Enough with the "Opta Supercomputer"

Statistics have their place in the game, but this ain't it.

If you’ve scrolled the boundless vertical landscape of football Instagram recently, you’ve likely seen it:

ARSENAL HAVE 67% CHANCE OF WINNING PREMIER LEAGUE TITLE, ACCORDING TO OPTA’S SUPERCOMPUTER

Usually, there’s some bold-face text superimposed on an image of Arsenal manager Mikel Arteta making a face. Maybe the North London club’s formidable defensive duo, Gabriel and William Saliba, are smiling widely while embracing near the corner flag.

The visuals may vary, but the message is absolute: Arsenal are the favorites, and to a quantifiable degree. The tone of these posts suggests that this information is of divine provenance, a gift from the machines, as if Moses just came down from Mount Sinai carrying a printout from … The Supercomputer.

But what the hell is even that?

It seems to be a marketing term for a statistical model, not unlike how any and every algorithm has now been rebranded as “AI.” Maybe the Supercomputer is also AI — why not. But how else can we describe it?

One way might be: “useless.” Around this time last year, Opta reports, the Supercomputer announced that there was a 75.3% probability that Manchester City would win the Premier League. Pep Guardiola’s team then romped to yet another English title, showcasing their perpetual dominance to the watching football world … ah, shit.

The Citizens finished third, and that was actually a serious recovery from a far worse position. The Supercomputer crowned them overwhelming favorites in November 2024, but by December 21, they were nine points off Liverpool’s pace¹ and spent the rest of the season scrapping to get into the Top Four.

Of course, the stats gurus would say this was just part of the 24.7% — the universe of scenarios not part of the 75.3%. But what is the point of these numbers?

“You have a 67% chance to win the league, unless something changes over the next seven months.”

City crashed spectacularly last season because a bunch of key personnel went Wile E. Coyote off the age cliff — a factor which apparently was not incorporated into the Supercomputer’s model — but mostly it was down to Rodri getting injured. Their performances didn’t immediately collapse after their Spanish midfield general did his ACL in September 2024, but that was what ultimately sent their season spinning earthward.

That’s why it doesn’t matter what exactly gets pumped into The Supercomputer before it spits out these predictions. It cannot possibly account for all the twists and turns in a single game of football — the inflection points and sliding-doors moments and thunderbolts of complete chaos that can strike at any time — let alone all the variables in play throughout an entire season. What if Declan Rice gets injured for Arsenal in their next match? We’ll just give The Supercomputer a Brogan Adjustment until it spits out 43%, then forget that whole 67% thing ever happened? Why bother with it in the first place when you can just say what’s true, that Arsenal are top of the league?

To be fair, some of this numerical resentment is probably tied up in my anxieties as a fan of the Arsenal. Three-straight second-place finishes, two where Arsenal led the pack before coming up short, will disabuse you of the notion that probabilities and percentages matter when the injuries and the red cards and the inexplicable individual mistakes hit. There’s also the knowledge that the football media is coded down to its DNA to build things up in order to tear them down, and that Arsenal are a target for trolling even from the Mainstream Objective Respectable Establishment Outlets on a level only comparable (maybe) to Manchester United.² The purpose of the 10 million posts about how Arsenal are favorites to win the title is not, I can assure you, to put the Gunners and their fans at ease.

But more than all that, this is the lump of coal at the bottom of the Statistical Stocking. There are a lot of fine and revelatory statistics out there, and Opta serves up many of them on a consistent basis. (They’ve got plenty in that article explaining why Arsenal are favorites!) But we are reaching a kind of bland, inscrutable, hyper-amalgamated endpoint for statistical analysis. I don’t care much about xG, but I can grasp on some cerebral level what people get from it. Personally, I’m just not moved by this stuff in general — it’s bloodless and boring compared to the stories of the game — but I try not to make my coverage a global jihad against statistics. If field tilt and PPDA get you going, so be it.

This Supercomputer business, though: It shall not stand.

