Betting at the Horseraces is half the fun of the sport. However, picking the right horse is not only hard but it is so hard that often even experts fail to pick the winner out of the 20 horses at the Derby. A new process can change the way we approach predictions of this nature; UNU’s swarm intelligence, which uses the collective decisive powers of a group of real people, have shown superior powers of predictability compared to both expert individual opinion and artificial intelligence. With the help of Swarm Intelligence, the people at Unanimous A. I. managed to turn $20 into $11,000 at the Kentucky Derby this year.
Horse racing in the USA still draws millions to the tracks every year. The biggest of all the races is the Kentucky Debry; not only does it kick off the Triple Crown, it also ranks first in North America and surpasses the attendance of all other stakes races in the country.
Betting at the Kentucky Derby
Betting at these races is a tradition as old as the races themselves and can be as much fun as it sounds. The common bets at the Derby are the Win-Place-Show Bets, which yield a victory if the chosen horses, win or make it in the top two or three. For more expert gamblers, the Exacta Bets offer more of a challenge; this requires the player to predict both the winning horse and the first runner up but in the correct order. The Trifecta can add more risk in the mix, as that requires the player to guess the top three horses to cross the finish line in order. Finally the Superfecta, the near impossible bet, where players have to guess the winner alongside the second third and fourth horses to finish the race in the correct order.
Experts and AI vs UNU
The Superfecta stumps even the best of specialists; this was proven when every expert at Churchill Downs was defeated by the Superfecta bet at this year’s races, when no one managed to name all four of the horses in any order, let alone in the correct one. Artificial Intelligence does not always fair better either, Bing Predicts managed to only name one of the four horses correctly, that too the heavily favoured Nyquist to win the race.
The UNU challenge
Given that it is so hard to correctly predict winners at these races, Hope Reese, a reporter for TechRepublic and the Atlantic, challenged Unanimous A.I. to use UNU to try and predict the Superfecta winners of the Kentucky Derby. Despite not being knowledgeable about horse racing and betting of this kind, Unanimous, confident in the ability of UNU’s Swarm Intelligence, decided to accept the challenge.
David Baltaxe, Chief Information Officer at Unanimous, said, “We were reluctant to take on this challenge; nobody here knows anything about horse racing, and it’s notorious for being highly unpredictable. Still, UNU surprises us again and again, so we recruited a swarm of volunteers through an online ad. The whole thing took 20 minutes.”
The first step, for the selected group, was to answer a set of questions, as a unified intelligent entity to narrow down the 20 horses to the top four; this was achieved in a 10 minute sessions. Next, the ‘swarm’ was asked to order the four horses in terms of rank. A week later, when the Kentucky Derby announced the post positions of the horses, which often impacts the potential outcome, the original ‘swarm’ was gathered and asked to make any changes they saw fit; within the boundaries of another 10 minute session, the group had come to their decision and replaced one of the four horses on the initial shortlist with an alternate option.
Reese, who came up with the idea of the challenge published the predictions in Tech Republic and on the day of the races decided to place a $1 bet on the UNU’s picks for the Superfecta, which, once the winners were announced, had turned into a whopping $541.
The people at Unanimous were a little more confident with their past successes with Swarm Intelligence, put down a bet of $20 on their own predictions for the Superfecta, which beat all odds to fetch them a grand winning of $10,842. The people at Unanimous also put a $20 bet on the Trifecta based on their predictions and won another $1,940. That is a massive 5400% profit on the original wager.
The experiment leaves us with the question – what makes Swarm Intelligence’ more accurate? Traditional artificial intelligence looks to mimic human intelligence with gadgets and software. Swarm Intelligence is a new form of A.I. that utilises technology that can leverage natural human instincts – intuitions, emotions and sensibilities, while making a decision – combining the best of both worlds. The UNU platform is a social environment that encourages groups to “think together” in real time; groups of people can therefore use UNU to gather their collective intelligence to make important decisions or just have some fun.
Countless species, in nature, have evolved to live in groups or swarms for a better chance of survival – amplify their intelligence in closed-loop systems that pool their insights and converge on optimal decisions, proving that many minds are indeed better than one and now UNU offers humans the same privilege of forming intelligent swarms.
Swarms Vs Standard Polls
Using the survey data from the individuals who participated in the Kentucky Derby swarm, Researchers found that if the group had resorted to a vote instead, they would have only succeeded in predicting the winning horse. Swarms outperform polls, surveys and markets because surveys use sequential prediction markets, whereas swarms tend to explore each given option in a more dynamic system and explore their options together to identify the optimal solution.
After the success of the Kentucky Derby Swarm Intelligence predictions, more people will definitely be keen to look toward UNU for accurate predictions for other big betting events; and given that it is possible to now be part of a future swarm, the next big win they successfully predict to land a 5400% gain could be yours.