The Empty Sheet and the 412 Raids: Where Kabaddi's Data Chain Breaks

May 2026, TransStadia Arena, Ahmedabad. Wearing a volunteer data-spotter's bib at the edge of the mat, I counted 412 raids by hand, one column at a time, while the coaching staff kept asking for tackle counts. My interest was the bonus-line economy. The sheet produced one pattern: when the right corner drifted early, raiders converted bonus points at 38 per cent; when the corner drifted late, 19 per cent. I emailed the sheet to the team analyst. The reply was one word: "interesting." That season Gujarat Fortunegiants lost the final to Patna Pirates.

Eight years later, another sheet is in front of me. Nine columns, a template in each, and one answer repeated in every cell: insufficient information. Zero information points. Zero named entities. Zero dates.

I counted 412 raids because I did not trust my own eyes. Here I met the inverse problem: a document with nothing in it to trust, which nonetheless looks complete, polished and professional. That is what happens to kabaddi coverage, only in reverse. The sheet looks full; the questions stay empty.

Context: the sport with the most data has the thinnest public record

Pro Kabaddi League began in 2026. Eleven seasons, twelve franchises, roughly forty raids and as many tackles per match, every action timestamped. On paper, India's second-densest body of granular action data after cricket. In practice, that data does not leave the room.

When I went looking for why my sheet came back empty, I found four break points in the information chain, and none of them are located in reporting. They are structural.

The door of the source. Big kabaddi stories sit behind paid walls, on JavaScript-rendered pages, often only in Devanagari or a regional script. Any Latin-first extraction fails here. That is the pipeline's fault, but the reader absorbs the result: "you can't find reliable data on kabaddi." You can. It is in the language the sport is played in.

Terminology. This is the costliest failure and the least noticed. English coverage uses "raider", "attacker" and "assaulter" interchangeably for one person. Hindi wire copy renders "do-or-die raid" as "nirnayak raid", which can apply to any raid; in fact it is only the third raid after two consecutive empty ones, and failure puts the raider out. "All out" becomes "sarvashoonya" in one report and "poori team out" in the next. The gap between a super tackle and an ordinary tackle — two points versus one, and only when three or fewer defenders are on the mat — simply vanishes in translation. "Cant", the continuous kabaddi-kabaddi chant, turns into "mantra".

This drift is not a language issue. When raid and attack collapse into one word, every successful-raid percentage collapses with it, because a successful raid requires the raider to return to his own half untouched. Touch and return is a raid. Touch and held is a tackle. One action, two opposite outcomes, one word in translation.

The rule variant. Kabaddi has a circle style and a standard style. Circle style has a round field, different boundaries, different scoring, and belongs to the rural wrestling tradition the belt press sells as the "national game". Standard style is what the Asian Games and PKL use. A report about a circle-style exhibition analysed as standard style produces numbers with no meaning. My sheet did not state which variant was in play. That is why it came back empty.

Time. Kabaddi news decays fast. Auction prices, injury status and standings all shift within weeks. An undated kabaddi report is a scorecard with the year torn off.

Core: rules, economics, and the arithmetic that was visible in 2026

The bonus line is a market, not a rule

A bonus line is drawn deep in the opponent's half. Touching it earns one point, but only when the defending side has six or more players on the mat. That condition rewires the whole game. The bonus line is an economy: with six defenders on the mat, you have opened a free-point door for the raider. Strong defensive units therefore deliberately hold the seventh defender out — trading tackle strength for bonus suppression. On television this reads as caution. On the mat it is a chosen weakness.

That was the 38-versus-19 pattern in my 412-raid sheet. When the right corner drifted early, the raider got the space to reach the bonus line. When it drifted late, the raider had to pause — and pausing means cant risk, the thirty-second clock, and two empty raids pushing the next one toward do-or-die.

The mechanism coverage almost never writes: a corner's drift is a signal of intent, not speed. A fast drift means the defence read the raider before the raid's length. A slow drift means the defence is waiting — for a bonus, perhaps, or for a super tackle. Both can be correct. You cannot read them the same way, and you cannot read them at all without a spotter.

