1 00:00:01,000 --> 00:00:05,000 There are some things in life that you just don't see coming. 2 00:00:05,000 --> 00:00:08,000 The world is random, unpredictable. 3 00:00:09,200 --> 00:00:12,160 Unlikely things can happen. 4 00:00:12,160 --> 00:00:16,040 So, are you blessed with good fortune? 5 00:00:16,040 --> 00:00:17,440 Or are you... 6 00:00:18,480 --> 00:00:21,000 ..one of the unlucky ones? 7 00:00:21,000 --> 00:00:24,400 But maybe there is a way through this mess of chance. 8 00:00:24,400 --> 00:00:29,840 Maybe the secret to being lucky is in trusting the maths. 9 00:00:56,240 --> 00:00:59,400 APPLAUSE 10 00:01:07,480 --> 00:01:10,520 CHEERING 11 00:01:12,720 --> 00:01:14,960 Good evening, everyone! 12 00:01:14,960 --> 00:01:16,520 My name is Hannah Fry. 13 00:01:16,520 --> 00:01:18,800 Welcome to the Royal Institution Christmas Lectures. 14 00:01:18,800 --> 00:01:20,840 And let's start as we mean to go on. 15 00:01:20,840 --> 00:01:24,720 So, who wants to be our first volunteer for the evening? 16 00:01:24,720 --> 00:01:27,200 Let's get you, there. If you want to come down. In the pink jumper. 17 00:01:27,200 --> 00:01:29,960 Round of applause as she comes to the stage! Thank you! 18 00:01:32,480 --> 00:01:34,360 What's your name? Megan. Megan. 19 00:01:34,360 --> 00:01:37,080 OK, Megan, how are you feeling today? You feeling lucky? Yeah. 20 00:01:37,080 --> 00:01:39,320 OK, good. All right. If you just want to come here, 21 00:01:39,320 --> 00:01:42,360 just stand just there. Now, Megan, above your head... 22 00:01:42,360 --> 00:01:44,320 AUDIENCE LAUGHS NERVOUSLY 23 00:01:44,320 --> 00:01:47,480 ..we've got a bag of gunge and there's just a little rope, 24 00:01:47,480 --> 00:01:49,720 just there. What do you reckon we're going to do? Megan? 25 00:01:49,720 --> 00:01:52,280 For some reason, I feel like it's going to fall on me! 26 00:01:52,280 --> 00:01:54,840 Well, we thought about it. We thought about it quite hard. 27 00:01:54,840 --> 00:01:56,520 But actually, it's not going to fall on you. 28 00:01:56,520 --> 00:01:59,080 You're going to be the one who's going up there to help us 29 00:01:59,080 --> 00:02:02,400 cut the ropes, so if you want to nip up there with Fran, just there. 30 00:02:02,400 --> 00:02:08,360 Now, instead, I am going to stand maybe not quite underneath it, 31 00:02:08,360 --> 00:02:11,920 but what we've got here, the bag of gunge is attached to a rope. 32 00:02:11,920 --> 00:02:15,160 And just at the other end of the rope there, there's a tiny, little 33 00:02:15,160 --> 00:02:16,840 weight that's balancing it there. 34 00:02:16,840 --> 00:02:19,720 And once Megan cuts the rope, the gunge is going to fall 35 00:02:19,720 --> 00:02:20,880 down this way. 36 00:02:20,880 --> 00:02:24,920 And then, either I've done my sums correctly, in which case the gunge 37 00:02:24,920 --> 00:02:28,680 is going to stop once it hits about here, or it's going to drop 38 00:02:28,680 --> 00:02:30,960 all the way to the floor, go splat, 39 00:02:30,960 --> 00:02:34,600 and then everyone in the first row is going to be covered in gunge. 40 00:02:34,600 --> 00:02:36,400 OK? Definitely one of those two things. 41 00:02:36,400 --> 00:02:39,640 Now, if we manage to get through this unscathed, I want 42 00:02:39,640 --> 00:02:43,280 you all to burst into spontaneous applause. 43 00:02:43,280 --> 00:02:46,560 And if this doesn't work, I'm just going to do the rest 44 00:02:46,560 --> 00:02:48,600 of the lectures covered in gunge. OK? 45 00:02:48,600 --> 00:02:50,760 All right? Sound good? All right, Megan, how are you doing? 46 00:02:50,760 --> 00:02:54,480 Yeah! OK. You ready to cut rope? 47 00:02:54,480 --> 00:02:57,440 Are you ready? Do want to give us a countdown? OK. 48 00:02:57,440 --> 00:03:00,360 Erm, five, four... 49 00:03:00,360 --> 00:03:04,080 Three, two, one... ALL: Two, one... 50 00:03:04,080 --> 00:03:06,440 Whoa! 51 00:03:06,440 --> 00:03:07,640 Ha-ha! 52 00:03:07,640 --> 00:03:10,120 APPLAUSE 53 00:03:13,880 --> 00:03:15,680 Thank you very much, Megan. 54 00:03:15,680 --> 00:03:18,440 You haven't ruined the lectures, so well done for that. 55 00:03:18,440 --> 00:03:20,280 That was great. OK. So what happened there? 56 00:03:20,280 --> 00:03:24,440 We had a bag of real gunge there, 57 00:03:24,440 --> 00:03:27,120 which genuinely almost did, 58 00:03:27,120 --> 00:03:29,840 in fact, splat all over us. 59 00:03:29,840 --> 00:03:31,360 Oh, no. 60 00:03:31,360 --> 00:03:32,560 LAUGHTER 61 00:03:32,560 --> 00:03:33,880 It's fine. 62 00:03:33,880 --> 00:03:36,320 Something quite special happened there, something quite 63 00:03:36,320 --> 00:03:38,560 extraordinary there happened, to stop that gunge 64 00:03:38,560 --> 00:03:41,120 from splatting everywhere. If you want to watch this back 65 00:03:41,120 --> 00:03:45,720 in a little slow-mo. So the weight was just enough to make the rope 66 00:03:45,720 --> 00:03:47,400 swing around that bar 67 00:03:47,400 --> 00:03:51,360 and wrap around enough times to cause the friction to stop 68 00:03:51,360 --> 00:03:53,640 the gunge from dropping. 69 00:03:53,640 --> 00:03:55,840 We've got another angle of this. 70 00:03:55,840 --> 00:03:57,320 So here's the second angle. 71 00:03:57,320 --> 00:04:01,120 Now you think it's going to drop all the way down. Just the last minute, 72 00:04:01,120 --> 00:04:04,120 there's just enough there for it to wrap around itself. 73 00:04:04,120 --> 00:04:06,800 Unwrapped itself there for a second. But we were safe, we were OK. 74 00:04:06,800 --> 00:04:10,280 We didn't get splatted. Now, here's a slightly smaller version here to 75 00:04:10,280 --> 00:04:11,600 explain what happened. 76 00:04:11,600 --> 00:04:17,280 Now, we calculated that if this weight here is 14 times lighter 77 00:04:17,280 --> 00:04:22,240 than the item on the other end, then that is enough to make sure 78 00:04:22,240 --> 00:04:24,880 that it would always stop itself from falling. 79 00:04:24,880 --> 00:04:26,880 Now, I knew that for sure, of course. You all knew, 80 00:04:26,880 --> 00:04:28,680 surely, that I knew that for sure. 81 00:04:28,680 --> 00:04:32,640 But still, you know, it's inevitable that you feel a little bit worried. 82 00:04:32,640 --> 00:04:35,200 And that's, I think, is something that you have to accept. 83 00:04:35,200 --> 00:04:37,520 You know, we're all humans. We are... 84 00:04:37,520 --> 00:04:39,920 We find it hard to override our instincts. 85 00:04:39,920 --> 00:04:42,640 We're not built for calm and rational thought. 86 00:04:42,640 --> 00:04:46,000 But that is the reason why we created mathematics. It's a way 87 00:04:46,000 --> 00:04:50,240 to step outside of ourselves and be objective, a way to calmly 88 00:04:50,240 --> 00:04:53,880 calculate something and be sure of the answer, rather than just rely 89 00:04:53,880 --> 00:04:56,560 on what our messy minds might tell us. 90 00:04:56,560 --> 00:05:00,400 And you know what? I'm a very big believer that maths can offer you 91 00:05:00,400 --> 00:05:01,600 a new way of looking 92 00:05:01,600 --> 00:05:05,440 at almost anything, because I think if you take the time to look, 93 00:05:05,440 --> 00:05:06,640 thank you very much, 94 00:05:06,640 --> 00:05:09,760 there are mathematical patterns hiding behind almost everything, 95 00:05:09,760 --> 00:05:15,280 even things that feel like they are very far away from being 96 00:05:15,280 --> 00:05:18,520 mathematical, even things that you think should be completely random. 97 00:05:18,520 --> 00:05:21,240 Now, we're going to come back to this one in a moment. 98 00:05:21,240 --> 00:05:24,480 But in the meantime, my fellow mathematician, Matt Parker, has got 99 00:05:24,480 --> 00:05:27,640 a perfect example for us, haven't you, Matt? Hello. Yes. 100 00:05:27,640 --> 00:05:31,120 I've got a huge group of people out here in the entranceway to the RI. 101 00:05:31,120 --> 00:05:33,720 If you'd like to come out and join me. Sure thing. 102 00:05:33,720 --> 00:05:36,000 So, thank you very much, everyone who came along. 103 00:05:36,000 --> 00:05:39,200 Bizarrely, all wearing red hats. That's fantastic. 104 00:05:39,200 --> 00:05:41,240 And all of you wearing yellow. Well, how fortunate! 105 00:05:41,240 --> 00:05:43,400 This must be the new fashion. 106 00:05:43,400 --> 00:05:47,320 So we're going to try an experiment with all of you in a moment 107 00:05:47,320 --> 00:05:49,720 to see what happens when you're moving through a crowd. 108 00:05:49,720 --> 00:05:52,000 OK. When you're moving in a crowd, 109 00:05:52,000 --> 00:05:54,480 I mean, everyone's making their own decisions. 110 00:05:54,480 --> 00:05:55,720 It's random, surely? 111 00:05:55,720 --> 00:05:57,840 It feels very random, very chaotic, unorganised. 112 00:05:57,840 --> 00:05:59,160 You're being bounced around. 113 00:05:59,160 --> 00:06:02,720 However, there are some patterns and, actually, we... 114 00:06:02,720 --> 00:06:05,360 WHISPERING: ..want them to walk naturally, so we're not going to 115 00:06:05,360 --> 00:06:07,920 tell them when the experiment starts. OK. Watch this. 116 00:06:07,920 --> 00:06:10,480 CONVERSATIONAL VOLUME: Actually, I'm really sorry, everyone. 117 00:06:10,480 --> 00:06:13,200 We've just realised, don't move yet, that we wanted the yellow hats 118 00:06:13,200 --> 00:06:16,720 on this side. We wanted, I know, the red hats on... It's not me, 119 00:06:16,720 --> 00:06:19,680 it's coming from upstairs. So I think we're clear. OK. If you can. 120 00:06:19,680 --> 00:06:22,600 Quick as you can. If you just swap over sides, exactly where you 121 00:06:22,600 --> 00:06:25,040 are but opposite across. OK, so you can see here, 122 00:06:25,040 --> 00:06:27,360 we've got an overhead camera rigged up so we can see. 123 00:06:27,360 --> 00:06:28,960 Come on, everyone, as quick as you can. 124 00:06:28,960 --> 00:06:30,840 Hurry up. Come on, faster, please. 125 00:06:30,840 --> 00:06:32,800 We've got a TV show to do. Here we go. 126 00:06:32,800 --> 00:06:35,720 So... They're not all jumbled up. It's not random, is it? No. 127 00:06:35,720 --> 00:06:37,600 Right. Look at this. So, because everyone... 128 00:06:37,600 --> 00:06:40,440 Having some mild issues here. Everyone's just following each other 129 00:06:40,440 --> 00:06:42,080 and you end up with these very clear lines. 130 00:06:42,080 --> 00:06:43,960 So if these were random, it'd be a complete mix. 131 00:06:43,960 --> 00:06:46,040 Instead, we've got these fantastic stripes. 132 00:06:46,040 --> 00:06:48,360 Come on, everyone. Let's keep it moving. 133 00:06:48,360 --> 00:06:51,120 You end up with these fantastic stripes. 134 00:06:51,120 --> 00:06:54,360 So, everyone. That was fantastic. Thank you so much. 135 00:06:54,360 --> 00:06:56,800 And that was the experiment. 136 00:06:56,800 --> 00:06:58,640 So, thanks very much for getting involved. 137 00:06:58,640 --> 00:07:00,040 Thank you. Thank you, Matt. 138 00:07:00,040 --> 00:07:03,120 Now, it turns out there are actually a lot of similarities 139 00:07:03,120 --> 00:07:05,160 between the maths of how people... 140 00:07:05,160 --> 00:07:07,080 AUDIENCE LAUGHTER 141 00:07:07,080 --> 00:07:10,160 Be quiet down there! ..how people flow through a corridor 142 00:07:10,160 --> 00:07:12,440 and how fluids behave. 143 00:07:12,440 --> 00:07:16,080 And if you can understand how people move when they're in a crowd, 144 00:07:16,080 --> 00:07:18,200 you can predict how they'll react 145 00:07:18,200 --> 00:07:20,960 in the case of an emergency and that's something that's incredibly 146 00:07:20,960 --> 00:07:23,680 important if you are designing buildings 147 00:07:23,680 --> 00:07:25,720 or stadiums or train stations, 148 00:07:25,720 --> 00:07:29,920 to make sure that your design is as safe as it could possibly be. 149 00:07:29,920 --> 00:07:33,600 And that, really, that is the point. That is what maths is all about. 150 00:07:33,600 --> 00:07:36,120 It's all about discovering those invisible patterns 151 00:07:36,120 --> 00:07:39,720 like this beautiful pattern left here by that swinging pendulum. 152 00:07:39,720 --> 00:07:44,800 Isn't that gorgeous? Much neater, I think, than you would expect. 153 00:07:44,800 --> 00:07:47,520 But, you know, maths isn't just about spotting patterns. 