600 Raids, One Empty Column: Where Kabaddi's Analytics Buried the Receipts

September 2026, a desk in Noida, and a Google Sheet open on my laptop with six hundred do-or-die raids logged in it. Every row carried the raider's name, the tackle type, the outcome, the minute. Fourteen columns were filled. The fifteenth was empty — and that empty column was the most honest thing in the whole sport. The column was labelled 'evidence', and where a claim had nothing behind it except a borrowed phone, I left it blank. The thread that ran that night carried a graph showing conversion climbing from 41% to 58%, a one-third collapse in dubki attempts, and not one line about the blank column. Four years later, looking back, it is obvious that the entire 'analytics revolution' in kabaddi has been built around that blank column. We started measuring, started counting, started bolding our conclusions — and quietly assumed that measuring something is the same as proving it. The phone was borrowed; the haircut was mine; the rumour did the rest.

To understand kabaddi you first have to understand that its structure is unusual in three places, and all three of them shape how the sport gets covered. The first is the thirty-second raid limit and the cant. The raider must keep chanting 'kabaddi-kabaddi' continuously; break the chant and the raid dies. It sounds like folklore until you realise it is the sport's hardest biomechanical brake — the moment a raider holds his breath, his explosive power drops, which makes the cant not just an honesty test but a fouling sensor. The second is the bonus line. Touching the deep mark scores one point, but only when the defence has six or more players on court. The rule itself pushes the raider toward risk and simultaneously compensates the defence for pricing that risk. The third is the all out. Wipe out a side and you get two points plus a full revival of your own dismissed players. That is why a kabaddi score never dies — it is the sport's prime-time trump card. And there is a fourth, which has generated most of the noise over the last five years: the do-or-die raid. After two consecutive empty raids, the third must score, or the raider himself is dismissed.

Everything in PKL sits inside that frame — two twenty-minute halves, seven a side, one point per touch. Standard style works exactly this way; circle style, played in Punjab's villages, runs on different court geometry and different scoring arithmetic. That difference looks small until an English-language report blends a circle-style result into a PKL data bucket. Then you realise the umbrella labelled 'kabaddi data' currently has two different sports standing underneath it, and nobody is willing to say which one they were talking about.

Now to the thing I actually counted. In 2026 — nineteen years old, second year of college, every live sport cancelled, stadiums looping empty — I rewatched PKL seasons 5, 6 and 7, roughly 180 matches, and hand-charted six hundred do-or-die raids. The result: over three seasons conversion climbed from 41% to 58%, while dubki attempts fell by a third. Lockdown made every raid do-or-die; the spreadsheet made every death a cell. That number was telling two stories at once, and we celebrated while reading only one of them. The first story: raiders began making better decisions under pressure. The second: the risk they abandoned was the dubki — kabaddi's most delicate, most stylish, least data-friendly weapon. It does not look good on a graph; it does not fit a model. So the system dropped it and called that efficiency.

This is where the real turn in the story comes, and it starts in Jakarta in 2026. That year India lost its first Asian Games gold in the sport — beaten 27-18 by Iran in the semi-final, having already lost a group game 24-23 to South Korea. At one in the morning I posted a twelve-tweet thread arguing that the defeat was a selection failure, not a capability failure: India had picked raiders on old form and old names and had no answer for Iran's wrestling-derived thigh-holds. The thread drew four hundred quote-tweets, of which 'traitor' was the most polite, and exactly two follows — both from working kabaddi reporters. In Jakarta I learned that a thread can find you before you find yourself, and I also learned to ask why the defence could not change when the opposition could. Iran did not build a kabaddi defence; it transplanted wrestling's thigh-lock and bridge onto a kabaddi court. Fazel Atrachali is the product of that school, and he is now the PKL's all-time leading tackle-points defender. That number belongs in the first line of every season preview, because it reminds you that rules do not adapt for anyone. Defences do.

