The Truth Behind 369 Raid Points: What the Kabaddi Scoresheet Won't Tell You

Pardeep Narwal's 369 raid points. Season 5, 2026. That figure is still the first sentence of every kabaddi preview. I reopened that season's scoresheet last week — the final of 28 October 2026, Patna's full home leg, and the matches in between that nobody rewatched. The problem isn't 369. The problem is what 369 was built out of.

I was in the Patna hall during that run, notebook open, and after every raid I wrote down one thing — which game-state the raid happened in. A point in the twelfth minute, when the seventh defender is still walking back into position, and a point in the first minute look identical on a scoresheet. On tape they are two different sports.

My hot take is flat: a raid point in kabaddi is not a performance metric, it is an accounting entry. Until you read the "when" column, that entry will mislead you.

Every preview of the current season is built on the same frame — total raid points, Super 10 counts, crores paid at auction. The frame is convenient because it is available. The Pro Kabaddi League scoresheet gives a raid-by-raid breakdown, fantasy platforms refresh the numbers every morning, and the broadcaster needs one number to talk about at half-time. 369 is that number.

Kabaddi's rule architecture breaks those numbers in three places, and none of the three ever shows up on your screen. First: All Out. When a side is fully dismissed, the opponent gets 2 points and all seven dismissed players return. Second: the Do-or-Die raid. After two consecutive empty raids, the third becomes mandatory-scoring — fail and the raider himself is dismissed. Third: the Bonus Line. Touching that deep line earns 1 point, but only if the defence has 6 or more players on court.

The sum of those three produces the tempo that puts kabaddi on prime time — fast-scoring, frequent comebacks, an open result until the last minute. The same sum produces the sport's biggest statistical mess, because all three rules manufacture moments where the defence is unstable for a few seconds — and points scored in those seconds get filed in the same column as every other point.

The window immediately after an All Out is kabaddi's weakest defensive state. Seven players are on the court, but seven players are not "set." The corner is still claiming his spot, the cover is still establishing his intercept line, the in-defender is still deciding whether to trail the raider or protect the bonus line. On tape that is a 40-to-70-second window. The raider who scores most in it is the one who raids first — the one whose own team has just completed the All Out.

The scoresheet never records that condition. So when you look at 369 at the end of a season, it is the sum of three things: raiding skill, the advantage of the post-All-Out window, and the ability of the raider's defence to force All Outs repeatedly. The third component is the defence's contribution, not the raider's — and it gets banked under the raider's name.

The second distortion is the Do-or-Die raid. After two empty raids, the third leaves the raider no option but risk. Banking two empty raids is a strategic investment, because on the third the defence has to guess where he will take the risk — and the defence usually guesses wrong. A raider converting above 70 percent on Do-or-Die is the genuine pressure player even if his raw totals trail the leaderboard.

Defensive accounting is inverted too. The real value of a defender like Fazel Atrachali is not in his tackle-point count but in how little the opposing raider converts on Do-or-Die while he is on the mat. Tackle points tell you how many opportunities a defender got; the conversion drop tells you how often he forced the raider into the wrong option.

The Bonus Line's condition is the least-read clause in the sport: 1 point only when the defence has 6 or more players. When the defence is down to 3 or fewer, the bonus line is dead — that game-state belongs to the Super Tackle, worth 2 points to the defence. A single raider's bonus points are decided by match state, not by his technique. The team whose defence keeps 6+ bodies on court keeps its raider in bonus-scoring situations.

I went looking for a three-peat and found a raider monopoly with a soft fixture list. Patna's Season 3-to-5 run ran through Pardeep, but the actual strength sat on the other side of the scoresheet — a defence that forced All Outs constantly and avoided them constantly. 369 there is the output of a system, not personal charisma.

My suggestion is to split raid points into three buckets: points inside 90 seconds of an All Out, points on Do-or-Die raids, and points in normal game-states. The third bucket is the real raiding skill. If a raider's price at the next auction is set purely on total points, you are paying a premium whose other half belongs to his defence.

Now the part where I could be wrong. My whole argument rests on the premise that defences are unstable after an All Out. That is true up to a limit. Top-four defences reset their rotation within 30 seconds of an All Out, and the window nearly vanishes. If that window genuinely shrinks across the last eight matches of the season, the three-bucket model weakens by exactly as much.

The second problem is the tape. The empty-stadium tapes don't lie, but they stay silent too — a raid that isn't on tape isn't in my count. And I don't trust a hot take until it survives a cold walk through the tape.

The third error risk is PKL tunnel vision. This whole frame is built on league data, which is the richest data available — but the league is not the map of the sport. Women's kabaddi has won gold at every Asian Games since 2026 and still has no franchise league; per-raid data there isn't public at all, so this analysis cannot be repeated on that side of the game.

My testable prediction for this season: of the raiders currently in the top three on total raid points, at least one will underperform his average in the playoffs — because post-All-Out raids arrive less often against slow, disciplined playoff defences. If that doesn't happen, my entire three-bucket model belongs in the bin.

Do the arithmetic, watch the tape, and tell me where I was wrong.