The Slingo Grand National Supercomputer has been hard at work making its predictions for this Saturday’s big race. According to its calculations, Intense Raffles, with his 11.4% chance of success, is the most likely horse to triumph at Aintree. How did the mysterious supercomputer reach this conclusion though and should punters pay any attention to it?
The Rise of Supercomputers in Sports
This prediction by the Slingo Grand National Supercomputer is yet another example of a supercomputer trying to predict an upcoming sporting event. They are becoming increasingly common across sports, in particular football, where they are used to forecast the outcomes of many leagues and cups.
The computers perform complex calculations and simulations to produce percentage chances of success, rather than making any definitive claims of who will win. The percentages they provide can rapidly change too when new information becomes available. In the case of football, supercomputers will tweak their calculations based on recent results, for example.
To give a clear example of this, at the start of the 2024/25 Premier League season, the Opta Supercomputer originally believed Manchester City had an 82.2% chance of winning the league. Pep Guardiola’s side were followed by Arsenal (12.2%) and Liverpool (5.1%). Fast forward to the end of March, however, and their calculations gave Liverpool a whopping 98.4% of success, Arsenal 1.6% and Man City 0%.
It would be a little unfair to say the original prediction was ‘wrong’. Although Opta believed Man City had the best chance, they never said anything about this being a certainty. A 4 in 5 chance of success does not equal success every time and we are witnessing the 1 in 5 exception. Unlikely things happen in sport and the computer helps us to highlight just how unexpected they are when they do happen. Any prediction about the future is just that, a prediction, and even a weather forecast that gives a 99% chance of rain does not mean that a dry day is impossible.
How Does The Grand National Supercomputer Work?
Moving back to the big race at Aintree, the supercomputer for this race ran 100,000 simulations of the race while taking into consideration market odds and ratings. As these were the only two data points considered, it is not the most complex of supercomputer predictions. A more advanced model might have looked at past trends and given some weight to factors like jumping success rate, course performance, jockey and trainer records, the weather, going and more besides.
That said, you could easily argue that key factors like these determine the odds of each horse and bookmakers, with their wealth of knowledge, do a fine job at pricing up runners. Therefore, simply using the odds and ratings (which impacts the weight a horse will be running with) might seem simple but it should still be an effective method that produces worthwhile results.
Grand National Supercomputer Predictions
This is what the Slingo Grand National Supercomputer concluded after its simulations. All runners not featured were given less than a 3% chance, the very lowest being Velvet Elvis (0.87%).
| Horse | Chance of Victory (%) |
|---|---|
| Intense Raffles | 11.4% |
| Iroko | 10% |
| Stumptown | 8.7% |
| I Am Maximus | 8.7% |
| Vanillier | 6.7% |
| Perceval Legallois | 5.7% |
| Hewick | 5.2% |
| Nick Rockett | 4.8% |
| Meetingofthewaters | 3.5% |
| Kandoo Kid | 3.3% |
| Grangeclare West | 3.1% |
The supercomputer believes Intense Raffles has the best chance of winning this year’s Grand National. However, it is also saying there is an 88.6% chance that a horse other than Intense Raffles will be first past the winning post. Therefore, it would be wrong to say anything like “Supercomputer predicts Intense Raffles will win”.
Are The Supercomputer Results Useful?
At first glance, you may be wondering what you can do with just a list of closely grouped percentages but they can help punters find value in the market.
| Horse | Chance of Victory (%) | Best Odds Implied Probability | Difference |
|---|---|---|---|
| Intense Raffles | 11.4% | 6.7% (14/1) | +4.7% |
| Iroko | 10% | 10% (9/1) | – |
| Stumptown | 8.7% | 10% (9/1) | -1.3% |
| I Am Maximum | 8.7% | 9.1% (10/1) | -1.3% |
| Vanillier | 6.7% | 9.1% (10/1) | -2.4% |
| Perceval Legallois | 5.7% | 7.7% (12/1) | -2% |
| Hewick | 5.2% | 9.1% (10/1) | -3.9% |
| Nick Rockett | 4.8% | 2.9% (33/1) | +1.9% |
| Meetingofthewaters | 3.5% | 4.3% (22/1) | -0.8% |
| Kandoo Kid | 3.3% | 4.8% (20/1) | -1.5% |
| Grangeclare West | 3.1% | 2.9% (33/1) | +0.2% |
By comparing the calculated chance of success with the odds, you can see which horses represent good value for money – if the computer is to be believed. In the case of Intense Raffles, the odds give him a 6.7% chance of winning yet the computer thinks his actual likelihood is 4.7% greater, making this a very good value bet. Conversely, a bet on Hewick would be poor value as you are betting on something with a 5.2% chance of happening at odds that imply just a 9.1% chance.
This all assumes that the supercomputer calculations themselves are accurate, but ultimately this is impossible to know. However, using a data-driven approach to find the best value in the market is far from the worst long-term betting strategy and this is an option worth considering – especially if this is the only time you bet on the horses or anything at all for that matter.