Which brings us to the most uncomfortable question. In six years we built four metrics for raiders — raid success percentage, do-or-die conversion, super raid count, average points per match — and two for defenders: tackle success percentage and super tackles. There is no metric at all for an Iran-style thigh-lock, or for the corner-cover coordination India runs. As long as a tackle is reduced to success or failure, the defender who shepherds a raider into the wrong direction and creates the tackle window will never score a point in the ledger. Cricket built assists and dot-ball percentage to escape exactly this trap. Kabaddi is still looking at defence in binary. My firmest claim is this: kabaddi's data measures offence and merely counts defence — and the gap between measuring and counting is enormous.

The most expensive consequence of that gap shows up on the auction table. In 2026, before season 9, the market anchor was Pardeep Narwal's record buy by UP Yoddhas, a benchmark that reset how franchises valued raiders. Before the auction I built a rupees-per-raid-point model on my six-hundred-raid sheet, ranked the pool, and published seventeen timestamped calls — six bargain raiders, three overpriced ones, one predicted final-round steal. Eleven landed. I live-tweeted all seventeen, including the misses. That was the day my followers stopped being people who agree with me and became people who check me.

But the real auction story is not the graphic on television. It is that railway clerks and police constables become crore-rupee lots — the best labour story in recent Indian sport economics. In the old semi-professional system a government job was the actual prize; PKL converted that prize into cash while leaving the entire risk on the player. Six hundred raids taught me more about football transfers than any fax machine: auction logic rewards last season's visible product, not next season's usable talent — and that is the engine of overpricing. A raider who has twice carried fifty raids in two seasons gets priced on a graph of old form, not on the metric that will survive the next season's defence.

And now the part every kabaddi growth story quietly skips. Women's kabaddi owns four Asian Games golds and not one franchise league. Men's PKL runs on broadcast rights, title sponsorship, ticketing and licensing; the women's game has the same sport, the same rules, the same tackles, and zero market structure. When someone says kabaddi is spreading across the world, ask which kabaddi. Overseas viewership and rights revenue are overwhelmingly diaspora-driven, and both Olympic bids — Tokyo and Paris — failed on participation spread and the women's-game criterion. Read those three facts together and you get the picture we keep selling as a global sport: a game whose women's team has won Asia four times and which has never dated its own women's league.

The governance layer makes it sharper. The IKF is the international umbrella, the AKFI runs the domestic game — with a long and turbulent relationship with controversy — and Mashal Sports operates PKL. One sport, three administrations, three agendas, three sets of interests. In that structure, regulation moves as slowly as PKL's commercial engine has moved fast. The biggest structural risk inside the sport is not the knee, ankle or shoulder — those are at least now visible in data. The real risk is that the regulator and the market are two captains on one ship.

Here is where I have to admit I might be wrong, and that admission matters. First: maybe the blank column is the sport's strength. Kabaddi came out of mud, and the moment we turn every dubki into a cell we lose the mud. The four-thousand-year claim may be romantic, but the drama of a circle-style village final cannot be imprisoned in a conversion rate — and I suspect that if a fourteen-year-old in Patna is learning a dubki today, he is not learning it from a spreadsheet. Second: maybe by calling Iran a co-favourite every year, we are falling into the exact trap I accused others of in 2026 — treating old results as proof of future ones. Iran's school works, but if India converts its passing and corner-cover coordination into measurable output next cycle, the slogan 'defence wins' will age as badly as 2026's 'the raider is everything'. Third, and least comfortable: maybe analytics was never needed. Maybe the sport's appeal is precisely that it is unpredictable, and every time we fit it to a model we peel off a layer. I mistook being noticed for being right, and the thread kept receipts — perhaps I owe the same accounting on my own sheet that I demand from others.

None of that means we should stop measuring. It means we should decide which question a measurement answers before we run it. My call for the next three years is simple: every auction round will shift from being judged on crores spent to rupees per raid point earned, and the franchise that does it first wins the next cycle. A second prediction, on the record: either a women's PKL is announced before 2028, or the issue disappears into silence — there is no middle path. And yes: every season preview should name Iran in its first line, not as a spoiler, but as a co-favourite. The day we accept that, kabaddi's data becomes honest for the first time.

Until then, that fifteenth column stays empty. And I think it should stay empty — because in a sport where a do-or-die raid itself survives on the player's own dismissal, there ought to be at least one column that says: this much is proven, the rest is story.