"Steve: Welcome to Science Talk, the weekly podcast of Scientific American for the seven days starting June 25th, 2008. I'm Steve Mirsky. This week on the podcast, Princeton astrophysicist J. Richard Gott talks about presidential electoral polling, of course, as well as some fun thoughts about the realities of time travel. If you've listened to the past few episodes you know that Scientific American editor Mark Alpert has a new novel out called Final Theory. J. Richard Gott was Mark's advisor in college. I had heard he was working on polling data analysis, so when I ran into him at a party on June 10th to celebrate the publication of Mark's book, I asked him to talk a little bit about the polling work as well as his time travel thoughts. Toward the end, you will hear a third voice that's Scientific American editor George Musser.
Steve: Dr. Gott, tell me about this research that you've done in, of all things, election results and polling data.
Gott: Well, Wesley Colley and I decided that we would like to try to apply median statistics to polling to try to predict the electoral vote in 2004, and the reason for this is that the median is very good when you have different results and one of them might be discrepant. So, Zeldovich, the famous astrophysicist, once said that in Russia they didn't make watches very well, so when friends got together they would compare the time; and so, one would say "it's 5 minutes to 10" and the other would say "it's 10 o'clock" and the other one would say "it's 4:30," so they say take the median, take the middle line.
Steve: This is like the statisticians who go duck hunting. You must know that story as well.
Gott: What is that story?
Steve: Four statisticians go duck hunting, one shoots high, one shoots low, one shoots left, one shoots right and they all say, "We got him."
Gott: Yes, shoot in the middle. Take the median, this is what we did. So we applied this in 2004 and the interesting thing was [that]it, a lot of states, crucial states like Ohio, Pennsylvania, and Florida were too close to call in an individual poll. So there would be results like, "Bush is leading Kerry 49 to 47 percent in Ohio, but it is a 4 percent margin of errors, so Ohio is too close to call." They said all those three states were too close to call, but we just kept adding up who was winning in those states over a period of a month. And then you would find out that Bush was significantly ahead in Ohio, significantly behind in Pennsylvania, and Bush was significantly ahead in Florida. So those states were really locked up about a month ahead of time and they never changed. There wasn't any trend to go in on. So we were in the end, our method of just saying, "Bush ahead in a given state", well the prediction has the most polls, who is ahead in the number of polls, forget about the margins of error, forget about the average, calculating the averagejust ask who has the most wins in a given state. And that works too over the previous months, to give us all the states correctly except for one, which was Hawaii. So we calculated the electoral vote very accurately, and whereas many people thought the election was too close to call or Kerry might even have a victory, we correctly called it in advance.
Steve: Now how do two nice astrophysicists wind up getting involved in the gutter world of politics?
Gott: Well we would use these median statistics to investigate the question of the acceleration of the universe. The universe is expanding faster and faster and a team of scientists, two teams of scientists, had found this result. But we wanted to check to see whether it was true using the median statistics. They measured brightness of supernovae and they said that these supernovae that were seen too bright or too fade[faint] given in a cosmological model. And they found that the cosmological model they[that] fit them the best and the average was the accelerating model. But we wanted to check and see if that was true for the median, because it might be that a fewer of those supernovae were way off[center]setter and they were pulling the average up, so what we found was that it was true for the median as well. In other words, you needed that accelerating universe to fit the data for most of the supernovae; in other words, most of the supernovae were at the wrong brightness if the universe's expansion wasn't accelerating. And now it is an important confirmation of what they previously found using the average.
Steve: Why don't we just real click [quickly] explain the difference between the average and median for anybody who hasn't taken their statistics classes for quite a few years?
Gott: Well, in that example I gave, 5 minutes to 10 o'clock and 4:30, the average is slightly afternoon isn't? [That's]The average of those three times, but the median, the one in the middle, that's 10 o' clock and that's probably right.
