Good Magazine ran an interview recently with a man they call The New Nostradamus. Bruce Bueno de Mesquita uses a mathematical model that is based entirely on game theory to predict the outcomes of political conflicts. He takes a very literal interpretation of the phrase “political science” and focuses his analysis strictly on issues of strategic interest, ignoring any cultural or historical aspects of the parties involved. He believes that the theory of rational choice can accurately predict the actions of any political actors as long as the data underpinning the determination of interests are correct. An analysis of his model’s predictive abilities done by the CIA found it to be accurate 90 percent of the time.

In the article a few of his predictions are discussed, but what is interesting is that he also makes a number of prescriptions. In fact, while there is a list at the end of the interview describing some of his accurate predictions, the discussion with him fails to clearly separate predictions from prescriptions. In the interview, he proposes a solution to the Israeli-Palestinian conflict and criticizes the outcome of negotiations with Kim Jong-Il of North Korea for not conforming to his model.

Perhaps not coincidentally, the recent agreement that the United States reached with the government of Pyongyang closely resembles the one that Bueno de Mesquita’s model suggested: Kim agrees to dismantle his existing nuclear weapons but not his existing nuclear capability. “He puts it in mothballs with IAEA (International Atomic Energy Agency) inspectors on site 24 hours a day, 365 days a year. And in exchange, we provide him with $1.2 billion a year, which we label ‘foreign aid,’ of course.” The “foreign-aid” figure published in the newspapers was $400 million, which concerns Bueno de Mesquita. “I read that and I said, I hope that’s not the deal because it’s not enough money. He needs $1.2 billion, approximately, to sustain the loyalty of his cronies in the military and so forth. It’s unpleasant, this is a nasty man, but we’re stuck with it. The nice part of the deal is that it’s self-enforcing. Each side has a reason to credibly commit to their part of the deal.”

It would appear that what he has actually developed is a highly sophisticated system of conflict mediation. His model assumes that people are selfishly rational and always gravitate toward very predictable terms in an agreement. It would be very interesting to show these predicted outcomes to two negotiating parties at the outset of their talks. Would they get to the same results faster?

Bueno de Mesquita acknowledges the power of what he is able to do with his work, which seems to play a big role in his approach. He will not call elections that he claims to know the outcome for because he does not want to influence them and he will not help organizations affect or manipulate government policy. Clearly, predicting the future is a complicated and controversial venture. It toys with our sense of continuity and our theories of causality, let alone the concept of free-will. It also seems that as people get better at it, we may be raising questions faster than we can answer them.

Share This ▼

We have recently resolved Bet 117 on Long Bets about the adjusted cost of energy. It was an interesting case where we had very specific criteria for who would win the bet, yet we could not adjudicate it when the time came. The bettors cited the Department of Energy published numbers to resolve their bet. However in the first quarter of 02006 when the DOE posted their numbers, they then quickly retracted them. It turns out they had several years worth of data incorrect due to the deceptive data from the Enron energy kerfuffle. It took the DOE over a year to straighten it all out. In any case we have yet another resolved bet. Congratulations to Steve Kurtz on winning the bet, and to both bettors who patiently awaited the results, and agreeing quite gentlemanly in the end.

Share This ▼

An amazingly in depth article by Long Now board member Paul Saffo on the intracacies of Forecasting appeared in the July 07 Harvard Business Review. The whole issue entitled Going The Distance is about long term thinking in the business realm and is highly recommended. Below are some highlights of Paul’s forcasting article…

The goal of forecasting is not to predict the future but to tell you what you need to know to take meaningful action in the present.

Rule 1: Define a Cone of Uncertainty
Rule 2: Look for the S Curve
Rule 3: Embrace the Things That Don’t Fit
Rule 4: Hold Strong Opinions Weakly
Rule 5: Look Back Twice as Far as You Look Forward
Rule 6: Know When Not to Make a Forecast

Cone of Uncertainty graphic from the sidebar in the article.

