that than the lady anyway uh quite a nice song a another uh inevitably will like uh the D says like this another dust Another dust Another bust do it if when you go home you can download it having paid legally on iTunes if you want um don't laugh And this guy is here saying: I will do it legally.
I'll play it backwards. If you think you hear what it is, don't say it, just see if you can figure it out and I will. Repeat the same thing over and over here we go, we're right, some people think they might have it. This is what we are going to do and this depends on you playing correctly. I'm going to divide the room in half, so down here. right in the middle there all this half here I want when I say don't go now you're going to close your eyes and this half I'm going to show you what he's saying so this half knows this half not all of you are actually the lucky side , be more amazing for yourselves, okay, ready, everyone on this side close your eyes now everyone on this side don't say it out loud, that's what he's saying, okay, this side open your eyes again.
I'm going to play it again if you can hear this raise your hand here we go can we hear that can you hear that here who can hear it on that side that means you can hear this all day these guys you guys are really F right now It's obvious today I think which is absolutely amazing that if you had that suggestion, you will hear it over and over again, if you haven't been prepared, you won't do it, and this comes up over and over again, yes, they tell you there's a pattern, you'll spot it, uh, and Now all of this starts to come together, so if you get enough random data, first of all it tells you that there will be a pattern that you will be looking for.
If you have enough data, you will find it and if you are not aware of the probability of it happening, you will think that it is surprising when in reality it is inevitable and it means that we constantly encounter people who will back up their arguments, uh, with with a pattern and with evidence which just doesn't hold up, but that's why we have science, that's why we have math, it's so that we can overcome our human intuition and get to the cause of the problem and really see if this is a significant finding or is it just noise? random, that's the beauty of math math exists separate from our brain math doesn't care how it works our brain exists but we can use it and we can learn to think in ways our brains normally wouldn't and we can use it to overcome this problem and for example, very quickly, I have a little bit of free time, there is a problem when you do surveys like food testing surveys and you don't know if the result is getting a significant or not, so if I wanted to decide between uh Coler a and Coler B sir, drink Cola B um, so you ask a group of people here to drink Cola, drink Cola B, tell us which one you prefer if one of them actually does it. knows better or whatever you are trying to find out, you should get more results saying Coler a over Coler b or whatever your research is, but in reality they are identical people who will still choose one or the other, so ask a caller . a caller B, they say that the caller a, you are very kind, you ask another caller, a caller B, they choose Coler a, you are doing well, it must be colay knows better or this is what It's called the null hypothesis, assuming there's absolutely nothing to it, it assumes both. the drinks are the same, suppose the medicine doesn't work, suppose there is no other planet around that star, whatever the zero case option is, assume that and see how long the result is likely to be now and if both colors they are identical, you can say well, if they choose perfectly at random, there is a 25% chance that you would have gotten this result and now that is not the case, there is a 25% chance that there is nothing to find, you say that if not there is nothing to find, there is a 25% chance that Getting these results is subtle, but that is the power of mass.
We can pin this down. It is generally accepted that if you reduce that probability to less than 5%, you will find something worthwhile that is greater than 5%. He needs to work a little more. more studies or maybe there is nothing and we can do these things properly, um, if people just select random locationsin the UK and makes triangles and picks the ones that fit and then publishes them in a potentially non-peer-reviewed journal. Send an email, then these things go straight into the public eye and this is where I'm going to go into my Call to Arms phase, which is why I ask for the close-up.
You'll notice that I also very nicely asked if they could. They arranged for me to have a piano on stage, which they did, which was going to be my closing number, but I don't think I'm going to have enough time to fit it in, but they were very good of them. and I said look, don't put it there, make sure you put up a protective barrier in case the crowd rushes forward, so thank you University of Birmingham, but I'm not going to use it. Where was I, oh yeah? Call to arms. Bear with me, if mathematicians don't talk and if scientists don't talk about what good science is and what good math is, then people will publish these crazy stories and it will seem like it's good because they will have evidence.
They're going to have data and unless people come out and say "well, actually this is how science works, this is how we do math," you can find these patterns or it's more likely than you think or it's because you were looking for it. , then the pseudoscience issue will continue and I am convinced that we need more people, mathematicians are prepared to go out into the media domain and do this, so, if I can receive a passionate plea, that is yes if we do not do physical education, people want to hear about scientific advances, they want to hear about good mathematics and if we don't do that, the crazy people will, so we have to go out and do it, and if the crazy people talk, we should say something and there is this guy that I I call it the fallacy of specialization in science, where scientists are so used to having their specialty and they don't talk outside of what's wonderful in science, but if you're the person who thinks oh, am I qualified enough to talk about this?