Steve: Right, right. So your papers are not actually going to come out until shortly before the election, but it is getting some attention already.
Gott: That's true. Well, Neil Tyson took our method and he applied it to the current situation. He was interested in seeing how Obama would do versus McCain and how Clinton would do versus McCain, and if you gave the states to who was ahead in the median poll, you put states where you had no polls or ties, just the same way that they went in 2004, they were 19 states with new polls, and what they showed was that currently at that time Obama had 252 electoral votes and Mrs. Clinton would have had 295. This is somewhat surprising to people, I think, and we look to the similar set of data from like April 26 to May 26 over a month period. For example Mrs. Clinton was leading McCain three polls to nothing in Florida and leading two polls to nothing in Ohio. Obama was behind two to one in Ohio and behind in Florida three to nothing. He was also behind McCain in Michigan by three to nothing and McCain was tied with Mrs. Clinton, and so that it had been a Democratic state before so it went to her. We just did a recent update on this and I think Obama has, I think, he went up to 256 electoral votes. The close states are really Virginia which is not a tie, McCain had been leading two to one in that state, Obama is winning in Colorado, in Iowa, and both he and Mrs. Clinton winning the popular vote. So Obama is ahead in the popular vote, the median statistics say, and the electoral vote, although he is behindif he picks up a few of these states he could be aheadbut he is currently behind in Michigan which is the state that Democrats won before. So this also tells you where the candidates might profitably spend their time. So Virginia is a close state, Missouri is a close state, New Mexico Obama is ahead in, and he is not ahead in Nevada; so it tells you where they might campaign. One of the things that we noticed, a difficulty if you will, with the electoral system is that we noticed that the close states are inordinately important. The states last time of Ohio, Pennsylvania, and Florida, the candidates spent a lot of time there and they made like 80 some odd trips to these states, you know; and so the states like California, New York, Illinois, Texas, they hardly visited these states, maybe except to raise money. But all the interest in the campaigning is focused on the few close states where it might make a difference. So if it was just by the popular vote, the candidates would campaign all over the country, you know; but here in the electoral system, it is very important what happens in the close states and in the end you might have only two or three states that are really truly up for grabs. And if people get better at polling, using our method and others, well then they can rightly and will focus their attention on those key states. Indeed Obama has been recently to Virginia and so he knows that's an important state. We're putting this up on Wesley Colley['s] Web site; it is called www.colleyrankings.com. He is one of the several people that do computer rankings for the football teams that decides which team gets to go play in the national championship, so thiswithin a week or soI'm sure [we'll] put up an automated method that will show you the results as of that day and show you how the candidates are doing. You can see which of the states are up in the polls and which they are tied in and we will keep the total of the electoral voters that goes along and we can see whether it stays the same or it changes.
Steve: And how do you spell Colley?
Gott: C-o-l-l-e-y.
Steve: Colleyrankings.com?
Gott: Colleyrankings.com.
Steve: Interesting, now I want to switch gears. You've done some kind of fun writing about time travel and general relativity. So what's the bottom line on time travel? Are there time travelers among us, and they're keeping themselves secret?
Gott: Well no! That's what[one of] the interesting things we found out. Time travel seems to be possible in Einstein's theory of general relativity. It has a number of unusual solutions including the black hole solution, which you've heard about, the expanding universe solution, which you've heard about, but there are also solutions to Einstein's equations that allow time loops to occur, where you can circle back in time and visit an event in your own past and whether or not you can actually realize those solutions, we may need to know the laws of quantum gravity. We know how gravity behaves on large scales, on macroscopic scales, but we may need to know how gravity behaves on microscopic scales. This would be new physics that we haven't done yet. But Einstein's theory of gravity, which is general relativity, which is the best theory of gravity that we know, it does allow these time travel solutions. These are only like wormholes or cosmic strings. These are only solutions that supercivilizations could really even contemplate. And one of the things we learned about these solutions is that if you build a time machine by twisting space and time so you can circle back and visit an event in your past, you can use the time machine to go back before the time machine was created.
