Wisdom in crowds

“Alone we can do so little, together we can do so much.” - Helen Keller


Thank you for joining me for blog 32, highlighting decision-making and the brain. This is my public exploration of what drives decision-making and how we can use that information to make better decisions, resulting in better outcomes.


We are currently exploring group dynamics and how that impacts decision-making. Today's question - 'when and why can a crowd help produce good decisions?'.


What is crowd wisdom?


We use wisdom of the crowds to make many important decisions. That is the basis of democracy, juries and markets. Here is a simple example from a book that I thought covered this topic really well ('Think Twice' by Michael Mauboussin).


- Michael asked 73 of his students to guess the number of jelly beans in a jar

- The average guess was 1,151 and the actual number was 1,116 (only 3% away!)


That's not the cool thing......

- The cool thing was that the average individual was 700 away so 60% of the actual number way (not close at all)

- Only 2 of the guesses were between the the average and the right answer


In essence, you took some guessers who weren't really that close to the right answer and combined their views to a pretty spot on number. How can it possibly work?


Why it works


1) Each person took their own approach to working out how many jelly beans there were, meaning they were not influenced by the ways others did it

2) Some approaches resulted in a higher number and others a lower number

3) Importantly there was an incentive (bragging rights) to get it right


For the more geeky of you, Michael uses the following equation to try and explain it:




So we can start off with individuals that are by no means experts but come to a really sensible answer due to their diversity of approaches. We will be covering incentives and diversity in coming weeks.


Many new technologies are based on wisdom of the crowds. We have tripadvisor for restaurants or amazon reviews for products to buy. People feel much more confident to visit a restaurant or make a purchase if they have a large sample of people that have positively reviewed it. We used to rely on our friendly video rental person (for those old enough to remember) for recommendations, but Netflix's recommendation technology has a huge amount of data and can use our previous history and if we input it our preferences.


Crowds vs experts


Michael Gove a British politician was famous for saying “I think the people of this country have had enough of experts with organisations with acronyms saying that they know what is best and getting it consistently wrong.” In a previous blog, we looked at how forecasting critical events like elections and geopolitical issues is too complex and reliability in forecasts is very low. In this blog we have observed that fairly simple things like 'how many jelly beans in a jar?' or 'is this product good or not?' can be done by crowdsourcing. So where should we get experts? The key use of experts is in trying to help us understand the rules of the game and how a situation works in practice. No matter how many people or technology you throw at a situation, you need to understand the domain you are operating in.


So what?


How is this all relevant to decision-making? Here are three take-aways I want to leave you with before we pick it up next week:


1) We use the wisdom of crowds in voting, juries and markets and most new technologies use it too.


2) It works when you can aggregate a lot of independent voices


3) For simple well-defined and understood problems, crowds will likely beat experts, but where the domain is more specialised and hard to understand, experts are necessary to bring us up to speed.


Thank you for joining. Next week - 'Incentives, Incentives, Incentives.'