Umpire's Call: How Two Seconds of Silence Rewires a Test Match

Umpire's Call: How Two Seconds of Silence Rewires a Test Match

My spreadsheet has 41 columns. Since 2026 it has logged 341 reviews — who asked, in which over, where the umpire was standing, how many seconds passed between the earpiece and the arm, whether it succeeded, and what the fielding side did in the three overs after losing one. Last winter I watched a single review nine times in a row.

Day four, final session. The spinner came around the wicket; the ball pitched on leg stump. The batter pushed the front leg across, the bat came down, then the pad took the impact. The on-field umpire raised his finger — out. The batter tapped his helmet and signalled for the review. Through the earpiece the third umpire said only, "Ball tracking coming."

Then two seconds. The on-field umpire dropped his right hand towards his waist and left it there, not raising the arm, not moving a finger. The crowd went quiet. On the stump mic there was nothing but wind. The tracking path appeared: the ball would have struck a small sliver on the inside of the stump. Umpire's call. The decision stood, the batter walked, the side lost its review.

Ninety minutes of silence can explain more than a thousand replays, but nobody keeps the ledger on those two seconds.

Context: What the law says, what the process does

DRS is simple in shape and not simple in logic. In Tests each side gets two reviews per innings, retained if successful; in ODIs and T20Is, one per innings. The batter has a 15-second window to signal, and if the side does not send word inside it, the review is gone. Since 1 June 2026, the ICC has removed the soft signal for catches; the on-field umpire's initial hunch is no longer treated as automatic evidence, and the third umpire judges directly from the footage.

More important is the umpire's call itself. It rests on the "clear and obvious error" standard. When ball tracking projects that less than half the ball would have hit the stumps, that does not meet the threshold of a clear error — so the technology's own uncertainty is not allowed to become the decision-maker, and the on-field call survives.

Treating that rule as a weakness is a mistake. It is an admission: ball tracking is a model too, a projection built from camera frame rates, revolutions on the ball, bounce and the pitch map. The model may be right in every frame, but between two frames the ball's position is an estimate.

Position first, law second

I start with a count, but a count never stands alone. I counted 118 decisions before I understood my own knee — a different sport, a different sum, the same habit. Where the umpire stands decides which line he can see and which one he will miss.

The bowler's-end umpire looks from beside the bowling shoulder. When a leg-spinner comes around the wicket to a left-hander, his sightline runs almost parallel to the ball's line, so he reads how far it is drifting in very well. But when the bounce is steep, judging height becomes hard. The striker's-end umpire is worse placed: from the leg-side angle he cannot cleanly see the inside edge of the pad or the gap between bat and pad. A ball that turns in only slightly does not give him pad-first contact.

That geometry is the biggest technical gap with spin. A seam-up delivery travels on a near-straight line, the seam behaves relatively consistently, and the projection is reliable. A spinning ball turns on its seam, changes direction between frames, and varies in bounce. That is why arguments flare around the deliveries of Nathan Lyon or Ravindra Jadeja. The argument is not about them; it is about the distance between the tracking model and the umpire's eye.

Audio shows what pictures hide

The stump mic is my favourite piece of evidence. On pad-first calls you often hear two sounds — one off the bat, one off the pad. Sometimes only one, meaning ball on pad with no bat. The problem is this: if the replay pictures and the audio timestamps differ by a fraction of a second, which one do you trust?

Neither, on its own. Audio is the first source, pictures are the verification tool, the scorecard is the final account. So for every review I line up three things: the umpire's position (what he could see), ball tracking (what the model says), and the audio (what the moment sounded like). When all three agree, I am satisfied. When they do not, I write: uncertain, and why.

I do not trust a pattern until it survives three different angles.

The clock: what a review actually costs

In my ledger the average review over the last two seasons ran 77 seconds. Tracking takes 20 to 30 of those; the rest is earpiece traffic, assembling the pictures and announcing the verdict. It does not sound long, but to a bowler standing at the top of his mark it is a great deal. In the three overs after a review, over-rate runs about 24 seconds slower — a figure from my own sheets, and an estimate bounded by the limits of a spreadsheet.

Sixty-four matches taught me that one spreadsheet is never enough. I have the number 341, but who actually lost what cannot be counted. Losing a review is not just losing an opportunity; it is losing a slice of control over a session, and nobody keeps an account of control.

Intervals: when heads change, numbers change

A Test day splits into drinks, tea and stumps. My sheets hint at something: reviews are requested more often in the ten overs before a break than in the first ten of a session. The easy explanation is fatigue lowering decision quality and the end of a session sitting heavier on the mind.

That is a hypothesis, not proof. There is a competing explanation — wickets turn more late in a session, spinners bowl more overs, and more turn naturally produces more reviews. Getting this past a hunch needs two or three more series. So it sits in the ledger marked: awaiting verification.

The contrarian angle: crowd anger and technological honesty

The crowd shouts "out, out", then stops when umpire's call appears. The argument sounds simple: the ball is hitting the stumps, so why is it not out? The demand follows — scrap umpire's call.

That demand is disbelief in the technology and excessive belief in it, held at the same time. The same person says tracking is not perfect, so it should not be the final judge, and also says whatever tracking shows is the last word. Both sentences cannot be true together.

The second problem is less comfortable. The review is now a momentum tool. Sides take it sometimes from conviction, sometimes to break a strike — a bowler has strung together three good overs, pressure is building, and what is needed is a pause. Even an unsuccessful review pays: two minutes of stoppage, rest for the quick, time to reset the field.

The decision itself has become a contest of reflexes — who lifts a finger fastest, who sends the signal quickest. Real skill lies elsewhere, in reading line, bounce and angle, not in reaction speed. If the call is settled purely by the speed of a hand, the game drifts from a contest of intelligence towards a contest of bodies. Sides that keep their review in hand can usually read the picture.

Takeaway: no claim without a count

My own ledger points to something clear. Arguments accumulate around every controversial decision, but the error usually is not in the technology, and not in the umpire's eye alone — the gap is in the process. There is no transparency on tracking; spectators do not know who is interpreting the images; and nowhere is there a record of how long a decision took.

If the ICC published a short review report after each series — how many reviews were used, how many were upheld, the estimated uncertainty on each tracking path, and the per-team retention rate — half the argument would settle itself. Technology does not end debate; it changes its language.

So next series, when "umpire's call" appears on the screen, ask one question: who lowered the hand, where was he standing, and how many seconds did it take? Answer those three and you will know what the game is actually resting on.

— Root: Referee