The Scorebook Is Silent, the Tape Speaks: An Audit File on Cricket's Record

March 2026, Bengaluru. Steve Smith signals for a review against an lbw decision at the M. Chinnaswamy Stadium. The fifteen-second clock is running — a clock that sits not in the Laws of Cricket but in the ICC playing conditions, and within which a batter must make his intention known. Ball-tracking arrives. The ball is shown to be going on to hit the stumps, but its inner clearance falls inside the error band that television compresses into a single phrase: umpire's call. The on-field decision stands. Somewhere inside that sequence Smith's eyes lift towards the dressing room, umpire Nigel Llong sends him back, and for a fortnight the television argument is about character rather than protocol. India win the Test by 75 runs. Virat Kohli says he will not call opponents cheats.

The broadcast cut that sequence down to nine or ten seconds. On my desk it ran in four steps: the player's signal, the umpire's confirmation, the replay sequence, and the ball-tracking projection with the tolerance band that television converts into a binary. The argument that ran through every living room ran on the fourth step; the first three never reached the table. The scorebook says one thing; the tape says another, and the column that stays blank is the one we fill with feeling.

A whole apparatus kept outside the law book

There is a distinction readers routinely skip. LBW, no-ball, appeal — these live in the Laws. Law 2 defines the umpires' role, Law 21 the no-ball, Law 31 the appeal. The review machinery — how many reviews, inside what clock, how wide the ball-tracking margin, and whose decision survives inside that margin — is not in the Law book at all. It sits in the ICC playing conditions, a document that can change series by series. The distinction matters because the single most argued-about object in the sport is not part of its permanent law.

The record question is older still. For decades the scorebook was the only evidence. Satellite replays arrived in the 1990s, and in 2026 ball-tracking was used on television in England for the first time. Since then two registers have run in parallel — the card and the tape — and no body routinely audits one against the other. A third register, the match referee's report, is not public. Three registers, three owners, no audit.

Reading numbers without a format tag is the profession's second disease. A T20 average and a Test average are not the same currency. For my own desk I keep rough bands: around 180 strike rate for a T20 finisher, around 5.2 economy for an ODI bowler, around 45 average for a Test batter. These are signals, not statutes. A reader who lifts an IPL strike rate to argue about ODI form does not have an argument; he has a translation error. One match, one season, one protocol — that is my scope. A reporter who never sets a scope disappears into the excavation of his own anomaly log and never files the piece.

The Motera register: 1,412 overs

In 2026, aged thirty-three, I abandoned a half-written MA dissertation to take a ball-by-ball scoring post at the Sardar Patel Stadium in Ahmedabad for 1,200 rupees a month. Across that Ranji season I charted 1,412 overs in nine Gujarat home matches, drew a wagon wheel for every innings and logged every leg-before appeal with three things: the bowler's release point, the batter's stride length and the umpire's initials. The file kept growing for eleven more seasons. Nobody ever asked to read it.

The value of that file was never in the file. It was in the baseline. When 1,412 overs have been logged in one hand, a discrepancy in somebody else's over log stops being an opinion and becomes a comparison. I kept counting overs because the thesis could not be counted. The habit survives — every claim has to be turned into a count before it is worth arguing about.

April 2026: how the anomaly log began

In April 2026 Hansie Cronje's confession broke; in December 2026 Mohammad Azharuddin received a life ban. Through that year I stopped writing opinion and started re-watching tape: sixty hours of 2026-2026 one-day internationals, from which I logged 214 no-balls and 41 abnormally slow scoring overs into a private spreadsheet. It was never published. In February 2026 that same spreadsheet put me on a broadcaster's third-umpire review desk for the India-Australia series, where at Eden Gardens I watched replays of V.V.S. Laxman's 281 instead of the cricket.

Two lessons came out of it. First, a pattern is an argument; a single clip is an anecdote. Second, a sequence is useless without its surroundings, so I began writing every decision in seven frames — three before, the decisive frame, three after. The anomaly log began as a place to put what did not fit. It is also necessary to state plainly that 214 no-balls are not evidence of a fix; they are a register of deviation, and a register does not assign cause. That year I stopped asking players about intent. Sources now exist mainly to confirm timestamps.

The four steps of the protocol, and the nine-second cut

In step one the batter signals within fifteen seconds; late and the right is gone. In step two the on-field umpire confirms and initiates. In step three a fixed replay sequence runs: front-on, side-on, split-screen, UltraEdge for bat or pad, and finally ball-tracking. In step four the projection arrives along with the tolerance inside which the technology is required to defer to the on-field call.

