UsableMarkets

I finally got a chance this weekend to go to the simExchange and check out what they’ve been up to. Like all good Web 2.0 sites these days it still retains the beta stamp, even it appears to be fully “locked and loaded”, you might say. Anyway, I found much here of interest. But good … bad …? Lets go to the video (game).

The Contracts
There has been some chatter on Midas Oracle about the structure of the contracts. I agree with Robin Hanson that there is a something a tad strange about not having any specified time associated with a contract. While the contracts certainly work – they respond to new information, etc. etc. etc. – there appear (to me anyway) to be two issues.

One, since the lifetime of each game is different, and, essentially unknown at the time of the IPO, the trader is being asked to make a very complicated prediction: how many games will be sold, over an unspecified period of time, and in an unknown number of countries. At the time of the IPO the company producing the game will probably not even have this information. Unless there is some sort of standard rollout, or expected lifetime of each video game, then I would imagine it must be very hard to predict this number.

More useful, it would seem, would be more specific contracts which deal with sales per year per country, or something like that. The simExchange has clearly decided not to follow the HSX model (you predict a movie’s profits for four weeks after its release), but there are many benefits to the HSX model, and I would bet that HSX produces more useful and relevant predictions.

Two, after a period of time, when a game nears the end of it’s life span, trading will necessarily fall off since the price will probably not change much and there is little profit to be made. Again, renewing contracts on a quarterly, or yearly basis, would keep trading invigorated, even after the game has been out for a long period of time.

This might also alleviate the need for a market maker to supply liquidity at the end of a game’s life span, when trading will be light.

The Trading
To start with, the simExchange does something very nice – which is to offer traders the opportunity to use the “basic” trading mode, or the “advanced” trading mode. I think this is a really neat innovation that I hope people will expand on as more prediction markets come to market.

That said, I found the basic mode to be confusing. It essentially tries to present an order book with the simplicity of a market maker. Even though it appears as if I can enter in any quantity I want, entering in the wrong quantity produces error messages.

Enter in any number you want (except those that don’t match an offer)

The trader is essentially required to understand the concept of an order book, without actually being presented with an order book. Confusing? I certainly found it so.

The same is true for short selling. Nothing has been done here to simplify the concept – like opposite outcome contracts at newsFutures – even though the interface has been simplified.

There is an important usability lesson here: Simplifying the interface does not necessarily lead to increased usability. Concepts also need to be simplified, or at least explained in meaningful ways, for usability to be increased.

And there are other problems with usability. Crucial language, for example, is not always clear. In the basic mode the order book states: “There is an offer for you to buy x shares @ x DKP.” Why not just say: “You can buy x shares @ x DKP.” Throwing in the term “offer” just confuses things.

And, finally, trades need confirmations. The updates in the Notifications area are just not clear enough, and traders need to have the chance to confirm that yes, this is what they intended.

As always, thanks for listening.
~alex

If you enjoyed this analysis please Digg it