Share This ▼

For the past few years Chris Hibbert has been working on Zócalo, an Open Source Toolkit for Prediction Markets. He writes, “my purpose in the project is to build prediction market software that people can use as a foundation for deploying many markets of this type. As I said in my proposal to CommerceNet, the plan is to start by building software to support economists doing experiments with prediction markets, and expand from there to internal markets within businesses, and eventually expand to the point where the software would be sufficient for public real-money markets in important questions on technology development, economic impacts, and other important issues” (Hibbert). Essentially, Hibbert believes that by making prediction market software open source and customizable, prediction markets will be more widely used. We are watching this closely as we are considering the same with our Long Bets software.  Build 2007.1 was released April 19, 2007; comments from users are most welcome.

(a screenshot of Zócalo via from sourceforge)

Share This ▼

photo of flooded Thames by elyob

The Long Now’s Long Bets project asks us, active bettors and wider public alike, to think more deeply and carefully about the medium- to long-term future than our assumptions (and busy schedules) might otherwise allow.

Nudging our culture towards assuming greater responsibility for addressing (and creating) possible futures is an ambitious undertaking, but the “bet” concept at the core of the project does seem to capture something fundamental about our predicament as creatures gifted — and burdened — with foresight. In some respect, every decision is implicitly a bet, a step taken despite the uncertainty of what comes next. In a 02006 article, Michael Chabon poetically put it this way:

[I]n having children—in engendering them, in loving them, in teaching them to love and care about the world—parents are betting, whether they know it or not, on the Clock of the Long Now. They are betting on their children, and their children after them, and theirs beyond them, all the way down the line from now to 12,006.

Even having kids is a bet, from this angle. Still, there are less lyrical, though deeply significant, bets being placed all the time — and not just on Superbowl Sunday. One relatively new arena for betting explicitly on future possibilities in the world of finance is the catastrophe bond. In April, London’s Financial Times reported a world first:

Investors have been given their first opportunity to bet against the possibility of a disastrous flood crippling business in the City and Canary Wharf, with the sale of the world’s first “flood” bond on capital markets.

The new $150m bond has been issued by an arm of Allianz, the German insurer, to protect itself from devastating claims that might arise from companies affected by such a flood.
…
In recent years, insurance companies have turned increasingly to the capital markets to spread their vulnerabity [sic] to catastrophes, and have already issued bonds whose value is linked to US hurricanes or earthquakes.

However, it has until now been impossible to create flood bonds, because most countries lack precise systems to measure damage.
…
The inaugural issue is worth $150m, but Allianz has in place a programme to issue up to $1bn in the future.

(Gillian Tett and Andrea Felsted, Insurers launch ‘London flood’ bond, Financial Times, 10 April 2007 [online]; 11 April 2007 p. 1 [print edition])

The report describes an interesting trend, of insurers spreading the financial risk associated with natural disasters to the capital markets (in the event of hurricanes, earthquakes, and now floods). A space to watch, in light of increasing concern over extreme weather events caused by climate change. But aside from underlying reasons for the shift (greater volatility?), the key point here is the enabling mechanism — increasing availability of ways to measure increments of disaster. What can be measured can be bet on. Those who follow Long Bets will know that adjudicability of bets is one of the key questions for long bettors. Rhetorically grand, but forensically fuzzy, claims are Long Bets’ bête noire. They run into the bar of adjudicability (and hence accountability) deliberately raised by the project. What we see here is that risks which can be specified and measured with sufficient precision become subject to the gaming of financiers. And it seems that, if and as technologies for quantifying natural processes improve and become more widespread, the pool of events we can rigorously bet on, in the world of finance or otherwise, will continue to grow.

Perhaps not just large-scale natural events, but other things (internal bodily processes? social dynamics of organisations or countries?) could become quantifiable, hence “bettable” — a site of unprecedented accountability. What might this mean in the long run for how risk is managed in society?

