Technology

Can AI Predict the Melbourne Cup Winner? I Decided to Find Out

Published November 5, 2024

The Melbourne Cup is one of the most celebrated horse racing events in Australia, drawing attention from fans and bettors alike. With technology making strides in various fields, it begs the question: Can artificial intelligence (AI) predict the winner of the Melbourne Cup?

This exploration led me to investigate various AI systems, particularly focusing on their ability to make accurate predictions in the context of horse racing. While AI can assist with diverse tasks such as drafting essays, generating meal plans, and even simulating therapy sessions, can it effectively tip a winner in a race known for its unpredictability?

My Quest for Horse Racing Insight

As a passionate sports fan, I admit that my tipping skills leave much to be desired. I tend to bet with my heart rather than my head, often choosing teams or horses based on emotional connections. For instance, I’m loyal to the Sydney Swans in AFL, and my selections for horse races often rely heavily on the names of the horses. In my heart, a horse named Fancy Man would catch my eye over others.

In light of my less-than-stellar track record, I decided to consult three different generative AI systems: ChatGPT, Google Gemini, and PunterGPT. My goal was to see which AI could provide a winning tip for the Melbourne Cup.

PunterGPT's Unique Approach

PunterGPT is developed by Michael Kollo and John Conomos. This tool takes a distinctive approach by analyzing text rather than numerical data. It has indeed proven its ability by accurately predicting last year's winner to finish in the top three.

According to PunterGPT, a relatively underdog horse, Onesmoothoperator, has strong chances in the upcoming Cup with odds at $8.50. PunterGPT highlighted the horse's impressive performance in the Geelong Cup, showcasing its stamina and ability to rally. The AI noted that Onesmoothoperator has a skilled jockey and appeared consistent in previous long-distance races.

On the AI's advice, I placed a bet on Onesmoothoperator, hoping luck was on my side this time.

ChatGPT and Its Selection

ChatGPT was more cautious in its betting advice, selecting Buckaroo as a strong choice for the Cup. ChatGPT based this decision on the horse's robust performance, particularly its runner-up finish in the Caulfield Cup, which is notable among competitors.

Since I'd already backed a somewhat risky option with PunterGPT’s choice, I decided to follow ChatGPT’s tip and placed a small wager on Buckaroo as well.

Google Gemini's Conservative Bet

In stark contrast, Google Gemini took a different route, expressing hesitation in providing outright tips. Its pick, Vauban, was based on pre-race analysis and expert opinions. Despite the analytics surrounding Vauban, I opted to place a bet on this favorite, even though favorites don't often win the Melbourne Cup.

Race Day Outcome

Sadly, when the Melbourne Cup took place, the results were disappointing for my AI-assisted bets. None of the horses I backed, including the AI's selections, made it into the top five. The winner, Knight's Choice, along with two others, took the lead. It appeared that while generative AI can assist with many tasks, predicting a horse race is not one of its strengths.

In conclusion, until AI becomes a more dependable tipster in horse racing, I may as well stick to my personal, heartfelt selections. After all, there's something to be said for betting on what you love, even if it doesn't guarantee success.

AI, MelbourneCup, Betting