Do-or-die: arithmetic, not psychology

Commentary sells the third raid as pressure. It is an insurance calculation. A team with two reliable raiders can absorb two empty raids because a fresh raider is available for the third. A team with one — the most common roster defect across eleven PKL seasons — cannot, because the third raid gambles its most valuable asset.

The real test of a do-or-die raid is not the raider; it is the bench. Teams with a third option play it. Teams without one avoid it, take the early bonus instead, and trade risk for return — which shows up as a one-point-per-raid gap that blooms into a "sudden" all out after the fourteenth minute.

The all-out maths is the clearest and least understood in the sport. Wiping out a side earns two points and revives all seven. So an all out is not two points; it is the return of seven players, which reopens the bonus line, which hands the raider a free point, which resets the shape of the next three raids. Commentary calls it a comeback. It is a reset, and the reset belongs to the team that forced it.

Volume versus efficiency: the distortion in every kabaddi report

Two hundred raid points do not say a raider is good. They say he raided a lot. The question is how many raids he attempted and how many succeeded. Raider A: 210 points in 22 matches. Raider B: 165 in 15. The first number looks larger. If A's success rate is 38 per cent and B's is 46 per cent, B is worth more per raid and is paid less at auction. That is where the money should have gone.

The same holds, more sharply, for defence. Tackle points are a team statistic; tackle success rate is a defender's. High 5s — five or more tackle points in a match — remain the most useful single measure, because they testify to an evening's work rather than a season's accumulation.

In 2026 I filed an Asian Games preview naming Iran co-favourites rather than spoilers. My only basis was a countable fact: across PKL Seasons 5 and 6 combined, Indian defenders had logged fewer High 5s than Fazel Atrachali alone. The editor ran it under the header "Hot Take". In Jakarta, India lost 23-24 to South Korea in the group stage and 18-27 to Iran in the semifinal. The piece resurfaced within hours and the comment section flipped.

In 2026 I learned that being early feels exactly like being wrong. Since then every preview carries a consensus check: name the co-favourite, then name the scoreline that would prove the take wrong. In 2026 my falsification condition was India's knockout tackle success rate above Iran's and an Indian High 5s tally above Fazel's. Neither happened.

The uncomfortable part is that none of that analysis exists in any public archive today — Indian or Iranian. What survives is 18-27. A scoreline reports an outcome, not a mechanism, and only the mechanism is reusable.

Auction premium: the sum nobody checks

PKL auction prices are public. Performance is public on scorecards. The two are never assembled in one place, in one format. So the question is asked every season and deferred every season: was the big bid correct?

My method is plain. Take the gap between base price and sale price, divide it by the player's realised output over the previous two seasons — successful raid rate and raid points per raid for a raider, tackle success rate and tackle points per match for a defender. If that ratio does not improve across two seasons, the premium was paid on reputation, not projection.

This produces its clearest verdict on volume kings. Naveen Kumar and Pawan Sehrawat are large on scoreboards and large at auction, and both can be true. But the scoreboard never says how many raids were lost in the raid, or how many were won for the team. Until that ratio is public, auction prices measure publicity, not performance. The perennial "next Pardeep Narwal" wave grows in exactly that gap.

The absent women's league is a pipeline leak, not a social complaint

Women's kabaddi exists in India — Asian Games medals, state circuits, full trials. A small three-team exhibition experiment ran in 2026 and was not converted into a league. Eleven men's seasons since; zero women's.

This is arithmetic, not sentiment. Reaching a platform like PKL requires four rungs: a domestic circuit, professional coaching, television, and wages. Men have all four. Women have the first, half the second, and neither of the last two. The result is a national team assembled every four years and then playing where no camera goes.

It damages the men's game too. A sport with half its talent outside the professional structure will always have half the supply — and that supply decides the next India-Iran final.

