The Wisdom of Crowds is the theory, written by James Surowiecki, about the aggregation of information in groups. His theory states that a larger group of diverse people can make better, more intelligent decisions, than any smaller collection of experts. In this post I investigate the extent that his theory applies to the betting exchange.
The Four Conditions Of The Theory
According to the Wisdom of Crowds theory, for a crowd to be considered “wise” — as opposed to irrational — four conditions should be met:
|Diversity of opinion||Each person should have private information even if it’s just an eccentric interpretation of the known facts.|
|Decentralisation||People are able to specialise and draw on local knowledge.|
|Independence||People’s opinions aren’t determined by the opinions of those around them.|
|Aggregation||Some mechanism exists for turning private judgments into a collective decision.|
The betting exchange (e.g. Betfair) is a platform for aggregating a crowd’s opinion on uncertain events in sport. Remarkable as it may seem, the betting public collectively forms the “fair” probability (odds) of an outcome through their opinions and actions.
Odds rise and fall, settling at values that incorporate all the information that has been consumed by the bettors. The price discovery is dynamic and never completely stationary because there will always be new information arriving onto the exchange, causing movements to occur.
1. Diversity Of Opinion On The Betting Exchange
Diversity is a solid basis for wisdom. Take an investor for example. He may have his money spread over several things — property, an ISA, shares, premium bonds, a private company, maybe some antique collector’s items, and whatever other assets. Individually it’s unlikely that all of his investments will come off. But the wisdom he has achieved through diversity will make his results smoother, and smarter, than if he only invested everything into only one of his ideas. Essentially, by incorporating different ideas he’s formed a type of wisdom that is more likely to succeed.
Diversity is also the basis for any competitive market and naturally provides collective wisdom. In prediction markets like sports betting on an exchange like Betfair, diversity is a prerequisite because of the uncertainty of the events and the typically large number of participants acting with different opinions, risk preferences and forecasting methods. Not everyone shares the same point of view — hence why one bettor will Back an outcome whilst another will Lay it.
Prediction methods vary. Some bettors try to determine the true probability of outcome i.e finding value. Others adopt a psychological approach (often analytical) and study trends and directions in the odds, intent on discovering biases. There are other types of prediction models too, such as: linear or nonlinear, static or dynamic, deterministic, probabilistic, and so on. Ironically, most approaches will be wrong — but the pooling of diverse ideas encourages collective accuracy on the betting exchange. As a result the crowd (the people betting) create odds formed from diverse opinions, that are on average, pretty accurate.
2. Independence & Herding Mentality
Independence of judgement is essential for the Wisdom Of Crowds, but difficult to achieve. Humans tend to herd, following opinions or strategies that appear rational rather than forming a justified opinion of their own. This is relevant in betting, because sometimes people are influenced by each other’s guess work, meaning there’s more chance that the odds will drift towards a misplaced bias. This is perhaps why, arguably, the betting exchange is not always wise.
Real Life Examples
Misplaced bias happens in some pretty serious real life situations. Has anyone watched the series “Making a Murderer” on Netflix?
The documentary encourages us to believe that an innocent man named Steven Avery, has been wrongly convicted of murder. A large part of the show centres around the negative media portrayal of Steven and his poor reputation within his local community.
What resulted was a thoroughly unfair trial, with both the local justice department and a jury heavily compromised by outside information. Indeed the public perception of Steven as a callous, cold-hearted murderer was almost impossible for the defence to overturn.
Whether or not Steven is innocent is questionable. But the fact that the trial lacked independence of opinion — thereby invalidating one of the four conditions of the Wisdom of Crowds theory — is undeniable. Naturally, those assessing in his case had great difficulty forming an unbiased view on his innocence.
The more we know of other’s opinions the more we tend towards a consensus. Yet in many cases this is detrimental to accuracy.
So Does the Betting Exchange enable independence of opinion? Does it satisfy the Wisdom Of Crowds theory?
The exchange involves a high number of diverse, anonymous participants from various locations, backgrounds & social groups, encompassing many different motivations — fans, professional bettors and Bookmakers do not necessarily share the same point of view. With this in mind, the exchange certainly provides a platform for independent opinions. But it cannot guarantee that prices aren’t at all influenced by “herding”.
Indeed, the exchange does sometimes misrepresent the odds of an outcome. This is particularly true if a high number of participants blindly follow early bettors rather than verify their judgement (e.g. over a long period of past data). Similarly, if a large number of individuals predominantly learn through pattern recognition techniques, the consensus could form inaccurate prices. This is because in actuality, many patterns are comprised of random noise, and good or bad luck often dominates outcomes.
I believe that the bettors acting independently, without reference to a past trends, help to counteract exaggeration, or ‘noise’, within the market. They keep people’s errors from becoming correlated. So when an observed ‘trend’ turns out to be random, it will often be cancelled out from others betting with a totally different perspective.
3 & 4. Forming Odds With Decentralisation & Aggregation
Decentralisation allows people to act freely and independently of one another, and still interact to produce aggregated results.
Examples Of Decentralisation
If you’re a computer geek then you’ll be very familiar with Linux. This is an example of a “decentralised system” with a central aggregator: Linus Torvalds. Linux is worked on collaboratively by many programmers throughout the world. It is said to be “decentralised” as it’s not acting under the influence of a top-down central authority.
Many companies focus their business on objectives to be achieved by units or modules. It is thought to increase efficiency and effectiveness. This is due to reduced congestion in communications, quicker reactions to unanticipated problems, improved ability to deliver services, and improved information about local conditions.
Wisdom of Crowds: Centralisation (A) Vs Decentralisation (B)
Decentralisation’s great weakness is that there’s no guarantee that valuable information uncovered in one part of the system will find its way through the rest of the system. Sometimes valuable information never gets disseminated — making it less useful than it otherwise would be. In the case of Linux, different people often come up with competing solutions to the same problem. So in hindsight it would have been more effective just to combine these concepts from the beginning, collaboratively. Decentralisation is therefore only of benefit if there exists a way of coordinating or aggregating together all the information.
Valuable information makes it’s way onto Betfair and can be seen publicly in the order book. If a huge sum is trying to back one outcome — perhaps due to knowledge — then the odds will fall as a result of an imbalance in the Weight Of Money. Although open to price manipulation techniques, there’s still a strong argument that aggregation is satisfied on the betting exchange.
Aggregation, the last of the four conditions, is where the odds are formed. The betting exchange achieves this by bringing together the combined average of the public’s opinion from various different channels. The odds for a horse came from the aggregate of all the information that formed every opinion about the likelihood of it winning, and is currently represented by the money placed in Back and Lay outcomes. Price convergence means the odds are continually and dynamically updated to reflect changes in supply and demand. Provided there’s sufficient liquidity, then [on average] the betting exchange will continue to show a reliable and accurate approximation of the fair odds.
… So, do you believe that the Wisdom of Crowds theory applies to betting markets?
I do. And i think it’s the most important realisation that I’ve ever had about sports betting.