When the Analysis Sheet Comes Back Blank: Where Vietnamese Table Tennis Loses Its Data

On Friday evening I reopened the deconstruction file for a domestic table tennis match and read it three times. Nine sections. Technique and tactics: blank. Player data and head-to-head: blank. Event system and ranking points: blank. Competitive landscape: blank. Rules and governance: blank. Coaching staff and talent pipeline: blank. Risk surface: blank. Public narrative and expectation: blank. Industry transmission: blank. One cell survived, on the last line: domain label — table tennis.

Seven years ago in Guangzhou I sat in front of a different sheet. An U19 match against the other academy in the city: I counted 18 pressing actions in the opponent's final third and only two successes. I redrew the whole match on the board, cross-checked it against positioning data, found the forwards drifting four to six metres off the central axis, and proposed narrowing the gap between the two wide midfielders from 28 metres to 22. We won the return leg 2-0. That sheet was so dense I had to draft it on paper before opening the laptop.

Today's sheet was so empty that I checked whether I had opened the wrong file. I opened it again. Still empty. Then it struck me: the emptiness was not the file's fault. It was data. It said that somewhere in Vietnamese table tennis's recording chain, a link had snapped before the match could be saved.

A modern table tennis match, if it is to be dissected properly, needs four groups of data. The first is the point-by-point ledger: who served, into which zone, how the opponent received, at which beat the rally ended. The second is the geometry of movement: foot position, racket angle at contact, the contact point along the ball's trajectory, the recovery distance after each stroke. The third is calendar and points: the WTT ranking system rolls on a 52-week cycle and counts a player's best eight results, so a ranking figure only means something if you know when it was captured. The fourth is equipment and regulation: under ITTF rules the total thickness of the racket covering may not exceed 4mm, plus surface flatness and volatile-solvent residue — the speed-glue ban took effect in 2026, the ball went from 38mm to 40mm in 2026, and the 40+ plastic ball replaced celluloid in 2026. Every ball change forces the entire technical system to be rebuilt.

At major events those four groups flow into one continuous stream. On Vietnamese domestic courts they usually break at the very first group. The national championship has referees, match sheets and set-by-set scores, but it rarely publishes a point-by-point ledger. Nobody records that in the fourth set a player served short to the left seven times in a row and the opponent pushed on only five of them. That does not mean it failed to happen. It means it does not exist in a verifiable form.

When I take on a match for deconstruction, I start by looking for three things: the date of play, the camera position, and the name of the person who recorded it. Without all three, every conclusion downstream is a guess wearing a good presentation. Friday's file had none of them. I closed it and decided to write about the gap itself rather than fill it with sentences that sound technical.

Without frames, a technical description is only a metaphor. The sentence "his backhand loop is better now" cannot be verified. Verification needs three things: racket angle at contact, the contact point along the ball's trajectory — rising, at the apex, or falling — and the recovery distance after the stroke. I once sat with a young player and showed him that his forehand loop was making contact about 5cm too far inside his body. The swing speed was unchanged. But the racket head travelled a shorter arc, the spin dropped, the ball landed 20 to 30cm shorter, and a stroke that should have applied pressure became a neutral placement. Those five centimetres are not small. They are the boundary between a point won and a counterattack conceded.

But to see those five centimetres you need a fixed camera, on the right axis, at a minimum of 60 frames per second. Domestic broadcasts usually have one high camera set behind the table at an angle — enough for viewers to follow the ball over the net, not enough to measure racket angle. I have tried measuring on a few such recordings. The error margin was larger than the movement I needed to measure. I abandoned it, because a measurement whose error exceeds its signal is worse than no measurement at all.

The first three beats decide most table tennis matches, and they are also the least recorded part. Table tennis differs from football in that a very high share of rallies end within the first three balls. The three most important indicators are therefore: win rate on rallies inside the first three beats, attack rate after receiving serve, and the distribution of rally length. With nobody tagging each rally, the coach is forced to use memory as the dataset. Memory is a biased sampler: it remembers the spectacular rally and forgets the repeated one. A player who faults three serves into the same zone across four sets will be remembered as "having a problem in the last set