One of the more irritating habits of our media in recent years has been its determination to target Celtic fans, and football supporters in general, with endless nonsense about supercomputer predictions and what AI supposedly thinks will happen.
Today we got another one of those stories. This time it came from the Scottish Daily Express, using Grok to predict which Celtic player will have the biggest impact on us winning the title. For good measure, it also predicted that we will win it.
There was a time, not so long ago, when the American statistical analysis company FiveThirtyEight was the go-to source for this kind of thing. It produced weighted averages based on past performance and current results. Newspapers loved it.
Most Read on The Celtic Blog
However, it was always garbage.
Statistical models cannot account for injuries, dressing room issues, tactical changes or simple human unpredictability. Yet, over the years, I have read countless breathless reports claiming that FiveThirtyEight had tipped the Ibrox club to win the league, only for reality to prove otherwise.
Now we have the latest version of the same nonsense. Sometimes it’s a “supercomputer”, which usually means someone running simulations on a laptop. Sometimes it’s Football Manager scenarios. Often it’s something as banal as people gaming outcomes on EAFC.
And now Grok is predicting title winners.
But who actually believes this stuff? Who clicks on it expecting insight? Who reads it and thinks they’ve learned anything useful?
The Daily Express would be better sticking to its usual obsessions rather than dabbling in fortune telling. Because that’s what this is. It has no real value. It’s simply a clumsy attempt to make AI sound like it can do something it cannot.
I work with AI every day. I use it to clean up dictated articles because speech-to-text struggles with accents, including my own. AI is excellent at certain tasks.
However, predictive analysis is not one of them.
Tools like Grok and ChatGPT are large language models. They work by analysing patterns in language and predicting what comes next based on existing data. They don’t think or reason, and they don’t foresee the future.
Yes, they can summarise data. Yes, they can produce probability models. But that’s no different from what FiveThirtyEight used to do. There’s no magic here, even though these articles try to present it as something close to voodoo.
That’s why the fundamental dishonesty of these stories is so irritating. They dress up guesswork as insight and speculation as science.
The Daily Record has been doing something similar with its three-person panel predicting results for the title contenders between now and the split. If you’re trying to forecast every game in advance, you might as well use tarot cards or a Magic 8 Ball.
In fact, you’d probably get results just as accurate by asking the first three people you meet in the street.
That panel includes Ryan Stevenson, Barry Ferguson and Chris Sutton. Stevenson and Ferguson both believe their teams won’t drop any points before the split. Sutton thinks Celtic will lose the derby at Ibrox.
None of them knows. None of them can know.
And this is the real issue. These aren’t predictions based on insight. They’re opinions filtered through bias. Even Sutton, who tries to be objective, is still guessing. That’s not a criticism of him. Nobody knows how this will unfold.
Everyone is searching for certainty. Everyone wants the piece of information that makes everything else fall into place.
But there is no such piece of information. No computer will find it. No AI will calculate it. And no pundit will intuit it. This title race will be decided where football is always decided: in individual games, in individual moments, in mistakes, in flashes of quality and in moments of pressure. That’s the reality.
We’re all on the roller coaster together. We’re strapped in. At this point, the only thing left to do is enjoy the ride. Or not.