The Empty Spreadsheet: The Limits of Modern Table Tennis Analysis
**Câu trả lời cốt lõi**: Phân tích bóng bàn hiện đại thất bại vì đo lường thứ dễ đếm. Chỉ số có sức dự báo cao nhất là tỉ lệ thắng điểm trong ba đường bóng đầu tiên; mọi thống kê khác phần lớn chỉ là hệ quả. **Dữ kiện chính**: - Bóng nhựa 40+ thay bóng celluloid từ năm 2014, làm giảm xoáy và đẩy trái tay thành vũ khí chủ lực. - Hệ thống xếp hạng ITTF dùng cửa sổ trượt 52 tuần, chỉ tính tám kết quả tốt nhất của mỗi tay vợt. - Tỉ lệ thắng trận nội bộ của tay vợt Trung Quốc thường lệch rất xa tỉ lệ thắng trận quốc tế. - Nhịp chân trước khi tiếp xúc bóng và đường di chuyển của người không đánh bóng trong đánh đôi không xuất hiện trên bảng thống kê thương mại. - Khoảng cách đọc tình huống giữa nhóm top 10 và nhóm ngoài top 50 gần gấp đôi, dù tốc độ tay tương đương. **Nguồn**: Phân tích chiến thuật của Phan Long, ghi nhận tại Thượng Hải, cập nhật ngày 13 tháng 8 năm 2026. **Hỏi đáp liên quan**: Q: Chỉ số nào dự báo kết quả trận bóng bàn tốt nhất? A: Tỉ lệ thắng điểm trong ba đường bóng đầu tiên gồm giao bóng, đỡ giao bóng và quả thứ ba. Q: Vì sao xếp hạng thế giới không phản ánh đúng sức mạnh tay vợt Trung Quốc? A: Vì xếp hạng chỉ tính điểm quốc tế, không tính các trận nội bộ có mật độ cạnh tranh cao hơn. Q: Khoảng cách giữa bóng bàn Việt Nam và khu vực nằm ở đâu? A: Nằm ở tầng mã hóa dữ liệu vi mô, không nằm ở kỹ thuật đánh bóng cơ bản.
The Empty Spreadsheet: The Limits of Modern Table Tennis Analysis
Opening
"Every revolution begins with a number left forgotten on a desk."
That night, on my desk in Shanghai, there was no number at all. Only an empty file. Twelve columns. Not a single row.
I had spent sixty-eight minutes in front of a screen watching a quarterfinal, pen in my left hand, slow-motion key under my right. I logged every serve, every opening footwork step, every racket angle at contact. I ended up with more than four hundred loose notes. When I opened the encoding software to turn them into data, I sat still for a long time in front of something very simple: not one of those notes could become a metric with predictive power.
Not because I recorded them badly. Because that match produced no new signal. The two players had met seven times in three years. They knew each other so well that every serve had become a conversation written in advance. I sat there, taking very diligent notes, and finished with a blank sheet.
That moment is the story worth writing. Not a story about a great match, but about an industry that produces millions of numbers a year, most of which measure nothing useful.
Context: table tennis wrapped in numbers that are easy to count
Since 2026, when World Table Tennis replaced the old World Tour, professional table tennis entered a fully reorganised era. Events were tiered clearly: Grand Smash at the top, then Champions, Star Contender, Contender, alongside the WTT Finals for the players with the best points haul of the year. The ITTF ranking system runs on a rolling 52-week window, counting each player's best eight results. Old points expire automatically, and they expire on the very day they were earned.
That mechanism creates a treadmill anyone following professional table tennis can feel: players must compete more, travel more, and above all be present at the right event, in the right week, at the right moment. Points have become a currency. A Grand Smash entry is not just a sporting opportunity; it is a financial decision.
Running alongside that treadmill is a second current: data. Every Grand Smash match now generates hundreds of data points. Ball speed off the racket. Revolutions per second. Win rate on serve. Win rate on receive. Which stroke ended the point. The final direction of the ball.
