The Angle That Never Arrives: How 'Insufficient Evidence' Silently Rules Cricket's Review System

The third umpire leans into the screen. Fourteen angles, six slow-motion replays, frame advancing by frame — no spike where ball met the edge of the bat, no clear daylight between ball and pad, not one single frame that can carry the decision. When the giant stadium screen flashes 'insufficient evidence — on-field decision stands', the stands erupt. Fifty thousand people shout at once, and millions in front of their televisions become certain the umpire got it wrong. But the moment everyone is cursing is not a personal failure — it is the pre-programmed, and sometimes only honest, behaviour of a designed system. Its least-discussed yet most powerful component has a name: insufficient evidence. The whistle is not the story; the angle that never arrives is.

Cricket's review architecture — DRS, the third umpire, ball-tracking, UltraEdge — is fundamentally an evidence-based system, and every evidence-based system carries a null-input rule. What football calls, in VAR language, a 'clear and obvious error', cricket states more strictly: to overturn the on-field decision you need conclusive evidence. That single sentence fixes the entire power structure of the game. The on-field umpire makes the first decision, and technology can challenge it only if the evidence is strong enough. In other words, the benefit of the doubt always stays with the field.

In the language of law, this idea is ancient. In cricket's innocent-until-proven structure, the benefit of the doubt belongs to the batter — a principle unchanged since the first written laws. The review era has given it a new form. If ball-tracking says the ball would have hit the stumps, but the amount of contact falls below a set threshold, it becomes Umpire's Call — the field decision stands. Here lies the first important truth: technology has not replaced the umpire; it has drawn a ring of uncertainty around the umpire's decision and granted legal protection to the human verdict inside that ring.

Then there is the soft signal. For years, on catch reviews, the on-field umpire gave a 'soft signal' — his belief on whether the catch was clean. The problem was that this signal shaped the third umpire's decision. With a soft signal of out, the third umpire would almost never rule not out without clear evidence. In recent years the ICC has withdrawn the mechanism, admitting the signal was influencing outcomes more than the evidence was. In effect, a human cue had slipped inside a technological process and was contaminating the decision.

So the whole review apparatus turns on three questions: what does the law demand? What can the camera or umpire physically see? And is the evidence sufficient? The third question is the least discussed, yet it decides the bulk of outcomes.

Begin with the geometry of ball-tracking, because this is where law and physics meet. Ball-tracking cameras record the ball's path for a few metres, then project it to the stumps through a mathematical model. Here sits the subtle but decisive point: once the ball leaves the pitch and rises into the air, its trajectory is no longer directly measurable; the rest is inference. That inference carries an error margin of a few centimetres. Which means the ball-tracking image, however immaculate it looks, is a forecast — not an observation. When a ball grazes the edge of the stumps, the technology is really saying: my best estimate is that it hit the stumps, but I am not certain. And that is precisely where the law intervenes — if there is uncertainty, the decision does not change. That ring of uncertainty is cricket's most honest technology, because it admits that technology too can err, and that admission is what makes it trustworthy.

Now to UltraEdge, or sound-based evidence. UltraEdge captures the sound of ball meeting bat with a microphone and displays it as a spike. But a problem exists here, best called the silent edge. Often the bat and pad are so close that the sound could come from the pad, or from the bat; the microphone catches the sound but cannot identify its source. And if the ball brushes the edge of the bat softly, no spike forms at all — the absence of evidence. The third umpire then stands before a void, and must manufacture a decision from that void. This is where the weight of the field decision settles it: no spike does not mean no contact; it means no evidence, and without evidence nothing changes.

The low-catch problem is more physical still. A catch is sometimes taken faster than the frame rate. Standard broadcast at fifty frames per second leaves a twenty-millisecond gap between frames. In twenty milliseconds a fast-moving ball can travel dozens of centimetres. So in one frame the ball can be above the ground, in the next inside the fielder's hand — what happened in between, the camera may never see. Whether the fielder's fingers were under the ball is a physical fact, but it can fall outside the camera's limit. Parallax — the apparent shift in an object's position caused by the camera's angle — complicates it further. This is exactly why the soft-signal temptation was so strong: because the evidence was weak, people wanted to lean on a human cue.