It's not my area. You are exactly the person. That thought process alone means you're someone who should be out there speaking up in defense of logic. rant about homeopathic remedies I'm not going to do now, but these things continue, if people are not prepared to say This is how we do science, this is how we know what works, it's a pattern that has no meaning behind it. that and I'm done with my rant that's it uh my opinions on homeopathic remedies are a

#### matt

er of public record um I'm being careful, I don't want this whole thing to go crazy Simon, sing me yes Anyway, I thought I'd end with one one of my favorite examples of all time: if you have enough data you can find whatever you want, actually some guessing, anyone want to guess what I haven't covered? what is the Bible, the Bible, if you write enough books, no, actually, you're right, no, he's right, it's the Bible Code.
I'd like to close to talk briefly about the Bible Code and the Bible Code says like this, you get part of the Bible. and you should do this in the original Hebrew or um uh not ancient Greek, New Testament Greek or whatever it was originally written, write it down, remove the spaces and I'll do it for the purposes of this uh in English, this is the NIV translation, this is First Kings 4:13-23 in case you are planning at home, and I'm not saying anything about the Bible as a religious text, I'm just saying if you take any book and remove the spaces, you end up with a grid of words.
Now, the way the Bible Code works is that you look at all these letters and ignore the original surface when reading the original text and you try to find hidden words by starting with one letter, jumping a certain amount to another letter, jumping a certain amount to the next letter and you spell hidden words, so I thought I'd check the Bible to see if there were any predictions about today because what people do is go back and say ah. if you skip every 17 letters, this spells a word and if this way another word and it was a prediction of someone's murder or a particular tragedy or whatever and they can, then all these words crossed and predicted that this was going to happen . happened so I checked this morning before I started if there's anything for today and some of you may have noticed that the word 20 is here and that's just in the text as such, the word 20 and it's strong, it's just a jump of one letter each. time, uh, but today is not the 20th, it's the 29th, so the word nine is up there.
N Jump. I jump and jump the same distance each time and because of the way I've wrapped the text, it means they line up because I have the same number of letters per line, so September 29th here we are. I got bored. So on September 29, everything came together in this part of the Bible in First Kings, chapter 4, and this must be today's part and and what's happening today, well, I'm here giving you a talk like it was, uh Dorothy Buck, and here's Matt M, it's 30 letters m, 30 letters, to 30 letters T, 30 letters, next T. Matt hits and some of you may have noticed.
Down here, Buck, here we go, so Matt and Buck are giving a talk about what you would want to know, so I took a closer look and Dorothy was talking about here, I can't, where is she? Let me see if I can. I can see it uh there it is there K jump one N Jump one or jump one t talking about a knot there you go and I'm talking about probabilities there we are probabilities that way I'm getting a little creepy now uh the more the more the OCD of you will have detected and I'm not even going to point this out, but up here, where is it? been math or mathematics or any other real word, it wouldn't have stood out well enough um and this is free, but it's exactly the same.
I was able to find today's date with the names of both speakers and the two topics we are talking about because there are so many letters for you, think about how many letters there are in the Bible and for each one of them, how many different jumps you can make from it , which becomes inevitable. I've done this talk several times and I can always find a prediction for I do the talk every time and it's just because of the sheer amount of data, so I want to finish with the three ways you end up thinking a pattern was detected, if you have enough data, you can find whatever you want. to find these patterns and we don't appreciate the likelihood that some of them are uh and I thought I'd just show you as a final point the predictive power of this because some of you are still looking for words that I can say uh and I'll save some of you the pain. to predict what will happen in the future because there is still something left from today who knows what could happen today we should check, well, the word drink appears there, this is more aimed at the organizers. and those of you over 18, but be patient, but drink what you're thinking, of course beer appears at the top so I hope it's inevitable.
I'm going to end up taking that e, taking that D in the middle exactly like an N so we have the word and there and it's there and it's there and and there and there and there and one of oh no, that's it there's actually 10 of them. where we are? So you should there uh there's six there's another four to find it if you get bored but I'm going to end with that you guys have been fantastic thank you very much we mentioned the evolutionary advantage of being able to detect patterns but surely there's also an evolutionary advantage of being able to calculate risks accurately.
