Squash Spent a Century Trusting the Eye. The Next Champions Will Trust Their Twin.

The slowest racket sport to embrace data is about to leap past analytics entirely. Here’s why we’re building digital twins and why squash is only the start.

I learned to read a squash court before I ever learned to read a spreadsheet.

I grew up in Egypt, where squash is less a hobby than a national apprenticeship, and I won my first British Junior Open title at thirteen. By the time I joined the professional tour, I could feel a match turning: the way an opponent’s drives start landing a foot short and the moment they stop volleying and start retreating to the back corners.

Every good player can feel these things. The problem is that feeling is not the same as knowing, and it is nowhere near the same as proving.

The sport that kept trusting the eye

For most of squash’s history, that gap didn’t seem to matter. The sport ran on the trained eye: the coach courtside, the player’s own memory of a match, and the shared folklore about who was dangerous from where.

Tennis built Hawk-Eye. Baseball rebuilt its entire front-office logic around data. Basketball put optical tracking on every possession. Squash, one of the most spatially intricate games there is, with two players sharing a single box and every shot defined by length, width, height, and pace, kept trusting the eye long after the other racket sports stopped.

I don’t think that was because squash players lacked an appetite for data. I think it was because the tools were built for sports with bigger budgets and bigger broadcast deals, and nobody bothered to build them for us.

I found out how much that gap costs the hard way. Then I found out how much closing it is worth.

Leaving the tour to learn the other language

In the middle of my professional career, I made a decision most players on the tour considered strange: I interrupted it to study computer science at Trinity College in Connecticut, the most successful program in the history of college squash.

I went 42–9 over my time there, captained the team in my final year, and learned to think about the game in a second language. Not the language of instinct, but the language of patterns, models, and evidence. Playing for a coach like Paul Assaiante teaches you that talent is common and preparation is rare. Studying computer science taught me that preparation, done right, is mostly a data problem.

When I returned to the tour, I treated my own matches as a dataset. I stopped asking “how did that feel?” and started asking “what actually happened, and how often?”

Where were my errors clustering? Which rally lengths was I winning, and which were quietly costing me points? What did an opponent do on the third ball after a hard cross-court every single time, not just the times I remembered?

In roughly a year, I climbed about five hundred places in the world rankings, from outside the top seven hundred to a career high of 179. I’m not going to pretend a graph hit a serve for me. But I’ll say this plainly: the players who saw their own game clearly improved faster than the players who only felt it.

Why I stopped calling it analytics

Here is the limitation I kept running into. Analytics, even good analytics, only describe the past. They tell you what happened. The harder and more valuable question is what would happen: against this opponent, with that game plan, when you are two games down and your legs are gone.

That is the leap we are building at Core Sports AI, and it is why I no longer call what we do video analytics. We are building a digital twin: a living, data-driven model of a player that you can test against before you ever step on court.

Feed the twin enough match data, and it starts to behave like the player it is modeled on, favoring the same corners and making the same decisions under the same pressure. A coach can run a game plan against an opponent’s twin the way an engineer stress-tests a bridge in simulation before anyone pours concrete. A player can watch how their own twin falls apart when it is tired and train that weakness out before a real match ever exposes it.

That is a different kind of tool. Analytics hand you a report. A twin hands you a sparring partner that never gets tired and never lies to you.

Who gets to use it

My conviction, and the reason I started building rather than just playing, is that this should reach everyone.

The junior coming up through a squash academy should be able to model her game the way a world number five models his. The club coach should be able to test a plan without relying on a memory that, however expert, is still just one human’s recollection of a blur.

The barrier was never the data. It was access, and access is an engineering problem, which means it’s a solvable one.

Data doesn’t replace instinct. It audits it.

People sometimes assume that putting this kind of technology into a sport this artful will flatten it, turning intuition into arithmetic. I think the opposite is true.

The best players I’ve ever shared a court with have an instinct no model will replace. A twin doesn’t compete with that instinct; it audits it. It tells you when your gut is right and, more usefully, when your gut is fooling you. It turns a vague sense that “I struggle in the front left” into a number you can train against.

The artistry stays. What disappears is the guesswork dressed up as feel.

Squash is where we prove it. Racket sports are where it goes.

We are starting with squash because it is the sport I know in my bones and because it has been starved of this kind of technology for too long. But a digital twin is not a squash idea. It is a racket-sport idea.

The same engine that models a squash player’s patterns models a tennis player’s, a paddle player's, and a badminton player’s. The court changes shape. The underlying problem, turning a blur of human decisions into something you can study and simulate, does not. Squash is where we prove it. Racket sports, all of them, are where it all goes.

The next generation of champions will not be the ones who feel the game best. Plenty of players feel it beautifully. They’ll be the ones who can also see it, clearly and honestly, and who can rehearse against a faithful model of themselves and their rivals before the first serve is ever struck.

I spent the first half of my life learning to read the court by eye. I’m spending the rest of it building the tools that let the next player do both.

I’m the founder and CEO of Core Sports AI, where we’re building the first digital twin technology for racket sports, starting with squash. I’m a former PSA World Tour pro (career-high #179), a three-time Egyptian national champion, and a former captain of the Trinity College men’s squash team. If you’re working on racket sports, sports technology, or computer vision, I’d love to hear from you.