Share This ▼

Philip Tetlock (screen shot from high-res Seminar video at members’ webpage)

Philip Tetlock recently presented a Seminar About Long-Term Thinking to the effect that confident forecasters ought to be ignored. Despite his research showing the profound unreliability of such speculation, it’s rare to find even a moderately systematic evaluation of political forecasts in the popular media. In this category comes an article from the online version of Radar magazine earlier this year, “The Iraq Gamble”, which scrutinises journalistic commentators’ predictions about the war in Iraq. In it, writer Jebediah Reed tracks the fortunes of predictions against those of the predictors, and is disturbed by what turns up.

[M]aybe something is amiss in the world of punditry. Are the incentives well-aligned? Surely those who warned us not to invade Iraq have been recognized and rewarded, and those who pushed for this disaster face tattered credibility and waning career prospects. Could it be any other way in America?
…
So we selected the four pundits who were in our judgment the most influentially and disturbingly misguided in their pro-war arguments and the four who were most prescient and forceful in their opposition.
…
Then we did a career check … and found that something is rotten in the fourth estate.

The suggestion is that American meritocracy seems to be in trouble when it comes to the careers of political pundits. Eight journalists’ profiles are presented under two headings: “Getting rich by being wrong”, and “Right but poor”. As one of those drafted into the latter category, Jonathan Schell, remarks: “There doesn’t seem to be a rush to find the people who were right about Iraq and install them in the mainstream media.” Meanwhile, certain confident voices have prevailed despite the question marks over their track record (a similar point is made in an article from FAIR.org last year, addressing the Iraq-related statements of columnist Thomas Friedman, who also comes into the firing line in the Radar piece).

Whether or not one agrees with the argument as applied to this politically contentious topic — and a larger sample would be required to draw firm conclusions — it seems clear that improving accountability (i.e., correspondence of reputation and record) for predictions made by public figures remains a major challenge in politics and the media alike. It’s an aspiration embedded in the Long Bets project, and any readers who think they can do as well as, or better than, these guys — on any topic of public interest — is encouraged to test their long-term prescience by placing a wager on the site.

Share This ▼

The online version of Popular Science magazine is offering a prediction market for science and technology. It uses token dollars instead of real money (in order to avoid gambling laws). Here is what they say about it:

Ready to bet on the future?
Join the PopSci Predictions Exchange.

Welcome to the PPX, the first place to bet on the future of science and technology. It’s easy and free: Log on, and we’ll give you POP$250,000 in our virtual PopSci Dollars. Use that money to buy propositions you think are likely to happen. If other traders also want to buy, that proposition’s price will go up, and you’ll make PopSci bucks. Expand your portfolio with bets on energy, space, consumer technology and extreme science, and compete against other players for prizes and bragging rights.

Here is a sample chart from bets on whether Internet Radio Survives

Share This ▼

This is a very short article on how economists are using prediction markets to predict weather at least as good as meteorologists, which is not very good.

Penn State Researchers Testing Futures Markets For Weather Forecasting

UNIVERSITY PARK, PA (March 21, 2007) – Economists at Penn State’s Smeal College of Business and College of Earth and Mineral Sciences are testing whether futures markets can be used to accurately forecast the weather, and, so far, they’ve found the markets to be just as accurate as major forecasting services.

The 60 participants in this predictions market experiment, which is in the midst of a two-year run at Smeal’s Laboratory for Economics Management and Auctions, are mostly students studying business or meteorology at Penn state. They use allotted funds to bet on what they believe the high and low temperatures will be in different U.S. cities on a given day. As the going rates for various temperatures fluctuate within the market, the researchers can weigh the market’s confidence in what temperatures will be reached.

(more…)

Share This ▼

It’s a bit rare that we see such clear and succinct predictions. These three by Ari Emanuel were published at the Huffington Post and were forwarded to me by Paul Saffo:

I have three predictions to make this morning: 1) John McCain will not be the Republican Party’s nominee. 2) Hillary Clinton will not be the Democratic Party’s nominee. 3) Before the end of Bush’s term, Condoleezza Rice will be Vice-President.

http://www.huffingtonpost.com/ari-emanuel/three-predictions_b_41851.html

Share This ▼

“What is it about politics that makes people so dumb?”

