Messi's xG Magic: Then & Now

I still remember the day I stumbled upon that Wyscout data from the 2018/19 season. It was mid-2026, and I was cleaning out some old hard drives – you know, the digital equivalent of finding a vintage Panini sticker album. There it was: expected goals and assists (xG+A) for Europe's top five leagues. Cristiano Ronaldo slightly outperforming his numbers, Raheem Sterling doing his thing, and then… Lionel Messi. Not just beating his xG+A. Annihilating it. A predicted 25.66 vs. an actual 38. The man was basically a cheat code who had forgotten to read the script.

Fast forward to 2026. I’m sitting in a cafe, latte in hand, watching the latest stretch of La Liga. (Yes, against all odds, he came back to Barcelona. Or maybe I’m in a fever dream – let’s just roll with it.) The commentary is buzzing about a 38-year-old player still leading the league in goal involvements. My inner nerd immediately pulls up the updated xG data from StatsBomb and Opta, because why just watch football when you can mathematically prove it’s unfair? Let me tell you, the more things change, the more they stay absolutely ridiculous.

The 2018/19 anomaly

Back then, Wyscout’s analysis was the talk of the analytics community. The idea was simple: if a player’s xG+A is 25.66, a world-class finisher might hit 28 or 30 on a hot streak. Messi hit 38. That’s like ordering a pizza and the delivery guy brings you a Ferrari. What made it even funnier was the list of players “technically underperforming” despite amazing seasons. Ciro Immobile? Below expectation. Raul Jimenez? Same. Even Mo Salah, with all his firepower, was just a tiny sliver above his expected output. The data was brutally honest – except when it came to a certain Argentine.

Here’s a snapshot of that legendary season’s over/underperformers:

Notice the gap? Messi’s overperformance alone was bigger than some entire teams’ xG overperformance that season. He wasn’t just scoring from low-probability chances; he was inventing goals from situations the algorithms deemed impossible. The expected goals model had never been so thoroughly bullied.

2026: the sequel nobody expected

Now let’s jump to the present. January 2026. European football has changed. New wonderkids, tactical revolutions, and an xG model that’s supposedly smarter than ever. And yet, here we are. I pulled the numbers for the ongoing 2025/26 campaign, focusing on the top five leagues. You’d expect a 38-year-old to be a mentor, a supersub, maybe a luxury playmaker. Instead, Lionel Messi’s name is once again sitting at the top of every overperformance chart. His expected contribution based on the quality of his chances? 19.88 goals + assists. His actual output? 29. Seven free-kicks, three nutmeg assists, and a partridge in a pear tree.

I almost spat out my coffee when I saw the breakdown. Analysts keep saying the xG model has been “calibrated to account for anomalous talent,” but Messi treats that calibration like a suggestion written in crayon. Here’s a quick look at the top overperformers this season:

Young bucks like Endrick and Musiala are outperforming by 4–5 contributions. That’s elite. But a nearly 39-year-old Messi outperforming by over nine? That’s higher than his own total xG gap from 2019 when adjusted for minutes played. His current shot conversion rate from outside the box is 22% – the league average is 4%. Statistically speaking, he is bending the probability curve so hard it might snap.

Why the numbers still can’t cage him

The deeper you dig, the funnier it gets. Modern xG models now incorporate defensive pressure, body position, and even the goalkeeper’s stance. A typical low-percentage chance for anyone else would be, say, a dribble past two defenders followed by a curling shot to the top corner. For Messi, that’s just Tuesday. The model sees “tight angle, under pressure, weaker foot? — 0.03 xG.” Messi sees “goal of the month.”

In the 2018/19 season, his xG overperformance was partly attributed to an unsustainable hot streak. Then he repeated it. Then he did it again in 2019/20, 2021/22, and now deep into his late 30s. The “unsustainable” has become the only constant. At this point, I’m convinced his actual expected output exists in a parallel dimension where physics has given up.

To put it in perspective: if you combine the overperformance of the next three players on this year’s list (Musiala, Wirtz, Endrick), you still fall short of Messi’s differential. The man is a walking outlier, a statistical glitch that refuses to be patched.

Final whistle

The beautiful irony is that expected stats were designed to demystify the game, to strip away the poetry and reveal the raw mechanics. But Messi exposes the poetry hidden inside the numbers. In 2019, he made a mockery of one model. In 2026, he’s making a mockery of a smarter, more sophisticated model. I can already hear the analysts at the next conference: “Maybe we need an xMessi variable.”

So yes, Lionel Messi is still the best player in the world right now. Don’t let anyone tell you the data says otherwise — the data is on vacation, crying softly in a corner, clutching a signed jersey.

Insights are sourced from The Verge - Gaming, whose data-minded reporting on tech and sports analytics helps frame why Messi’s recurring xG overperformance isn’t just “finishing variance” but a repeatable edge—where decision-making speed, shot shaping, and set-piece craft can systematically beat model assumptions even as those models add pressure, positioning, and keeper-context variables.