The Rule of Insufficient Evidence: Who Carries the Burden of Proof in Cricket's Review System

I was watching the match, and it was clear the problem did not sit in Aleem Dar's eye — it sat in the gap in the rule. March 19, 2026, Melbourne Cricket Ground. A World Cup quarterfinal, Bangladesh against India. Rohit Sharma was on 90. Rubel Hossain's delivery climbed above waist height, Rohit pulled, and a catch went up at deep midwicket. The umpire standing at the bowler's end did not call no-ball. Bangladesh's fielders raised their hands, the signal for a review came — and that is exactly when the real question landed. Can a non-call of no-ball be reviewed? Under the International Cricket Council's playing conditions that day, there was no answer. The technology existed, the cameras existed, the slow-motion replay existed; but the window for submitting evidence did not. Past that single delivery Rohit reached 137, India posted 302, and Bangladesh lost by 109 runs.

That night I understood the controversy was not about umpiring quality. It was about the definition of evidence. Who carries the burden, how much evidence flips a decision, and which way the default leans when evidence is absent — the answers to all three are written down in advance, and nobody on the field reads them.

A review system is really a law of evidence, not a technology showcase. From the start of DRS in July 2026 at the Sinhalese Sports Club in Colombo, through its gradual expansion into ODIs and T20Is over the following decade, what changed was not the number of cameras but the threshold of proof. The phrase the International Cricket Council uses for what the third umpire needs to overturn an on-field verdict is 'conclusive evidence.' That phrase sounds innocuous. It is not. It is a threshold, and a threshold means somebody decided where the line falls.

On November 6, 2026, at the Arun Jaitley Stadium in Delhi, Angelo Mathews walked out to bat in Bangladesh versus Sri Lanka, and a broken helmet strap ran the clock down. Bangladesh appealed, and Mathews became the first player in international cricket to be dismissed timed out. That incident teaches the same lesson from the opposite direction. Where the rule is explicit — a two-minute limit, a list of exceptions, a batter's obligation to be ready — no review system is even needed. The precedent ledger answers instantly. Where the threshold of proof is vague, the same argument returns year after year.

To see the difference, look at another sport. At the 2026 World Cup in Russia I logged 29 reviews across the tournament during football's VAR debut, among them the 58th-minute Griezmann penalty in France versus Australia, the first VAR-awarded penalty. Football's standard was 'clear and obvious error.' Cricket's is 'conclusive evidence.' Different words, but both answer the same question: who bears the burden of proof? Football said the error must be obvious, cricket said the proof must be conclusive. Both are rules designed to protect doubt.

This is where cricket's most unspoken truth hides. Cricket's laws have long operated a tiered evidentiary standard; nobody has simply named it.

In run-outs and stumpings there is no benefit of the doubt for anyone — the third umpire looks at the information and decides. Doubt does not favour the batter. In catches and edge cases, if conclusive evidence is missing, the on-field decision stands. And in LBW, ball-tracking creates a numerical door — whether the projection hits the stumps is determined by whether the centre of the ball falls inside the outer edge of the stump. Inside means 'hitting,' and the call flips. Hanging outside means the call survives; that is umpire's call.

This must be admitted: it is cricket's own null-handling protocol. When data is missing, the system returns a fixed answer — the on-field decision. When an umpire's doubt, estimate or experience takes the place of evidence, that is not weakness; that is engineering. In a data pipeline, if a system filled an empty input with an invented value, we would call it a bug. Cricket does not do that. Cricket returns empty-handed to the on-field call.

I opened the abandoned-match file, and the precedent ledger answered. In September 2026, the Chittagong Abahani versus Sheikh Russel Krira Chakra match was abandoned in the 89th minute after a referee was assaulted. The Bangladesh Football Federation deducted three points and fined five lakh taka. I spent three weeks then reading the match commissioner's report, Article 63 of the federation's disciplinary code, and two prior 2026 rulings. In those documents the question of evidence is arranged differently — no video, but a report, testimony and a timeline. The rule stays the same; the source of proof changes.

This brings up another reform from 2026. In May that year the International Cricket Council scrapped the soft signal, effective from the World Test Championship final in June. The common assumption is that reviews became more accurate. My reading differs. Removing the soft signal did not improve accuracy; it changed the default value. Previously the on-field umpire's hunch set the starting value. Now, if evidence is missing, the default is not out. The burden of proof is unchanged; only the empty cell leans the other way.

Hence my second realisation: the real cost of a review system is not the wrong verdict, it is the unexamined default. A default is invisible, so nobody audits it. Who set it? In which meeting? In which draft? Nobody asks, because watching the ball's projected path on a screen is far more thrilling.

Bangladesh's case is different. Large parts of domestic cricket have no DRS. So the default of proof there is the umpire's eye, and the ledger is the match commissioner's report. When technology is absent, insufficient evidence is not a temporary state — it is a permanent one. And whoever decides where cameras are installed indirectly decides which decisions will be questioned and which will not. The absence of technology is also a decision.

Now to the point where many will disagree with me. Cricket audiences remember Melbourne 2026 as an injustice. My precedent ledger remembers it as a gap. Those are two different things. An injustice can be appealed to the umpire; a gap has to be appealed to the lawmaker. The first hour of any controversy is always spent in freeze-frame, and a freeze-frame is never complete evidence. A still image stops time, but the law does not stop time — the ball's path, the batter's position, the state of the match are all in motion. I have watched this scene for eighteen years: two people read the same clip two ways, and both call their reading 'data.'

The referee's eye is trained on angles, not applause. The most instructive part for me is that better cameras have not reduced argument. They have increased it. When you show a threshold clearly on screen, viewers start questioning the threshold itself — and that is healthy. What is unhealthy is when the threshold is invisible but the verdict is visible.

A limit must be drawn here. 'Insufficient evidence' is an honest answer, but an honest answer and a neutral system are not the same thing. If a board decides how many cameras a match gets, it indirectly decides where evidence reaches and where it does not. In domestic competitions this limit often becomes an administrative question rather than a sporting one. I opened the protocol document, and the rule itself gave testimony: a system that cannot produce evidence cannot demand it either.

So what should the next reform be? Not another camera or another version of ball-tracking. It should be transparency — publishing, for every review, which standard was applied, what evidence was found, and where the on-field default did the work. If the International Cricket Council kept a review ledger — every out, every umpire's call, every record of insufficient evidence — then in five years we could say by calculation rather than guesswork which thresholds work and which survive only by habit. Precedent before reaction.

The question now is simple: the next time the screen says 'evidence insufficient,' will we treat it as a failure, or ask — who drew the limit of this evidence, and why?