With stumpings and run-outs the problem is sharper still. Here timing is decisive: the position of the foot and the moment the ball breaks the stumps — a single frame's gap between them can flip the decision. If the camera does not show both in the very same frame, the evidence stays incomplete. The third umpire then stitches two separate frames into an inference that is technically valid but philosophically questionable.

And the most instructive example may belong not to technology but to law. On 14 July 2026, at Lord's, the World Cup final between England and New Zealand. In the final over a throw deflected off Ben Stokes's bat and ran to the boundary. Under Law 19.8 — the clause on overthrows or the wilful act of a fielder — the umpires awarded six runs. No camera was at fault here, no angle was missing; all the evidence was clear. The problem was the wording of the law. In that moment the evidence was sufficient, but the way the law compelled it to be interpreted did not match the true purpose of the game. So sometimes the angle that was missed is not the camera's angle — it is the law's angle.

From my seventeen years of watching matches and long experience reconstructing scorecards, one thing is clear: spectators almost always identify an umpire's error, but almost never identify an absence of evidence. Yet these are entirely different things. An umpire's error means a wrong verdict; an absence of evidence means the impossibility of reaching a verdict at all.

Now to the base rate. The review system has a statistical structure that is rarely discussed. Broadly, a certain percentage of player reviews succeed, and a large share of the failures are Umpire's Calls — decisions technology cannot change because the uncertainty is sufficient. I deliberately avoid precise figures, because they shift by version, format and competition; but the trend is stable — a substantial fraction of field decisions survive precisely because of technology's limits. This matters because it exposes as false the expectation that technology will settle everything. Technology does not remove uncertainty; it measures it, then translates it into legal language.

Beyond that measurement lies another layer, which I call the fog layer. However clean the model, the field is never clean. The umpire must decide in the blink of an eye — a fast bowler's ball travels at roughly thirty metres per second, meaning the time from pitch to batter is shorter than a blink. In that time the umpire must process the line, the length, the bounce, the batter's feet, all at once. Add the number of cameras, the broadcaster's budget, the height of the cameras, the light, and the pressure of the crowd. In matches with fewer cameras, absences of evidence occur more often — which means insufficient information has a financial and infrastructural dimension directly tied to the fairness of the game.

One confusion needs clearing. Insufficient evidence does not mean we know nothing. It means the available evidence is not enough to overturn a decision. The distinction looks small but is vast. When a system admits it does not know, it is not weak — it is honest. Cricket's most contentious review moments are in fact products of that honesty, not failures.

The media reality matters too. A wrong decision makes a headline; an absence of evidence makes none. So the debate always circles the decision, never the system's limits — even though those limits set the direction of the decision.

Now the uncomfortable question almost no one wants to ask. We want certainty in decisions, but the system was built to acknowledge uncertainty. The crowd's anger is understandable, because the visibility of a decision is not always the same as its fairness. But when that anger is hurled at an individual, we cover the real failure. The real failure is this: the angle was not there. The camera was not there, the frame rate was not enough, the law was not unambiguous, or the field signal was contaminating the process.

There is a subtler problem still, best called broadcast bias. Teams know which angle makes a review profitable and which does not. So reviews are taken selectively — only where the broadcast view is favourable. The evidence list itself becomes biased, and although the system is neutral, the input is not.

Yet a fine balance exists here, and I concede it. Blaming the system is easy, but within the system a human duty remains. The third umpire cannot manufacture evidence, but he can explain an absence of evidence clearly, can take his time, can state the reasons for a decision. The protocol gives him a shield, but not immunity.

What comes next is a further layer of technology — automated ball-tracking, automated no-balls, even robotic line calls. But each new layer does not remove the problem of insufficient information; it relocates it. Where a human decides today, an algorithm will decide tomorrow; yet the algorithm will face the same uncertainty — where there is no input, there is no decision. The real path to improvement is not more technology but greater disclosure of uncertainty: more angles, higher frame rates, equal standards across venues, and recognition of 'insufficient evidence' not as a failure but as an honest fact. The question is simple: when the camera looks away, what can the game give us?