Fame League & Prediction Markets

Review

I enjoyed Cedric's presentation on the Fame League. It was the first time I had heard of the term "Prediction Market". The gist of the presentation was clear: Fame League incorporated a prediction market into their game, unfortunately, the execution was not done well. Throughout the presentation, Cedric continuously emphasised on the potential of a prediction market and why we should be concerned about it.

There's going to have to be a compromise in a 7 minutes presentation. At the end of it, I wasn't too sure what exactly it was that Fame League failed in or messed up but thanks to Cedric's emphasis, I knew what to look out for. I first wiki-ed about the Wisdom of the Crowds, Prediction Games and Prediction Markets: 

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

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

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

And then I checked out Fame League to see what it's all about. At first sight, it was confusing. After a while later, it was still confusing. I can understand why Cedric said his group took a long time to figure what the game was about. If there was any aspect of a Prediction Market in Fame League, it was most difficult to find. A large portion of the game was dedicated to being a stock exchange more than anything else. What made it confusing was that a lot of the stocks involved betting on celebrities or cars or basically things we would probably not see being bet on in a stock exchange. Maybe I'm totally missing on something about the game or I just don't understand the game mechanics. I think that real world stock exchange is easier to understand than this game (not necessarily easier to win money) because the real thing makes more sense.

The parts of the game that involved a Prediction Market appeared to be but just a small section compared to the rest of the game. Well, at least I did kind of understand this part of the game. For anyone who wishes to get a good idea of what a Prediction Market is about, take a look at the following site:

http://www.intrade.com/

The site gives a clear cut idea of what a Prediction Market is. So if you're trying Fame League now and see a section that bears similarity to what is shown on intrade.com, that's a Prediction Market.

To give Fame League credit, the notion of creating it is probably for players to gain an understanding on how trading on an exchange works. To focus on the fact that they are not fully exploiting the potential of a Prediction Market may be unfair towards them because that may not be what the game is about. Still, for a first timer, I feel very confused and would probably not be playing this alone unless I had other friends to play with. There are still a lot of features that I have yet to explore in Fame League but perhaps I may not have the patience of doing so.

Improvements to Fame League

I think there can be some improvements made to the game. Firstly, while the graphics are very pretty, I think that the interface can be improved. I feel overwhelmed upon my first entry into the main page. So much information and at the same time, I don't really know what to do yet. Secondly, the game can afford to become more newbie-friendly. Have a structured "safe-zone" where players can first learn without fear of loss about how the game operates before being thrown into the big sea to get eaten by the sharks. If learning is part of the aim of the game, then the game should be designed in a manner to teach rather than to obfuscate the player.

Opportunities

I think I can see where Cedric and Prof Ben is coming from when they say that there is a huge potential in the Prediction Market as a game. Game wise, it is rather intellectual and requires that players at least have some knowledge or interest on current affairs, sports and entertainment, technology or whichever aspect they're interested in. Commercial wise, the data derived from the players' betting can be sold as information.

Here's an idea of how I see a Prediction Market game could possibly be implemented. Please keep in mind though that this is just a bit of brainstorming on my part and I have yet to do pure solid research on this. I would use a similar model of how Channelnewsasia.com categorises its news:

Asia Pacific, Singapore (Local News), World, Business, Sports, Technology, Entertainment, Health, Legal.

I think the terms speak for themselves quite easily. Prediction options are then available on various issues under each section. To make the game realistic, a similar trading format as that of intrade.com can be adopted using virtual currency.

Limitations & Challenges

One reason that Prediction Markets work is partly because traders are betting their own money on the likelihood of a particular event's occurrence. No one is going to try and mess up one's own bet by making a bad decision (subjective to the self). However, in a game where real cash is not involved, players may be less inhibited in their playing style and take risks or perform actions that deviate from their behaviour in the real world. Or they make take the game as what it is: a game and not treat it seriously. This can be both a good or bad thing.

Suppose a scenario whereby an event is touted by reports to be highly unlikely to happen. A trader may then decide to allow general consensus to affect his or her judgement. However, if no money is involved, the trader may feel less inhibited to make a decision based on intuition which, steers in the opposite direction. In this situation, having a game system will output a different result from an actual prediction market. Both information are likely to be as valuable.

A challenge to the game is in making it addictive or creating widespread mass appeal. Why would players opt to play this over others? A game such as this requires patience of which the player would have to wait for the prediction to either fall or come through. This can take anything from weeks to months. The sports section will no doubt be more active given how much is happening everyday while issues on policies or politics will require that the trader wait months before knowing the actual result.

Game mechanics is very important to get right. Perhaps my initial suggestion of using virtual currency is not a good option at all. Simulating realistic conditions may be appealing to some but may not be appealing to all. At the end of the day, players may be more easily addicted to games of which they can see continual progress in levels. 

One key importance of implementing such a game is in making sure that the data on the predictions is reliable and useful for study. If such an end can be achieved simultaneously while providing fun for the user, then it does not matter whether or not the game simulates real world trading. Admittedly, one might question how true such predictions might be if the game is not an accurate simulation. Truthfully, no one knows. Unless there has already been prior research done, no one can say for sure that such an idea or model will not work.

Conclusion

All in all, I actually find this idea of a game very appealing and am glad that Cedric and team shared Fame League and such an idea with us. Thanks guys! Thank you to Kok Wee for assigning me Fame League too. Haha! I enjoyed thinking and writing about it.

I think there is definitely lots of potential for it as a final project but at the same time, the scale of it appears to be really massive and complicated. Since I don't have any final project ideas yet, this seems quite a good option except it gets a suicidal rating. If anyone wants to work on a project like this, just holler and let me know. I would be keen. =)