Governance: the risk that never appears on a scoreboard

AKFI's administrative turbulence is the oldest, least-written story in Indian kabaddi — suspensions, reconstitutions, election disputes. The player-level damage runs through national camps, selection meetings and international eligibility. I should separate what I saw from what I inferred. I read agency reports of suspensions and reinstatements. I did not see, anywhere in public, which player's selection stalled in those windows. I infer damage occurred; I have no sheet on it.

Two structural risks go untouched: age fraud and anti-doping, recurring integrity problems in South Asian kabaddi, where birth certification is weak at the grassroots; and the government-job and award incentive system — railways, police, army, the Arjuna Award — a real safety net whose distribution is opaque. Both act on the one variable nobody measures: a player's career curve.

Broadcast concentration: the largest commercial vulnerability

For most of the league's life, one group has carried the bulk of the broadcast. That was a gift early — prime time, production quality, a cricket audience. It is a vulnerability now, because a league with one buyer has a rate card at that buyer's discretion. The leak shows up first where we do not look: the archive. Whoever airs the match also owns the raid-level log. When the contract changes, who owns the old seasons? Nobody has answered.

Three months, an empty stadium, and a crowd that turned out to be a player

In 2026 the pandemic pushed Season 8 off the calendar and killed the arena work with it. I took a nine-thousand-rupee gig hosting a spectator-less state athletics meet at a Gujarat Sports Authority track and turned it into a lab: crowd beds recorded at three volume tiers, a second-screen stat feed, and a deliberately quiet announcing style for field events.

Hosting taught me that the pause is the loudest part of the call. Watching Season 8 behind closed doors in Bengaluru confirmed the corollary: home advantage was never a statistic. It was a production decision. Who chose the noise, who paid for the silence — none of it reaches a scorecard. Three months in an empty stadium, and the crowd turned out to be a player.

I started four newsletters that year and killed three. Serial-starting is my own disease, and kabaddi coverage has it too: an enthusiastic first week every season, no follow-up.

Contrarian: the empty sheet is not the scandal. The full sheet is.

The easy move here is to say kabaddi's data is so weak that the pipeline returned nothing — comfortable, consensus-friendly, and consistent with my falsification-first wiring.

I think the truth is inverted: kabaddi holds far deeper data than its public record, and that ownership is the problem. PKL generates more tagging than most academic analyses assume — timestamps, defender positions, bonus attempts, do-or-die tags. The data exists, on someone's server, in someone's spreadsheet.

My own sheet is the evidence. The reply was one word: "interesting." That word is Indian kabaddi's institutional memory. The team never asked for the sheet because it wanted tackle counts. I never published it because I did not know where such things go. I own that half. Four hundred and twelve raids of counting were lost from both ends.

This is not an argument that kabaddi coverage is uniquely bad. Cricket's backstage is just as opaque. The difference is forty years of public scorecards to fill the dark with. Kabaddi has eleven seasons of spreadsheets with no named owner.

So here is the condition that would prove me wrong. If a public, machine-readable kabaddi portal appeared with season-wise raid-level logs, per-player successful raid percentage, tackle success percentage and High 5s tallies, my entire argument collapses. I have looked. I have not found it. I also accept my search was limited, and that this is a conclusion to re-run monthly, because institutions never accept a finding that permanently comforts them.

A stronger alternative deserves stating: perhaps this is not kabaddi's fault but the fault of every sport built on the broadcasting model. Television treats data as property, not content. I find that reading persuasive precisely because it sharpens my argument rather than softening it — but it stays unproven until PKL is compared against a league of the same age, and nobody has run that comparison.

Takeaway: the question is not who wins; it is who keeps the file

I will be asked what happens to India at the next Asian Games. That is the second question. The first is whether India will own its own raid log before those Games, or repeat 2026 — selection on reputation, remorse on the scoreline.

The day kabaddi gets a public raid log, coverage of this sport changes, and I will not mind my argument going stale. That is the best kind of defeat in this trade.

Until then, after every raid there is a sound from the mat, slightly ahead of the scoreboard — a raid failing, close enough to hear before the scoreboard knew. Anyone who has heard it knows the rest is just data.

I will keep counting. This piece is only a record of that.