You look at the winding path up the mountain of a title-winning campaign, all the glorious uncertainty in every one of the 29 games to come, the creeping frost of November and the boisterous belligerence of Boxing Day and the trench warfare of January and February and the terrifying opportunity of Winning Time in April and May, and you assign it a number? It’s like paving a highway up Everest, or the far more treacherous K2: It isn’t possible, and it’s an ugly goddamn idea in the first place. It’s disrespectful to the very concept of this thing. What good will your probabilities do you when you reach The Bottleneck?

Anyway, here’s a number: There’s 76% of the league season still to go. Put your calculators away and enjoy it.

The podcast is off this week, but here are…

A few matches worth your time this weekend:

NOTTINGHAM FOREST vs MAN UNITED

Sat @ 11am ET, 3pm UK

The visiting Red Devils are suddenly resurgent under Ruben Amorim, even on points with Manchester City and sniffing around the Top Four. Forest are in dire straits, now on their third manager of this young season in the figure of Sean Dyche. The wily old Premier League veteran may be just what the Tricky Trees need to climb out of the relegation zone, and at home, they could disrupt United’s flow state.

As an alternative, at Dyche’s spiritual home of Turf Moor, BURNLEY face ARSENAL in the same match window. It’s the first of three away matches in a week for the 67% favorites…

TOTTENHAM vs CHELSEA

Sat @ 1:30pm ET, 5:30pm UK

Here’s a proper London derby featuring two clubs who don’t like each other one bit. Tottenham have taken Chelsea’s preordained place in this league, sitting third after nine matches, while the Blues — “Champions of the World” and preseason title challengers — have slipped to ninth. Enzo Maresca needs a result here, but Thomas Frank seems to have hit on something real with Tottenham so far.

There’s also REAL SOCIEDAD vs ATHLETIC CLUB in the same window, and the Basque derby is a special one. Beyond that, there’s also an erstwhile Bundesliga title bout in BAYERN MUNICH vs BAYER LEVERKUSEN.

LIVERPOOL vs ASTON VILLA

Sat @ 4pm ET, 8pm UK

Here’s a peach of a fixture to close out your English Saturday. Liverpool are in free fall after cranking out the points to start the season: They won all of their first seven matches, but they’ve lost six of their last seven in all competitions.

Arne Slot’s army will retreat to their citadel, Anfield, where they’ll hope to fight off a resurgent Aston Villa. The season so far is on an opposite trajectory for Unai Emery’s bunch, as they’ve followed up a poor start with a fine run that culminated with a 1-0 victory over Manchester City last weekend.

PLAYOFF GOAT WATCH

Sat @ 7:30pm ET

Lionel Messi and the crew are looking to close out their first-round playoff series against Nashville Soccer Club.

It’s the first of many hurdles that stand between history’s greatest player and the biggest prize of them all: the Philip F. Anschutz Trophy.

MAN CITY vs BOURNEMOUTH

Sun @ 11:30am ET, 4:30pm UK

Here’s a genuine top-of-the-table encounter, if a somewhat unlikely one. City took that aforementioned L from the Villans last weekend and slipped to fifth in the table, one point behind Sunderland. The Black Cats are a great story, but so are Bournemouth. They’re second in the league. Andoni Iraola has done a phenomenal job on the South Coast, no doubt about it, but City will jump above his team if things go the wrong way for the Cherries at the Etihad.

MILAN vs ROMA

Sun @ 2:45pm ET, 7:45pm UK

This is fourth hosting second in Serie A, as each of these grand old clubs enjoys a renaissance. Massimiliano Allegri wasn’t a universally beloved appointment among the AC Milan faithful, but he’s steered the Rossoneri back into contention, battling the likes of Napoli and Inter at the business end of the table.

The story at Roma is perhaps even more spectacular, as Gian Piero Gasperini has journeyed down from a hugely successful spell with Atalanta in Bergamo to build on Claudio Ranieri’s legendary tenure with Roma last season. The capital side are joint-top with Napoli on 21 points after nine matches.

1. The Reds also had two games in hand.

2. This is because it’s good for business. Online Arsenal fans reliably engage with this stuff, boosting it in the algorithm and driving views.