Puyo Puyo Champions: A 2-2 Record Advanced, but the Ledger Reads 9-7

Four rows on the score sheet: 3-0, 3-2, 0-3, 3-2. Late on Thursday night I opened the Nagoya results and went to the match column first. Two wins, two losses —.500. In my ledger, a.500 record is an exit number, not a survival number. At the 2026 Asian Games Puyo Puyo Champions event, India's Paritosh survived anyway. What kept him alive was not the player. It was the rule that the top three in each group advance.

There is no clutch round here, no last-second finish. There is one structural fact that converts a four-match record into a tournament identity: the quota. In a five-player group, 60 percent of the field reaches the knockout. In that format a.500 record is not an anomaly; it is the expected outcome.

A second column surfaced later, one I had skipped at first. At game level Paritosh is 9-7 positive. Add 3-0, 3-2, 0-3 and 3-2 and you get a two-game positive margin. The match column says fringe; the game column says no worse than third-best outside the group's top two. Same four matches, two contradictory stories.

Puyo Puyo Champions is Sega's versus puzzle title, sold in Japan as Puyo Puyo eSports. Character selection in this discipline is effectively cosmetic — there is no champion power hierarchy, so there is no pick-ban meta and no patch-driven meta either. The real skill sits in chain construction (GTR, stair chains) and in garbage-sending pressure. This report contains no patch or version information at all. Anything written under the heading of patch analysis would have been speculation, and speculation does not enter the ledger.

The event runs under the Olympic Council of Asia umbrella as part of the esports programme at the 2026 Asian Games in Aichi-Nagoya. Group B held five competitors, each playing four round-robin matches in a first-to-three series. Group play was on Thursday, the quarter-finals fall on Saturday — roughly a two-day gap. The venue is Aichi Sky Expo Hall, Nagoya.

Paritosh's four opponents: Munkhzorig Ganbold of Mongolia (3-0 win), Koh Mun June Justin of Singapore (3-2 win), Tzu-ang Kuo of Chinese Taipei (0-3 loss), Tanarak Wongkitikun of Thailand (3-2 loss).

The presence of this title at the Asian Games is itself a data point. Puzzle disciplines sat outside multi-sport events for years; their entry into the Olympic framework means the discipline list is widening beyond MOBA and FPS. Paritosh's result is the smaller story, but read together they form a larger picture: South Asian competitors are now entering disciplines with no domestic league pyramid behind them.

The sequencing is the largest variable here. Two wins came in the first two matches, two losses in the last two. The further the session went, the worse the result. There are two explanations, and four score lines cannot separate them. One is fatigue or a mental fade — what the report called a momentum shift. The other is opponent-quality escalation: the group's stronger players may simply have fallen later in the schedule.

The 0-3 loss to Tzu-ang Kuo is the most decisive result in the group — not a single decider game was played. Wongkitikun won 3-2, meaning the final match was genuinely close. Together those two facts say this: Paritosh's last two matches came against the group's stronger half, and the last of them was competitive.

My proxy metric here is not xG but game differential. In MOBA analysis we look at team-fight win rate and objective conversion. The equivalent indicators in a puzzle title are three: game differential, decider rate, and the within-match performance curve. The first is known — 9-7. The second is known — two of four matches went to a decider, one won, one lost. The third is unknown, because there is no round-level data: who completed a chain first in which round, whose stack rose higher, what the harassment volume was. None of those numbers is in hand.

That is exactly why I treat the two 3-2 matches most carefully. Two deciders, one win, one loss. Flip any single game and the entire qualification story flips — win one decider and Paritosh sits in the top two; lose one and he is out. Across a four-match sample, those two games carry the whole weight of the narrative. Where the sample is thinnest, the story is loudest — the ledger remembers what the highlight reel forgets.

One more structural point: scouting in a small pool is symmetric. In a five-player group everyone has seen everyone. A debutant's natural edge — no prior tape for opponents to study — effectively expires by match three.

Had the quota been two, Paritosh would have exited on a.500 record. Imagining that rule change is not entertainment, it is a model test: same performance, different structure, different outcome. The distance between third and fourth in that table was made by the rules, not by the play.

Across the past few years of keeping a South Asian esports ledger, I have built one habit: I split every result into two buckets first — structural context and performance attribution. In this match the first bucket holds the quota rule, the small pool and the two-day gap. The second holds 9-7, the two deciders, and the finality of the 0-3.

Now the reverse side of my own reading. The eye test says momentum shifted, and on sequencing the eye is right. He lost the last two matches; that is a fact. But momentum is a description, not a cause. Without chain-level data there is no way to separate fatigue from opponent escalation. So in my ledger this stays at provisional tier, not firm.

The opposite rigidity would be just as incomplete: the format saved him, therefore the result carries no weight. That conclusion is wrong. A net positive 9-7 game margin came against a group containing Chinese Taipei and Thailand, two disciplined regions. That is not an empty number; it is a floor. This is where my own trap waits: the person who always checks the numbers starts expecting the correction more than the finding. The eye test here said he survived, and the eye test is correct. My addition is only an increment: survival and form are two different measurements.

A third angle belongs here too, even though in my columns it is usually the easy alibi. Writing about South Asian competitors, I routinely raise the ping floor and the device tier. It applies less here: a versus-puzzle title is low-bandwidth, tolerant of input latency, and not server-dependent the way MOBA or FPS is. So a smaller share of any performance gap can be pushed onto infrastructure — and the weight on performance attribution rises instead. I am firm on the mechanism, provisional on the magnitude.

One caution sits attached to this piece. The source is single and largely unattributed — group size, quota rule and series length are all inferred from score lines. Until official regulations are checked against it, this entire analysis stays at the provisional tier.

Saturday's quarter-final is the next data point. I am pre-registering the triggers now, because writing them afterwards turns them into explanations. One: if he again loses the last match of the session, the fade model gains a second point — and at n=2 that is still not a pattern, only a second observation. Two: if he keeps winning the decider games, the fade hypothesis weakens.

The bracket draw is the other large variable. Given a 0-3 loss on record, finding the group's most consistent competitor on the opposite side of the bracket would materially ease his path.

My baseline is not patch-based, because no patch data exists — it is opponent-pool-based. If group size and quota rule are confirmed officially, the baseline itself needs updating; change the group size and the 9-7 read changes with it.

In a discipline with no xG, no champion pool, and only four matches, what do you actually measure? Probably the first entry in the ledger should be the format. The player comes after.