Techcrunch 50 - and the winner is Otherin.......er, sorry, Yammer!

While many have been watching the TC50 livevlogging, I've been watching Hubdub's prediction market with a lot more interest. In this case the market did not predict the winner. The eventual winner was not the predicted OtherInaBox, but Yammer, here predicted as a 5th place play - the TechCrunch piece notes that:

Yammer is Twitter with a business model. Created by an existing company, Geni, to scratch its own itch, Yammer takes the familiar Twitter messaging system and applies it to internal corporate communications. There is such a huge demand for this type of service that 10,000 people and 2,000 organizations signed up for the service the first day it launched on Monday. Anyone with a corporate email can sign up and follow other people in their company. But if a company ants to claim its users, and gain administrative control over them, they will have to pay. It’s a brilliant business model.

Runners up were Atmosphir, FitBit, Grockit, GoodGuide and Swype - the last 2 were also in the top 5 Hubdub predictions.

Why the difference?

One option could be that Hubdub was being gamed (looking at the OneInaBox scoring, and possibly Akoha, that is very probable) - but another equally likely explanation is that the judges are not using the same criteria as the "wisdom of crowds" prediction market. The really interesting thing to me is that the prediction market throws numerical evidence into the light on the gap between the "market" and the "closed committee" outcomes. (The comments on the TechCrunch piece also seem to side with the market)

(Update - Hubdub's Nicel Eccles has commented that the first 2 results were very likely gamed, thus the predicted "last 5" would have had Yammer in 3rd place, after Goodguide and in front of Swype. So 3 out of 5 ain't bad! However, looking at comments in the TC50 and Twitterverse space, I'd still say that hypothesis 2 - a difference between market and judges' criteria - held true)

(Update 2 - another reason a closed result may differ from a prediction market is if the result has been "gamed" - ie rigged in some way. Valleywag is wondering aloud as to whether this may be the case)

It will be very interesting to fast forward a year or so to see how the Prediction Market fares against the judges' decisions.