Vietnamese fans can now open a phone and know where Fan Zhendong stands, how many points Wang Chuqin must defend next month, how many times Sun Yingsha has faced Japanese opponents and how many she has won.
That is real progress. But it creates an illusion that is very hard to detect: the feeling that once you have enough data, you understand the match.
The first three seconds of a match do not lie. The rest is just how we fool ourselves.
I learned this in 2026, when I left the commentary booth to join the data analysis unit of a professional table tennis club in Shanghai. My first task was to encode the team's entire season, shot by shot, frame by frame. Four months later, when we ran the first model, the result silenced the room for several minutes.
The only indicator with a clear correlation to match outcome was the point-win rate across the first three strokes: serve, receive, and the third ball. Everything after that, however meticulously recorded, was largely consequence. A player winning 62 percent of points inside the first three strokes almost certainly wins the match, regardless of what he does afterwards.
That is why I started counting everything from the first bounce. And why I gradually realised that most of what this industry measures is only the tip of an iceberg.
The three-stroke system: where the match actually happens
A world-class table tennis point, from the moment the ball leaves the tossing hand to the third bounce, lasts under two seconds. In that window, three decisions are made almost simultaneously: which spin, which placement, which tempo.
When encoding elite matches, I do not count total points. I count how many distinct serve "families" a player uses in a match, then cross-reference them against the points won from each group. In one Grand Smash match I tracked in Singapore, the winner used fourteen different serve variations across five games. Grouped by spin family, the real number was three. Nine of the fourteen were variations of the same family.
The interesting part lay elsewhere. He won 71 percent of first-game points with the first spin family, then was read in game two. He switched to the second family. In game three, the opponent adjusted. By game five, with both families decoded, he returned to the first — but with an entirely different placement pattern.
A standard data sheet would record: "fourteen serve types, 63 percent win rate." That number is technically correct and tactically useless. It does not say what changed, when, or why the opponent could not adjust in time in the final game.
What the audience sees on screen is not the real match. The real match happens in the gaps the camera forgets.
The largest gap is the footwork before contact. Broadcast cameras follow the ball. When the ball travels, the lens travels with it. But the decision is made before the ball leaves the opponent's racket, at the moment the receiver lowers his centre of gravity and plants his left foot in exactly the right place. Across more than two hundred matches I have encoded, I have never seen a commercial data platform record that moment.
In 2026, when the pandemic paused the sport and it restarted without crowds, I had a rare chance to test this. I re-downloaded every pre-pandemic match and every restart match and placed them side by side. My initial hypothesis was that players would serve more conservatively without crowd noise. I was wrong.
What I found ran the other way: without spectators, lower-ranked players attacked the third ball harder and more decisively, and won more points on it. The stadium pressure disappeared, and what disappeared with it was caution. I rewrote my conclusion in the middle of the analysis, and that made the piece more trustworthy rather than less.
An empty stadium does not strip a match of its soul. It only removes the makeup and reveals the tactical skeleton.
The invisible players: doubles and the movement problem
If there is one format modern data has abandoned entirely, it is doubles.
In doubles, only one player hits the ball on each contact. The other does not touch it. And in most stat sheets, the player who never touches the ball does not exist.
I once spent an entire tournament doing one thing: mapping the movement paths of the non-hitting player in women's doubles pairs. I recorded their position at four moments: when the partner prepares to serve, when the partner contacts, when the opponent receives, and when the partner prepares for the next stroke.
A clear pattern emerged. Strong pairs move along something close to a V-shape: the non-hitting player drops back toward the left corner to open space for the partner, then advances diagonally as the partner strikes. Weaker pairs move in straight lines, and the distance between the two players barely changes all match.
That difference decides a great many points, yet it appears in no column of any stat sheet. The sheet records who scored and who erred. It does not record who created the space for the other to score.