154 00:07:47,520 --> 00:07:50,000 It's also about using them to your advantage. 155 00:07:50,000 --> 00:07:54,280 So here is something that might look as though it's random. 156 00:07:54,280 --> 00:07:56,520 Best part of Christmas, obviously. Who gets the toy? 157 00:07:56,520 --> 00:07:58,360 Who gets to wear the hat? 158 00:07:58,360 --> 00:08:00,760 Now, here you go, if you want to have a go on this, 159 00:08:00,760 --> 00:08:03,040 because you only do this, go for it... 160 00:08:03,040 --> 00:08:05,560 ..a few times a year, 161 00:08:05,560 --> 00:08:07,840 you may not notice, there we go, you can have that one... 162 00:08:07,840 --> 00:08:11,000 ..you may not notice that there is, actually, a pattern to this. 163 00:08:11,000 --> 00:08:13,440 But luckily, I have done hundreds of these. 164 00:08:13,440 --> 00:08:16,640 And there is a knack to winning. But I can explain to you. 165 00:08:16,640 --> 00:08:19,280 So what you have to do - I'll teach you, right, then I can show you. 166 00:08:19,280 --> 00:08:21,920 So what you want to do, you want to hold your end lower than the other 167 00:08:21,920 --> 00:08:25,440 person, use two hands with a really sort of steady, firm hold, 168 00:08:25,440 --> 00:08:28,560 and you don't want to do any twisting or pulling. 169 00:08:28,560 --> 00:08:30,720 You don't want to really tear this. 170 00:08:30,720 --> 00:08:33,160 So what you're trying to do, ultimately, you're trying to let 171 00:08:33,160 --> 00:08:36,040 the other person do all of the work so that then... 172 00:08:36,040 --> 00:08:38,360 Ahh! ..you end up winning. 173 00:08:38,360 --> 00:08:40,600 Well done! You can have the hat, now. Enjoy! 174 00:08:42,800 --> 00:08:44,840 Now, I've got to be straight with you. 175 00:08:44,840 --> 00:08:46,520 This isn't going to work every single time. 176 00:08:46,520 --> 00:08:49,040 Especially not if your opponent knows the same trick as you 177 00:08:49,040 --> 00:08:51,320 and you're just steadily trying to get lower and lower 178 00:08:51,320 --> 00:08:52,400 than the other person. 179 00:08:52,400 --> 00:08:55,280 But the maths of prediction isn't about saying what's definitely going 180 00:08:55,280 --> 00:08:58,280 to happen. It's about considering all of the possible outcomes and 181 00:08:58,280 --> 00:09:00,120 working out what's likely to happen. 182 00:09:00,120 --> 00:09:01,960 So, to show you what I'm talking about here, 183 00:09:01,960 --> 00:09:04,760 join me in welcoming University Challenge mathematician 184 00:09:04,760 --> 00:09:07,480 and schoolteacher Bobby Seagull. 185 00:09:09,840 --> 00:09:13,680 Hey, Bobby! It's very good to see you! 186 00:09:13,680 --> 00:09:16,000 APPLAUSE DROWNS HANNAH'S SPEECH 187 00:09:18,600 --> 00:09:20,360 What have you been doing there, Bobby? 188 00:09:20,360 --> 00:09:24,080 So I've been flipping coins hundreds of times. 189 00:09:24,080 --> 00:09:28,320 OK. On a sort of fair, unbiased coin. 190 00:09:28,320 --> 00:09:31,560 You'd expect the chance of a head or tail to be 50/50. Yeah. 191 00:09:31,560 --> 00:09:35,600 So I thought, let me try and test this by flipping a coin six times 192 00:09:35,600 --> 00:09:38,840 in a row. So you'd expect, six times in a row you'd expect three heads, 193 00:09:38,840 --> 00:09:40,240 three tails, right? Yes. 194 00:09:40,240 --> 00:09:44,240 But sometimes things don't quite work out the way we want them to. 195 00:09:44,240 --> 00:09:47,080 So, have you been flipping a coin six times in a row, 196 00:09:47,080 --> 00:09:48,240 many times in a row? 197 00:09:48,240 --> 00:09:51,720 Yes, Mr Seagull's a true maths hero teacher, flipping it 2,000 times. 198 00:09:51,720 --> 00:09:54,760 No, no, no. I had to enlist the help of school students. 199 00:09:54,760 --> 00:09:56,320 Nice. So we went across London, 200 00:09:56,320 --> 00:09:59,160 found students from Westminster Academy... Thank you. 201 00:09:59,160 --> 00:10:01,640 ..and students from my very own school, Little Ilford. 202 00:10:01,640 --> 00:10:04,520 And that's these guys here, right? Yes, exactly. These students here. 203 00:10:04,520 --> 00:10:07,960 So we got them to line up, all with a coin in their hand, and flipping 204 00:10:07,960 --> 00:10:10,320 the coin six times. OK. 205 00:10:10,320 --> 00:10:14,040 If they get a heads, they take one step in that direction. 206 00:10:14,040 --> 00:10:16,920 And can you guess what happens when they flip a tail? 207 00:10:16,920 --> 00:10:18,960 I'm going to go this direction. You're right. 208 00:10:18,960 --> 00:10:21,800 And let's see what happens when they flip the coins. 209 00:10:21,800 --> 00:10:25,320 So this is like a visual representation of the six flips. 210 00:10:25,320 --> 00:10:27,600 OK, so that's the third flip there, is it? 211 00:10:27,600 --> 00:10:28,840 Yep. And that's the fourth. 212 00:10:28,840 --> 00:10:31,280 You can see wind blowing the ties about. 213 00:10:31,280 --> 00:10:34,840 So, spreading out, every extra flip, they're spreading out more and more. 214 00:10:34,840 --> 00:10:36,080 Exactly. 215 00:10:36,080 --> 00:10:39,760 And now, if we get the students to move down towards the graph... 216 00:10:39,760 --> 00:10:41,080 Oh, it's like a human bar chart! 217 00:10:41,080 --> 00:10:43,680 ..you start to see a pattern emerging. 218 00:10:43,680 --> 00:10:46,560 Ah, those are very good. So there's way more people in the middle there. 219 00:10:46,560 --> 00:10:50,000 Yes. And roughly the same number on the tails and the head side. 220 00:10:50,000 --> 00:10:53,520 And a few have got six tails and six heads. 221 00:10:53,520 --> 00:10:56,320 So these are three heads, three tails on this side. Exactly. 222 00:10:56,320 --> 00:10:59,720 All here, I mean, these guys must have thought they were very lucky. 223 00:10:59,720 --> 00:11:01,560 And down here, all tails. Exactly. 224 00:11:01,560 --> 00:11:03,680 But you didn't just do this once either, did you? 225 00:11:03,680 --> 00:11:06,840 I mean, this is like nested, lots of levels to it. 226 00:11:06,840 --> 00:11:10,360 Exactly. So many classes repeating it many times 227 00:11:10,360 --> 00:11:12,760 to get us a better dataset. OK. 228 00:11:12,760 --> 00:11:15,280 Here we see, this is Westminster Academy students. 229 00:11:15,280 --> 00:11:18,960 So, I mean, pretty much every single time we're seeing most people 230 00:11:18,960 --> 00:11:21,320 being in the middle. But you always get these bits at the end. 231 00:11:21,320 --> 00:11:22,920 You always get the extremities. 232 00:11:22,920 --> 00:11:26,760 But the beauty comes when we try and combine all the information. 233 00:11:26,760 --> 00:11:30,400 So here we see all these little dots, they're a student, 234 00:11:30,400 --> 00:11:32,760 a proud, hardworking maths student. 235 00:11:32,760 --> 00:11:34,640 And you put them together and we get what looks 236 00:11:34,640 --> 00:11:36,720 like a normal distribution curve. Ahh! 237 00:11:36,720 --> 00:11:39,560 And this is what the maths would have predicted from the beginning. 238 00:11:39,560 --> 00:11:41,600 Exactly. Very lovely. Thank you very much. 239 00:11:41,600 --> 00:11:43,000 Bobby Seagull. Thank you! 240 00:11:43,000 --> 00:11:44,520 APPLAUSE 241 00:11:46,160 --> 00:11:47,960 The point that we're trying to make here 242 00:11:47,960 --> 00:11:50,000 is that life is full of randomness. 243 00:11:50,000 --> 00:11:53,440 But even in the midst of chance, maths can still tell you what might 244 00:11:53,440 --> 00:11:56,000 happen and also tell you just how likely that might be. 245 00:11:56,000 --> 00:11:59,080 Which, of course, is precisely what predictions are all about. 246 00:11:59,080 --> 00:12:01,280 So, OK. Let's take a prediction about the future 247 00:12:01,280 --> 00:12:03,120 that we are all very familiar with - 248 00:12:03,120 --> 00:12:05,520 what the weather is going to be like tomorrow. 249 00:12:05,520 --> 00:12:08,720 So I got this from my phone earlier. This is my weather app. 250 00:12:08,720 --> 00:12:13,760 And this here, it says that tomorrow the chance of rain is 20%. 251 00:12:13,760 --> 00:12:18,800 Question is, what does that 20% actually mean? When it comes 252 00:12:18,800 --> 00:12:23,160 to talking about the chance of rain, what does that 20% actually mean? 253 00:12:23,160 --> 00:12:25,960 Who wants to kick us off? You want to make a guess as to what you think 254 00:12:25,960 --> 00:12:27,640 20% means in that context? 255 00:12:27,640 --> 00:12:31,440 The chances are probably 20%, it's going to cover. 256 00:12:31,440 --> 00:12:33,400 OK, yeah, 20% is going to cover. 257 00:12:33,400 --> 00:12:37,320 So you think it's about 20% of the space will get covered in rain? 258 00:12:37,320 --> 00:12:40,680 What do you reckon? 20% of what? Do you want to give me a guess? 259 00:12:40,680 --> 00:12:43,240 Anyone here want to give me a guess? Yeah, go ahead. 260 00:12:43,240 --> 00:12:45,680 How much of the country is going to have rain in it? 261 00:12:45,680 --> 00:12:47,640 OK, good. Good response. OK. OK. 262 00:12:47,640 --> 00:12:49,360 Well, let's ask someone who really knows. 263 00:12:49,360 --> 00:12:53,160 I want to introduce you to someone who, as you will find out, their 264 00:12:53,160 --> 00:12:55,880 life depends on correctly predicting the future. 265 00:12:55,880 --> 00:12:59,360 I'd like you to join me in welcoming Professor Chris Jackson. 266 00:13:02,640 --> 00:13:05,520 Hey, Chris. How are you doing? Hi. Good, thank you. 267 00:13:06,480 --> 00:13:08,480 APPLAUSE DROWNS HANNAH'S SPEECH 268 00:13:09,800 --> 00:13:12,800 So how did they do? In terms of understanding that 20%, how did 269 00:13:12,800 --> 00:13:15,320 our audience do? They did pretty well, yeah. They did very well. 270 00:13:15,320 --> 00:13:16,440 What does it actually mean? 271 00:13:16,440 --> 00:13:17,760 It means that if I was to live 272 00:13:17,760 --> 00:13:20,120 the same day over and over 100 times, it would rain 273 00:13:20,120 --> 00:13:21,160 on 20 of those days. 274 00:13:21,160 --> 00:13:24,200 Does it tell you whether you should bring an umbrella or not, though? 275 00:13:24,200 --> 00:13:26,600 I'm a pessimist. You should always bring an umbrella. 276 00:13:26,600 --> 00:13:29,080 Yeah. I'm the same. No-one likes getting wet, right? No. 277 00:13:29,080 --> 00:13:32,080 No-one likes getting wet. But you don't predict the weather, do you? 278 00:13:32,080 --> 00:13:34,480 No, I don't. Something much more risky. 279 00:13:34,480 --> 00:13:36,920 Yeah. I try and understand when, how volcanoes behave 280 00:13:36,920 --> 00:13:38,280 and when they might erupt. 281 00:13:38,280 --> 00:13:40,840 And you use those predictions to actually go in volcanoes? 282 00:13:40,840 --> 00:13:43,240 Yes, I do. And we've got some of your photos here. 283 00:13:43,240 --> 00:13:45,360 I mean, that's pretty close, right? 284 00:13:45,360 --> 00:13:47,840 Yeah, it's pretty close. But it's important work we're doing 285 00:13:47,840 --> 00:13:50,880 as scientists. You know, we handle the risk. 286 00:13:50,880 --> 00:13:53,320 And so you're calculating this risk at all times, are you? 287 00:13:53,320 --> 00:13:56,600 Yeah. We have visual observations on the volcano, how active it is. 288 00:13:56,600 --> 00:13:58,720 But also, we have an idea of what the likelihood is 289 00:13:58,720 --> 00:14:00,880 that this volcano may erupt while we're there. 290 00:14:00,880 --> 00:14:02,480 Have you ever had a close call? 291 00:14:02,480 --> 00:14:04,600 I've never had a close call. Not yet, at least. 292 00:14:04,600 --> 00:14:05,680 OK. All right. 293 00:14:05,680 --> 00:14:08,200 Well, of course, because this is the Christmas lectures, 294 00:14:08,200 --> 00:14:10,560 we've got a model of a volcano to demonstrate. 295 00:14:10,560 --> 00:14:12,600 So you can tell us a bit about how your predictions 296 00:14:12,600 --> 00:14:14,440 work, using this thing here. OK. 297 00:14:14,440 --> 00:14:17,760 Yeah. So, the thing with volcanoes is they're very difficult to see 298 00:14:17,760 --> 00:14:18,840 inside and underneath. 299 00:14:18,840 --> 00:14:21,920 So we have to rely on a number of observations while we're looking 300 00:14:21,920 --> 00:14:23,480 at the volcano. 