 East Bay, number 44. I'm co-author of the East Bay, Rick Kronisky. Thanks for coming out, this is going to be fantastic. Let's get started real-way. So I like the next talk. It lets me talk about this, coin flips. So coin flips are amazing stuff. Not only do they decide how many of our sporting events start, they're apparently used all the time for how we name the corporation. So instead of Packard Hewlett, we got Hewlett Packard just based on the coin flip. Same thing with Baskin Robbins, it's pretty amazing. Portland, Oregon. Do you know this shit could have been Boston, Oregon? That does not sound right. I found our next speaker mostly because of my obsession about coin flips. He wrote a follow-up to this paper. So in this paper, they wanted to know whether or not flipping the coin actually gets you 50-50 odds. And certainly you think it's this sort of mechanical system. You know the way to the coin. You presumably figure out what angle you're flipping it at. It should be predetermined. And in fact, you could build a machine that always flips the same way. But at least these guys said that there is a bias and it's about 1%. So 51-49% that they would think for the same side that it was up. Now their model only took into account this rotating coin. So David Alders at his undergrad actually tests this a few years back. So some poor student taking his mathematics class apparently spent one hour per day for an entire semester tossing it going 40,000 times. Whether or not they found evidence of this bias, I'll leave it to you to read their research findings. So I have to say it was not quite as strong as the original paper proclaimed. So coins can also land on their edge. So important things have been cited this way as well. Jeremiah Clark was a very, very sad composer and wanted to commit suicide, but he didn't know whether to drown himself or to hang himself. So he flipped a coin, but he did so in the mud and it landed on its edge. And therefore he shot himself. Here's a lovely quote about... And apparently these statisticians have since figured out the odds of a coin landing on section on a perfectly flat surface for a coin that has the same aspect ratio of a nickel. It should happen about one in every 6,000 times. So I'm looking forward to the next set of undergraduates looking at thicker coins and figuring out how many land on their edges. David Ollis's students have blocked the cool projects. This is just this year's batch, but I'm really pleased to introduce him tonight because he'll talk to you more about all the usual occurrences of probability in every life. Here's David Ollis. Microphone! Oh hell. Normally people complain I talk too loudly. He probably took a statistics course. Yeah, he probably took a statistics course in college and it happened for a bit of probability. But frankly what we see there is pretty boring. So some time ago I teach at Berkeley. I teach a course on what's not in other courses. I do this by giving 20 lectures on totally different things, as different as I can think of to do with probability. And at the end of the course I ask students to promote the like and dislike on the different classes. And this is the sort of like minus dislikes here. So again, that's what I talk about. And in this talk I'm just sampling a few of the things from the Berkeley course. On the top one, we're seeing different things on there and on here as some of you've done. But over there one of the most favorite ones is Everyday Inception of Chance. So just as an outsider textbook in everyday life, in what context do you actually think in terms of chance or probability? In the field an idea where that would be very hard to figure out. I mean I could ask you this but you wouldn't really know how to answer. Nowadays we have all sorts of data we can get. You can look at blogs, you can look at tweets. I happen to be able to get, so just as the search engine thing had come online, I happen to get asked for and get a file of every query ever made to being containing the traces, probability or chance of. And there are 100,000. Google probably has 10 million by now. The search engines save every query it's ever made. You may or may not have thought of that but they do. So when I'm giving this talk at sort of our length, I sort of ask the audience to try and guess what other people interested in when they're asking a search engine for the chance of how probability are. And academics are mostly pretty bad at doing that. So in a 20 minute talk I'm going to ask you to give suggestions right on the board. I'll just show you. So what I actually did is I had enough patients to sample for the 670 that they used and manually put them into piles of 10 similar ones and picked out representative ones from each sample. So on my web page you can see a list of 67, some of which I'll show you. So here it goes. This is what people are interested in. So about half of the total is sort of health related and half of those, that is about a quarter of a total, that's sort of pregnancy contraception, et cetera, over there. And then there are other sort of health related ones. And then when you get away from health it's sort of all over the place. So there's a bunch here, probability of birthpeak in Chicago, lots of chance to become an adult, et cetera. Because the people here are I think striking naive users of search engines to believe you can actually sort of answer some of these questions. But anyway. So the rest of the bottom line to this, I think of it poetically as sort of a map of the world of chance. I'm trying to understand, again, what context people think in terms of chance. So as you can imagine, you know, what I want to get from tweets is very different. People don't tweet so much about their concerns about being pregnant. Or not. That's why you get a different set of everyday examples here. Anyway, this just sort of makes my initial point that the things you see about probability and statistics, textbooks, which are sort of a lot of different from what people perceive in everyday life. Moving on, these are sort of great personal concerns that people think about. Moving on to the bigger picture. Prediction. Often when we think about probability, something in the future where a lot happened, lots of chance that it happened. Again, we can think of that in personal life. But on the bigger picture, I think of geopolitics here. It's a funny world. I know probability. So to be self-evident, I think about who our next president is going to be. That's not all about Hillary. Well, they're probabilities. You don't know. You want to think in terms of probabilities. Discussion of such future things on television isn't done in terms of probabilities. People just say, I think this is going to happen. I think this other is going to happen. But there is a smaller world where people do think of in terms of probability here. Because I don't have a magic way of figuring out probabilities for real-world events. But there are other things that one can do math way. So there's a notion of a prediction tournament. Somehow, okay, this is not turning up on the screen or the way I was hoping it was going to. But you can roughly see some of those. They're just turning up better over there. So there's an ongoing game, if you like, called the Move Judgment Project which you can sign up for and play. And at any time you have a bunch of questions of the format, you will have a certain thing happen before a certain deadline. Will any member of China's Politburo design be suspended or be removed during 2016? Will there be an appointment to be secretary-general? UN, et cetera? So do you get questions about politics, economics, or those other worlds? What you ask to do is to just predict the probability of this happening as of today. And then, of course, it updates as time goes on as relevant news comes in. So the figures being shown on the screens are a kind of average probability of consensus but you score points in terms of where you can predict these things better. What's interesting here is sort of how you actually score who's good at this and who's bad at this. Again, the thing to emphasise is that I guess we're asking people to predict yes-no on each of these events. And at the end of the day, you just counter the questions they got right. That seems to be the score. We're specifically not asking them to predict yes-no, we're asking them to write down a probability. So you have a bunch of different questions. If I'm saying this game, I have a probability of each question and I'm now ignoring the fact that we update as time goes on and then we find out whether they happen or not. Now it's still quite obvious how much scores and how many people figure out if I'm good or bad at making these predictions. OK, let's go off the stage here. To try to match the other, I'm going to have a few algebra symbols on the next slide and no one else. But the bottom line behind algebra is that in this type of competition, we get...