The Wisdom of Crowds
August 16, 2005, [MD]
This was on my reading list for a long time, finally picked it up at the Denver Public Library and read it in one sitting. Here are my as usual cryptic notes.
- First example: put 750 marbles in a glass and have a group guess independently. The average will be very close to the truth.
- What is needed for a group to work: some individual information (could be bias, different approach etc), not leaning or depending too much on the others in the group, being able to express/agregate their information (voting, markets - not reaching consensus).
- Price markets work well because of this. Stock markets not so well because everyone want to predict what others think, not what the value of companies are.
- Examples of why corporations exist, very much in line with “The Efficient Society” by Joseph Heath.
- My thought: can the last be transferred to the elections in the US? The two candidates that run for election are not presenting voters with a great deal of choice, so the primaries are very important. But with the last Democratic primaries, people did not try to vote for the person they thought would best do the job, but rather the person that they thought other people thought would best do the job (ie the person who was most “electible”), and thus their choice was perhaps wrong.
274 p.
Stian
Stian Håklev August 16, 2005 Toronto, Canada comments powered by Disqus