301 00:14:23,480 --> 00:14:25,920 So one thing we can use is analysis 302 00:14:25,920 --> 00:14:28,880 of the gas that comes out of the volcano. 303 00:14:28,880 --> 00:14:32,360 So by looking at the amount of gas and the type of gas, because gas 304 00:14:32,360 --> 00:14:33,840 is contained in magma, 305 00:14:33,840 --> 00:14:36,280 if we measure that, we can have an idea if magma's moving 306 00:14:36,280 --> 00:14:37,880 into the volcano and the volcano 307 00:14:37,880 --> 00:14:39,240 might be about to erupt. 308 00:14:39,240 --> 00:14:42,400 And so if you see this kind of gas, you're calculating the risk 309 00:14:42,400 --> 00:14:44,240 of an eruption all the time? Yes, exactly. 310 00:14:44,240 --> 00:14:46,920 How much gas and what type of gas may actually kind of give us 311 00:14:46,920 --> 00:14:50,040 an indication of what the nearness of an eruption is, yeah. 312 00:14:50,040 --> 00:14:52,240 And are there other things that you're looking out for? 313 00:14:52,240 --> 00:14:53,960 Yeah. So another thing we can look for is, 314 00:14:53,960 --> 00:14:56,400 magma moves into the volcano, it pushes against the rock, 315 00:14:56,400 --> 00:14:59,400 the rock fractures and releases energy in the form of earthquakes. 316 00:14:59,400 --> 00:15:02,280 So, if we can measure the earthquakes, where they are in the 317 00:15:02,280 --> 00:15:04,920 volcano, then we can actually measure the earthquakes in terms 318 00:15:04,920 --> 00:15:07,120 of how strong they are as well. 319 00:15:07,120 --> 00:15:09,600 With the gases together, we may be able to use that in some 320 00:15:09,600 --> 00:15:12,600 sort of like forecasting sense, so, what's the likelihood? 321 00:15:12,600 --> 00:15:16,640 Are there some signs that are sort of absolute no-nos, where you won't 322 00:15:16,640 --> 00:15:18,960 enter a volcano after that point? 323 00:15:18,960 --> 00:15:22,840 Yeah, I think if, you know, there's lots of gas, lots of earthquakes, 324 00:15:22,840 --> 00:15:25,880 maybe even lava coming out the top of it, that's probably time 325 00:15:25,880 --> 00:15:28,280 to leave! Time to... 326 00:15:28,280 --> 00:15:30,720 Time to take a bit of a step back. Time to step back, yes. 327 00:15:30,720 --> 00:15:32,320 Maybe let's do that! Let's do that. 328 00:15:32,320 --> 00:15:34,800 So this one looks like it's imminent. OK? 329 00:15:34,800 --> 00:15:36,360 NERVOUS LAUGHTER 330 00:15:36,360 --> 00:15:38,960 Thank goodness you've got big glasses on then, huh? 331 00:15:38,960 --> 00:15:41,640 Have you ever been at an actual volcanic eruption? 332 00:15:41,640 --> 00:15:43,960 I have. Yes. Because this stuff is really serious, right? 333 00:15:43,960 --> 00:15:46,160 This is not just playful... Absolutely serious. 334 00:15:46,160 --> 00:15:48,600 It's serious for the people who work in and around volcanoes. 335 00:15:48,600 --> 00:15:49,760 But it's even more serious 336 00:15:49,760 --> 00:15:52,080 for the people who have to live with them every single day. 337 00:15:52,080 --> 00:15:53,560 Yeah. 338 00:15:53,560 --> 00:15:55,280 So this one looks like it's imminent. 339 00:15:55,280 --> 00:15:56,720 This one looks like it's imminent. 340 00:15:56,720 --> 00:15:59,880 I'm going to say, you know, probably 90% chance at the moment 341 00:15:59,880 --> 00:16:01,200 it's going to erupt. OK. 342 00:16:01,200 --> 00:16:02,400 NERVOUS LAUGHTER 343 00:16:02,400 --> 00:16:03,440 Whoa! 344 00:16:03,440 --> 00:16:04,600 AUDIENCE GASPS 345 00:16:04,600 --> 00:16:07,360 That's 100% chance! Yeah. Impressive! 346 00:16:07,360 --> 00:16:09,920 AUDIENCE WHOOPS 347 00:16:15,680 --> 00:16:18,840 Fantastic! Thank you very much! Cheers. Thank you! 348 00:16:21,480 --> 00:16:25,600 There is something that's important to say about a 90% prediction. 349 00:16:25,600 --> 00:16:29,720 10% of the time that volcano won't explode. 350 00:16:29,720 --> 00:16:33,080 And while it sounds like I'm saying something quite obvious there, 351 00:16:33,080 --> 00:16:35,440 when you are dealing with the messy world of uncertainty, 352 00:16:35,440 --> 00:16:38,480 you need to understand that being wrong is sometimes part 353 00:16:38,480 --> 00:16:43,440 of prediction, because sometimes errors can have unexpected 354 00:16:43,440 --> 00:16:46,120 consequences. Now, let me show you what I'm talking about here, 355 00:16:46,120 --> 00:16:51,640 because we have just had a delivery of 100 Christmas presents 356 00:16:51,640 --> 00:16:54,960 and the rumours are that, hiding amongst these hundred presents, 357 00:16:54,960 --> 00:16:57,440 there are five brand-new smartphones. 358 00:16:57,440 --> 00:16:59,680 And one of you gets to come down and open one 359 00:16:59,680 --> 00:17:01,520 of these presents at random. 360 00:17:01,520 --> 00:17:03,840 So, who wants to come down? 361 00:17:03,840 --> 00:17:05,080 LAUGHTER 362 00:17:05,080 --> 00:17:07,240 What a surprise. Perfect. 363 00:17:07,240 --> 00:17:09,040 All right. Let's see who we can find. 364 00:17:09,040 --> 00:17:11,480 If you want to come down there, yeah. Perfect. Round of applause. 365 00:17:18,560 --> 00:17:20,480 What's your name? Yee-ling. Yee-ling, perfect. 366 00:17:20,480 --> 00:17:23,400 OK, Yee-ling. All right. So if you want to stand just over there. Now, 367 00:17:23,400 --> 00:17:26,080 you're going to get to pick one of these presents completely at random. 368 00:17:26,080 --> 00:17:29,560 Five of them are phones, the other 95, they're kind of rubbish presents 369 00:17:29,560 --> 00:17:32,200 like socks and satsumas. No-one wants any of them. 370 00:17:32,200 --> 00:17:34,680 And because you only get to pick one present, we want to give 371 00:17:34,680 --> 00:17:38,280 you the best possible chance at getting a phone. 372 00:17:38,280 --> 00:17:41,040 But luckily, Matt Parker has invented 373 00:17:41,040 --> 00:17:44,800 a very special present scanning machine to help you. 374 00:17:44,800 --> 00:17:46,240 Haven't you, Matt? 375 00:17:46,240 --> 00:17:49,080 Look at this! All right. Yeah. Amazing! 376 00:17:49,080 --> 00:17:54,280 This is my Xmas Ray Detector .80. 377 00:17:54,280 --> 00:17:56,360 And what we can do is we can put these presents 378 00:17:56,360 --> 00:17:59,000 through the scanner. Actually, do you want to come with me 379 00:17:59,000 --> 00:18:01,560 around over here? If you want to stand behind like the conveyor belt 380 00:18:01,560 --> 00:18:05,200 over here where they come out, if I turn this machine on and I start 381 00:18:05,200 --> 00:18:08,480 putting the presents in the top here, it will try and detect if 382 00:18:08,480 --> 00:18:09,960 there's a phone in them. 383 00:18:09,960 --> 00:18:13,440 If there's no phone, it just spits them out the same way they went in. 384 00:18:13,440 --> 00:18:17,080 If it thinks there's a phone, it rewraps it. Look at that! 385 00:18:17,080 --> 00:18:19,040 Ooh! That's a winner. 386 00:18:19,040 --> 00:18:21,720 So, if it says 'phone', you put it on the table. 387 00:18:21,720 --> 00:18:23,600 If it doesn't say 'phone', we don't care. Right? 388 00:18:23,600 --> 00:18:25,800 Just down there somewhere. Is that OK? Are you ready? 389 00:18:25,800 --> 00:18:27,840 I'm going to start piling them in here. 390 00:18:27,840 --> 00:18:29,680 You sort them out. Here they come. 391 00:18:29,680 --> 00:18:32,440 You seem pretty happy with this machine, Matt. I'm very proud. 392 00:18:32,440 --> 00:18:34,160 It is a... How accurate is it? 393 00:18:34,160 --> 00:18:38,720 This machine is 80% accurate. OK. 80% accurate. That's pretty good, 394 00:18:38,720 --> 00:18:40,160 Matt, 80% accurate. 395 00:18:40,160 --> 00:18:42,280 OK. So, then, one of these, then, 396 00:18:42,280 --> 00:18:44,000 that's rewrapped as a phone, 397 00:18:44,000 --> 00:18:46,200 what do you reckon are the chances that this is a phone? 398 00:18:46,200 --> 00:18:47,960 Shout out what you think it is. 399 00:18:47,960 --> 00:18:49,280 AUDIENCE OFFERS SUGGESTIONS 400 00:18:49,280 --> 00:18:51,720 OK. All right. Let's think this through, then. 401 00:18:51,720 --> 00:18:54,000 So, your machine, how accurate is your machine? 402 00:18:54,000 --> 00:18:56,480 80% accurate. OK. 403 00:18:56,480 --> 00:18:59,320 So your machine is 80% accurate. 404 00:18:59,320 --> 00:19:01,800 There are five phones in total. 405 00:19:01,800 --> 00:19:05,960 80% accurate means that it's only going to find four of them, right? 406 00:19:05,960 --> 00:19:07,440 Which does, Matt, that does mean one 407 00:19:07,440 --> 00:19:09,120 phone's going to end up on the floor. 408 00:19:09,120 --> 00:19:13,560 Yes. The downside to 80% accurate is it's 20% inaccurate. Ahh! 409 00:19:13,560 --> 00:19:16,240 But on the upside, 80% accurate. OK. 410 00:19:16,240 --> 00:19:18,320 So, I mean, you make a good point. 411 00:19:18,320 --> 00:19:20,040 Thank you! OK, one phone on the floor. 412 00:19:20,040 --> 00:19:22,920 It's not the end of the world. Still gives you a really good chance 413 00:19:22,920 --> 00:19:24,120 at finding a phone. 414 00:19:24,120 --> 00:19:26,840 There's something going on there, though. 415 00:19:26,840 --> 00:19:29,160 Hang on, hang on, hang on, hang on. Stop the machine. 416 00:19:29,160 --> 00:19:31,640 Whoa-whoa-whoa. Whoa-whoa-whoa. OK. Hang on. Matt... 417 00:19:31,640 --> 00:19:34,520 Yeah? There are way more than four phones on that table. 418 00:19:34,520 --> 00:19:37,320 Yeah. It's 80% accurate 419 00:19:37,320 --> 00:19:39,640 at both detecting a phone 420 00:19:39,640 --> 00:19:41,880 and detecting NOT a phone. 421 00:19:41,880 --> 00:19:44,560 Are you saying, hang on, are you saying this machine is taking socks 422 00:19:44,560 --> 00:19:46,920 and satsumas and wrapping them as phones when they're not? Yeah. 423 00:19:46,920 --> 00:19:48,520 20% of them. 424 00:19:48,520 --> 00:19:51,040 I don't know if I brought this up earlier, Hannah, 425 00:19:51,040 --> 00:19:52,960 but it's 80% accurate. 426 00:19:52,960 --> 00:19:54,480 AUDIENCE CACKLES 427 00:19:54,480 --> 00:19:58,040 Yeah, but Matt, there's 95 socks and satsumas here. Yeah. 428 00:19:58,040 --> 00:20:02,840 20%, that's not, that's 19 rubbish presents in that pile of phones. 429 00:20:02,840 --> 00:20:05,880 Yeah, but 80% of them are down here. 430 00:20:05,880 --> 00:20:09,000 So, the problem is like, this is 80% accurate, 431 00:20:09,000 --> 00:20:11,840 it doesn't mean these are going to be 80% phones. Yeah. 432 00:20:11,840 --> 00:20:16,600 It just means, I'd like to recap, that my machine is 80% accurate. 433 00:20:16,600 --> 00:20:19,640 OK, OK, OK. But right, by the time you're finished, you're going to 434 00:20:19,640 --> 00:20:21,920 have, what, 23 presents over here? 435 00:20:21,920 --> 00:20:24,120 Only four of them are going to be phones. 436 00:20:24,120 --> 00:20:27,800 That gives you a four in 23 chance of finding a phone. 437 00:20:27,800 --> 00:20:29,640 It's about 17%. So it wasn't... OK. 438 00:20:29,640 --> 00:20:31,360 Well, do you want to pick one at random? 439 00:20:32,920 --> 00:20:35,920 I mean, you can open it, have whatever's inside. 440 00:20:35,920 --> 00:20:39,280 It's overwhelmingly likely to be a satsuma. 441 00:20:39,280 --> 00:20:42,080 Well, you can keep it, though. Merry Christmas. 442 00:20:42,080 --> 00:20:45,120 Thanks for your machine, Matt. My pleasure. Amazing! 443 00:20:46,800 --> 00:20:48,400 80% accurate! 444 00:20:53,800 --> 00:20:57,440 That kind of error, mislabelling satsumas as phones, 445 00:20:57,440 --> 00:20:59,720 it's something that's called a false positive. 446 00:20:59,720 --> 00:21:02,360 And it goes to show how maths can sometimes really prove 447 00:21:02,360 --> 00:21:03,880 your intuition is wrong. 448 00:21:03,880 --> 00:21:05,640 Now, false positives, they're everywhere. 449 00:21:05,640 --> 00:21:07,640 You see this every time you go through an airport. 450 00:21:07,640 --> 00:21:11,000 Think about all of the people that are pulled over from the scanner 451 00:21:11,000 --> 00:21:14,080 for having lip balm and hair straighteners in their luggage. 