Unlike the How long will the Writers Guild of America strike last? market I set up within the public Inkling Markets, which uses a series of contracts (one month or less, two months, three months, etc.), the Media Predict market is using Inkling’s “Predict a specific date” structure.
When the price goes up, the date moves later, when the price goes down, the date moves earlier.
See by yourself:
It’s the liquidity, stupid. 
—
UPDATE: Mike Giberson…
To avoid confusing the casual reader, I should probably point out that there is no “Inkling-Giberson conglomerate” (I’m just an Inkling user and fan who happens also to be an economist and blogger).
Chris, you aren’t clear why you think Inkling’s “Predict the date” is a better design than my use of Inkling’s multi-outcome structure. You hint it has something to do with liquidity. I need more hints. I don’t get it.
I do see that all the trading gets focused as buy or sell of a single contract rather than spread out across, in this particular example, nine contracts. But my approach allows traders to express both a point estimate and a spread (a trader can buy both “two months” and “three months” and “four months” if he feels that the strike will last more than one but less than five months, without a strong feeling about when during that middle three months things will end). In the predict a date approach, if the current prediction is in the middle of the third month (our trader’s average expected value), the trader can’t trade. I am merely asserting some advantages, but don’t think anyone has researched the relative performance of the two designs (except maybe internally at Inkling). (Anyone at Inkling want to speak up?)
To avoid confusing the casual reader, I should probably point out that there is no “Inkling-Giberson conglomerate” (I’m just an Inkling user and fan who happens also to be an economist and blogger).
Chris, you aren’t clear why you think Inkling’s “Predict the date” is a better design than my use of Inkling’s multi-outcome structure. You hint it has something to do with liquidity. I need more hints. I don’t get it.
I do see that all the trading gets focused as buy or sell of a single contract rather than spread out across, in this particular example, nine contracts.
But my approach allows traders to express both a point estimate and a spread (a trader can buy both “two months” and “three months” and “four months” if he feels that the strike will last more than one but less than five months, without a strong feeling about when during that middle three months things will end).
In the predict a date approach, if the current prediction is in the middle of the third month (our trader’s average expected value), the trader can’t trade.
I am merely asserting some advantages, but don’t think anyone has researched the relative performance of the two designs (except maybe internally at Inkling. Anyone at Inkling want to speak up?).