At Southeast Asian level, the doubles gap is usually not about stroke technique. It is about movement. Reviewing many doubles matches from recent SEA Games, the same pattern repeats: losing pairs lose points where both players stand still and watch the ball. Nobody makes an obvious error. But nobody moves either.
That is the kind of defeat that leaves no trace in the data, and therefore never gets fixed.
The forgotten number: the domestic win rate
There is one metric I consider the most important in men's table tennis today, and it appears on almost no international outlet: the domestic win rate of Chinese players.
The world ranking measures accumulated points at international events. It cannot measure what happens in internal training sessions, in the Chinese national championship, or in the selection rounds that only about twenty people in the world are good enough to enter.
The result is a paradox any long-time follower knows: some players ranked outside the world top 30 sit inside China's strongest eight. At an international event they exit early because they draw a teammate. At a domestic event they beat the reigning world champion.
Based on my experience tracking matches, I always cross-check two figures before making any judgement: the international win rate and the domestic win rate. For most players in the world, the two are close. For Chinese players, they diverge sharply, and the harder number is always the domestic one.
This means the world ranking, to a certain degree, measures the wrong thing. It measures accumulated results, not real strength. Over the past twelve months, when I compared rankings against direct head-to-head results among leading players, the correlation held in roughly two-thirds of cases. The remaining third is where the ranking lies.
That is a large enough share that anyone using the ranking as a prediction basis should pause.
Equipment: the plastic ball and the slow disappearance of spin
You cannot discuss modern table tennis without one physical change that reshaped the entire sport.
In 2026, the ball diameter increased from 38 to 40 millimetres. In 2026, celluloid was fully replaced by plastic, the so-called 40+ ball. Both changes sound minor technically. Both rewrote the textbook.
The plastic ball is larger, softer at the surface, and above all generates significantly less spin than a celluloid ball of the same size. Less spin means a straighter flight, and a straighter flight means more force is needed to create a safe trajectory over the net.
The direct consequence was the rise of the backhand. In the celluloid era, a heavily spun backhand could create enough curve to be safe at close range. With plastic, the backhand became a weapon with a far better risk-reward ratio. And as the backhand grew stronger, players used it more. In some elite matches I recently encoded, backhand usage on decisive strokes has overtaken forehand usage.
On the equipment side, the world split more sharply than ever. Traditional Chinese rubbers have a tacky surface, allowing high spin with a small motion range, especially on the forehand. European and Japanese rubbers go the other way: less tack, more elasticity, suited to away-from-table play and heavy counter-looping.
Sponge hardness is a variable fans routinely ignore. For Chinese rubbers, 39 to 41 degrees is the common range among professional men. A one-degree difference sounds tiny, but it changes the delay of the stroke, and the delay changes contact timing, and contact timing changes the entire rhythm of the third ball.
Blades follow the same logic. A five-ply all-wood construction gives clear feel but lacks power from distance. Seven plies give more power but reduce close-range control. Blades with carbon layers close to the outer wood deliver speed on immediate contact, suiting close-to-table play. Blades with carbon buried deeper feel softer at low speed, suiting players who need more processing time.
Each of those choices creates an adaptation period, usually lasting months. During that window, a player's data becomes meaningless because it reflects someone who has not finished converting. I have seen many analyses conclude a player is "declining" when in fact he has simply changed rubbers.
Three technical schools colliding
Men's table tennis currently hosts three technical currents, and their collision produces most of the matches worth watching.
The first is the efficiency school, tied to the Chinese training system. Its hallmark is extreme optimisation of the first three strokes. Players are trained to end points within four contacts or fewer, and everything else is a means to that end.
The second is the physicality and counter-looping school, tied to Europe and Brazil. Here points last longer, and physical power becomes the decisive variable. Hugo Calderano is the clearest example: he can strike from far behind the table while retaining speed and spin that most Asian players lose once they retreat.
