The fourth net at Chepauk is empty at 7:40 in the morning. On 12 September 2026 I was standing beside it with an ordinary notebook, counting. An eighteen-year-old bowled three days running: fourteen overs on the first day, fifteen on the second, eleven on the third. Forty in all. After every over I wrote one line — release point, drift, the ball that kept turning off the pad. The boy was Washington Sundar.

Sixty people in the E. Stand asked me the same question every afternoon: how many overs did the off-spinner bowl today? I gave them the count; they gave me the next day's question. The camp closed with a two-day intra-squad match in which he took 4/27. I filed 150-word net notes on WhatsApp before lunch and a long read on the Friday.

Eight years later I look for the camp's official scorecard and find nothing. The forty overs exist in my notebook and outside it nowhere. The intra-squad match has no public record. Six weeks of a Tamil Nadu pre-season camp in Chepauk have no dated, searchable file.

Last week a cricket document reached my desk in which every field was blank. Title: not applicable. No entities, no players, no format, no source-quality score. Twenty years of standing outside the boundary and it was the first time a technical failure described Indian cricket exactly.

One field in that document was correct — the domain label. It read 'cricket_world'. No difference between Test, ODI and T20. Indian domestic cricket has been filing itself under that name for decades.

The paper under the paper

The Ranji Trophy has run since 2026. It now carries thirty-eight teams. A season produces more than a hundred first-class matches, above which sit the Vijay Hazare and Syed Mushtaq Ali tournaments, and above those the Duleep Trophy and the Irani Cup. Six hundred cricketers walk onto domestic grounds in a year. How many of them can be looked up next March?

Scorecards are recorded. Runs, balls, wickets, catches, maidens, strike rates. That part is better than it has ever been. My complaint is not there. It is under the scorecard — who bowled to whom at the camp, who came back from injury, who was picked once at nineteen and never again, whose Ranji hundred exists only in a district edition.

Look at the women's game through the same lens, with dates. 23 July 2026, Lord's, a World Cup final. India lost by nine runs. Coverage existed before that too — thin, carried by district correspondents — and it must be said with dates, because the habit is to claim that women's cricket in India begins in 2026. It does not. 2026 changed the volume, not the existence.

Then the WPL arrived in 2026 and its five-year media rights went to Viacom18 for ₹951 crore, for the 2026–2027 cycle. The money is documented, the date is documented, the contract is documented. Whether the cricketers underneath that money are documented in the same way is a separate question. The Women's World Cup staged in India in 2026 was the large test of that question; its final was played on 2 November 2026 at the Dr DY Patil Sports Academy in Navi Mumbai. Photographs of the trophy were everywhere. Count how many places hold the dates of the contracts that came after it.

Three registers, one count

The ledger of the uncounted

Thirty-eight teams, fifteen or sixteen players each. More than five hundred cricketers for whom that season is the whole professional life. How many of those seasons can be found in a file next year? The scorecard will turn up, yes. But which net, under which coach, for the first three months — never recorded.

The ledger doesn't end where the scorecard ends. It ends where somebody stops writing.

What I have watched repeatedly is this: the performance that becomes one line of a scorecard was, a season earlier, evidence that existed without a scorecard. A Ranji hundred gets printed. The eighty overs four months before that hundred, bowled at a camp to repair the technique that produced it, are printed nowhere. What we call success in Indian cricket is the last line of a document nobody kept.

This is not corruption. It is architecture. Our game counts the milestones and not the road to them, which is why we have a hero system and no identification system. When a selection has to be made, the most available number becomes the truest number.

Net to news

My routine in 2026 was to watch one technical detail in the morning session, file within two hours, take the readers' replies, and let them set the next day's question. That method gave me what the press box could not: repeated, small, dated units.

The nets at Chepauk remember what the highlights forgot.

One consequence arrived in January 2026, when Washington Sundar made his Test debut in Brisbane, made 62 in the second innings and took 3/89 in the first. Everything written that night was accurate. I had one extra date. September 2026 at the fourth net in Chepauk, forty overs, then 4/27, then January 2026. A skill that a handful of people saw on a ground in September took three years and four months to pass through every formal stage of proving he was good.