452 00:21:14,080 --> 00:21:17,440 And the number of people who are false positives massively 453 00:21:17,440 --> 00:21:20,640 overwhelms the number of real weapons that the security team 454 00:21:20,640 --> 00:21:21,720 are looking for. 455 00:21:21,720 --> 00:21:24,960 But there is a really dark side to this kind of error, too, 456 00:21:24,960 --> 00:21:28,320 because imagine if, instead of scanning for presents, 457 00:21:28,320 --> 00:21:30,560 we were screening for cancer. 458 00:21:30,560 --> 00:21:34,240 Now, even if a cancer screening test, like a blood test 459 00:21:34,240 --> 00:21:38,600 or a mammogram, the ones that work with 89% accuracy 460 00:21:38,600 --> 00:21:43,680 because they are not perfect, there will always be false positives. 461 00:21:43,680 --> 00:21:47,480 There will always be people who believe that they have cancer 462 00:21:47,480 --> 00:21:49,920 when, in reality, they actually have nothing to worry about. 463 00:21:49,920 --> 00:21:51,480 This is kind of present scanning, 464 00:21:51,480 --> 00:21:54,520 but from the perspective of the gift box and you can imagine 465 00:21:54,520 --> 00:21:58,840 just how much pain and anxiety it might cause to be mislabelled. 466 00:21:58,840 --> 00:22:01,160 Now, this doesn't mean that you shouldn't be screened, 467 00:22:01,160 --> 00:22:02,800 but it's really important to understand 468 00:22:02,800 --> 00:22:05,240 what these results mean. 469 00:22:05,240 --> 00:22:09,080 But OK, if you can accept that being right isn't always possible, 470 00:22:09,080 --> 00:22:13,520 if you can really understand these numbers, then you can still use 471 00:22:13,520 --> 00:22:15,760 them to your advantage, because sometimes luck 472 00:22:15,760 --> 00:22:17,640 really is a matter of life and death, 473 00:22:17,640 --> 00:22:20,360 or in this case, zombies. 474 00:22:20,360 --> 00:22:23,800 So to explain this properly, I would like you to join me in giving 475 00:22:23,800 --> 00:22:26,080 a warm welcome to an epidemiologist 476 00:22:26,080 --> 00:22:30,160 from the London School of Hygiene and Tropical Medicine, Ros Eggo. 477 00:22:41,120 --> 00:22:44,560 OK, Ros, you study the mathematics of disease, right? 478 00:22:44,560 --> 00:22:46,400 How much of it is luck and chance? 479 00:22:46,400 --> 00:22:48,160 Well, there's a lot of chance that's involved 480 00:22:48,160 --> 00:22:49,880 in the transmission of disease. 481 00:22:49,880 --> 00:22:52,280 So if there's an epidemic happening at the moment, the chance 482 00:22:52,280 --> 00:22:54,560 of you getting it depends on how many other people 483 00:22:54,560 --> 00:22:57,720 there are that have the infection at the time. 484 00:22:57,720 --> 00:23:00,600 And then if you happen to meet one of those, which is more likely 485 00:23:00,600 --> 00:23:03,160 when there's a lot of people who have it, 486 00:23:03,160 --> 00:23:05,440 there's also a chance you'll get it from them or not. 487 00:23:05,440 --> 00:23:08,360 And if you have some pre-existing immunity, maybe 488 00:23:08,360 --> 00:23:10,480 you've been vaccinated or you've had it before, 489 00:23:10,480 --> 00:23:13,240 then the chance of them passing it to you also goes down. 490 00:23:13,240 --> 00:23:16,960 So there's a lot. So, then, how can maths help you? Mmm. 491 00:23:16,960 --> 00:23:19,720 It's really difficult to predict, on an individual level, if somebody 492 00:23:19,720 --> 00:23:23,200 will get it. But on a population level, with these nice, big numbers, 493 00:23:23,200 --> 00:23:25,840 there's some things we can predict - when the epidemic will go 494 00:23:25,840 --> 00:23:28,640 up and when it will come down. We try and predict the peak. 495 00:23:28,640 --> 00:23:31,520 And we try and predict which groups might be at risk of infection. 496 00:23:31,520 --> 00:23:35,040 Well, talking of epidemics, there are some rumours 497 00:23:35,040 --> 00:23:37,760 of a zombie apocalypse about to hit. 498 00:23:37,760 --> 00:23:41,000 OK, so, everyone here, underneath your seats, 499 00:23:41,000 --> 00:23:42,880 you should have a zombie mask 500 00:23:42,880 --> 00:23:45,560 and you should have some ping pong balls. 501 00:23:45,560 --> 00:23:48,240 Now, these are your zombie germs. Mm-hm. 502 00:23:48,240 --> 00:23:53,160 So how it works is, if a zombie germ touches you, you become a zombie. 503 00:23:53,160 --> 00:23:57,240 You put up your zombie mask and then you take your zombie germs 504 00:23:57,240 --> 00:23:59,360 and you throw them straight up in the air as high 505 00:23:59,360 --> 00:24:01,000 as you possibly can. OK. 506 00:24:01,000 --> 00:24:03,160 And in a moment, I'm going to start off this infection. 507 00:24:03,160 --> 00:24:06,160 So everyone start with your mask down to kick us off, 508 00:24:06,160 --> 00:24:09,200 but you can make a prediction about what might happen here. Um. 509 00:24:09,200 --> 00:24:12,080 So over here, nobody has any protection against zombies, 510 00:24:12,080 --> 00:24:13,280 zombie infection. 511 00:24:13,280 --> 00:24:15,880 So we're going to expect a lot of cases and possibly quite quickly. 512 00:24:15,880 --> 00:24:17,840 OK. All right. Let's give it a go. Is everyone ready? 513 00:24:17,840 --> 00:24:20,480 Everyone got their balls in their hand? All right. We're going to 514 00:24:20,480 --> 00:24:22,880 start this infection. Here's the zombie germs coming at you. 515 00:24:24,000 --> 00:24:26,760 I'm going to... I'm directly targeting people. 516 00:24:26,760 --> 00:24:28,800 OK. Oh, already. 517 00:24:28,800 --> 00:24:31,920 So, we've got little, actually, a few different patches here. 518 00:24:31,920 --> 00:24:33,040 We do, yeah. 519 00:24:33,040 --> 00:24:35,720 In fact, they're starting to get a little bit frightening around here. 520 00:24:35,720 --> 00:24:40,520 And it's travelling backwards through, and now a big cluster of 521 00:24:40,520 --> 00:24:42,840 zombies there, right in the middle. Uh-huh! 522 00:24:44,720 --> 00:24:47,880 How realistic is what they're doing, compared to how normal, real 523 00:24:47,880 --> 00:24:49,880 diseases spread? 524 00:24:49,880 --> 00:24:53,000 Well, obviously, for real diseases, it's not quite so frightening 525 00:24:53,000 --> 00:24:54,880 as what's happening up here. 526 00:24:54,880 --> 00:24:59,560 But we use similar type of methods, using chance to pass on infection 527 00:24:59,560 --> 00:25:02,400 to understand how real diseases spread around. 528 00:25:02,400 --> 00:25:05,160 But for zombie infections, nobody ever recovers. 529 00:25:05,160 --> 00:25:07,240 Nobody, OK. And that's, in the real world, 530 00:25:07,240 --> 00:25:09,160 people usually recover. 531 00:25:09,160 --> 00:25:11,120 I think there are, in fact, OK, 532 00:25:11,120 --> 00:25:13,640 so if you... 533 00:25:13,640 --> 00:25:15,800 ..it's still going! It is still going! 534 00:25:15,800 --> 00:25:20,360 And if you are not a zombie, 535 00:25:20,360 --> 00:25:23,280 could you stand up for us? How many people are there? 536 00:25:23,280 --> 00:25:25,040 Oh, we've got some! 537 00:25:25,040 --> 00:25:30,160 Three people who managed to escape the zombies! 538 00:25:30,160 --> 00:25:33,320 You also can't target people directly with your zombie germs. 539 00:25:33,320 --> 00:25:35,960 That wasn't part of the rules. OK. Everybody, sit down. 540 00:25:35,960 --> 00:25:39,240 Thank you very much. Well done. Now we can do this again, but this 541 00:25:39,240 --> 00:25:43,360 time, we can use some kind of zombie defences. Exactly. Yeah. 542 00:25:43,360 --> 00:25:45,880 So over here, there's going to be some protection against 543 00:25:45,880 --> 00:25:48,840 infection, and these people are going to be completely protected. 544 00:25:48,840 --> 00:25:51,680 OK. So those of you who have, you've got protective masks 545 00:25:51,680 --> 00:25:53,240 under your seat and this makes you 546 00:25:53,240 --> 00:25:56,000 completely immune from zombie infection. It does, yeah. OK. 547 00:25:56,000 --> 00:25:59,240 So if the zombie germs touch you and you're wearing a face mask, 548 00:25:59,240 --> 00:26:01,960 you're wearing a protective mask, don't worry about anything. 549 00:26:01,960 --> 00:26:05,840 Don't put up your zombie mask, don't throw any germs around. 550 00:26:05,840 --> 00:26:08,680 You're completely immune. And we'll see what happens now. 551 00:26:08,680 --> 00:26:11,800 And what's your prediction about what will happen in this case? 552 00:26:11,800 --> 00:26:14,880 Well, there's a lot of protection up here. So I think the epidemic will 553 00:26:14,880 --> 00:26:18,000 be over very quickly and maybe not many people will be infected. 554 00:26:18,000 --> 00:26:19,720 OK, let's give it a go. Here we go. 555 00:26:21,960 --> 00:26:23,400 OK. 556 00:26:23,400 --> 00:26:25,840 So, it's still started. We've got a few. 557 00:26:27,440 --> 00:26:29,720 Oh! 558 00:26:29,720 --> 00:26:33,360 OK, I mean, that stopped almost immediately. 559 00:26:33,360 --> 00:26:38,240 OK. So those of you who are not wearing a protective mask, but also 560 00:26:38,240 --> 00:26:39,960 not a zombie, could you stand up? 561 00:26:41,080 --> 00:26:43,360 Oh, goodness me. Look at that! Wow, that's a lot. 562 00:26:43,360 --> 00:26:45,480 And there's a point to it. Thank you very much, everyone. 563 00:26:45,480 --> 00:26:47,960 You can sit down. There's a point to all of this, right? 564 00:26:47,960 --> 00:26:50,760 Yeah, exactly. So the protection that we had in the population 565 00:26:50,760 --> 00:26:54,480 from the people wearing masks who are vaccinated against infection 566 00:26:54,480 --> 00:26:57,200 has protected everybody in this part of the audience 567 00:26:57,200 --> 00:26:58,520 from getting infection. 568 00:26:58,520 --> 00:27:00,160 And is this the same thing that happens 569 00:27:00,160 --> 00:27:02,000 when we're vaccinated against diseases? 570 00:27:02,000 --> 00:27:03,400 Exactly. It is. 571 00:27:03,400 --> 00:27:06,400 So we have here, kind of, this community 572 00:27:06,400 --> 00:27:08,720 has immunity from infection. 573 00:27:08,720 --> 00:27:11,080 And so, even the people who weren't themselves vaccinated 574 00:27:11,080 --> 00:27:12,520 have been protected by the protection 575 00:27:12,520 --> 00:27:14,000 in the whole community. 576 00:27:14,000 --> 00:27:16,600 So how much does the maths of modelling disease 577 00:27:16,600 --> 00:27:19,240 make a difference back in the real world? 578 00:27:19,240 --> 00:27:22,680 So we take experiments kind of like this, but in the computer, 579 00:27:22,680 --> 00:27:25,280 take things that we know about how infection spreads 580 00:27:25,280 --> 00:27:27,720 and about the population in general, 581 00:27:27,720 --> 00:27:32,680 and the goal is to use the computer, use our simulations, to figure out 582 00:27:32,680 --> 00:27:35,000 ahead of time what would be the best interventions. 583 00:27:35,000 --> 00:27:38,320 What should we do to decrease the number of cases in everybody, 584 00:27:38,320 --> 00:27:39,920 without it having to happen? 585 00:27:39,920 --> 00:27:43,360 So it doesn't matter whether you're talking about flus or Ebola 586 00:27:43,360 --> 00:27:46,200 or, I don't know, TB in cows, you can use these mathematical ideas. 587 00:27:46,200 --> 00:27:48,760 Exactly. Yeah, for human diseases, for animal diseases, even for 588 00:27:48,760 --> 00:27:51,120 plants, it's really useful. And demonstrates that vaccines 589 00:27:51,120 --> 00:27:53,160 really do make a difference. It does. It does. 590 00:27:53,160 --> 00:27:56,000 Would you rather live over here with all these zombies or over here? 591 00:27:56,000 --> 00:27:57,840 Definitely over there. Definitely over here. 592 00:27:57,840 --> 00:27:59,640 Ros, thank you very much. Thank you. 593 00:28:03,480 --> 00:28:06,040 There is a really simple point in all of this. If you've got maths 594 00:28:06,040 --> 00:28:09,360 on your side, it's not just about spotting patterns. 595 00:28:09,360 --> 00:28:12,600 It's about using what the numbers tell you 596 00:28:12,600 --> 00:28:15,080 to bend the world to your will. 597 00:28:15,080 --> 00:28:19,240 And there is one game of luck, skill and stats that actually captivates 598 00:28:19,240 --> 00:28:22,600 millions of us every single week, because believe it or not, 599 00:28:22,600 --> 00:28:26,280 the Premier League is awash with mathematicians. 