Live odds on the UK election | Hacker News Reader

Live odds on the UK election

smarkets.com

45 pointsvide0star25 comments

Graphs over time: http://politicszone.betfair.com/

To interpret the numbers do 100/(decimal odd) to get percentages as shown in the graphs.

For those that haven't been following UK politics, the UK's first ever televised prime ministerial debate was in mid April. Many of the dramatic changes in the graphs result from the Liberal Democrats doing well out of that encounter.

the smarkets dudes are really smart guys, i work across the floor from them at another startup. you should keep an eye on smarkets and the guys behind it, they're always doing/discussing some technologically cool shit.

... or just a link to a gambling site :-(

It's a prediction market - different than a pure gambling website. The prices are set in realtime by the users so there is 'wisdom of the crowd' information for the public.

"wisdom in the crowd" seems like a contradiction. Nice phrase.

http://en.wikipedia.org/wiki/The_Wisdom_of_Crowds

"a book written by James Surowiecki about the aggregation of information in groups, resulting in decisions that, he argues, are often better than could have been made by any single member of the group."

Check the reference to Delphi Methods in above wikipedia link. They are used for estimating efforts in software development.

There's a big difference between prices emerging and a crowd that discusses a question to come up with a "group answer".

and as far as I know they are the best mechanism to predict the outcome of an election. It's an honest opinion since they are betting their own money independently of their personal preferences.

Are you suggesting that people will make bets in a purely rational way, and only by considering (in this case) what they know about the UK election? I suggest that plenty of people will either bet emotionally, choosing the option they prefer, or psuedo-rationally (i.e. reasoning poorly). I expect many will also use the odds provided on the website to guide their choices.

Possibly not a site like this. But predictions made by bookmakers should in theory be accurate, since the bookmaker has a financial incentive to make them accurate and no political incentive to misrepresent them.

The irrational bettors will be driven out of the market, and made money of, by the rational punters.

If you truly believe this, why don't you take their money by betting against them?

Have you looked at the site? What should I bet against? There are a lot of choices :-)

I don't imagine a clear divide between 'rational' and 'irrational' betting (for want of a better description), I'm just saying that I would expect there to be a lot of both going on.

If you're wondering why I believe this, the book 'Predictably Irrational' by Dan Ariely springs to mind as a good read about the subject. Also 'The Tiger That Isn't' (Michael Blastland/Andrew Dilnot), not about human rationality as such but rather, how easily we can be foxed by information depending upon how it is presented to us (in the context of 'reasoning poorly' about the election).

The user pays money in the hope of winning more back - it's gambling however you slice it.

There is no information of any value here outside the context of the site itself - unless you're claiming that your audience is a representative sample of the UK population?

That's not how prediction markets work. People aren't supposed to bet on whom they want to win. People are supposed to be searching a profit, that is find differences between the odds given in the current market and what they perceive as the true odds.

People who do this consistently over time, tend to accumulate more money, and thus accrue more influence as time goes on. Also people who bet foolishly, e.g. just bet on who they like to see winning, create opportunities for other people to make money of their back.

> The user pays money in the hope of winning more back - it's gambling however you slice it.

That describes investment as well.

Indeed - what's your point? Since both generally involve using guesswork (however educated) in an attempt to increase wealth, it seems reasonable to draw a parallel between the two.

Depends on what you define as gambling. E.g. playing crap or betting on fair coin-flips is decidedly different from prediction markets or investment; because everyone agrees on the probabilities already, so there's no information aggregation and discovery.

It's the same basic process though isn't it? Making a bet on the basis of a guess about the future. Of course, it may be a very well educated guess or just a shot in the dark, but unless you can actually correctly predict the future, there is always some element of uncertainty involved.

It's not about the uncertainty. But about aggregating information about that uncertainty through prices (in markets) or using that information to up your chances in, say, Poker.

In `pure' gambling, like betting on a coin flip, everyone already has all information. So it's less interesting, and warrants a different category of game.

It's by far the most accurate prediction of the result. People putting their money where their mouth is - generally a lot more reliable than somebody with a clipboard in a shopping center asking voters, or the opinions of a TV reporter.

>It's by far the most accurate prediction of the result.

Do you have any evidence for this? E.g. comparisons of polling vs betting predictions from previous elections.

The title says "odds"—what did you expect?

Reply on news.ycombinator.com