Umpire's Call: When the Calibration Margin Becomes the Law

Hook

The final over at Lord's, 14 July 2026. Trent Boult releases, Ben Stokes dives in to complete the second run, Martin Guptill's throw ricochets off the bat and runs away to the boundary. The two umpires at the ground, Marais Erasmus and Kumar Dharmasena, raise their arms for six: two runs plus the four-run overthrow. That single run eventually swung the final.

No one questioned whether the ball had touched the bat. No camera failed, no technology was blamed. The question sat elsewhere: under Law 19.8, which runs get added to an overthrow that reaches the boundary, and why the question of whether the batsmen had crossed when the throw was released enters the calculation at all. A year or so later, Dharmasena himself admitted in an interview that it had been a judgement error.

Slow the frames down and one thing becomes clear. The problem was not eyesight; it was the gap between the precision of the measurement and the language of the law. That gap sits inside cricket's most argued-about term today: umpire's call.

Context

In 2026, at the very end of the nineties, I covered the Wills Cup in Dhaka for Prothom Alo, and that is where the writing began. Reports were built from the scorecard then, and arguments lived in the pavilion. The habit of settling arguments with the law book in hand came much later.

Then came 2026, and age 59. A disputed 89th-minute penalty in a Chattogram Abahani versus Dhaka Abahani match set the whole city arguing. What I started that day on Facebook Live, with a paused replay and a walk through Law 12 and the referee's positioning, I named 'Referee's Eye'. The first episode drew 3,200 views. Within three months the page had 12,000 followers. I also set up a WhatsApp group with 15 local referees where every disputed decision was re-examined frame by frame. I wrote my first legal explainer then, citing IFAB Law 11 and the 2026-17 Laws of the Game.

The following year, 2026, brought the VAR debut in Russia. Watching from Chattogram, I tracked every VAR review of the tournament. The penalty for Griezmann in the 58th minute of France versus Australia sat near the start of that list. Of the 29 reviews, 17 overturned the on-field decision. I spoke by email to two retired FIFA referees to understand how they carried the pressure of an overturned call. I wrote a 12-part series called 'VAR and the Death of Referee Intuition'; it was shared 8,000 times.

Football's VAR and cricket's DRS should not be treated as the same animal. Football has no formal acknowledgement of uncertainty. Cricket does, and its name is umpire's call.

Core: What the Tape Shows When Slowed Down

An LBW review is not one decision but three separate questions. The first is pitching: where did the ball pitch, in line with the stumps or outside leg? The second is impact: did the ball strike the pad in line with the stumps? The third is projection: had the pad not intervened, would the ball have hit the stumps?

The first question is binary. Ball pitched outside leg means not out; there is no room for inference, the system simply computes the trajectory. But the second and third questions bring in the projection, which is prediction, not observation.

Under the ICC playing conditions, if the projection shows the ball hitting the stumps but the centre of the ball is inside half a ball's width from the edge of the stump, the on-field decision stands. That is umpire's call. Where the umpire gave it out, it stays out. Where the umpire gave it not out, the batsman survives.

This design does not hide the technology's error. It translates that error into statutory language. Football's VAR has no such boundary; offside is decided by millimetres, yet the system's own margin of error was never given formal standing. Cricket is the only major sport that has written the machine's uncertainty directly into the rule.

When the Bangladesh Premier League was cancelled in 2026, I retreated indoors and watched 500 match tapes from 2026 to 2026, building a private database of 1,200 referee decisions, each tagged by law, minute and outcome. The pattern is clear. Umpire's call incidents are not random; they cluster at the very edge of the band, where the centre of the projection grazes the edge of the stump.

There is also a calculation almost nobody carries in their head. Under the playing conditions of many competitions, a review that ends in umpire's call is retained. The challenge on the field stops being an expression of belief and becomes a probability calculation, an asymmetric balance of cost and refund. Where the law declines to punish error and instead converts error into a valid outcome, part of captaincy becomes probability theory.

There is another layer of technology with none of this complexity. Front-foot no-balls are now a binary decision on the third umpire's screen; whether the foot landed beyond the line is not arguable. In 2026 the ICC removed the soft signal, taking away the obligation to lean on the on-field umpire's assumption. Cricket is steadily handing binary decisions to the machine, and compromising with the human on probabilistic ones.

A binary decision can be delegated wholesale to technology. A probabilistic decision forces the law to compromise with the human, and umpire's call is precisely that compromise.

That compromise also carries a match-control cost, visible in another sport. At the 2026 World Cup in Qatar, semi-automated offside checks averaged 25 seconds, and those 25 seconds became a new variable in the rhythm of a match. In the same tournament I watched Morocco's 4-1-4-1 low block concede only one open-play goal across five matches while holding 38 percent possession. Referees allowed 14 tactical fouls per match without a card. The foul threshold and the projection threshold are the same kind of decision: where the line is drawn, settled in advance.

Contrarian Angle: The Umpire Is Not the One Being Protected

The conventional view is that umpire's call protects the umpire, guarding his authority. Slow the tape down and the truth is more uncomfortable.

Umpire's call does not protect the umpire. It protects the protocol. Decisions that end in umpire's call are not counted separately as correct or incorrect in any official statistic. An umpire's accuracy therefore becomes impossible to measure honestly: where the machine says the ball would have hit, the responsibility dissolves into the name of the law regardless of whether it would have.

And the real benefit sits elsewhere. The margin the rule rests on, half a ball's width, is not a measure of the tracking system's actual error; it is a substitute number. No broadcaster or provider publishes the per-delivery calibration margin. A law that leans on a number never disclosed is only partially transparent, and that inflates rather than reduces public distrust of the technology.

I watch Bangladesh's matches from Chattogram every day as well. But the referee's eye imposes one condition: the same frames, the same law, the same threshold for both sides. Change the country and the frame does not change.

The same confusion drives the shouting. Projection, calibration margin and on-field call are three distinct things, collapsed into one and then blamed on the machine. Slow the tape down and the three layers separate again.

Looking Forward, Not Summarising

A published calibration ledger, recording each series' tracking error margin and software version, is not a large ask. And replacing the review-retention rule with a fixed challenge budget, a set number per innings with no refunds, would make captaincy decisions cricket decisions again. The question is not whether we trust the technology. The question is how far we will force it to show its sums.