Social dynamics and the ability to solve problems are of concern. We seem unable to make decisions in large groups. Surveys indicate that disruptive innovations come from very small groups and individuals. Why is this so?
Social behaviour is unlikely to be ergodic. An ergodic system is where probabilities of an expected outcome are independent of sample size, time of sample and other variables. The penny is an example of an ergodic process. A single penny, flipped over time, will yield a 50:50 ratio of head to tail. A barrel of pennies dumped on the floor will have approximately a 50:50 ratio of heads and tails. Social groups are different; particularly humans. We instinctively feel this to be true, but we attend large meetings anyway to satisfy false instincts.
A real world example occurred last year. I am a 10-year veteran of the Congress for the Future of Engineering Software (COFES). Every year, we get together to discuss the latest hot and high-tech software in Scottsdale, Ariz. Most meetings need some fun. Us geeks deserve it. Last year, we decided to have an IQ test for the group. Brad Holtz is the ringmaster of this conference and suggested that I make up the IQ test, one aimed at the community of attendees – a geek IQ test, if you will. We came up with four questions here at the barn and presented them to Brad. The answers bounced back quickly and correctly. We could assume, therefore, that a significant minority of attendees would get the right answers.
At the last big dinner of the conference, we set up the competition. There were more than 200 attendees. They were sitting about six to a table, and each table had a copy of our IQ test. The first table to get the correct answer would be presented with a wonderful present – maybe even a printed certificate. Any resource could be used to find a solution. Going on the web with an iPod was legal.
An hour after the test began, no table had finished. What? The rule most could not solve was this three-word test. They had to come up with the next entry in the following sequence. Solid, liquid, gas…?
There were a lot of arguments from the crowd. When we told them the correct answer – plasma – the audience mostly agreed with it.
Was the audience dumb? No. The problem was the interaction within the groups.
A year later, as we planned the next meeting, we were thinking about the situation dynamics. We knew from reading the literature that ergodic behaviour is unlikely in social instances. The likelihood of a small group behaving the same as a large group is not probable. No table had the answer to our questions. But we estimate that at least 10 percent of the individuals in the audience had the right answer. Looking back, I remember some individuals telling the other members of the table, “See, I was right.” If one in 10 of the individuals at the conference knew the right answer, each table could have had the right answer inside the smaller group. But the group would not accept the individual solution.
We’ve all been in meetings like that. Democracy seems to stamp out innovation. There are no evil people; it’s just a matter of the math. Remember, the relationships between people are links. If you have two people in the group, there is one link; three people generate three links; four people results in six links; and so on. The complexity of the group is link-dependent, not node-dependent. It would almost have been impossible to have a minority vote in a multi-linked situation survive. The table probabilities and individual probabilities were different.
At a later meeting, we were in a 12-man group trying to summarize a particular problem. We had an hour. The first 15 minutes were unsatisfactory. So, taking our IQ test supposition to heart, we broke up into small groups. Solutions appeared. By George, it works. What this says is that meetings are for information, not for decisions. Only 10 percent (an approximation) of disruptive innovations come from big companies. The rest come from people outside of the domain – i.e., individuals or very small groups. And now it makes sense.
The answer is to keep it small. Applications of this philosophy are in software, risk, team management and IQ testing. The optimum group size is three to four people. We could discuss this in more depth, but I have to go to another meeting.
Dick Morley is the inventor of the PLC, an author, speaker, automation industry maverick and a self-proclaimed ubergeek. E-mail him at firstname.lastname@example.org.