Bruce Bueno de Mesquita Uses Games To See the Future

parallel_prankster writes "Bruce Bueno de Mesquita is a professor of politics at New York University and a senior fellow at the Hoover Institution at Stanford University in California. In his new book, The Predictioneer (The Predictioneer's Game in the US), he describes a computer model based on game theory which he — and others — claim can predict the future with remarkable accuracy. The website also has a game page where he provides an online version of the game and information on how to play." The (semi-paywalled; may need to register) New Scientist has a story on de Mesquita, too; a snippet: "Over the past 30 years, Bueno de Mesquita has made thousands of predictions about hundreds of issues from geopolitics to personal problems. Overall, he claims, his hit rate is about 90 per cent."

Comments

If he isn't already rich then he's lying

[pizza_milkshake]

But then, he knew I'd say this.

Re:If he isn't already rich then he's lying

[red_blue_yellow]

From TFA:

How is such accuracy possible? What Bueno de Mesquita is not doing is predicting random events such as lottery draws. Nor does he claim to be able to forecast the movement of stock markets, the outcome of general elections or the onset of financial crises - events where millions of people have a small influence, but none is able to move the market on their own.

Rather, he confines himself to "strategic situations" where relatively small numbers of people are haggling over a contentious decision. "I can predict events and decisions that involve negotiation or coercion, cooperation or bullying," he says. That includes domestic politics, foreign policy, conflicts, business decisions and social interactions.

Now, that's not to say that he couldn't make money using the predictions, but maybe he's actually more interested in the science/mathematics side of it than the business potential?

Re:If he isn't already rich then he's lying

[BigSlowTarget]

So I guess he doesn't need funding then? Oh and doesn't have any good causes he feels could benefit from donated earnings?

As for not predicting the movement of stock markets, if you can predict business decisions you can predict selected market movements. Will XYZ get the contract? Their stock will go up. Will Mr. suchandsuch decide to buy company z? You can bet company Z stock is going to rise. When you make or lose money based on your predictions every prediction is documented - if you're up you're right and if not you're wrong.

The value of a "90 pct hit rate" can only be reasonably compared to a combination of other forecasting methods and random chance. Documentation of every prediction, wrong and right is essential.

Re:If he isn't already rich then he's lying

[zmollusc]

You used 'lose' instead of 'loose' when the correct word is 'lose'. Is the interweb broken? Next we will see someone in a car analogy slowing the vehicle by applying the 'brakes'.

Re:If he isn't already rich then he's lying

[Anonymous Coward]

Not if they specify Toyota.

Re:If he isn't already rich then he's lying

[wizardforce]

His model seems to do well in cases where a relatively small number of people have significant influence on the outcome and since accurate predictions about the stock market require a model that both accurately predicts events in industry/production as well as the influence of large numbers of stockholders it probably wouldn't be wise to apply his model to that kind of situation; he even admits as much, that it's only good for fairly small groups.

Re:If he isn't already rich then he's lying

[SerpentMage]

I read part of his book since his book uses what one would call behavioral economics. A very interesting field. What I don't like about him is that he manipulates (read the book). And as such believes that manipulation is everything. I stopped reading half way because I hated his style. That is his flaw since not everybody can be manipulated. People can act irrationally. That he does not account for in his model, and is what I would think is a major flaw. It is sort of like saying, "Greece will be bailed out, Greece will be bailed out" Why? Because that the logical choice and would be according to this guy who does predictions.

What they all forget is Germany... Germany is now saying, "hey Greece head over to the IMF" It is at that point the financial community says, "ooops..." But if you are investing money it is a big f****g oops! And that big f****g oops is what causes financial companies to loose money...

Re:If he isn't already rich then he's lying

[SerpentMage]

In my early investing career I asked in a forum, "if somebody has a money making strategy why would they not use it themselves?"

Two answers:

1) They don't have enough capitalization and are trying to attract money.
2) Snakeoil...

1 does happen, albeit not that often, but it does. 2 is the more common answer.