The third, and the most notable of the past three years, is the hybrid generation. Truls Moregard plays in a way that almost breaks standard technical rules, with flat hits, unusual contact points and unpredictable rhythm. Felix Lebrun brings back the left-handed penhold grip, but attaches it to a modern backhand the previous penhold generation never had. Tomokazu Harimoto represents another model: close-range blocking at high speed, converting defence into attack within a single beat.
Notably, all three currents share one trait traditional data cannot capture: decision speed. When I measure the interval from the ball leaving the opponent's racket to the receiver's first foot movement, the figure among the world's elite is usually under 0.2 seconds. Outside the top 50, it is nearly double.
The gap between the two groups is not hand speed. It is reading speed.
Where Vietnam sits on this map
Vietnamese table tennis has names that have secured a regional standing: Nguyen Anh Tu, Dinh Anh Hoang, Nguyen Khoa Dieu Khanh, Mai Hoang My Trang. They are a generation with solid fundamentals and stable competitive temperament.
But when their matches are placed beside continental-level matches, a gap appears, and it is not in stroke technique. It sits at the encoding layer.
Over years of tracking Southeast Asian table tennis, I have noticed that the region's stronger teams began recording micro-metrics years ago: foot position on receive, recovery time to the neutral position after each stroke, backhand usage on the third ball. The rest are still recording total points won and total errors.
Those two approaches produce two different development speeds, and the gap compounds annually.
What is worth saying is that micro-encoding does not require expensive technology. It requires one person in front of a screen, a notepad, and the discipline to record the same thing the same way across hundreds of matches. I did that with free software for the first three years of my career.
The counterintuitive angle: data as a shield
Once data becomes widespread, it stops being a tool and starts being a shield.
A coach wanting to defend a decision will choose the option with the most data, not the best one. A player wanting to avoid criticism will choose the safe style, because the safe style can be justified with numbers and the risky style cannot.
This is why I increasingly distrust the total backhandisation of men's table tennis. At first glance it looks like technical progress. Look closer and it is often risk avoidance. A player who develops a strong backhand gets a clean metric, a stable point-win rate, low variance across matches. A player who develops a creative forehand gets some brilliant matches and some terrible ones.
The industry rewards measurable stability and punishes hard-to-measure creativity. That is a skewed incentive system, and it is gradually homogenising young players into identical versions of one another.
Data does not replace the intuition of an old coach. It only hands him a more accurate map.
And there is one thing this entire system will not let anyone say: "I don't know."
That night, when my data file came up empty, my first reaction was to reopen the match and find more things to count. I tried to turn a match with nothing new into a match with something. It took about two hours before I accepted that the correct output was a blank sheet, and that a blank sheet is itself a finding.
In an industry measured by output, nobody wants to publish an empty finding. But an empty finding told me these two players had reached perfect informational equilibrium, and that any next change would come not from technique but from psychology or fitness. That is a valuable judgement, and it could only be made by accepting the emptiness instead of filling it with meaningless numbers.
What to verify at the next event
At the next round, I will track four things, and I suggest anyone wanting a deeper read of the game track them too.
First, each quarterfinalist's point-win rate across the first three strokes. This figure predicts outcomes better than total points won.
Second, the number of recovery steps back to the neutral position after each stroke in the fifth game of long matches. This is a hidden fitness metric, and it usually decides who wins the closing points.
Third, the position of the non-hitting player in the semifinal doubles pairs. A V-shaped movement path versus a straight line will reveal which pair has been coached at the micro level.
Fourth, and perhaps hardest to track, whether a player emerges from the internal selection rounds with a low world ranking but a high domestic win rate. If so, that is a candidate for a major upset, and the ranking will not tell you.
The data revolution in table tennis is not over. It has only just reached the point where people begin to realise that most of what was counted over the past decade was the easiest thing to count, not the thing that decides outcomes. The next step is not to count more, but to count better.
And sometimes, the most correct thing is to accept that in a particular match, there was nothing new to count at all.