Steve: Really. That's pretty interesting.
Gott: So if you bend space and time up in the year 3000 so as to create a time loop, you might use it to go from 3002 back to 3001, which you can use it to come back to here, because that's before the time machine was invented. So the reason, it is a sufficient explanation for why you don't see any time travelers at the Kennedy assassination, filming that event in [a] white spacesuit, you know, you don't see that. You don't expect any time travelers to go back before the first time machine is even invented. We haven't invented one yet. So you don't expect to see any.
Steve: About 100,000 years from now, the place might be lousy with time travelers from a 100,003.
Gott: Yes, one of the things that can occur is that even though you inspect the past very carefully for time travelers and never find any, when you enter a time machine, suddenly a time traveler to the future can show up and say hello!
Steve: Interesting. Now in a real but possibly trivial sense, we are all time travelers right now, right?
Gott: Well, [in] one waywe are all are going towards the future. So, you can time travel towards the future fast. We know that this is possible. Einstein's theory of special relativity which he did in 1905, show[s] that moving clocks ticks slowly and so astronauts, the greatest time traveler we have so far, is an astronaut named Sergei Krikalev.
Steve: He has been circling very fast for months and months...
Gott: About 800 dayshe was in orbit going 17,000 miles an hour. So, he is right now a one forty-eighth of a second younger then he would have been if he had stayed home. So, that means that when he came back to the earth, he found that one forty-eighth of a second to the future of where he expected it to be. So this man has actually time traveled one forty-eighth of a second to the future.
Steve: Now we all do that to an incredibly infinitesimal event when he[we] takes a jet plane across the country.
Gott: That's true. It's just, you might get a few nanoseconds, you knowif you go around the world, on a trip around the world, thenthey've actually done this where they've sent a jet plane around the world, and when they come back it's about 50 billionths of a second to the future of where as it has aged 50 billionths of a second less. Now if you would go fast, faster: Let's say you went at 99.995 percent the speed of light, 500 light-years out to the [a] star [unclear 15:17], and then came back finally, the earth would be a 1,000 years older, and you will have aged 10 years. So this is how you can visit the world a 1,000 years [from] now. It's just very expensive.
Steve: And it's a one-way trip.
Gott: Oh! No, no, no! You can go out and come back.
Steve: No, I mean it's a one-way trip into the future.
Gott: Oh yes, yesto come back then you need one of those time loops. You actually need to twist spacetime, and this is possible as space and time are curved. So Magellan's crew left Europe and they went west, west, west around the earth and came back to Europe. This is impossible if the earth is flat but it is possible if the earth is curved. So with Einstein's theory of curves[d] space-time, so its loops coming back are actually possible; [but] we haven't built any one of those yet.
Steve: Good stuff. What are your, you have any personal time travel plans?
Gott: Well, people often ask me, "Well where would you go and visit?" Well, I would say, the past, what we already know [about] what happened in the past, so if you wanted to know what Cleopatra looked like, well actually there is picture of her on a coin, you know. So I would visit the future, so, and we know you can do that if you go fast enough. So, I would like to go about 200,000 years in the future and see what had happened to the human race, whether we are still around, whether we are extinct, or what we are up to!
Steve: And the first words you will say are "Get your hands off me, you dirty damned apes."
Gott: You could be in for a few real surprises!
Musser: Out of curiosity, if you then did that initial experiment where you went 1,000 years into the earth's future and you only age 10 yearsyou could take the entire earth, send it to [unclear 17:10] [the same star] and back and then you would finally be in sink [sync] again.
Gott: That's true. One of my students once wrote a science fiction story. I had them write science fiction stories based on real science and he said, "Well there would be [a] Broadway play: In the first act of [it took] place [in the present] and the second act took place 20 years later. They reassembled the actors, they too sent the audience off on a spaceship very fast and back so they only aged in days, and then they would see the second act and the actor is old. [Then] they would send the audience away again and bring them back, and then in the third, act the actors were 40 years older, so you could use it for that.