Re:If he isn't already rich then he's lying

[Imsdal]

As MAD Magazine said years ago: "The best things in life are free. The expensive part is paying for the dinner and movie that comes first".

Re:If he isn't already rich then he's lying

[fractoid]

Yeah. Like that 'rich dad poor dad' guy. He thinks he's a financial guru but what he really did was make a modest fortune by sniping mortgage foreclosure auctions (in essence, ripping off the struggling families who'd just lost their homes), and then when that stopped working, he switched tactics to making money by selling "how to make money" books and giving seminars.

Re:If he isn't already rich then he's lying

[dargaud]

Yeah, it's like the turkey predicting that tomorrow it will go to sleep well fed. It will be wrong only once after all. Just before thankgiving. Not a bad score, right ?

Re:If he isn't already rich then he's lying

[jandersen]

Why would you assume that the first thing anybody with a good idea would do, is to go and make money from it? Things like altruism and curiosity for its own sake are arguably some of the traits that make humans "human"; and there are many things that are much more interesting and satisfying than money.

That aside, the purpose of game theory as such is to predict the behaviour of systems, so it isn't so surprising that they achieve some success. The big problem, as far as I can see, is to create a model that is realistic enough, pretty much like when you predict the weather, although the method is different. And then you have to know which questions to ask, since the answer may well be something like "42".

Prediction is not actually that difficult in itself; you do it every day when you say things like "Tomorrow I will go to a meeting" - and sure enough, next day you do go to a meeting. This is trivial, of course, but that is all there is to it; the rest is down to how many data you have and whether you are able to take it all into account.

And that, funny enough, is why using a techique like Tarot or I Ching can sometimes be amazingly effective. When you lay out Tarot cards, it is of course entirely random, but that is exactly why it works; because when you try to think about the consequences of some important problem, you are likely focusing too strongly on only a part of the available data, and trying to interpret the random set of images you've laid out breaks you out of the box you have created for yourself - it is, in effect, a form of brainstorming.

Re:If he isn't already rich then he's lying

[Jiro]

First of all, there are so many things that money can buy that it's difficult to think of how someone who isn't a monk or equivalent wouldn't want some more. Does he have a mortgage? Children or grandchildren interested in going to Stanford? Even assuming his personal needs are met, enough money could, oh, endow a university and create a whole research center just to study his theories.

But perhaps a more important reason is that using his theories to make money is hard to fudge. Making money would prove that he's right, and surely he's interested in proving he's right even if he's not really interested in the money--it would show that he's not just making easy predictions, or fudging the predictions to match the events, because that kind of thing can make the predictions look successful but won't get him any money.

Re:If he isn't already rich then he's lying

[JesseMcDonald]

Human behavior isn't random, but it is chaotic in the context of competition. One can predict the result easily enough when there are no conflicting self-interests, as in your example, but that predictability disappears as soon as you try to influence the outcome in a way one or more of the individuals involved perceives as not being in their favor.

If all you want to do is model non-coerced and purely cooperative human behavior then that won't be an issue, but most behavioral economists find themselves employed in the development of political policies which are intended to change others' behavior. The chaotic nature of competition makes the long-term outcome of such measures impossible to predict; this phenomenon is observed all the time, and goes by the name "Law of Unintended Consequences".

Already been done.

[dynamo52]

Hari Seldon invented psychohistory.

Re:Already been done.

[Forthac4]

I saw an interview with him on The Daily Show and he made mention of the parallel with Seldon.

Re:Already been done.

[buchner.johannes]

http://www.thedailyshow.com/watch/mon-september-28-2009/bruce-bueno-de-mesquita

When you predict enough, you gotta be right

[Opportunist]

At least sometimes. If I make a thousand predictions (the more they contradict each other, the better) and only publish them AFTER the results are in, I can easily claim that I can predict the future. It's a simple magician's trick. Ask a person to think of a number between 1 and 10 (or pick a card, or whatever), then hand him a sealed envelope telling him you knew he'd pick that number (or hand him an envelope containing the card). You couldn't write it down and give him that envelope after he chose, so you have to be able to predict it, else you could not have written it down before the show, right?