Because? Are we that bad at that? The question was, I mentioned an evolutionary benefit of the screening pattern, but wouldn't there be an advantage in being able to calculate the risks? And I'm not an evolutionary biologist or a behaviorist or anything else. uh, but uh, from my unprofessional understanding of this, you want to overestimate your risk because there's nothing wrong with running away if there's no danger, but there's a big price to pay if you don't run away and there's danger and so we're programmed to overestimate the odds of something happening because a false positive is dramatically safer than being eaten by a mammoth that sentence became less correct as it went on.
I'll be honest, complete your own hilarious prehistoric animal, well, is it there? Is there any chance you actually play the piano? No, no, not only is it small, it's zero, that's it, um. I was wondering if you've heard of some sort of pattern that was detected in this text in the source uh on the computer called windings where you wrote it, you've heard of it, don't you have the source? Yeah, yeah, technically write the lyrics seeing how we're all being pedantic here tonight, yeah, and you wrote 911 and then the plane number CR okay, this was if you get the Wind Windings Font SL Mar, uh, and you mention it and what Microsoft thought would be a good way to have a lot of them, was like a precursor to clip art, so it was a typeface that was just characters instead of letters, uh, it.
It comes in as the second most annoying typeface, slightly behind Comic Sand, possibly the best titled typeface of all time. And if you put 911, it shows a plane flying in two rectangles. Some would say it's not in good taste. But the problem. That's after 9/11, after a tragedy of that magnitude, it's etched into the public consciousness, even people who have been born since then are aware of it and all the visual aspects, and the notion of it and everything that it entails. with that and if there's something out there that accidentally looks like that or seems like a coincidence, you'll find out that the same thing happened when Michael Jackson died, because people started filming everything around his house and sure enough, some footage turned up with a shadow that looks like Michael Jackson's ghost is because when something that big happens, people look everywhere for a strange part and I would be surprised if there was ever something on that scale and then something like that didn't appear.
The problem is when you see that because it was an email that Ford passed around a couple of years ago, you see it and you get scared but you never see the thousands of people who are at home and have gotten tired of listening. to backward records so they decide to look for other meaningless patterns and you don't see that you only see the surprising result a longer answer than required but I enjoyed it, there's that guy who seems too erudite for my taste, do you have? Is there something to be said about data mining and the supermarket and its purchases?
Do I have anything to say about data mining and supermarkets and shopping? And personally, maybe I don't think life is too short for loyalty cards to be completely honest and I think it is anyway, it's a bit of a false economy, no it's not, it's a tax on people like me who don't bother carrying one with them, but if you collect enough data and process it well, amazing things happen, many of our modern medical advances are due to our ability to collect a lot of meaningful data and process it in a meaningful way and get proper results, and Supermarkets aren't stupid, Amazon isn't stupid, Google isn't stupid, they have huge amounts of data and those are amazing things you can do. do really really strange things with it um so uh the Amazon book where it predicts a book that you might like becomes scarily accurate uh and it's because they're using the data correctly um personally, I have some uh Liberties hypothetical issues regarding how it's does but I like the concept and I think to be completely honest it's more beneficial than harmful, but that's yeah.
Divination is a good thing, but it can be used for evil, just like the Internet. I've heard that people need to get on the trains. some of you have houses to go to, so we may have to finish this pretty quickly, just a very quick and simple one, so would you play a lottery ticket with 1, two, 3, four, five, six? Would you play a lottery ticket with 1? 2 3 4 5 six uh, I play the lottery uh because a non-zero probability can still happen, no

#### matt

er how small, winning the lottery is still a little more likely than playing the piano, um, but the problem with 1 2 3 4 5 is that I forget how many people play that same number if you want to win the lottery and not share it with anyone else, you want to choose numbers that are greater than 31 because people like to choose birthdays and you want them to be distributed randomly and that's how it is .
It's harder than you think to do that, but if you win, then you have less chance of sharing it with anyone else, so I wouldn't play either. I often play with prime numbers, to be honest, if you want to know my Top or Secret Questions of someday. I'll hang around afterwards if people have questions about overly personal masks or anything we haven't been able to come see me for.