Do-or-Die on the Auction Table: The Real Ledger of Kabaddi's Transfer Market

It was two in the morning and the laptop had one column open — the outcome of the do-or-die raid. August 2026, Jakarta Asian Games, men's kabaddi semi-final: India 18-27 Iran. A few days earlier South Korea had beaten India 23-24 in the group stage. Between 2026 and 2026 India had won seven straight Asian Games golds. In Jakarta the sequence broke, and what was on television could be dismissed as a bad day or coded as a pattern.

I re-watched every Indian raid off the broadcast and hand-coded 61 of them: cant length, raid type, and above all the do-or-die outcome. Nineteen of 61 converted. India came out of their do-or-die raids at under a third, while Iran's tackle points clustered on ankle holds. The thread went up at three in the morning; by eight it was in an editor's inbox.

The ledger said 2 a.m. The city said otherwise.

Six years later kabaddi's transfer market runs without that column. On the auction table a raid point has a price, a tackle has a price, a broadcast graphic has a price. Do-or-die conversion has none. This is a report about the empty column.

In July 2026 the first Pro Kabaddi League auction was held in Mumbai. Before it, the elite kabaddi player was an employee — Indian Railways (RSPB), state police, the Army. Income was a clerk's salary, departmental leave, a tournament allowance and job security. The auction repriced those bodies, but only the ones inside the camera frame. The player who spent twelve years at district level and never got on television still earns the old salary.

Kabaddi's transfer mechanism does not work like football's. There are no two windows. One auction day, a retention list published before it, and a Final Bid Match card in between. Contracts are annual. There is no club-to-club transfer fee and no loan system. So the most honest transfer-window document kabaddi produces is the release list — who was let go, published before anyone bids. Letting go is the biggest decision in this market.

Consider the FBM card. A team can match the top bid for a player it released. That means price discovery is blocked precisely for the players teams want most. The thin part of the market sits where demand is highest. What is left is the real auction; what goes into FBM is a bookkeeping story.

The salary cap is the second invisible hand. A squad has a fixed pool. One expensive raider means two or three mid-price defenders who are not bought. That is why the release list must be read as squad shape, not as names.

National selection sits adjacent to this market, run on reputation, designation and the memory of the last tournament rather than published data. Jakarta 2026 was the sample.

One correction first. A 2026 RTI reply confirmed India has no officially declared national sport. The line appears in almost every kabaddi report, always unsourced. Remove it and the next question writes itself: what exactly are we calling a national priority — the gold, or the ledger that has to exist before the gold?

The do-or-die raid is the market's cheapest and most expensive action. A failed do-or-die gives the opposition a point and removes the raider. The visible cost is one point; the real cost is numerical imbalance across the next two minutes. Yet scouting treats the stat as noise, because a raider gets two or three do-or-die raids a match. Over a season that is 40 to 60; over three seasons, above 150. The problem is not the sample. The problem is the column: do-or-die conversion never appears in the broadcast graphic.

Before the 2026 auction I built a valuation sheet weighting do-or-die conversion, super-tackle rate and raid frequency against expected price. It flagged a raider near base price converting 44 per cent of his do-or-die attempts. The room bought a headline name.

The arithmetic is simple and therefore awkward. A headline raider scores 7.5 raid points a match at 30 per cent do-or-die conversion; a base-price raider scores 6.8 at 44 per cent. The season average favours the first. Close matches are not decided by average raid points but by the do-or-die column, because tight games turn on two or three points and do-or-die raids cluster in the last five minutes. The market buys the average and sells the moment.

The stat that prices a defender is not public. The league's stat line shows tackle points. It does not show how many opposition raids that defender faced. Two tackle points off nine raids is better than 1.8 off six; the scoresheet files both in one column. Super-tackle rate slips away the same way — a three-man tackle is defence's do-or-die, because failure hands the opposition a revival.

Iran's Jakarta defence reads clearly here: ankle holds, low-body control, wrestling-derived balance. It is a technical premium the auction never prices, because the league's highlight economy rewards the dash and the thigh hold. Iranian defenders found a durable market because their work is legible on television. The Indian defensive work that creates space for a teammate — closing the escape route, arriving from behind — never is.

The tackle that reads on camera sells; the tackle that makes space for a teammate ends up on the release list.

The third piece is the wage bill. Read a release list by squad shape. Dropping one big raider buys two mid-price defenders and a backup raider. The question is not who is the better player but who carries the risk inside a one-year contract. Annual contracts mean the franchise carries injury risk for exactly one year, then the same body returns to market, a year older. That is why teams prefer a 24-year-old at base price to a 30-year-old at a premium — option value on the next contract, not performance.

And the old job returns. Railways or police employment gave a floor: minimum income, leave for injury, a pension. The auction supplied a ceiling and removed the floor. One good season, a crore bid, a release, the depot.

The clerks had day jobs. India had a team sheet.

The auction makes a player a crorepati; the release list makes him a clerk again.

Jakarta was not a selection failure so much as a measured gap: Iran's wrestling-derived defence against reputation-led Indian selection, converging on 18-27. Why does a raider trusted with do-or-die raids all season suddenly get that trust in the last five minutes of a semi-final?

Here is the honest turn, and it cuts against my own work. I went looking for a scandal and came back with a spreadsheet. My 2026 model was not vindicated, and the room was not stupid. I had predicted a variable whose per-player, per-season sample was too weak to defend: a 44 per cent rate on 40 raids carries a confidence interval so wide that no bidder should gamble on it. The room priced what it could verify. My model had outrun its evidence.

Second, the public data is context-poor. Raid quality is not recorded — how many defenders the raider faced, from what angle, at what fatigue. Defender positioning is not recorded. With inputs that thin, the market's apparent irrationality is rational: it prices what it can check and discounts the rest.

Third, the stat is endogenous. A team with a weak raiding unit pushes its raider into more do-or-die raids, so a high conversion rate sometimes signals fewer alternatives. That was the hole in my 2026 log: I measured the player and not the team's constraint.

Fourth, the biggest transfer-window move is the one not made — retaining a cheap defender, using the FBM card properly, keeping a player with zero camera time. Those decisions show up in the table and never in the paper. I found the model in a folder nobody opened, and my own error was inside it.

What comes next needs no revolution, only three published numbers. One: do-or-die conversion with sample size attached — 44 per cent on 40 raids is not 44 per cent on 120. Two: tackle points per opposition raid faced, so defensive workload is visible. Three: a loan mechanism, so a 21-year-old defender can play eight matches elsewhere instead of burning a contract year on the bench while his next price falls.

None of it costs money. It costs one column, printed honestly.

When a raider's price falls at the next auction, the question will not be what his do-or-die conversion was. It will be whether anyone in the room had looked at it — and whether, having looked, anyone was listening.