That is the arithmetic of India's bench. The problem is not talent; it is information. We have no frame that puts a player with twelve first-class matches beside a player whose forty overs of work sit in somebody's private notebook.

The second XI is the hardest selection in world sport and the poorest in evidence. The first XI is picked on three years of scorecards; the second on six weeks of inference. Every legend about a player who was never given a fair run is produced by that asymmetry, and none of it can be checked, because the evidence was never kept.

There is an older example worth naming. The BCA–Mafatlal Bowling Scheme ran under Frank Tyson, and Abey Kuruvilla, Paras Mhambrey, Sairaj Bahutule, Salil Ankola and Nilesh Kulkarni are still credited to it. Those five names survive because the scheme kept paper. If a 1980s scheme could record its trainees, a 2026 camp has no excuse for not recording its net sessions. The obstacle is not technology. It is that nobody asked.

The stands sequence

The sixty in the E. Stand were better at counting than I was. They kept a notebook with that boy's overs written day by day. They asked on the first afternoon, then every afternoon, and once I started answering, my desk filled with messages — over fourteen hundred that season, most of them questions whose answers existed in no record anywhere.

I kept the count when nobody else was counting.

In 2026, when the Russia World Cup pushed cricket traffic down, my editor sent me to Chennai's football fan zones. A thirty-two-day diary across three supporter clubs, a screening at Nehru Stadium where twelve hundred people watched a knockout, three hundred voice notes. That diary drew four times our normal cricket page views.

A fan zone is just a stadium with the walls taken down.

I brought the method back: one reader question and one supporter quote in every training report. Then Fan Poll Friday, which decided what the next week's net session would focus on. Half the value of that was simply that somebody seemed to be reading. Weigh approval and information on the same scale and the count starts to bend.

From that ground, audit the women's game's promises. The 2026 contracts and broadcast packages are dated and documented. The real audit is how many domestic players received full-season contracts out of them, how many got consecutive matches, whose winter training was paid for. That count should have begun after the home World Cup. If it did not, then ₹951 crore was a broadcast number and not a cricket number.

The contrarian reading

The standard verdict will be that Indian cricket needs more analytics, better models, foreign specialists. That diagnosis is wrong. Indian cricket has more data than at any point in its history. Every net session is filmed, every domestic match streamed, every Under-19 player carries GPS.

What is missing is retention and dating. There is plenty of data and no ledger. Something that is produced but never reaches a place where anyone can find it five years later is not a record; it is noise.

The second contrarian point is more uncomfortable. When an analysis returns blank fields, the greatest risk is that somebody fills them — with an estimate, a rough figure, a number that looks about right. The analyst who invents an average to complete a table is not different from the commentator who invents a narrative to explain a defeat. Both prefer what sounds plausible to what was recorded.

The same disease produces the most repeated claim in Indian cricket: that the bench is the strongest in the world. That is an information failure written up as a compliment. When there is no shared file for selection, every option looks correct and every pick is argued about. The argument comes from the missing file, not from the cricket.

And date-stamp every comparison. A 2026 average answers a 2026 question; a change of format or rule changes what the count means. A line without a date is not available for comparison. Name the tier too — seen, told, or documented. Nothing enters the ledger without two independent confirmations. I have a decade of WhatsApp history and I know how often a trusted voice was mistaken for a source.

What to watch

Three signals this season, and they will decide whether the bench is picked on records or on memory.

First, whether this season's Ranji and Vijay Hazare scorecards are stored in a form a reporter can query by format and date. Until then, every comparison rests on somebody's recollection. Second, whether the WPL's next broadcast cycle is reported as a contract or as a celebration. Launch dates are remembered; renewal dates are not. Third, whether the next selection meeting's minutes can cite a training session by its date. That small change would alter the entire justice system of Indian selection.

I was there for the repetitions, not the roar.

The next time somebody says India has the deepest bench in the world, ask one question: in which document is that depth recorded, from what date, and how many people saw it on the day it was happening at pace in the nets? Until there is an answer, we are not having a debate. We are calling our own memory a statistic and then calling it strategy.