What you don't hand him are the other envelopes containing the other numbers/cards.

Re:90% Accuracy

[red_blue_yellow]

Well, actually, they say 90%. From TFA:

According to research by the CIA, Bueno de Mesquita's model is more than 90 per cent accurate (British Journal of Political Science, vol 26, p 441).

Is that independent enough for you?

Ahmadinejad

[Anonymous Coward]

He "predicted" that Ahmadinejad wouldn't be reelected to the presidency of Iran, because he's group had no popular support. We all know how that ended.

Re:Ahmadinejad

[wizardforce]

We all know how that ended.

Well technically, he probably wasn't wrong on that point. It is quite possible, if not likely, that Ahmadinehad wasn't actually elected by the people of Iran but remained in power because the elections were rigged and didn't really matter anyway.

Re:90% Accuracy

[ShakaUVM]

>>Is that independent enough for you?

Then that should have been posted in the summary instead. Whenever I read self-written claims of a model's accuracy, my bullshit meter goes off.

In related news, Miss Cleo predicts ongoing instability in the Middle East, conflicts over water rights, and people being unhappy with those jokers in government.

Not "the Future"

[plasticsquirrel]

He doesn't claim to be able to simply "predict the future." Accurate information is only given in situations where a limited number of people are making a decision, and where accurate information is available on them for input. The key is basically that it assumes that serious decisions are made primarily according to the players' own interests (a reasonable assumption). Given the limited problem set, it doesn't seem too unrealistic to believe that one could make a very simple, basic model with some level of accuracy. Even without elegant theories, if accurate inputs and outputs from past events were available, a statistical model could probably be generated automatically.

I wonder if eventually every government will spend significant time consulting these machine-oracles? It reminds me of the various mathematical methods of prediction that still exist in China and India. Some of the Chinese models still require a significant amount of abacus shuffling, and a large set of reference books for all the possibilities. These were probably formed from similar basic methods of trying to gather data, compare it, and map inputs to outputs.

What's he predicted?

[Mr. Freeman]

I can believe a 90% hit rate. I can predict the future, and so can you, with 90% accuracy. See, if you don't claim to be able to predict EVERYTHING then you can easily "predict" obvious things.

I predict that tomorrow someone will die in the world.
I predict that tomorrow at least one person will spend $100 on a TV somewhere in the US.
I predict that tomorrow the temperature will be higher than 32 deg F in California.

Tomorrow look up these details and see how many I get right. I CAN PREDICT THE FUTURE!!

This guy claims to be able to predict "only certain things" which really means he's predicting things obvious even if it's not obvious that they're obvious. For example, he claims to be able to predict foreign policy. Did he predict foreign policy, or did he just watch the news and make some predictions based off of what all the political analysts are saying?

From TFA "These [predictions] include whether or not North Korea's supreme leader, Kim Jong II, would dismantle his nation's nuclear arsenal" How stupid do you have to be to believe that he "predicted" this? Everyone and their fucking aunt is watching the news, everyone is reporting on it, the government is doing fucking insane amounts of research and analysis as to what foreign leaders' views are regarding nuclear weapons. It's not that hard to make a guess as to what's going to happen when you have that much information available to you.

This method is NOT 'future seeing'

[nido]

From the /. story headline (emphasis added):

Bruce Bueno de Mesquita Uses Games To See the Future

Having read the fine links, it seems Mr. de Mesquita doesn't actually "see the future". He gathers data and throws it into his computer, which applies game rules to determine the most likely outcome. To me, "seeing the future" implies predicting the unpredictable - assasinations, a meteor taking out a major area, the abdication of a king (so he could marry his American sweetheart), etc.

Indeed, here's a quote from the New Scientist article:

According to political scientist Nolan McCarty of Princeton University, this is the real strength of the approach. "I suspect the model's success is largely due to the fact that Bueno de Mesquita is very good on the input side; he's a very knowledgeable person and a widely respected political scientist. I'm sceptical that the modelling apparatus adds as much predictive power as he says it does."

