Prediction Markets

• Intrade.com : The most popular prediction market in the US. Due to the current laws on using money for betting, intrade allows predictions on everything save sports.

• Iowa Electronic Markets : The Iowa Electronic Markets is an academic market examining elections where positions are limited to $500.

• Qmarkets : Qmarkets allows websites to create their own prediction markets and provides idea management services for Enterprises.

• Betfair.com : A great site for betting on sports, unfortunately Betfair is only available outside the US because, yes, it uses real money.

• Lumenogic : The company formerly known as Newsfutures, one of the earliest providers of prediction markets to both the general public and private companies. It is now called Lumenogic.

• Hollywood Stock Exchange : A web-based, multiplayer game in which players use simulated money to buy and sell “shares” of actors, directors, upcoming films, and film-related options.

• Consensus Point : Used by companies such as Best Buy and Motorola, Consensus points uses “predictive analytics” to predict what companies may need.

• Zocalo : An open source tool kit for building your own Prediction markets.

• Inkling : Inkling allows you to download software also to set up your own prediction markets. It is also a paid service, so you get detailed information regarding the markets you have set up.

• PredictX : PredictX is a non-commercial prediction market for academic research which allows users to trade virtual money on future events.

• Smarkets : Another sports-betting website currently not accessible to participants in the US.

Polling is becoming harder and harder to perform and is becoming more and more of an art. One obvious reason is cell phones. With many more Americans owning cell phones, it is harder and harder to conduct surveys and polls over land lines. 25% of Americans now only have a cell phone (myself included). This is a problem because calls are much more easily screened and blocked, but also because most people keep their number when they move (I have a New York area code, despite living in Atlanta), thereby making demographic information a nightmare.

Also, the so-called robocalls are prohibited from calling cell phones, so real people need to be hired, which spikes up the costs for pollsters. The average response rate for polls has now decreased to 15%, from 35% in the 1990′s, thereby increasing the number of costs that need to be made in the first place.

So what is a viable alternative? Well, despite there not being as much liquidity in the house and senate elections as there is in the presidential ones, we can clearly see that right before the election results were out, the Iowa Electronics market gave the house to the republicans and the senate to the democrats, which is exactly what happened.

We also refer you to the prediction market results right before the 2008 presidential election. It seems as though prediction markets might be a better indicator than polls after all.

We have not yet, however, reported on Zocalo Prediction Markets. This is an open source tool kit for building your own Prediction markets. What we like about it is that since it is open source, unlike Predictify, you can see the underlying information for the markets, rather than just the results. You can, obviously, also set up your own predictions markets (unlike HSX or Consensuspoint). So basically you get the best of both worlds. Of course, once you have done so, it is up to you to find participants and to collect and analyze the results.

Another site we have not reported on is Inkling. This site allows you to download software also to set up your own prediction markets. It also promises more liquidity since it appears to be very wide-spread. It is also a paid service, so you get detailed information regarding the markets you have set up. On the other hand, you probably can get none based on prediction markets of others. The currency is also not real (it is based in the US, so this wouldn’t be allowed), which makes us question the subsequent accuracy of the markets. If anyone has had experience using either of these markets please do let us know what you thought.

If we want to make the power of prediction markets more mainstream, it will be once the US allows real currency to be used, and it will undoubtedly be through channels such as these.

Unfortunately, this will be a short post. Since I am currently residing in the United States (and they have markets in sports) I am not allowed to have an account with Smarkets. I will defer to a video they have posted (although even this is truncated), and if anyone out there has tried using them I would love to get in touch. Suffice it to say that their manual odd-setting feature seems very interesting, as well as their open source API. As always, we are curious about its effect on forecasting, prediction accuracy and liquidity.

Watch the (half?) video here.

Luckily, we then found out that it had merely migrated to a new and (we think) improved site. We were so impressed that we wanted to share it here. The H1N1 (or influenza formerly known as Swine flu) market can be found here. As of the writing of this article, there is a 96% chance that at least half the influenza cases this season will be H1N1.