600 00:28:26,280 --> 00:28:30,160 So, here's the staff photo from Liverpool Football Club. 601 00:28:30,160 --> 00:28:32,960 You've got the team manager there, obviously a very important job, 602 00:28:32,960 --> 00:28:35,280 but here, this is a group of mathematicians. 603 00:28:35,280 --> 00:28:38,760 These guys are the unsung heroes of Liverpool football team 604 00:28:38,760 --> 00:28:43,360 and it's their job to make the team as lucky as possible. 605 00:28:43,360 --> 00:28:46,800 So please join me in welcoming Tim Waskett. 606 00:28:53,080 --> 00:28:55,920 So, that's you in that photo, Tim? Yeah. That's me, just there. 607 00:28:55,920 --> 00:28:58,680 All right. So how on Earth do you go about changing a football game 608 00:28:58,680 --> 00:29:00,600 into something to do with numbers? 609 00:29:00,600 --> 00:29:05,280 So the primary currency that every football game is based on is goals. 610 00:29:05,280 --> 00:29:07,440 Obviously, that's the most important thing. 611 00:29:07,440 --> 00:29:10,560 And it's our job to turn every action on pitch, every pass, every 612 00:29:10,560 --> 00:29:14,480 throw in, every tackle, every shot, into a goal probability. 613 00:29:14,480 --> 00:29:17,600 And I think we've actually got an image from your team. 614 00:29:17,600 --> 00:29:20,480 So this is showing you the darker the colour, 615 00:29:20,480 --> 00:29:22,720 the darker the red, the more likely you are to score? 616 00:29:22,720 --> 00:29:25,160 I mean, from there, it's pretty easy, right? Yes. Exactly. 617 00:29:25,160 --> 00:29:27,920 So this is taken from literally hundreds of thousands of shots 618 00:29:27,920 --> 00:29:30,320 through major leagues all over the world. 619 00:29:30,320 --> 00:29:34,280 And by looking at where the shots take place, how often they became 620 00:29:34,280 --> 00:29:37,360 a goal, gives us a probability of a shot from a similar situation, 621 00:29:37,360 --> 00:29:39,200 ultimately ending as a goal. 622 00:29:39,200 --> 00:29:42,320 Now, I understand that there's a little game that you play 623 00:29:42,320 --> 00:29:45,560 with all of this, of trying to guess what the maths says 624 00:29:45,560 --> 00:29:48,920 is the probability of a particular shot going in. Yeah. 625 00:29:48,920 --> 00:29:50,480 I think we have a little clip. 626 00:29:50,480 --> 00:29:52,880 So, we call this "the expected goals game". OK. 627 00:29:52,880 --> 00:29:54,760 So what game are we watching here? 628 00:29:54,760 --> 00:29:58,760 So this is a friendly between Tranmere Rovers and Liverpool, 629 00:29:58,760 --> 00:30:01,960 this is a pre-season friendly. And you paused the footage just... 630 00:30:01,960 --> 00:30:03,960 Yep. So, he's about to do a header. 631 00:30:03,960 --> 00:30:06,480 So he's about to do a header and he's right in front of the goal, 632 00:30:06,480 --> 00:30:10,480 there, very close. So what do we think the probability of this shot 633 00:30:10,480 --> 00:30:11,880 turning into a goal is? 634 00:30:11,880 --> 00:30:15,160 What does the maths calculate the chances of this turning into a goal? 635 00:30:15,160 --> 00:30:17,880 We have some options. Do you reckon it's 50%, 636 00:30:17,880 --> 00:30:21,200 75% or 99%? OK. 637 00:30:21,200 --> 00:30:23,120 Shout out your answers. What do you think it is? 638 00:30:23,120 --> 00:30:25,360 What do you reckon the maths says is the chances? 639 00:30:25,360 --> 00:30:27,120 AUDIENCE RESPONDS ENTHUSIASTICALLY 640 00:30:27,120 --> 00:30:29,120 Oooh, 75 came out quite clearly there. 641 00:30:29,120 --> 00:30:30,840 All right, what did the maths say? 642 00:30:30,840 --> 00:30:33,440 Well, the actual answer here is 99%. 643 00:30:33,440 --> 00:30:34,880 SIMULATED CHEERING 644 00:30:34,880 --> 00:30:38,920 So, that means in exactly the same position with exactly the same 645 00:30:38,920 --> 00:30:42,840 surroundings, if you reran these 100 times, it would only not 646 00:30:42,840 --> 00:30:44,480 result in a goal once. Exactly. 647 00:30:44,480 --> 00:30:47,280 So, 99 shots of a similar sort of position will end in a goal, 648 00:30:47,280 --> 00:30:50,360 and only one will get saved by the goalkeeper. Oh, OK. All right. 649 00:30:50,360 --> 00:30:53,800 Let's try one more. We've got another clip for you. 650 00:30:53,800 --> 00:30:54,840 Here we go. 651 00:30:57,160 --> 00:30:59,840 Oh, OK, so he's much farther out here. Much further out, 652 00:30:59,840 --> 00:31:02,280 and he's also not central onto the goal, he's off to one side. 653 00:31:02,280 --> 00:31:03,880 So we know it's going to be a lower chance. 654 00:31:03,880 --> 00:31:05,520 It's going to definitely be a lower chance. 655 00:31:05,520 --> 00:31:07,920 But what does the maths actually calculate it to be? 656 00:31:07,920 --> 00:31:09,440 So we've got some options for you. 657 00:31:09,440 --> 00:31:11,600 All right. Do you reckon 658 00:31:11,600 --> 00:31:15,560 the maths says that this is 4%, 7% or 10% chance of getting a goal? 659 00:31:15,560 --> 00:31:17,320 What do you reckon? Shout it at me. 660 00:31:17,320 --> 00:31:18,760 AUDIENCE RESPONDS 661 00:31:18,760 --> 00:31:20,560 Oh, everyone's going in the middle there. 662 00:31:20,560 --> 00:31:23,200 What's the answer? You're spot-on. It's actually 7%. 663 00:31:23,200 --> 00:31:25,320 And let's see what actually did happen. 664 00:31:26,600 --> 00:31:28,800 Ah, a miss! One of the reasons 665 00:31:28,800 --> 00:31:31,040 that that was only 7% and not any higher was because he shot 666 00:31:31,040 --> 00:31:32,960 it straight at the goalkeeper. 667 00:31:32,960 --> 00:31:35,280 I mean, that's always a big mistake. 668 00:31:35,280 --> 00:31:38,920 But it's not just about strikers that you're translating 669 00:31:38,920 --> 00:31:40,440 into numbers, is it? No, exactly. 670 00:31:40,440 --> 00:31:43,080 So this is the easiest thing that we can do, but we can do 671 00:31:43,080 --> 00:31:45,640 a similar calculation for every other event on the pitch. 672 00:31:45,640 --> 00:31:47,680 So any time the ball is passed, for example. 673 00:31:47,680 --> 00:31:50,040 OK. So to help us to demonstrate this, 674 00:31:50,040 --> 00:31:53,560 I'd like to invite onto the stage Bertie and Jamie 675 00:31:53,560 --> 00:31:55,720 from a London youth team. 676 00:32:00,800 --> 00:32:04,960 Thank you very much, guys. Thank you. So... 677 00:32:04,960 --> 00:32:07,680 Bertie and Jamie are just going to have a bit of a kickabout. Yeah. 678 00:32:07,680 --> 00:32:11,920 Tell us how this works, then. So for roughly 200 games per weekend, 679 00:32:11,920 --> 00:32:16,800 we get data involving every single ball touch in the game. 680 00:32:16,800 --> 00:32:21,120 So, the way the data is collected, every time that a player passes 681 00:32:21,120 --> 00:32:23,520 the ball, they'll mark it on the pitch and they'll say 682 00:32:23,520 --> 00:32:26,400 this is the position on the pitch and the player who made that pass, 683 00:32:26,400 --> 00:32:28,720 and then there'll be somebody watching for the other team 684 00:32:28,720 --> 00:32:29,880 making their passes. 685 00:32:29,880 --> 00:32:32,280 So two people will be going backwards and forwards, marking 686 00:32:32,280 --> 00:32:35,600 up all of these events. So roughly every second or so, there's a new 687 00:32:35,600 --> 00:32:37,240 pass, so it's pretty frantic. 688 00:32:37,240 --> 00:32:39,800 So you just have someone there to kick and kick and kick. It doesn't 689 00:32:39,800 --> 00:32:42,240 sound like the most exciting way to watch football. No. Well, 690 00:32:42,240 --> 00:32:45,240 thankfully, it's not us who has to do this work. We actually have 691 00:32:45,240 --> 00:32:48,680 a data supplier who provides us with these data, these files for us. OK. 692 00:32:48,680 --> 00:32:51,640 Bertie and Jamie, thank you very much for your help there. Thank you. 693 00:32:56,640 --> 00:32:59,560 So what do you end up with once you've done all of this? 694 00:32:59,560 --> 00:33:01,080 So for every game we get 695 00:33:01,080 --> 00:33:03,800 approximately 2,000 ball touch events. 696 00:33:03,800 --> 00:33:06,760 And that tells us the position of the player who makes the pass. 697 00:33:06,760 --> 00:33:09,760 But what that doesn't tell you is where all of the other players 698 00:33:09,760 --> 00:33:11,440 are on the pitch at that moment. 699 00:33:11,440 --> 00:33:13,320 But you can get hold of those numbers? 700 00:33:13,320 --> 00:33:15,320 So we can. For the Premier League games, we get 701 00:33:15,320 --> 00:33:16,560 what we call tracking data. 702 00:33:16,560 --> 00:33:20,280 And so this is a set of cameras all around the stadium and that's 703 00:33:20,280 --> 00:33:22,800 monitoring the position of all of the players. 704 00:33:22,800 --> 00:33:25,480 So, 22 players plus the position of the ball. 705 00:33:25,480 --> 00:33:29,320 And it does that for 25 frames a second for the full 90 minutes. 706 00:33:29,320 --> 00:33:32,440 So you end up with approximately 1.5 million data points. 707 00:33:32,440 --> 00:33:34,320 That is a lot of numbers! 708 00:33:34,320 --> 00:33:37,360 And in fact, we've got here on the gigantic spreadsheet, essentially. 709 00:33:37,360 --> 00:33:40,040 So, this is just one game. And you're watching a game 710 00:33:40,040 --> 00:33:41,320 of football here, basically. 711 00:33:41,320 --> 00:33:42,840 Yes, exactly. Through numbers. 712 00:33:42,840 --> 00:33:45,840 OK. Interesting. I mean, maybe less exciting than watching 713 00:33:45,840 --> 00:33:48,000 Match of the Day. Well, it depends on your perspective. 714 00:33:48,000 --> 00:33:49,920 But you're not just collecting these numbers, are you? 715 00:33:49,920 --> 00:33:51,200 What do you do with them? 716 00:33:51,200 --> 00:33:54,960 So this data can give us a goal value for every position 717 00:33:54,960 --> 00:33:56,520 and for every player on the pitch. 718 00:33:56,520 --> 00:33:58,040 And what does it end up looking like? 719 00:33:58,040 --> 00:34:00,440 I think we've got a little... Yeah, we have a little animation. 720 00:34:00,440 --> 00:34:02,680 OK. So tell us what we're looking at here. So this is what we call 721 00:34:02,680 --> 00:34:06,680 pitch control. So you can see the players are in the circles there, 722 00:34:06,680 --> 00:34:09,560 and the arrows represent the direction and speed 723 00:34:09,560 --> 00:34:10,760 that they can travel in. 724 00:34:10,760 --> 00:34:12,720 And you've got a blue team and a red team. Yep. 725 00:34:12,720 --> 00:34:15,280 So the red team here is actually Liverpool. 726 00:34:15,280 --> 00:34:19,120 And the areas in red are places that the Liverpool players can get 727 00:34:19,120 --> 00:34:22,440 to sooner than the players who are in blue. Ahh! 728 00:34:22,440 --> 00:34:24,640 Based on how quickly people run. Yes. 729 00:34:24,640 --> 00:34:26,600 And based on where the ball is. Exactly. 730 00:34:26,600 --> 00:34:29,120 So, for example, the ball, in this particular example, 731 00:34:29,120 --> 00:34:30,760 is that yellow dot there. 732 00:34:30,760 --> 00:34:34,160 And if this player, who is currently in possession of the ball, 733 00:34:34,160 --> 00:34:36,160 this is number ten, this is Sadio Mane, 734 00:34:36,160 --> 00:34:39,240 his best option at this stage, is to probably pass to one 735 00:34:39,240 --> 00:34:42,160 of these red areas. And, in fact, he ends up passing the ball up into 736 00:34:42,160 --> 00:34:44,640 this red zone here to be picked up by player 66, 737 00:34:44,640 --> 00:34:46,240 which is Trent Alexander-Arnold. 738 00:34:46,240 --> 00:34:50,080 So this is a real game and we can play on watching the game 739 00:34:50,080 --> 00:34:51,640 through this heat map here. 740 00:34:51,640 --> 00:34:53,840 So, this thing here, what's this telling us? 741 00:34:53,840 --> 00:34:56,680 So, this is what I was talking about before, about turning 742 00:34:56,680 --> 00:34:59,640 everything into a goal probability. So this value now, 1.3%, 743 00:34:59,640 --> 00:35:04,400 this is the probability that a goal will be scored with the ball 744 00:35:04,400 --> 00:35:06,680 in this position within the next 15 seconds. 745 00:35:06,680 --> 00:35:08,280 It's quite hard from there. Really. 746 00:35:08,280 --> 00:35:10,840 From there, yes. Because you're so far back. You've got a long way to 747 00:35:10,840 --> 00:35:12,720 go before you reach the goal. It's very accurate. 