The Third Raid's Arithmetic: Does the Kabaddi Raider Choose the Risk, or Does the Rotation?

There is a file on my laptop called do_or_die_v4.xlsx. It holds 1,842 raids: five Pro Kabaddi League seasons between 2026 and 2026, and only those raids that fell into the do-or-die category — the ones where, if a point did not come, the raider himself went off.

Of those 1,842, the raider's first movement went toward the bonus line in 402. The success rate there was 47 percent. In raids where the raider charged straight at a corner, that rate was 31 percent. Sixteen percentage points separate the two — and yet, on the mat, the raider bolts for the corner on nearly every third raid.

This piece is an examination of that gap. And the examination concludes that the fault does not belong to the boy who walks out for the third raid.

A kabaddi court is 13 by 10 metres, seven a side, two twenty-minute halves, thirty seconds to a raid. The raider must chant "kabaddi-kabaddi" continuously; pausing to breathe voids the raid. That is the basic architecture, and every spectator knows it.

What is less understood is how three rules lock into one another.

First, the bonus line. Touching the line drawn deep in the opponent's half earns one point — but only when the defence has six or more players on the court.

Second, the super tackle. If the defence is playing with three or fewer players, every completed tackle is worth two points instead of one.

Third, the do-or-die raid. After two consecutive empty raids, the third becomes mandatory; fail to score and the raider is dismissed.

Read separately, these are three rules. Read together, they are a market in which the defence decides which wager is even available to the raider. At six men the bonus is alive and the super tackle is dead. At five the equation flips — the bonus is gone, the super tackle is live. The defender count changes, and the raider's entire risk menu changes with it.

That is why kabaddi scores swing so violently, and why the sport sits so well inside a short prime-time window. An all out gives two points and revives the entire dismissed side — a team-wipe reset that makes defeat technically never final.

The most interesting pattern in my 1,842-raid log is not the do-or-die raid. It is the two empty raids before it.

Two consecutive empty raids are frequently not a raider's failure; they are a concession the defence chooses to make. The defence hands over the bonus, or lets the raider walk back untouched, because it knows the pressure will swing onto the third raid and the burden of risk will invert. This is a transfer of risk — every empty raid loads debt onto the next one, and the final instalment falls due on the third.

The second pattern is more uncomfortable. Of those 1,842, the raiders who took the do-or-die were, 58 percent of the time, the men who had raided least in that half. The third raid goes to whoever the rotation delivers there, not to whoever is best suited to that moment. The team's best raider has usually already been spent on the first and second.

I should also record what my coding cannot see. It does not measure positioning — which corner was leaning which way. It does not measure fatigue — how much breath the raider had left on the third raid. And it does not measure the coach's instruction at all, which is often what a timeout settles. The 47-versus-31 gap is a signal, not proof.

But a signal carries a price, and the price is paid in slots, not points. A failed do-or-die does not merely remove a raider; it brings the defence to six on the very next raid, opens the bonus window, and hands the opposition a free point. One raid's error becomes the next raid's discount.

Now the part where the comfortable story breaks.

After every defeat one raider is named. We are told his head went under pressure. My entire argument runs the other way: a do-or-die is not a mental examination, it is the output of a selection process. Whoever the rotation sends there goes there. A coach who will not change the order at a timeout cannot blame the raider.

Apply the same logic at national level. At the 2026 Jakarta Asian Games, India lost the group game to South Korea 23-24, then the semifinal to Iran 18-27 — the first Asian Games gold ever missed. Around that camp I filed a fourteen-page note on Iran's wrestling-derived ankle holds and South Korea's dash angles. Nobody read it. Selection went on reputation. And that night a 21-year-old left corner was blamed online, while Fazel Atrachali was already rewriting tackle-point records.

Calling Iran a spoiler is inherited Indian-preview language, not fact. In the men's game Iran is a co-favourite, and in the women's game Chinese Taipei and South Korea are genuine contenders — India's four Asian Games golds do not erase that.

The real blind spot is that we inspect the third raid and skip the first two. The empty raid is where the match is decided; the do-or-die is only where it becomes visible.

Technique changes, and the change can be costed. The fading dubki, the rising dependence on the bonus line — that is adaptation with a measurable bill, not the loss of a soul. Pardeep Narwal's 369 raid points for Patna Pirates in Season 5 looked astonishing; the raid-zone mapping put the bulk of them through the right-corner channel, not through the celebrated dubki. The mat's claim was never that he could not be stopped; it was that he could not be stopped from there. Highlights and space are two different things, and the auction still bids on highlights.

Do not watch the third raid. Watch the second empty raid — who let it happen, what the defender count was, and whether the coach used a timeout to change the rotation or let the order stand.

I am writing this before the result, not after: teams that take a timeout and change the raider before a do-or-die will finish the next season above average on conversion. If they do, that file of 1,842 rows knew more than any full stadium. If they do not, the file was still asking the right question — and next time the question gets asked first.