Methinks Mr. de Mesquita's method works because he meticulously gathers excellent data. If his data was sloppy, his rate of successful "predictions" would be much lower than it is.

Sometimes events which are 'unpredictable' happen. In retrospect we say, 'oh yes, this event was the only logical event to have taken place'. But such an event is typically unthinkable before it happens. Mr. deMesquita's model doesn't allow for the unpredictable, and is therefore NOT 'future seeing'.

I have a book on seeing the future. Here's a quote from the first couple pages that I typed up for a 2008 election prediction poll on K5 a while back:

Your Nostradamus Factor, by Ingo Swann

Chapter 1: Jumping The Time Barrier

Like many others, I've had good reasons during my life to assume that the future can be seen. But if I had any doubt it would have vanished as a result of an astonishing forty-five seconds when I found myself in Detmold, then in West Germany, in the spring of 1988.

Detmold is near the beautiful Teutoburger Forest and a famous pre-Christian shrine, Horn-Externstein, which is a pile of towering rocks riddled with sonorous caves. Until the time of Charlemagne it is said that Nordic kings came to Horn-Externstein to consult seers about the future.

I was invited to Detmold by Herr Manfred Himmel in April 19988 to give a series of lectures about psi research. This was Herr Himmel's fifth "esoteric" conference, and it was well attended by several hundred people. Herr Himmel was ardent about psychic matters, and the talks of his other speakers were interesting to me. Some of these speakers were also practicing psychics who were busy giving individual "readings" and making predictions about the future.

I was billed as the famous American superpsychic who had "astonished scientists" since my first formal laboratory experiments in 1970. But I have never given individual "readings," and I never made predictions about the future.

Many of Herr Himmel's conference attendees were visibly disappointed that I did not give the expected readings and did not foresee the future. Although I had studied "prophecy" and predicting for many years and had even experienced some novel insights about it, I was well aware that most predictions turn out to be wrong. I felt I had a scientific reputation to protect, which would be damaged if I accumulated a list of erroneous predictions. Moreover, I didn't view myself as a future-seer in any professional sense, and I though that predicting should be left to those who were or at least tried to be.

I gave several lectures and workshops at the conference, as well as the keynote address. I had worked hard at preparing this address, entitling it "Revising Psychic Research Methods and Expectations in the New Age," and even gave the opening statements in German before continuing

more points from the fine article

[nido]

The last few paragraphs of the fine article validate my post above. Here's a quote:

So what of the future? Another of Bueno de Mesquita's recent predictions addresses the future of climate change negotiations up to 2050. Depressingly, he predicts that although the world will negotiate tougher greenhouse gas reductions than in the Kyoto protocol, in practise these are likely to be abandoned as Brazil, India and China rise in power in relation to the European Union and the US.

One 'black swan' that Bueno de Mesquita's prediction dosen't take into account are technological developments which solve the energy problem.

Suppose a backyard inventor develops a thermodynamics-compliant engine/transmission that gets 3x better fuel economy with 1/100 as many parts (conventional reciprocating piston engines have 1000's, this invention has 25 or so), and allows for mechanical storage of 95% of a stop's kinetic energy in a hydraulic pressure tank.

Or what if there is a worldwide surge of volcanic activity in the 2030's, which makes human production of CO2 insignificant?

I, for one, am Not impressed. 'tis time for me to go to bed - maybe I will have some prophetic dream tonight.

Re:The future

[TapeCutter]

"As for how to reduce emissions to avoid such property damage, this is less clear cut."

Yes, in the long run the tradgedy of the commons is a more destructive failure of politics than war.

Re:Irrelevent - English is dead

[Jedi Alec]

Oh, a kingdom for the option to mod a post ironic, no matter how irrelevant it may be.

Re:Irrelevent - English is dead

[HungryHobo]

http://www.google.ie/search?q=predictioneer

Results 1 - 10 of about 57,900 for predictioneer.