From his perspective as a pyschology researcher, Philip Tetlock
watched political advisors on the left and the right make bizarre
rationalizations about their wrong predictions at the time of the
rise of Gorbachev in the 1980s and the eventual collapse of the
Soviet Union. (Liberals were sure that Reagan was a dangerous idiot;
conservatives were sure that the USSR was permanent.) The whole
exercise struck Tetlock as what used to be called an
“outcome-irrelevant learning structure.” No feedback, no correction.

He observes the same thing is going on with expert opinion about the
Iraq War. Instead of saying, “I evidently had the wrong theory,” the
experts declare, “It almost went my way,” or “It was the right
mistake to make under the circumstances,” or “I’ll be proved right
later,” or “The evilness of the enemy is still the main event here.”

Tetlock’s summary: “Partisans across the opinion spectrum are
vulnerable to occasional bouts of ideologically induced insanity.”
He determined to figure out a way to keep score on expert political
forecasts, even though it is a notoriously subjective domain
(compared to, say, medical advice), and “there are no control groups
in history.”

So Tetlock took advantage of getting tenure to start a long-term
research project now 18 years old to examine in detail the outcomes
of expert political forecasts about international affairs. He
studied the aggregate accuracy of 284 experts making 28,000
forecasts, looking for pattern in their comparative success rates.
Most of the findings were negative— conservatives did no better or
worse than liberals; optimists did no better or worse than
pessimists. Only one pattern emerged consistently.

“How you think matters more than what you think.”

It’s a matter of judgement style, first expressed by the ancient
Greek warrior poet Archilochus: “The fox knows many things; the
hedgehog one great thing.” The idea was later expanded by essayist
Isaiah Berlin. In Tetlock’s interpretation, Hedgehogs have one grand
theory (Marxist, Libertarian, whatever) which they are happy to
extend into many domains, relishing its parsimony, and expressing
their views with great confidence. Foxes, on the other hand are
skeptical about grand theories, diffident in their forecasts, and
ready to adjust their ideas based on actual events.

The aggregate success rate of Foxes is significantly greater, Tetlock
found, especially in short-term forecasts. And Hedgehogs routinely
fare worse than Foxes, especially in long-term forecasts. They even
fare worse than normal attention-paying dilletantes— apparently
blinded by their extensive expertise and beautiful theory.
Furthermore, Foxes win not only in the accuracy of their predictions
but also the accuracy of the likelihood they assign to their
predictions— in this they are closer to the admirable discipline of
weather forecasters.

The value of Hedgehogs is that they occasionally get right the
farthest-out predictions— civil war in Yugoslavia, Saddam’s
invasion of Kuwait, the collapse of the Internet Bubble. But that
comes at the cost of a great many wrong far-out predictions— Dow
36,000, global depression, nuclear attack by developing nations.

Hedgehogs annoy only their political opposition, while Foxes annoy
across the political spectrum, in part because the smartest Foxes
cherry-pick idea fragments from the whole array of Hedgehogs.

Bottom line… The political expert who bores you with an cloud of
“howevers” is probably right about what’s going to happen. The
charismatic expert who exudes confidence and has a great story to
tell is probably wrong.

And to improve the quality of your own predictions, keep brutally
honest score. Enjoy being wrong, admitting to it and learning from
it, as much as you enjoy being right.

–Stewart Brand

(Iraq footnote. I asked Tetlock to opine on which experts were most
right about how things have gone in the Iraq War. He said the most
accurate in this case were the regional experts, who opposed the
invasion, and what they are predicting now is a partition of Iraq
into Kurdish, Shia, and Sunni areas.)

Stewart Brand
Seminars & downloads: http://www.longnow.org/projects/seminars/

Share This ▼