Three Columns Nobody Records in Vietnamese Table Tennis Notebooks

On the spectator stand, a coach opens a notebook and rules two columns. The left column holds player names; the right holds the score of each game: 11-7, 9-11, 11-5, 11-8. When the match ends he closes the notebook, clips the pen to the spine, and walks down to the coaching area. That notebook holds exactly the amount of information a results table on the organisers' website holds — and not one word more.

I stayed behind for a long while after that session. Across three days of a national youth tournament, I counted 41 matches, 14 coaches, and precisely three people writing something other than the score. One drew small squares on graph paper. Two used private shorthand and crossed it out when a game ended. Everyone else wrote the score. I do not blame them. Recording the score is the umpire's job, and the umpire had already done it.

By the end of the season, though, when the coaching staff had to finalise a squad for an international event, the only thing on the table was those columns of scores. Everything else had vanished with the applause.

HOW MANY DATA CELLS DOES ONE MATCH PRODUCE

A game to 11 points, plus the two-point margin in the deciding phase, usually lands between 18 and 22 points. A three-game match generates roughly 60 rallies. A five-game match passes 100. Each rally in turn carries at least five pieces of information: who served, the serve type, the receive, rally length, and how the point ended.

Multiplied out, a five-game match is about 500 data cells. A national tournament of 200 matches is 100,000 cells. That is a modest volume — small enough to fit in a spreadsheet one person could finish in four evenings.

Yet most of those cells are left blank. Vietnam's table tennis weakness lies in data custody: information is generated and then abandoned inside the arena, so nothing remains to analyse afterwards.

The world federation's ranking only counts international events. Which means a Vietnamese player may play 15 to 20 international matches a year and more than 100 domestic ones, but only the international group exists on paper. The rest is an empty file.

I still remember opening an empty dataset. In 2026, when global football stopped, I sat down with data from 50 V.League 2026 matches to build an "effective pressing" spreadsheet — possessions recovered within five seconds of losing the ball. The result surprised me: the team reputed to press best was not the one with the highest recovery rate, but the one that transitioned fastest. I drew a lesson that had nothing to do with football: without data, people default to what they remember. And memory always favours the winner.

THE MINIMUM FIVE COLUMNS

If I had to start over, I would propose one sheet of paper per match, with five columns.

A column for the sequence of points. Not just the final score of each game, but every point in the order it happened. Losing a game 9-11 with four straight points conceded at the end is not the same as losing 9-11 with those four points scattered.

A column for who served on each point. It is so simple it gets skipped, and yet it is the only column that shows who controls the rhythm.

A column for serve type: topspin, backspin, sidespin, short, long. Four symbols are enough.

A column for how the point ended: service error by the opponent, receive error by the opponent, forehand winner, backhand winner, or an error from the fifth ball onward.

A column for rally length, counted quickly by the number of times the ball touches the table.

Five columns. One page. One match. Each column written once, costing perhaps twenty extra seconds per game.

I have applied this principle in my commentary work. For every match I follow, I redraw the movement pattern of both players in the deciding game, then attach three lines of footnote beneath it: who lost the initiative on which point, why, and what would change if it happened again. Every hand-drawn diagram is a story the numbers cannot tell. The reverse holds too: a column of scores is a story a diagram cannot rescue.

WHAT THOSE FIVE COLUMNS UNLOCK

Once the sequence of points exists, patterns the scoreboard hides start to surface.

A player winning 60% of matches may be losing on his own serve. Another losing more often may still hold a high winning rate on serve, meaning the problem sits in his receive rather than his attacking technique. From the outside the two look alike: one wins a lot, one wins little. From the spreadsheet they are entirely different problems.

I call this statistic the "first three seconds" — points that end within the first three ball contacts after the serve. In table tennis it is the closest available proxy for the quality of the serve and the receive. A player with a high rate usually wins without entering long rallies. A player with a low rate is forced to extend rallies, and the longer a rally runs, the higher the chance of error.

The second metric is the winning rate in rallies from the sixth contact onward. That column measures stamina and consistency. Combine the two and you get four player types: good serve and durable, good serve but fading, weak serve but resilient, and weak across the board. Those four types need four different training plans. Without a spreadsheet, all four are treated the same.

I once spent nearly four days redrawing the defensive system of a team at a World Cup, cross-checking 200 defensive phases across five matches, only to reach one simple conclusion: they conceded the middle, not the flanks. Four days for one sentence. But the sentence was right, and it could not have been guessed by eye.

The principle I set for myself from then on: never assert a tactical claim without at least three independent data sources behind it. Applied to table tennis, those three sources can be simply the sequence of points, the serve column, and the position diagram from the deciding game.

THE BLIND SPOT IS NOT THE CAMERA

The first reaction to any call for record-keeping is usually: there is no shortage of cameras, no shortage of phones. Everyone has a phone. Every arena has someone filming.

But unannotated footage is just memory with colour. A twenty-minute clip will not tell you whether the seventh point of the third game was a sidespin or topspin serve, and certainly not on which contact the player chose to change direction. To extract that from a clip you have to rewatch and take notes — which is the exact work you set out to avoid.

A pressing spreadsheet does not create a coach, but it creates a reader of matches. Data does not replace the coach's eye. It forces that eye to be more precise, to say why, to stop leaning on vague feeling.

A subtler blind spot: people record when they lose and stop recording when they win. But a 3-0 win contains errors too — they simply went unpunished. If records are kept only in defeat, the dataset skews from the very first cell, and every conclusion drawn from it leans pessimistic.

The remaining blind spot is sample size. Eight matches prove nothing. That is why I always place a hand-drawn diagram beside every number. The diagram supplies context; the number supplies frequency. Remove either one and the conclusion becomes a guess wearing the costume of analysis.

I look at the diagram first and the reputation second. Names like Nguyễn Anh Tú, Trần Tuấn Quỳnh, Mai Hoàng Mỹ Trang or Nguyễn Khoa Diệu Khánh can make an audience remember a match for years. But knowing which rhythm they genuinely own takes a notebook, not an ovation.

WHAT TO DO NEXT SEASON

I am not proposing a large system. No software, no hiring. Just one rule: one match, one page, five columns.

If every team adds those five columns to every match next national season, then within a year Vietnamese table tennis will hold something nobody currently has: a real form record instead of a list of scores.

And with that record, people will discover the most interesting thing about this sport — that the smallest detail at the table has a reason. A short serve at 9-9 is not random. Half a second of retreating footwork has a cause.

I will do my part. Next season I will try to log the full five columns at a youth tournament, publish the spreadsheet, and check it against what I have previously claimed. If I am wrong, I will say which column I got wrong. Prediction is an art; data only paints the background. But a background painted wrong cannot produce a correct picture.