New Influenza Market site here.

In a nutshell, this would have a been a market in which people could trade futures on specific terrorist-related activities, such as a bombing in Lebanon, an attack in the Green Zone, a deposition of some regional leader, etc. etc. The idea was that this market, using the Dumb Agent Theory, would help analysts in garnering the wisdom of crowds in order to determine what is likely to happen. The complaint was that people would profit from terrorist attacks, which was unethical and immoral.

If you would like to know about this market in (much) more detail, we recommend reading Thoughts on the Theory and Practice of Speculative Markets qua Event Predictors by Mason Richey. We should warn that this was obviously written with an academic target market, so nothing is watered down or skipped. On the other hand, it goes into great detail and gives much background to the situation.

For the record, we at Dumbagent did have reservations about this market, although these had nothing to do with the market’s morality. People trade on events on betting websites regardless. In fact, if there is a way to harness this information in order to prevent attacks we are enthusiastically in favor. Our main problem is that the market itself would be self-defeating. If more people were to expect an attack on a certain country in one month’s time, shares in this event would increase in value, but since this would be spotted, more efforts would be made to prevent said attack, in which case the probability of this attack would decrease. This means that the more efficient the market is, the less likely it is to work, or, as Richey puts it: “a market’s supply-demand logic cannot tolerate the act of a securities purchase contributing to that act’s unprofitability.”

Definitely worth a read. Although PAM did not work out, it was a step in the right direction and we hope the idea did not die with it.

Full article here.

Most of our readers will be familiar with, if not users of, Twitter. Many of our readers will also be itching to find new Prediction Markets to invest with (well, that might just be us). So why not try Pretweeting.com?

Of course, we should specify that no real money seems to be involved. Also, the predictions deal with the number of occurrences of certain words. While we won’t discount this off-hand, we find it hard to see any scientific utility in the knowledge of these trending words (already available, by the way, as trending topics on the Twitter website). Business, on the other hand (and depending on the type of business), might benefit by keeping up with the real-time popularity of certain terms, or testing out the popularity of certain campaigns. Let’s see how this pans out.

As a nice segue, we would also like to remind our readers you can follow our Dumbagent Twitter account @dumbagent, as well as our Current Events Twitter account @DACurrentEvents.

It is therefore with a heavy heart that we announce the closing of Predictify.com. According to them:

Due to the tough economic climate, we are planning to cease operations and shut down the company in the near future. As such, we are no longer approving or scoring questions and apologize for any inconvenience this may cause.

This doesn’t really explain much. We would like to add that, as we’ve mentioned on our Twitter feed, we don’t think they milked their site as much as they could have. The Iowa Electronics Market does more or less what they do, but had less liquidity. On the other hand, it was, and still is, used by many people as an indicator and an educational tool to learn about prediction markets. As far as we know, Predictify never once tried this.

That having been said, we will miss it. In four days it closes its doors.

We also like the fact that anyone can ask a question which then anyone can respond. Predictify has been following this model for several years now, so we asked them if they had any raw data or statistics regarding the accuracy. We have been rebuffed in this regard, but we suspect they do (their own site advertises their accuracy in political questions). If anyone has a method of obtaining these results, or if any intern would like do scan through them for extra credit, we would love to see some of this data. For example, is there a minimum number of predictors in order to achieve acceptable levels of accuracy in predictions? Are questions that offer cash incentives really always more precise? Are predictors more accurate in sports than politics? In Current events than pop culture? Are predictors from certain demographics more accurate than others?

These are all titillating questions which Predictify has hitherto refused to supply us an answer with. We will keep trying, however, and let you know as soon as we are successful.

Prediction markets have been working surprisingly well for firms which have used them, especially with forecasting. They have been used internally, mostly along with other tools, in several companies in order to make major decisions. It is a wonder, therefore, why more companies have not adopted them.

An article in The Economist discusses the possible reasons why, such as that it’s a “pretty evangelical business…”