Broadcast shows step four as a graphic and step one as a two-second clip. The two middle steps stay off camera. The consequence is that every controversy is fought over the only part that was shown, and that part is the last and most technical fragment of the protocol. In Bengaluru the decision took longer than the replay — the replay was nine seconds, the procedure was four steps. In 2026 I built a forty-one-minute video about that sequence simply because people were shouting about step four and had never heard of the first three. It drew 11,400 views in its first month and I answered roughly three hundred comments myself. The most common question was which side umpire's call favours — the wrong question, because it is asked about intent rather than about procedure.

Tennis publishes millimetres; football keeps a default

One comparison earns its place here, not as decoration. Tennis Hawk-Eye handles the same class of decision as cricket's ball-tracking: a predicted path, and a confidence limit on that path. The difference is disclosure. Tennis gives the player a limited number of challenges and puts the margin on screen — the ball was two millimetres in, and the viewer can read it. The manufacturer's published accuracy claim for Hawk-Eye is within 3.6 millimetres, and that claim is itself part of public argument. In cricket the same technology produces a binary on screen: hitting, umpire's call, or missing. The margin is never shown.

Football's video referee runs the other way. The standard there is clear and obvious error, the referee on the pitch remains the decision-maker, and the technology advises. Cricket inverts the default: inside the band, the technology must defer to the on-field call. Which means cricket's controversies are generated not by technical failure but by a deliberate choice of opacity. The controversy-generating mechanism is designed, not accidental — and a design can only be criticised once its thresholds are published. Until the margin appears on screen, the argument will keep being about bias instead of about procedure.

What never reached the card

The sanction table for slow over rates exists in the code of conduct, but which innings drew how much of a penalty is not published in any checkable form. So the claim that over rates are enforced is a claim nobody can verify. Net run rate arithmetic is the same shape: a tournament position is decided by a formula whose working is never shown in the table.

Older limits deserve the same honesty. Ranji overs that were never televised exist in the record as numbers only. Until the late 2010s, several women's matches never entered the public record with a full scorecard. Plenty of pre-2026 careers were built along routes for which no franchise record exists at all. Every era's record was true within its own boundary. Today's data desk reproduces a narrower version of the same problem: the ball-by-ball feed is deeper than ever, while the governance layer — referee reports, hearings, pitch inspections — stays off camera.

The discipline of the empty cell

This week an audit file reached my desk in which the title field was blank, the source blank, the summary blank, and the column that matters most — the list of discrete facts — entirely empty. One field had content: a domain tag, cricket_asia. Ask under audit rules what can be said about any match from that, and the answer is nothing. Not the format, not the venue, not the players. The correct output is a completed scaffold with insufficient information entered in every cell. The temptation is to fill the gaps with plausible paragraphs — one on Asian cricket's revenue concentration, one line on IPL capital networks. Every such sentence would be unfalsifiable, and therefore decoration rather than journalism. When the column is blank, printing the blank column is the only honest output. The empty file is itself a signal: an extraction failure, not an article with nothing in it. Log it, re-run the process, do not publish the guess.

Where my own bias hides

A record-corrector carries a particular weakness: correction becomes an identity, and an identity starts producing corrections for its own sake. After four decades in this work I am also obliged to write, in print, that the standard account was right wherever the tape showed it was right. That is why the anomaly log stays open in both directions.

Second, the anti-intent standard applies to suspicion as much as to praise. Forty-one slow overs are an anomaly, not a cause. I will not write that those overs meant anything, because the record does not establish meaning. Third, counting is not analysis. A no-ball counted without length, line and field placement is an ornament. The crowd's anger is legitimate as an experience of unfairness; it is not an argument about procedure. The real question is not who benefited. It is how wide the tolerance was, and who holds that number. A referee's eye is just a habit of distrusting first angles — and the habit has to be turned on one's own angles too.

What comes next

Three demands, all documentary. Every published number should carry a format tag — Test, ODI or T20 — and untagged figures should not be publishable at all. The ball-tracking margin should appear on screen in millimetres, not merely the binary. Blank columns and match referee reports should be made public so that over-rate and conduct claims can be checked. The next time umpire's call appears on a screen, the question should not be who gained. It should be how many millimetres the margin was, and whose desk that figure is sitting on. The anomaly log stays open; it has just taken a new entry — the file in which everything was blank.