Greg Lang, RAIN
19 November 2020

The wisdom of crowds and group captives

There’s a sign on a wall in my home that reads: “Never underestimate the power of stupid people in large groups.” James Sworlecki’s 2004 bestselling book “The Wisdom of Crowds” contends that the wisdom of the many outweighs the wisdom of the few, whether expert or otherwise. Sworlecki’s book is filled with examples to support his group theory.

One such story is that of Francis Galton, a scientist who conducted a breeding study at the annual West of England Fat Stock and Poultry Exhibition in 1906. At the event there was a competition in which 800 people estimated the weight of an ox and recorded it on a ticket. Galton wanted to prove democracies do not work and borrowed the 800 tickets for his study, assuming the simple average of all the tickets would be wildly inaccurate, thereby proving the stupidity of common people.

“All phases of group problem-solving are equally important in supporting a healthy group dynamic.”

Galton reviewed 787 tickets in total; 13 were illegible. To his surprise, the average guess was 1,197 pounds, and the winner guessed 1,198 pounds. It was a good day for democracy.

This group intelligence is at work in many places in our world today. It is the reason Google can scan a billion web pages and ensure most-searched items always come to the top.

Having trouble making money betting on football games? How do the bookmakers always pick the correct spread? The line moves based on how the crowd bets. It is why stock market pricing usually works—and sometimes does not.

The market swings resulting from the COVID-19 pandemic are a good example of group intelligence not working. Panic sometimes outweighs logic, but group logic usually wins in the end. However, most consumers and investors believe valuable knowledge is still in very few hands.

Experts call this group-think “collective intelligence problem-solving”.

There are three types:

  1. Cognition problems: where some answers are better than others. When will the world economy return to normal?
  2. Coordination problems: coordinating behaviour. How do we get people to exit a stadium in an orderly fashion, get traffic to merge on to a highway, or people to board a bus or subway successfully?
  3. Cooperation problems: getting the self-interested and distrustful to work together. How best to engage experts with competing interests to solve global issues such as pandemics, climate issues, and paying taxes?


Let us look at a cognition problem. In 2011, the Journal of Food Processing Engineering reported that shorter frying time and longer centrifuging maximises oil removal from a potato chip’s surface, with an 81 percent reduction.

There were also too many broken chips in the centrifuge. How would you improve this process? You could try better oil, varying the speed of the centrifuge, vacuum frying, or maybe something else entirely. I’m guessing something else. Did anyone guess Topcoder?

Topcoder is a crowdsourcing company with an open global “community” of designers, developers, data scientists, and competitive programmers. Topcoder sells its services to large, mid-size, and small-business clients. It pays community members for their projects: it is freelancing with a twist.

Freelancers compete to solve problems for real companies, with the winners rewarded and others getting nothing. Anyone with an internet connection can use Topcoder, meaning anyone can learn and compete.

Back to the oil problem and the chips. Did anyone guess music? One of the engineers was also a concert violinist. Using the same principle as an opera singer breaking a glass, vibration can be used to remove oil from chips—you just need the right pitch.

The 2014 movie “The Imitation Game” is the story of mathematician Alan Turing and the Hut 8 team at the UK Bletchley Park codebreaking centre during World War 2. Turing’s multi-skilled team cracked the German Enigma code to help the Allies to victory; it took the Hut 8 team about a year. One of Topcoder’s engineers using current technology was able to crack the Enigma code in about two hours.

The wisdom of group risk-sharing

This same collective problem-solving applies to group captives, including risk retention groups. I have four group clients that share some common traits but have others that are unique in the group. The combination makes each one successful.

  1. Cognition problem-solving: often the best answers to the challenges of group captives are solved using a professional service team: underwriting, actuarial, claims, reinsurance, etc. These resources can be resident or outsourced. Every captive has a unique skillset and personality.
  2. Coordination problem-solving: peer pressure to do the right thing when no-one is looking helps keep group captives profitable. Risking capital and premium to solve problems and share risk is unique to insurance. Some professions such as researchers share data. Pooling financial resources takes it to a higher level. It takes significant trust and coordination by the members.
  3. Cooperation problem-solving: applying group-think to an industry’s unique exposures takes some level of cooperation too. In some aspects of their business these captive owners compete. The best groups share expenses for new insurtech to help compete in the future.

All phases of group problem-solving are equally important in supporting a healthy group dynamic. The success and profitability of group captives bears this out—truly “wisdom in crowds”.

However, remember the sign in my house: make sure you are running with the right crowd.

Greg Lang is the founder of the Reinsurance and Insurance Network. He can be contacted at: