The Limitations of AI with Sports Betting Picks

January 14, 2025 by Staff

We live in an era where AI is becoming the dominant technological force. The most prominent companies in the world are putting it front and center of their strategy; it's driving the stock market, and experts claim it will revolutionize how we work and socialize. Of course, not everyone is sold on the tech, but it has felt somewhat ubiquitous since ChatGPT took the world by storm a couple of years ago.

Naturally, AI has made an impact on sports. Many positive use cases exist for the technology in analytics, sports science, etc. Yet, we also see AI tools being touted for sports betting picks. You'll see these in social media adverts. Some promise the moon, offering fool-proof gateways to riches. You should be very wary of these as there is almost certainly a catch, perhaps even a scam. Others are more sober in what they claim to offer. While the latter may have some benefits, some limitations should be acknowledged.

AI is not an oracle

The first thing to say is that AI is not an oracle. Its ability to predict has arguably been misunderstood. It is adept at organizing and analyzing data, but it lacks the human ability to leap from point A to point B without all the information. In short, it cannothonestly reason and create, despite what its proponents say. Critics further claim that artificial intelligence is a misnomer, less like sci-fi movies and more like dressed-up algorithms. But that debate is for another day.

Nevertheless, we can illustrate what we mean with a simple example. If you have been tracking the Super Bowl odds all season, you'll know that sportsbooks cooled their backing of the Kansas City Chiefs as the regular season progressed. They started the season as favorites, and rightly so, but their lackluster performances led sportsbooks to drop them as outright favorites, with the Detroit Lions topping the market instead. This is easy to understand for sports fans and is part and parcel of any football season. For clarity, we should point out that the sportsbooks didn't dismiss the Chiefs' chances, just cooled a bit. But it's significant in betting terms.

Yet, the data can be tricky. The Chiefs had the best record in the NFL regular season, alongside the Lions. So, if you are asking an AI to make an assessment, it all depends on what data it receives. The 15-2 record, the No.1 seed, the previous Super Bowl wins? That makes things all look rosy for the Chiefs. AI cannot watch the games like you or I, so it couldn't see the sloppy play that infiltrated the Chiefs' game this season. Of course, you can start feeding the AI better data to get a more rounded picture, including that the Chiefs almost made an art form out of winning narrowly during the regular season.

Stats can be misleading

The point, as such, is that statistics can be misleading. A lot of data from a football team can be given to AI, but you must be sure that it can contextualize it all. Moreover, you should be aware that AI – at least the consumer models most of us have access to – limits how it can search the web for live data. Many websites will block AI; some have lawsuits against AI companies to prevent access. The fact the bots will be missing chunks of data can be crucial in sports betting picks.

Finally, there is the human element – let's call it intuition. If we return to the Chiefs, many of us feel that this is a mentality team that can step up in the Playoffs. Humans recognize these factors – being clutch, winning ugly, raising your game – that go beyond data. It's about more than gut feeling; it's about understanding the mentality in sports. You can see that in teams like the New England Patriots in the 2010s, and the Chiefs have it now. It does not show up in a bunch of numbers.

AI will undoubtedly evolve and improve, and there is a good chance it will become a tool every sports bettor uses. But we should stress again that it is not an oracle. You can use it to assess large troves of data, but we would say that the analysis should always be tempered with human reasoning. Be cautious of anyone touting guaranteed success with AI, as that goes against every aspect of the sports experience: bad teams win, good teams lose, and sometimes we know it will happen despite what the data says.