748 00:35:12,720 --> 00:35:15,840 But we can play on this game and see how it evolves. 749 00:35:15,840 --> 00:35:17,960 So, the ball does get passed over. 750 00:35:17,960 --> 00:35:19,640 Yeah. It gets successfully received. 751 00:35:19,640 --> 00:35:21,920 Still not a very high chance of scoring a goal from there. 752 00:35:21,920 --> 00:35:24,040 No. But now, Trent Alexander-Arnold, 753 00:35:24,040 --> 00:35:26,560 his best option is to run into this position, here. Ahh. 754 00:35:26,560 --> 00:35:28,360 So he's now dribbling the ball forward. 755 00:35:28,360 --> 00:35:32,440 So, are you using this information to look at what did happen and work 756 00:35:32,440 --> 00:35:35,480 out what should have happened? So, we use this in a number of ways. 757 00:35:35,480 --> 00:35:37,160 The main way we use this is 758 00:35:37,160 --> 00:35:40,200 to evaluate player performance after the game. 759 00:35:40,200 --> 00:35:43,000 And in this particular game, if we play on one more time, 760 00:35:43,000 --> 00:35:44,880 what was the result of this sequence? 761 00:35:44,880 --> 00:35:47,640 So if we pause it right about now, 762 00:35:47,640 --> 00:35:49,680 you can see Trent Alexander-Arnold 763 00:35:49,680 --> 00:35:51,720 now has the ball very close to the goal. 764 00:35:51,720 --> 00:35:54,520 He's in a good position, but his best option now is to pass 765 00:35:54,520 --> 00:35:57,240 into this red area here, where it can either be received 766 00:35:57,240 --> 00:36:01,080 by Mane or Shaqiri, who is number 23 down here. 767 00:36:01,080 --> 00:36:03,320 One of these two players is most likely to get 768 00:36:03,320 --> 00:36:04,920 to this red zone. 769 00:36:04,920 --> 00:36:07,560 And in actual fact, what happens in this particular situation, 770 00:36:07,560 --> 00:36:09,080 is that Mane receives the ball 771 00:36:09,080 --> 00:36:11,240 right round about here, and he scores a goal. 772 00:36:11,240 --> 00:36:13,280 Which is exactly what you want. Exactly. 773 00:36:13,280 --> 00:36:15,800 So are you using this stuff to just analyse your team 774 00:36:15,800 --> 00:36:18,160 or are you using it to analyse other teams, too? 775 00:36:18,160 --> 00:36:21,720 So the advantage of this is we see all of the players at the same time, 776 00:36:21,720 --> 00:36:24,040 which means that we can analyse all of the players 777 00:36:24,040 --> 00:36:27,960 within the Premier League and a large number of players in all sorts 778 00:36:27,960 --> 00:36:30,920 of other leagues around the world, using the ball touch event data. 779 00:36:30,920 --> 00:36:34,400 And that gives us some really good information on which players 780 00:36:34,400 --> 00:36:37,240 are doing well and who we might be able to sign in the future. 781 00:36:37,240 --> 00:36:39,920 Ultimately, to give you the best chance possible at beating your 782 00:36:39,920 --> 00:36:41,680 opponent. Exactly. Amazing. 783 00:36:41,680 --> 00:36:44,520 Tim, thank you so much for joining us. Thank you. 784 00:36:44,520 --> 00:36:47,480 APPLAUSE 785 00:36:47,480 --> 00:36:48,880 That's the thing about being lucky. 786 00:36:48,880 --> 00:36:51,040 Sometimes it's not just about what YOU do, 787 00:36:51,040 --> 00:36:53,800 but also about who you're up against, 788 00:36:53,800 --> 00:36:57,600 and if winning is what you're after, something rather intriguing happens 789 00:36:57,600 --> 00:37:01,320 when you start to look at the maths of competition. 790 00:37:01,320 --> 00:37:05,120 OK, so for this, I would like two volunteers who are willing 791 00:37:05,120 --> 00:37:07,160 to compete against one another. 792 00:37:07,160 --> 00:37:09,520 OK, perfect. If you want to come down here. 793 00:37:09,520 --> 00:37:13,320 That's one, and we'll get someone from over here. 794 00:37:13,320 --> 00:37:15,360 Round of applause. Welcome to the stage. 795 00:37:15,360 --> 00:37:16,680 APPLAUSE 796 00:37:16,680 --> 00:37:20,280 What was your name? Nat. Perfect, OK, Nat. 797 00:37:20,280 --> 00:37:22,680 Thank you. And what was your name? Jasmine. Hm? 798 00:37:22,680 --> 00:37:26,120 Jasmine. Jasmine. OK, Nat and Jasmine, right, OK. 799 00:37:26,120 --> 00:37:29,640 This game is called Goodie or Baddie, 800 00:37:29,640 --> 00:37:33,280 right, and the reason why is because in every round you have two choices. 801 00:37:33,280 --> 00:37:35,440 You can either decide to be a goodie, 802 00:37:35,440 --> 00:37:37,720 in which case you put on your red hat, 803 00:37:37,720 --> 00:37:40,440 or you can decide to be a baddie, 804 00:37:40,440 --> 00:37:43,160 in which case you put on your purple hat. 805 00:37:43,160 --> 00:37:46,720 And what we're going to do is we're going to play four rounds, 806 00:37:46,720 --> 00:37:49,160 and in each round you have a chance to win some points. 807 00:37:49,160 --> 00:37:52,880 If you get to 12 points in those four rounds, 808 00:37:52,880 --> 00:37:58,080 then you win an amazing goody bag, OK? I mean, it's pretty special. 809 00:37:58,080 --> 00:38:00,000 So this is the way that it works, OK? 810 00:38:00,000 --> 00:38:03,120 If both of you decide to be a goodie, 811 00:38:03,120 --> 00:38:06,280 then you will end up getting three points each. 812 00:38:06,280 --> 00:38:08,640 If both of you decide to be a baddie, 813 00:38:08,640 --> 00:38:10,440 you will end up getting one point each. 814 00:38:10,440 --> 00:38:13,440 But if one of you decides to be a goodie 815 00:38:13,440 --> 00:38:15,400 and the other person decides to be a baddie, 816 00:38:15,400 --> 00:38:16,960 then the baddie takes everything. 817 00:38:16,960 --> 00:38:20,440 They get five points and the other person gets nothing. 818 00:38:20,440 --> 00:38:23,680 OK? Remember, four rounds. You have to get 12 points. 819 00:38:23,680 --> 00:38:24,920 You happy? OK, all right. 820 00:38:24,920 --> 00:38:27,360 Turn around, so you can't see each other's choices. 821 00:38:27,360 --> 00:38:29,040 Here we go. OK. Here we go. 822 00:38:29,040 --> 00:38:31,240 So, red for goodie, purple for baddie. 823 00:38:31,240 --> 00:38:34,120 Make your choices now. 824 00:38:34,120 --> 00:38:35,960 Oh! 825 00:38:35,960 --> 00:38:37,680 Turn around and have a look at each other. 826 00:38:37,680 --> 00:38:40,840 Oh! OK. All right. Five points, five points there. 827 00:38:40,840 --> 00:38:43,000 One, two, three, four, five. 828 00:38:43,000 --> 00:38:45,800 OK. All right. Round 2. Round 2. Here we go. Here we go. 829 00:38:45,800 --> 00:38:47,240 OK. Round 2. 830 00:38:47,240 --> 00:38:48,840 He stole some points off you that time. 831 00:38:48,840 --> 00:38:50,040 Let's go for Round 2. 832 00:38:50,040 --> 00:38:51,360 Go. 833 00:38:51,360 --> 00:38:53,120 Oh! Payback! 834 00:38:53,120 --> 00:38:54,600 Turn around, turn around. 835 00:38:54,600 --> 00:38:57,000 Only one point each. You blocked him there. 836 00:38:57,000 --> 00:38:59,360 OK, right. Round 3, Round 3, here we go. 837 00:39:01,240 --> 00:39:03,480 Make your choice for Round 3. 838 00:39:05,440 --> 00:39:07,800 And turn around and have a look at each other. 839 00:39:07,800 --> 00:39:09,160 Argh. Argh! 840 00:39:09,160 --> 00:39:10,960 You got him back. You got him back. 841 00:39:10,960 --> 00:39:13,040 OK. There's still one round left. 842 00:39:13,040 --> 00:39:15,360 One round left, turn around. 843 00:39:15,360 --> 00:39:16,840 Here we go. OK. 844 00:39:16,840 --> 00:39:19,320 Do you want to take your hat off and make your choices? 845 00:39:19,320 --> 00:39:21,080 Go for it. 846 00:39:21,080 --> 00:39:22,600 LAUGHTER 847 00:39:22,600 --> 00:39:23,960 What an unsurprising ending. 848 00:39:23,960 --> 00:39:26,000 Turn around, have a look at each other. OK. 849 00:39:26,000 --> 00:39:29,960 So neither of you... I'm afraid neither of you got up to the line. 850 00:39:29,960 --> 00:39:32,400 Neither of you got into our points. 851 00:39:32,400 --> 00:39:34,800 But you know what's strange, though? 852 00:39:34,800 --> 00:39:36,480 There were four rounds. 853 00:39:36,480 --> 00:39:38,840 12 points was all you needed to win a prize. 854 00:39:38,840 --> 00:39:41,520 If both of you had just played goodie every single time, 855 00:39:41,520 --> 00:39:44,040 then you both would have walked away with an amazing prize. 856 00:39:44,040 --> 00:39:46,120 The fact you played baddy, you blocked each other. 857 00:39:46,120 --> 00:39:47,240 But, you know what? 858 00:39:47,240 --> 00:39:49,280 It's understandable, because, OK, let's say... 859 00:39:49,280 --> 00:39:51,240 Mind if I take this hat off your head for a second? 860 00:39:51,240 --> 00:39:56,240 Let's say your opponent had chosen to be a baddie. 861 00:39:56,240 --> 00:39:58,520 In that situation, you've got two choices, right? 862 00:39:58,520 --> 00:40:02,080 You can be a goodie, in which case you get no points at all. 863 00:40:02,080 --> 00:40:04,600 Or you can be a baddie, in which case you get one point. 864 00:40:04,600 --> 00:40:06,080 So, if your opponent's a baddie, 865 00:40:06,080 --> 00:40:08,640 being a baddie is definitely the best thing to do. 866 00:40:08,640 --> 00:40:10,920 But what if... Do you mind switching your hats for me? 867 00:40:10,920 --> 00:40:13,560 If your opponent instead had chosen to be a goodie, 868 00:40:13,560 --> 00:40:16,160 let's think about what was available to you. 869 00:40:16,160 --> 00:40:18,280 Be a goodie - you get three points. 870 00:40:18,280 --> 00:40:20,600 Baddie - you get five points. 871 00:40:20,600 --> 00:40:24,320 So even in this situation it's best off for you to be a baddie. 872 00:40:24,320 --> 00:40:28,320 So it turns out it doesn't matter what your opponent does. 873 00:40:28,320 --> 00:40:32,640 In any one round, it is always best for you to play the baddie, 874 00:40:32,640 --> 00:40:35,680 even though if you both just played goodie, 875 00:40:35,680 --> 00:40:38,200 you both would have ended up getting a prize. 876 00:40:38,200 --> 00:40:40,800 So thank you very much to my volunteers. 877 00:40:40,800 --> 00:40:42,320 You're both very bad! 878 00:40:42,320 --> 00:40:44,240 APPLAUSE 879 00:40:48,560 --> 00:40:50,920 Not very Christmassy, is it? Being bad to each other. 880 00:40:50,920 --> 00:40:54,000 Actually, this is a very famous game among mathematicians 881 00:40:54,000 --> 00:40:57,280 and it goes to show that winning isn't always about luck, 882 00:40:57,280 --> 00:41:00,960 but it also, I think, highlights one of the great tragedies of humanity 883 00:41:00,960 --> 00:41:04,600 that sometimes the tempting thing to do right then and there, 884 00:41:04,600 --> 00:41:07,120 doesn't actually lead to the very best outcome overall. 885 00:41:07,120 --> 00:41:09,720 And actually, you see this time and time again, 886 00:41:09,720 --> 00:41:12,800 the very tiny little selfish choices that we all make, 887 00:41:12,800 --> 00:41:16,360 they add up to mean that eventually we can all lose out. 888 00:41:16,360 --> 00:41:20,200 So it doesn't matter here whether you're talking about climate change 889 00:41:20,200 --> 00:41:23,640 or plastic waste or North Sea fishing. 890 00:41:23,640 --> 00:41:26,880 We often get sidetracked by the really small choices in front of us 891 00:41:26,880 --> 00:41:30,560 rather than holding the big picture in our minds. 892 00:41:30,560 --> 00:41:33,600 And part of that is a mathematical reason, as we've just seen 893 00:41:33,600 --> 00:41:36,520 with that game, when the incentives aren't set up 894 00:41:36,520 --> 00:41:38,840 to encourage really good behaviour. 895 00:41:38,840 --> 00:41:42,000 But I think it's also worth remembering that as humans, 896 00:41:42,000 --> 00:41:45,840 we are not these perfectly rational objects. 897 00:41:45,840 --> 00:41:48,440 If you really want to be lucky, 898 00:41:48,440 --> 00:41:52,680 then you have to take all of the weirdnesses of humans into account. 899 00:41:52,680 --> 00:41:56,800 So, OK, let's have a look at some of your weirdnesses as an audience. 900 00:41:56,800 --> 00:42:02,320 Underneath your seats, you should have a whiteboard and a pen. 901 00:42:02,320 --> 00:42:05,520 What we're going to do, if you get those out and get them ready... 