Steve: That's a Tony Award winner, no questions.
Musser: Absolutely.
Steve: J. Richard Gott also has done some thinking about using probability theory to estimate the length of time a particular thing or species will survive. There's a lot of interesting material on that in his Wikipedia entry. The June 6th Neal Tyson article that Gott mentioned was in The New York Times; just go to their Web site and search for Richard Gott and it comes right up; the article is called "Vote by Numbers". By the way if you go to the colleyrankings [site] you'll see that in the days after Gott and I spoke againwe spoke on June 10ththere was a big change in the electoral vote analysis based on swings in polls in Virginia and Ohio. Again the polling data is at www.colleyrankings.com.
Steve: Now it's time to play TOTALL....... Y BOGUS. Here are four science stories; only three are true. See if you know which story is TOTALL....... Y BOGUS
Story number 1: University of Florida Professor has designed what should be a working flying saucer.
Story number 2: Researchers have pinpointed the date of a solar eclipse described in Homer's Odyssey.
Story number 3: Half of the world's 24 million known chemical compounds fall within only 143 basic shapes.
And story number 4: The German cockroach is making a big comeback in urban environments.
Time is up.
Story number 1 is true. Florida engineering professor Subrata Roy has designed a prototype flying saucer. Electrodes on the surface would ionize the surrounding air into plasma; passing an electric current through the plasma would create force, which would in turn get you lift. So there would be no moving parts. You'll still lose your luggage.
Story number 2 is true. Researchers have used other astronomical clues in the Odyssey to pinpoint an eclipse mentioned in the epic to April 16, 1178 B.C. They published their report in the Proceedings of the National Academy of Sciences. For more, check out JR Minkel's article on our Web site. The Odyssey of course tells the story of Odysseus, [who] attempts to find Ithaca, which isn't easyjust ask anybody who went to Cornell.
And story number 3 is true. Half the known chemical compounds share only 143 basic shapes. That's according to a study in the Journal of Organic Chemistry. The author suggests that by evaluating compounds with shapes that fall outside the group of 143, chemists might discover novel rings, linkages, and atomic groupings with potential applications in medicine or materials.
All of which means that story number 4, about the German cockroach making a big come back is TOTALL....... Y BOGUS, because the German cockroach didn't have to make a big comebackit's going strong. It's the roach commonly found in homes, however; a couple of days ago I evicted an American cockroach from my house. The Humongous American cockroach keeps German cockroaches as pets. I captured the roach in a large jar and put it out in the street where [it] carjacked the Mini Cooper some Web searching turned up that fact that the American cockroach really prefers the South for its hot weather and that citing [sightings] in New York City, where I live, though not rare or not [aren't] all [that] common either. So with global warming; we can perhaps look forward to gigantic roaches in more of our homes.
Now I'll tell you another cockroach story. There was a great researcher named Berta Scharrer. She used the American cockroach as her research subject and the American cockroach from the tip of the antenna to the back end, it's got to be a good three inches longand I based that on my personal observations this week. So Berta Scharrer was working with American cockroachesthis is back in the 1940sand there was a cockroach infestation in the building in which she worked. It was an infestation, however, of German cockroaches. Nevertheless, the all-male faculty blamed Berta for the infestation despite the fact that the species that she worked on was not the species that infested the building, just one of the things that she had to deal with being a woman scientist way back then.
(music)
Well that's it for this edition of the weekly SciAm podcast. Visit http://www.SciAm.com for the latest science news, videos and slideshows and sign up for the daily digest at http://www.SciAm.com/daily, I did. For Science Talk, the weekly podcast of Scientific American, I'm Steve Mirsky. Thanks for clicking on us."
Posts
No comments:
Post a Comment