Umpire's Call: How Cricket's Review System Actually Decides — and How Teams Get It Wrong

March 19, 2026, Melbourne Cricket Ground. A World Cup quarter-final, Bangladesh against India. The 40th over, Rubel Hossain to Rohit Sharma, on 90. A catch behind, the on-field umpire's finger goes up — out. Then the review. For the better part of two minutes the big screen runs the angles: behind the wicket, front-on, snicko, slow motion. Then the decision turns over. Rohit stays, India's innings swells, and the margin at the end of the night is 109 runs.

Half a subcontinent split that night. Some said the ball had hit the ground, some said glove, some said the camera frame rate was never sufficient. I had no notebook then. It would arrive three years later, in the summer of 2026, at a radio desk in Dhaka, where I time-stamped and logged all 22 VAR reviews of that football World Cup. But the question from that night in 2026 never left: we are not arguing about the truth of the ball. We are arguing about the boundary of proof — and the law drew that boundary long before the camera arrived.

Context: the grammar that runs a review

The Umpire Decision Review System began as a trial in July 2026, during Sri Lanka's home Test series against India. Early on it was almost a courtesy: whether you wanted it or not was a matter of consent. Then it became permanent, and with it the grammar of how cricket writes its sentences changed.

In Tests today each side gets two reviews per innings; past the 80th over, if a team has fewer than two left, it is topped up to two. In limited-overs cricket the ledger is tighter — one review per side. Time is also a currency: a review runs the replays and generally freezes the match for 60 to 90 seconds. In a T20 innings that lasts 80 minutes, 90 seconds is a lot of ground.

But the foundation sits elsewhere. The ICC introduced the system with a precise mandate: to eliminate the obvious umpiring error. The system is built for obviousness, not for truth. That is exactly why every review lands on 'conclusive evidence' — if you want a decision overturned, your proof has to sit outside every shadow of uncertainty.

In 2026 another brick came out of that architecture. In March the ICC Men's Cricket Committee recommended abolishing the soft signal, effective from 1 June 2026; the World Test Championship final at The Oval was the first major stage under the new reality. The soft signal meant an on-field umpire gave a weak indication on a doubtful catch, and the TV umpire needed certain proof to overturn it. The protocol is gone. The question has simply moved: decisions are still made on the weight of proof, and there is always an invisible layer underneath.

My 2026 method was rudimentary: start and end time of each review, a drawing of the official's signal, how many seconds each angle was shown. The subject was football, but the discipline serves cricket better — because cricket has a finite number of replay angles, and every angle carries its own blind spot.

Core analysis: four angles, one verdict

Angle one — ball tracking is a forecast, not a recording

The biggest misconception about ball tracking is that it is video. It is not. Up to the stumps, what you see is recorded. Behind the stumps, what you see is a prediction — a model that reads pace, spin, bounce and pitch type and projects a future. When the ball hits the pad and stops, the rest of the path never happened; it is inferred. So the question in front of the TV umpire is not 'where did the ball pitch'; it is 'how much can we trust the model in this situation'.

Umpire's Call is not a technological failure; it is a deliberate allocation of error between a machine and a human. When ball tracking predicts the ball hitting the stumps but that projection grazes the very edge of the model's own margin, the system hands the final word back to the on-field umpire. And the reviewing side keeps its review, because the decision was not 'clearly wrong'. That is the grammar.

Angle two — Law 36 is itself an unequal document

No camera can repair the asymmetry of the law. LBW has three conditions: the ball must pitch in line with the stumps, impact must be in line, and the ball must be predicted to hit. But if a batter is playing a shot and impact is outside off, it is not out. If the ball pitches outside leg, it is not out. Those two exceptions are the standing grievance of a generation of spinners, and they have nothing to do with technology. Much of what the subcontinent calls 'a review injustice' is in fact a statutory exception — and review makes it more visible, not less.

There is a strange consequence here: the more review is normalised, the more the law's imbalances become news. The camera does not indict anyone. It only shines a light.

Angle three — a review is a fixed asset

Every review is an investment decision, and most teams treat it as an emotion rather than a probability sum. Working a VAR desk taught me that requesting an on-field review is a risk: you have just spent one hand. In cricket the risk is more exposed — losing a review means no avenue against a clear error for the rest of the innings.

In my notebook I logged who asked, at what score, at what stage. Cricket showed the same pattern: a captain who burns a twelfth-over boundary-line LBW chasing a twentieth-over caught-behind has sold his rights for the next sixty overs. The gap between a captain's instinct and a dressing room's probability table is what turns matches.

Angle four — the blind spot is budgeted too

Technology's limits are not only mathematical, they are financial. Not every match carries the same number of high-speed cameras, snicko units or ball-tracking rigs. Some fixtures are million-dollar broadcasts with three or four high-frame-rate angles; some international duty is covered by two cameras, where you cannot tell whether the seam has rotated. Same law, same doubt, different territory of proof. Where the threshold for conclusive evidence is tied to a broadcast budget, the word 'system' means different things to different boards.

A latency note

From Doha to a remote desk, distance is measured in latency, not miles — and that is truer in cricket. The feed arrives eight to forty seconds late; audio and video drift out of sync; stump mic is not always on. So a remote vantage point has to list its own limits. Watching behind-closed-doors football in 2026 taught me that in an empty ground, every stump-mic whisper becomes a confession — an umpire's 'not out' and his hesitation land on the same recording. But what the man standing at the ground hears, I can never fully hear. Distance is a limitation, not a credential.

Contrarian angle: the problem is not the machine

The easiest conclusion is the wrong one: that scrapping Umpire's Call fixes everything. Remove the margin, and the entire burden of decision migrates to a private model that has its own error bar — one nobody will see any more. The more 'accurate' cricket tries to become, the more authority it hands to an invisible arbiter. A system's legitimacy comes from its transparency, not from the number of decimal places.

Second, the real inequity is not sides against the system but board against board, which maps directly onto access to evidence. Third, fan memory is biased: we never remember the reviews that were rightly won, and we carry the lost ones for a decade. If every 'tragic review' since the middle of the last century were put in one ledger, more than half would be weak appeals nobody discusses.

Takeaway: waiting for the review coach

Before the next major tournament I want to see one thing in a dressing room: a person on the bench who is not the batting coach or the bowling coach, but who runs the review desk. Where a TV umpire takes a hundred seconds, someone should own the eight-second probability call.

My Dhaka notebook has more questions than answers, and that is the point. The next time the big screen flashes 'Wickets: Umpire's Call', will your captain be making a decision — or taking a chance?