902 00:42:05,520 --> 00:42:08,920 What we're going to do is when I say "go", 903 00:42:08,920 --> 00:42:12,600 I want you to pick a number completely at random 904 00:42:12,600 --> 00:42:14,320 between one and ten. 905 00:42:14,320 --> 00:42:17,160 I want you to write on your whiteboard when I say go, 906 00:42:17,160 --> 00:42:18,960 and then we'll hold them up. OK. 907 00:42:18,960 --> 00:42:21,120 Number between one and ten, completely at random. 908 00:42:21,120 --> 00:42:22,280 Go. 909 00:42:22,280 --> 00:42:24,160 And then hold it up when you're finished, 910 00:42:24,160 --> 00:42:26,040 and let's have a look at what you've got. 911 00:42:26,040 --> 00:42:28,480 OK, OK, let's have a little look at what we've got. 912 00:42:28,480 --> 00:42:29,840 OK. What can I see here? 913 00:42:29,840 --> 00:42:31,160 OK, OK. All right. 914 00:42:31,160 --> 00:42:35,120 I'll tell you what, if you wrote down a one, could you stand up? 915 00:42:36,560 --> 00:42:37,960 Oh! Hardly... 916 00:42:37,960 --> 00:42:40,280 Certainly not 10% of the audience, that, is it? 917 00:42:40,280 --> 00:42:41,760 Thank you very much, sit down. 918 00:42:41,760 --> 00:42:44,280 If you wrote down a ten, could you stand up? 919 00:42:44,280 --> 00:42:46,080 Oh, again, only a smattering. 920 00:42:46,080 --> 00:42:47,920 How intriguing. All right. Sit down. 921 00:42:47,920 --> 00:42:51,000 If you wrote down a seven, could you stand up? 922 00:42:51,000 --> 00:42:53,400 Ooh! I see! 923 00:42:53,400 --> 00:42:55,240 All of a sudden... 924 00:42:55,240 --> 00:42:57,560 Thank you very much. You can sit down. 925 00:42:57,560 --> 00:43:01,360 Now, in fact, actually, if you play this with big groups of people, 926 00:43:01,360 --> 00:43:04,640 it almost always happens that seven is the most common number, 927 00:43:04,640 --> 00:43:07,760 to be chosen. Huge swathes of people will choose a seven. 928 00:43:07,760 --> 00:43:09,880 And there's a kind of strange reason for that. 929 00:43:09,880 --> 00:43:12,760 If you're picking a random number, one feels like it's too small. 930 00:43:12,760 --> 00:43:14,120 Ten feels like it's too big. 931 00:43:14,120 --> 00:43:15,840 Five is kind of too much in the middle. 932 00:43:15,840 --> 00:43:17,320 Two's even - can't choose that one. 933 00:43:17,320 --> 00:43:18,960 Eight's sort of too neat. 934 00:43:18,960 --> 00:43:21,680 All the other numbers fall away, and you're left, 935 00:43:21,680 --> 00:43:23,960 if you're picking a number at random, 936 00:43:23,960 --> 00:43:27,560 with only seven really feeling like the random number. 937 00:43:27,560 --> 00:43:30,400 I guess the point in all of this is that you have to remember 938 00:43:30,400 --> 00:43:33,720 that human behaviour really isn't actually random. 939 00:43:33,720 --> 00:43:36,680 Humans are not very good logical machines. 940 00:43:36,680 --> 00:43:40,160 And what that means is that if you really want to make yourself lucky, 941 00:43:40,160 --> 00:43:43,160 you've got to go beyond the world of maths, 942 00:43:43,160 --> 00:43:45,560 because, you know, there is actually some evidence 943 00:43:45,560 --> 00:43:48,720 that shows that just thinking you are blessed with good fortune 944 00:43:48,720 --> 00:43:51,400 means you can actually make your own luck. 945 00:43:51,400 --> 00:43:53,920 And there is someone who knows quite a lot about this. 946 00:43:53,920 --> 00:43:56,920 This is Dr Michael Gervais in Los Angeles. 947 00:43:56,920 --> 00:44:01,440 He's known as the secret weapon of top athletes around the world. 948 00:44:01,440 --> 00:44:03,280 Hello, Michael. How are you doing? 949 00:44:03,280 --> 00:44:05,600 I'm fantastic. Great to be here. Thank you, Hannah. 950 00:44:05,600 --> 00:44:08,080 You have worked with a number of amazing people, 951 00:44:08,080 --> 00:44:09,280 haven't you, Michael? 952 00:44:09,280 --> 00:44:11,520 I've been fortunate, yes, for sure. 953 00:44:11,520 --> 00:44:13,800 What is it that you do? 954 00:44:13,800 --> 00:44:15,720 Well, by trade and training, 955 00:44:15,720 --> 00:44:18,120 I am a sport and performance psychologist. 956 00:44:18,120 --> 00:44:22,280 So you're trying to get people in the right mind-set to win? 957 00:44:22,280 --> 00:44:26,480 Yeah. There's three things that we can train as humans. 958 00:44:26,480 --> 00:44:29,480 We can train our craft, our body and our mind. 959 00:44:29,480 --> 00:44:32,040 And world-leading thinkers and doers 960 00:44:32,040 --> 00:44:34,960 are not leaving one of those three up to chance. 961 00:44:34,960 --> 00:44:38,600 And the science is informing us of best practices to be able 962 00:44:38,600 --> 00:44:41,920 to train our mind, to be able to adjust to the unfolding, 963 00:44:41,920 --> 00:44:44,120 unpredictable, unknown. 964 00:44:44,120 --> 00:44:46,560 So do you have any top tips for us 965 00:44:46,560 --> 00:44:51,000 as to how to train our minds to get the best out of ourselves? 966 00:44:51,000 --> 00:44:54,680 I wish I had tips and tricks and hacks and short cuts, 967 00:44:54,680 --> 00:44:56,520 and there really aren't many, 968 00:44:56,520 --> 00:44:59,000 and so if we can learn from the best in the world 969 00:44:59,000 --> 00:45:03,240 how to fundamentally organise your life, to strengthen your mind, 970 00:45:03,240 --> 00:45:05,000 to deal with the unknown. 971 00:45:05,000 --> 00:45:08,320 And what we've come to find out is that there are a handful of skills 972 00:45:08,320 --> 00:45:10,920 that people practice - mindfulness being one of them, 973 00:45:10,920 --> 00:45:12,560 optimism being another. 974 00:45:12,560 --> 00:45:14,000 And those are trainable skills. 975 00:45:14,000 --> 00:45:16,200 Confidence is a trainable skill. 976 00:45:16,200 --> 00:45:20,280 Being able to be calm in any environment is a trainable skill. 977 00:45:20,280 --> 00:45:24,360 So there's a handful of a few that are most employed by most. 978 00:45:24,360 --> 00:45:27,840 I hear that you've worked with a few daredevil skydivers. 979 00:45:27,840 --> 00:45:31,680 I think we have a little clip of one of the people you've trained. 980 00:45:31,680 --> 00:45:34,760 Just tell us what Luke Aikins did. He's here in the green. Absolutely. 981 00:45:34,760 --> 00:45:39,280 So, Luke Aikins is one of the most extraordinary base jumpers, 982 00:45:39,280 --> 00:45:42,520 aeronautical flight folks in the world. 983 00:45:42,520 --> 00:45:45,200 And what he did - he was the first ever to do this - 984 00:45:45,200 --> 00:45:49,680 is that he jumped from 30,000 feet, which in and of itself is a big deal 985 00:45:49,680 --> 00:45:53,360 because you need oxygen masks to be able to carry enough oxygen. 986 00:45:53,360 --> 00:45:55,400 But he did it without a parachute. 987 00:45:55,400 --> 00:45:57,640 Without a parachute?! He was the first person... 988 00:45:57,640 --> 00:45:59,880 Yeah, and he's the first person to do that, 989 00:45:59,880 --> 00:46:03,560 and he jumped into a 16-storey net that he and his team built. 990 00:46:03,560 --> 00:46:06,480 And so you were behind the scenes in advance of this event, 991 00:46:06,480 --> 00:46:09,960 trying to get him to focus in the right way? 992 00:46:09,960 --> 00:46:13,560 Yeah. When your life is on the line and the stakes are high 993 00:46:13,560 --> 00:46:15,520 and consequences are real, 994 00:46:15,520 --> 00:46:18,000 nobody leaves very much up to chance. 995 00:46:18,000 --> 00:46:20,560 And so training your mind is one of the ways to help somebody 996 00:46:20,560 --> 00:46:23,400 have great command of themselves under duress. 997 00:46:23,400 --> 00:46:25,840 And so absolutely, yeah, this is one... 998 00:46:25,840 --> 00:46:28,480 This is an incredibly dangerous project. 999 00:46:28,480 --> 00:46:30,080 Do you think he's mad? 1000 00:46:30,080 --> 00:46:31,920 LAUGHTER 1001 00:46:31,920 --> 00:46:33,120 I get that question a lot. 1002 00:46:33,120 --> 00:46:36,840 No, he is actually just like some of the greatest mathematicians 1003 00:46:36,840 --> 00:46:40,080 in the world, just like some of the greatest historians, 1004 00:46:40,080 --> 00:46:42,960 is that they go to the edges of their potential. 1005 00:46:42,960 --> 00:46:45,480 And when somebody goes to the edges of their potential, 1006 00:46:45,480 --> 00:46:48,000 they're taking the next natural step. 1007 00:46:48,000 --> 00:46:49,960 And in some cases, it changes humanity. 1008 00:46:49,960 --> 00:46:53,120 In other cases, maybe it changes a family legacy. 1009 00:46:53,120 --> 00:46:56,720 But this is what the greats do, is they add to their body of work 1010 00:46:56,720 --> 00:46:58,800 by extending their capabilities. 1011 00:46:58,800 --> 00:47:00,720 Now, in those extended areas, 1012 00:47:00,720 --> 00:47:03,280 in that place where they're not quite sure 1013 00:47:03,280 --> 00:47:05,040 if they have what it takes, 1014 00:47:05,040 --> 00:47:09,200 that's where luck... We want to diminish the amount of luck involved 1015 00:47:09,200 --> 00:47:12,080 and make sure that we're increasing them out of skill. 1016 00:47:12,080 --> 00:47:16,480 And so is a lot of this about helping them overcome their nerves? 1017 00:47:16,480 --> 00:47:20,480 Well, certainly the emotional component to exploring one's 1018 00:47:20,480 --> 00:47:23,560 potential towards mastery, towards high performance 1019 00:47:23,560 --> 00:47:27,320 and achievement and success - emotions are part of it. 1020 00:47:27,320 --> 00:47:30,480 And emotions and thoughts and environment 1021 00:47:30,480 --> 00:47:32,920 are the three legs to the stool. 1022 00:47:32,920 --> 00:47:36,600 And while we might not be able to ever really manipulate our 1023 00:47:36,600 --> 00:47:38,960 environment, we can manage our thoughts 1024 00:47:38,960 --> 00:47:40,600 that influence our emotions. 1025 00:47:40,600 --> 00:47:42,160 So those are the three components 1026 00:47:42,160 --> 00:47:44,160 that we want to make sure we're investigating. 1027 00:47:44,160 --> 00:47:47,640 That's amazing. Dr Michael Gervais, thank you very much for joining us. 1028 00:47:47,640 --> 00:47:50,080 APPLAUSE 1029 00:47:55,800 --> 00:48:00,720 Feeling lucky - just feeling lucky - can change your chances of success, 1030 00:48:00,720 --> 00:48:02,560 and as Dr Michael Gervais was saying there, 1031 00:48:02,560 --> 00:48:04,280 it's all about fighting your instincts, 1032 00:48:04,280 --> 00:48:05,560 about overriding your gut, 1033 00:48:05,560 --> 00:48:08,080 stepping outside of yourself and being objective, 1034 00:48:08,080 --> 00:48:11,760 and I think there is something in there that maths can empathise with. 1035 00:48:11,760 --> 00:48:14,240 OK, how about those people who are really lucky? 1036 00:48:14,240 --> 00:48:17,000 So how about the ones... Not just ones who have maths and bravery 1037 00:48:17,000 --> 00:48:20,320 on their side, but how about people who walk away with the really big 1038 00:48:20,320 --> 00:48:23,280 prizes, the ones who stand out from the crowd, right? 1039 00:48:23,280 --> 00:48:25,320 The real one-in-a-millions? 1040 00:48:25,320 --> 00:48:26,760 Well, I'll tell you what. 1041 00:48:26,760 --> 00:48:29,320 Let's try and find one, because what we're going to do 1042 00:48:29,320 --> 00:48:32,120 is we're going to whittle down all of you 1043 00:48:32,120 --> 00:48:35,000 to find the luckiest person in our audience. 1044 00:48:35,000 --> 00:48:36,960 You know how we're going to do it? 1045 00:48:36,960 --> 00:48:38,520 We've got a game show. 1046 00:48:38,520 --> 00:48:40,520 APPLAUSE 1047 00:48:46,320 --> 00:48:49,240 So, are you feeling lucky? 1048 00:48:50,560 --> 00:48:53,360 I'd like to welcome my glamorous assistant, Matthew Parker. 1049 00:48:53,360 --> 00:48:56,360 Now, OK, everybody, if you would like to stand up 1050 00:48:56,360 --> 00:48:59,040 underneath your feet, you should have two hats. 1051 00:48:59,040 --> 00:49:01,120 One is yellow. One is blue. 1052 00:49:01,120 --> 00:49:03,440 What we're going to do is we're going to run a series 1053 00:49:03,440 --> 00:49:05,360 of semi-random rounds. 