Monday, July 17, 2006

So just when you thought it couldn't get worse, there's a full scale disaster going on. :-(

And I'm kind of tired of arguing with all these kids with their enthusiasm for total war in the Middle East.

The more shocked and hurt and confused they get, the more they call for a cathartic release that will just burn the enemy off the planet for once and for all. "Make it go away. We don't care how! We just don't want to have to deal with it any more."

How do you start a conversation there? With no shared world-view and no shared moral convictions?

Well. There is one thing : to appeal to facts. What does the evidence actually show us? Do much trumpeted military operations have the expected effect?

The problem is that the evidence can always be spun. Can always be reinterpreted. There is no pure experience data that isn't shaped by our conceptual framework.

So how do you find common ground for communication?

Well maybe markets can help. Markets simplify communication to minimal binary oppositions : profit / loss, buy / sell. Surely there is a basic world-view so simple that some kind of common understanding can be reached, a shared channel through which some kind of signal can be squeezed.

The simplification of communication in markets is the source of both their great good and their great evil. The good is that the simplified communication protocol lets the market scale. Only markets succesfully incorporate and co-ordinate the work of billions of humans. And those co-ordinated billions achieve great things together.

At the same time, the simplicity strips those interactions of almost everything that is worthwhile about being human. There's no room for empathy or rich experience or tenderness or laughter or recognition of the other as a full person. That narrow bandwidth is filled with nothing more than simple : worthwhile / not worthwhile signals.

Markets are usually praised by their supporters for the first. And damned by their detractors for the second. But we can surely recognise that both claims are true, and that markets can, but should, be used pragmatically.

I'm suddenly very excited by prediction markets and particularly this StrategyPage one.

Prediction markets are traditionally celebrated for their power to integrate information from many sources. Now I wonder if their simplification might also help break through the barriers of communication. Disagree with someone's opinion? Why waste time arguing past each other when you could both just buy shares on opposite sides of a prediction? After a set amount of time, you'll know who was right.

It's very easy to write a blog post ranting about how X is going to happen or Y won't. And next year, who's gonna remember? Not you. Probably not your fans. Nor your opponents who gave up reading you because you were obviously wrong the one time they bothered to look at your blog.

But in a prediction market, things are different. It creates a fairly cheap and straightforward sort of accountability for the opinions you express. It keeps track of your predictions and how they actually turned out.

Your bets are open for everyone to see. Even though there's no real money, there's a reputation reward for being right. In fact, statistics could be produced showing all sorts of interesting things about your positions. How closely does your behaviour tended to follow particular trends or social clusters? Are you easily led astray by certain kinds of claim? You, and everyone else, can see whether biases are leading you to make many erroneous predictions. And, if so, what those biases are. The market won't let you hide this from yourself or others.


John Powers said...

You always come up with ideas that make scramble to make connections, perhaps some of them not really relevant to the ideas you are actually presenting.

On the subject of prediction markets it seems to me good thing that they be primarily based on reputation value and not monetary gains. I'm not sure why I think that, but the Pentagon's Policy Analysis Market (PAM) ran into similar reaction by Congress

The signal "This is play" is something that Gregory Bateson paid a lot of attention to both while wearing his ethologist hat and his cybernetisist hat. What I get from reading Bateson is that "play" isn't so simple and is revealing of the fundamentals of communication.

The problem that comes to mind about the prediction markets with money at stake is deception. Some businesses use market games to identify promising new projects
2006/03/idea-markets.html Rite Solutions has "Mutual Fun" using "play" money. Real money rewards for "correct" market plays within the game are only available within the context of the actual development of the project.

Michael Shermer has a post in "Scientific American Online:"

The Political Brain:A recent brain-imaging study shows that our political predilections are a product of unconscious confirmation bias

Your idea that prediction markets may be useful in starting conversation among people "with no shared world-view and no shared moral convictions" seems a promising feature of a peer-review system for politicians the SA article suggests is needed.

John Powers said...

Have you ever thought of putting a tip jar on your blogs and use the money to finance your trades at intrade?

The only other blogger I know who posts about prediction markets in the context of their regular blogging is Billimon. I would love to hear you the two of you discuss the using prediction markets. Maybe a podcast;-)

StrategyPage really gets it right with the handy comments button for each prediction. I'm just too cheap to subscribe.

I noticed that Dan Abbott didn't take the bait as far as prediction markets go. I think you asked the right question in your comments

"What are the criteria of something being legitimately credited to one side or the other? Is mere causal involvement sufficient, or do we have to suppose that Israeli strategists explicitly visualized and took steps intended to bring about that moment of PLO defeat by Hamas? Visualized how clearly?"

My suggestion is for you to find some predictions already existing on the StrategyPage and go back to Dan with those.

Composing said...

Actually dan tdaxp messaged me a couple of links today to things he's written, which can play the role of predictions. I don't have time to get more into it because it's end of semester marking crunch.

Yep, was interesting to see Billmon using them as well. I think I'd feel immoral betting real money on something like Strategy Page. It's already an uncomfortable feeling when you know you're hoping for bad things to happen to prove that you're right. Adding money to the mix would make it worse.