1054 00:49:05,360 --> 00:49:08,800 All you need to do is put on your blue hat or your yellow hat, 1055 00:49:08,800 --> 00:49:12,280 whichever you think is most likely to win. 1056 00:49:12,280 --> 00:49:16,120 If you are unlucky and you get one of the semi-random rounds incorrect 1057 00:49:16,120 --> 00:49:18,800 I'm afraid you then have to sit down. 1058 00:49:18,800 --> 00:49:20,560 You are out of the game. 1059 00:49:20,560 --> 00:49:23,000 If you get it correct, you stay standing up. 1060 00:49:23,000 --> 00:49:24,800 You're through to the next round. 1061 00:49:24,800 --> 00:49:26,480 OK, are you ready to play? 1062 00:49:26,480 --> 00:49:27,680 Let's go. 1063 00:49:30,280 --> 00:49:31,960 Round 1. 1064 00:49:31,960 --> 00:49:35,440 Which of these two balloons is going to explode 1065 00:49:35,440 --> 00:49:37,520 in a ball of fire? 1066 00:49:37,520 --> 00:49:40,040 Is it yellow or blue? Make your votes now. 1067 00:49:42,960 --> 00:49:44,320 Everyone's voted. 1068 00:49:44,320 --> 00:49:45,760 It's quite loud. 1069 00:49:45,760 --> 00:49:47,240 Cover your ears. Here we go. 1070 00:49:47,240 --> 00:49:49,720 Ready? The blue one is... 1071 00:49:49,720 --> 00:49:52,200 BALLOON POPS ..not explosive. 1072 00:49:52,200 --> 00:49:54,400 Which means the yellow one... 1073 00:49:54,400 --> 00:49:57,000 DEEP BOOM ..is! 1074 00:49:57,000 --> 00:49:59,200 Blue, you are wrong. Sit down. 1075 00:49:59,200 --> 00:50:01,400 Yellow, stay standing up. 1076 00:50:02,600 --> 00:50:05,000 Everybody, hats off. 1077 00:50:06,080 --> 00:50:07,680 Round 2. 1078 00:50:09,320 --> 00:50:11,360 APPLAUSE 1079 00:50:15,280 --> 00:50:17,280 Round two is a race. 1080 00:50:17,280 --> 00:50:21,080 If you think the blue channel cockroach will win, 1081 00:50:21,080 --> 00:50:22,560 put on your blue hat. 1082 00:50:22,560 --> 00:50:25,800 If you think the yellow cockroach will win, yellow hat. 1083 00:50:25,800 --> 00:50:27,200 Everyone ready? 1084 00:50:27,200 --> 00:50:28,840 And they're off. 1085 00:50:29,880 --> 00:50:31,600 LAUGHTER Come on! 1086 00:50:31,600 --> 00:50:33,960 SHOUTS OF ENCOURAGEMENT 1087 00:50:35,800 --> 00:50:37,560 FROM AUDIENCE: Come on, blue! 1088 00:50:37,560 --> 00:50:38,600 Blue, blue, blue! 1089 00:50:41,120 --> 00:50:42,440 Oops. 1090 00:50:47,720 --> 00:50:49,160 And the winner is... 1091 00:50:49,160 --> 00:50:50,760 Blue! 1092 00:50:50,760 --> 00:50:52,400 APPLAUSE 1093 00:50:54,840 --> 00:50:56,440 Yellows, sit down. 1094 00:50:56,440 --> 00:50:58,480 Blues, stay up. Hats off. 1095 00:51:00,480 --> 00:51:02,200 Round 3. 1096 00:51:02,200 --> 00:51:03,280 APPLAUSE 1097 00:51:04,720 --> 00:51:06,400 OK, everyone. 1098 00:51:06,400 --> 00:51:08,080 I've got a pancake here. 1099 00:51:08,080 --> 00:51:10,560 One side of it is yellow. One side of it is blue. 1100 00:51:10,560 --> 00:51:14,720 Once I flip the pancake, which side will land face up? 1101 00:51:14,720 --> 00:51:17,000 Yellow or blue? Make your votes now. 1102 00:51:17,000 --> 00:51:19,040 Here we go. Ready? 1103 00:51:20,600 --> 00:51:22,200 Oh! SHE CHUCKLES 1104 00:51:22,200 --> 00:51:23,840 It's yellow! 1105 00:51:23,840 --> 00:51:26,360 Blues, you lose. Sit down. 1106 00:51:26,360 --> 00:51:28,360 Yellow, you are through to the next round. 1107 00:51:28,360 --> 00:51:30,320 Everybody, hats off. 1108 00:51:31,360 --> 00:51:33,080 Round 4. 1109 00:51:38,840 --> 00:51:41,080 I found this person backstage. 1110 00:51:41,080 --> 00:51:43,800 Is their name Dave? Is it Tom? 1111 00:51:43,800 --> 00:51:46,080 If you think it's Dave, blue hat, Tom, yellow hat. 1112 00:51:46,080 --> 00:51:48,440 Look at him. That's the facial hair of a Dave, 1113 00:51:48,440 --> 00:51:50,680 but the shirt of a Tom. What do you reckon? 1114 00:51:50,680 --> 00:51:52,560 OK. 1115 00:51:52,560 --> 00:51:53,880 Hats on. 1116 00:51:53,880 --> 00:51:55,200 And the correct name is... 1117 00:51:57,200 --> 00:51:59,360 It's Tom! 1118 00:51:59,360 --> 00:52:01,160 Blues, sit down. 1119 00:52:01,160 --> 00:52:02,760 Yellows, stay up. 1120 00:52:02,760 --> 00:52:04,560 Hats off. 1121 00:52:06,920 --> 00:52:08,840 Round 5. 1122 00:52:08,840 --> 00:52:10,800 APPLAUSE 1123 00:52:13,160 --> 00:52:17,640 Who is tonight's extra special celebrity guest? 1124 00:52:17,640 --> 00:52:21,760 Is it Operation Ouch's Dr Xand van Tulleken 1125 00:52:21,760 --> 00:52:25,560 or Operation Ouch's Dr Chris van Tulleken? 1126 00:52:25,560 --> 00:52:28,720 Yellow or blue. Make your bets now. 1127 00:52:32,960 --> 00:52:34,840 And the answer is... 1128 00:52:34,840 --> 00:52:37,160 Hey, everybody! 1129 00:52:37,160 --> 00:52:38,960 Which one are you? 1130 00:52:38,960 --> 00:52:41,840 What?! We've known each other for years. 1131 00:52:41,840 --> 00:52:44,360 It's Xand van Tulleken! OK. 1132 00:52:44,360 --> 00:52:46,480 Yellow, you are through to the next round. 1133 00:52:46,480 --> 00:52:48,480 Thanks, Xand, that's all we need from you. 1134 00:52:48,480 --> 00:52:50,400 Blues, sit down. Everybody, hats off. 1135 00:52:52,160 --> 00:52:53,440 Round 6. 1136 00:52:53,440 --> 00:52:55,680 APPLAUSE 1137 00:52:56,960 --> 00:52:58,520 All right. 1138 00:53:00,560 --> 00:53:02,600 Am I wearing blue socks? 1139 00:53:02,600 --> 00:53:04,640 Am I wearing yellow socks? 1140 00:53:04,640 --> 00:53:07,120 Vote now with your hats. Blue or yellow socks? 1141 00:53:07,120 --> 00:53:09,640 Blue hats for blue, yellow for yellow. 1142 00:53:09,640 --> 00:53:11,880 Yellow, blue, blue, blue, yellow. 1143 00:53:11,880 --> 00:53:13,520 OK. Are you ready? 1144 00:53:13,520 --> 00:53:15,000 They are... 1145 00:53:15,000 --> 00:53:16,840 Have you ever...blue? 1146 00:53:16,840 --> 00:53:18,080 Ready? 1147 00:53:18,080 --> 00:53:20,120 Yellow! 1148 00:53:20,120 --> 00:53:22,880 Blues, you're down. Yellows, stay up. 1149 00:53:24,360 --> 00:53:26,440 Who've we got? One over there. 1150 00:53:26,440 --> 00:53:28,400 We've got two over there. 1151 00:53:28,400 --> 00:53:29,840 We've got one over there. 1152 00:53:29,840 --> 00:53:31,720 OK, everybody. Hats off. 1153 00:53:33,800 --> 00:53:35,440 Round 7. 1154 00:53:35,440 --> 00:53:37,000 APPLAUSE 1155 00:53:40,280 --> 00:53:44,120 Which one of my two party blowers is going to be longer when I blow it? 1156 00:53:44,120 --> 00:53:46,520 Is it going to be blue or yellow? 1157 00:53:46,520 --> 00:53:47,960 Make your vote now. 1158 00:53:49,240 --> 00:53:51,400 OK. They've made their bets. All right, here we go. 1159 00:53:52,360 --> 00:53:53,960 PARTY BLOWERS SQUEAK 1160 00:53:53,960 --> 00:53:55,800 And the answer was blue. 1161 00:53:55,800 --> 00:53:58,040 Yellow, you lose. 1162 00:53:58,040 --> 00:54:00,080 Blue, you're through to the next round. 1163 00:54:00,080 --> 00:54:01,720 Final Round. 1164 00:54:02,920 --> 00:54:04,920 APPLAUSE 1165 00:54:06,880 --> 00:54:08,120 OK. 1166 00:54:08,120 --> 00:54:10,160 We've got three people left. 1167 00:54:10,160 --> 00:54:13,200 This one is a straight run between me and Matt 1168 00:54:13,200 --> 00:54:16,440 as to who can get the balloon on their head to pop first. 1169 00:54:16,440 --> 00:54:18,040 Who wants to go... We need to... 1170 00:54:18,040 --> 00:54:19,800 You need to pick a team this time. 1171 00:54:19,800 --> 00:54:21,600 So who wants to go for me? 1172 00:54:22,840 --> 00:54:24,920 Thank you. And who's going for Matt? 1173 00:54:24,920 --> 00:54:27,920 Thank goodness that music doesn't get annoying. 1174 00:54:27,920 --> 00:54:30,360 We have got... Let's have a look. 1175 00:54:30,360 --> 00:54:33,040 We've got two blue and one yellow. Interesting. 1176 00:54:33,040 --> 00:54:34,520 Interesting. OK. 1177 00:54:34,520 --> 00:54:36,880 OK, ready, Matt? Ready. 1178 00:54:36,880 --> 00:54:38,600 Three, two, one. 1179 00:54:38,600 --> 00:54:39,680 Go. 1180 00:54:41,280 --> 00:54:43,120 AUDIENCE SHOUTS ENCOURAGEMENT 1181 00:54:47,160 --> 00:54:48,560 BALLOON POPS 1182 00:54:48,560 --> 00:54:50,160 CHEERS AND APPLAUSE 1183 00:54:51,960 --> 00:54:53,720 Right. 1184 00:54:55,560 --> 00:54:59,800 Tie breaker, will the blue/yellow coin land blue or yellow? 1185 00:54:59,800 --> 00:55:01,280 You've gone for yellow. 1186 00:55:01,280 --> 00:55:03,160 You're going to go for blue. Are you ready? 1187 00:55:03,160 --> 00:55:04,760 And the winner is... 1188 00:55:05,960 --> 00:55:09,360 Yellow! You're the luckiest person in the room! 1189 00:55:09,360 --> 00:55:10,400 The luckiest person! 1190 00:55:10,400 --> 00:55:12,040 APPLAUSE 1191 00:55:15,400 --> 00:55:17,320 Amazing! 1192 00:55:23,480 --> 00:55:24,880 The winner! 1193 00:55:24,880 --> 00:55:27,440 Abi, the luckiest person in our audience, 1194 00:55:27,440 --> 00:55:29,680 we've got a crown for you. 1195 00:55:29,680 --> 00:55:31,320 We've got a trophy for you. 1196 00:55:31,320 --> 00:55:33,360 We're not going to give these to you just yet 1197 00:55:33,360 --> 00:55:34,920 because we really want you to prove 1198 00:55:34,920 --> 00:55:36,960 that you are the luckiest person in the audience. 1199 00:55:36,960 --> 00:55:38,800 So what you've got to do, Abi, 1200 00:55:38,800 --> 00:55:42,240 is you have got to just land a ping pong ball into that, OK? 1201 00:55:42,240 --> 00:55:43,880 Now, we know you're lucky. 1202 00:55:43,880 --> 00:55:45,720 I mean, you just beat everyone in this room, 1203 00:55:45,720 --> 00:55:47,720 so we're not going to make it too easy for you. 1204 00:55:47,720 --> 00:55:49,520 Come and follow me. 1205 00:55:49,520 --> 00:55:52,440 We're going to shoot the ping pong ball from up here. 1206 00:55:54,280 --> 00:55:55,960 Let's see how lucky you really are. 1207 00:55:55,960 --> 00:55:57,480 We want you to get that crown. 1208 00:55:57,480 --> 00:55:59,200 We want you to get that trophy. 1209 00:56:00,280 --> 00:56:02,360 So if you just want to stand in there. 1210 00:56:02,360 --> 00:56:06,560 Your job now, Abi, is to try and fire it into there. 1211 00:56:06,560 --> 00:56:08,360 Let's give her some encouragement. 1212 00:56:08,360 --> 00:56:09,760 Yes! Come on! 1213 00:56:09,760 --> 00:56:11,520 CHEERS AND APPLAUSE 1214 00:56:13,520 --> 00:56:15,360 Come on, you're the luckiest person. 1215 00:56:15,360 --> 00:56:17,280 Not bad. 1216 00:56:17,280 --> 00:56:18,760 OK, we'll give it another go. 1217 00:56:18,760 --> 00:56:20,640 Come on. We really want to get the crown. Come on. 1218 00:56:21,880 --> 00:56:23,680 Oh, it's pretty close. 1219 00:56:25,040 --> 00:56:28,480 It's almost like they weren't lucky and someone just had to win! 1220 00:56:28,480 --> 00:56:32,960 We can speed this up if we bring in the string. 1221 00:56:37,520 --> 00:56:39,520 Thank you for the string! 1222 00:56:41,160 --> 00:56:43,640 Of course, because the more times that you do something, 1223 00:56:43,640 --> 00:56:45,440 even if it's really unlikely, 1224 00:56:45,440 --> 00:56:48,320 eventually it becomes a mathematical certainty. 1225 00:56:48,320 --> 00:56:49,680 Here we go. 1226 00:56:52,160 --> 00:56:55,320 Abi, we're going to put some safety glasses on you. OK. 1227 00:56:55,320 --> 00:56:58,320 We're going to give you a scary-looking glove. 1228 00:57:02,560 --> 00:57:04,120 OK, so this sets fire. 1229 00:57:04,120 --> 00:57:06,680 Can you give us a test click, make sure it's working? 1230 00:57:06,680 --> 00:57:08,960 OK. You're going to, in a second, we'll count down. 1231 00:57:08,960 --> 00:57:11,480 You're going to set fire to the string. Are you ready? 1232 00:57:11,480 --> 00:57:15,480 Count down from... Everyone. Three, two, one, go! 1233 00:57:15,480 --> 00:57:17,320 Whoa! 1234 00:57:36,440 --> 00:57:39,160 Surely one's going to go in! 1235 00:57:39,160 --> 00:57:40,200 Come on! 1236 00:57:50,600 --> 00:57:52,120 Whoa. 1237 00:57:52,120 --> 00:57:54,680 Hang on, hang on, hang on, hang on. 1238 00:57:54,680 --> 00:57:56,880 There's one in there! 1239 00:57:56,880 --> 00:57:58,360 CHEERS AND APPLAUSE 1240 00:57:58,360 --> 00:58:00,560 You're the luckiest person in the audience! 1241 00:58:03,040 --> 00:58:04,560 So, OK, everyone. 1242 00:58:04,560 --> 00:58:07,600 We have just discovered how we can use our understanding of maths 1243 00:58:07,600 --> 00:58:10,080 to get lucky, and in the next section, we're going to explore 1244 00:58:10,080 --> 00:58:12,680 how we can up our chances of winning at even bigger challenges, 1245 00:58:12,680 --> 00:58:15,480 and learn that bending the rules can sometimes be good. 1246 00:58:15,480 --> 00:58:16,880 Goodnight. 1247 00:58:16,880 --> 00:58:18,520 The luckiest person! 1248 00:58:18,520 --> 00:58:21,000 CHEERS AND APPLAUSE