Taolu Has No Win Rate — and Vietnam's Martial Arts Media Is Using the Wrong Ruler

On the evening of 8 August 2026, I sat in a café on Lạch Tray Street in Hải Phòng, replaying a sports programme. The on-screen graphic carried a line I had to rewind three times: “Win rate: 62%.” Beside it was the name of a wushu athlete, taolu division.

Taolu has no opponent. There is nobody to beat.

A taolu routine unfolds on a floor whose dimensions are set by the International Wushu Federation: one athlete, one routine, roughly 1 minute 20 seconds to over two minutes depending on the event. Nobody stands opposite to throw a strike. Nobody can be hit. Yet the graphic still handed that athlete a win rate, as though she had beaten ten opponents in a single night.

What made me sit back was not that the number was wrong. It was how confidently it was presented — clean, tidy, coloured, numeric. Confident enough that a casual viewer would accept it as information.

I got EURO 2026 wrong, not because I lacked information, but because I trusted the crowd. Since then I have kept one professional habit: whenever a metric looks too smooth, I go looking for the ruler that produced it. In Vietnamese martial arts coverage, that ruler is being misused — and misused systematically.

Three families, three rulers

Combat sports with opponents — boxing, MMA, kickboxing, Muay Thai, sanda — are the family where win rate means something, though it means something different in each. Olympic and amateur boxing abandoned electronic clicker scoring in 2026 and moved to the professional 10-9 system, where judges reward clear effectiveness and damage. Unified-rules MMA also scores 10-9, but weighs damage, positional control and pace. Sanda scores on clarity and effectiveness: a clean head kick or a clean throw is worth far more than a jab to the body. Muay Thai is harder still, because international scoring under world federation rules differs substantially from traditional scoring at the big Bangkok stadiums, where balance, rhythm and control can outweigh raw strike counts.

The second family is performance — taolu. No opponent, only a judging panel. Points split into three panels: A for movement quality on a 5.0 scale, B for overall performance on 3.0, C for difficulty on 2.0. Total: 10.0. The athlete performs alone, and the only comparable output is a score.

The third family is the mixed one, and the most easily confused. Sanda and taolu sit inside the same federation, the same national team, the same competition programme at major Games. Because they share a door, media collapses them into one block — and once collapsed, the ruler of one family gets applied to the other.

That is the root of the problem. Not laziness. Collapse.

Three months of my own log

From 1 May to 31 July 2026, I logged 148 Vietnamese-language martial arts media items that crossed my path: TV bulletins, news articles, long social posts, podcasts. I did not sample by taste; I logged whatever surfaced in my own feed, including work I thought was poor.

The result: 63 of the 148 used at least one combat-sport metric — win rate, number of wins, knockout rate, unbeaten streak — to describe an athlete or event in the performance family. Of those 63, only 9 noted that the subject was a performance discipline.

Put plainly, close to half the items I encountered used the wrong ruler, and more than 85 percent of those had no idea they were doing it.

Vietnam is not a stranger to taolu. Nguyễn Thúy Hiền won the world changquan title in 2026, one of the earliest markers that a Vietnamese athlete could stand at the top of a discipline scored by absolute points. In later years, Dương Thúy Vi was the familiar taolu face of the national team in nandao and nangun at continental level. We have a tradition in exactly the family the coverage misreads most.

The consequences are concrete. It reshapes what audiences treat as value: tell a viewer an athlete has a 62 percent win rate and they will instinctively rank her below one with 85 percent, even though neither figure exists and even though the real currency is score and score variance. It bleeds into investment decisions: a sponsor reading two articles about the same athlete — one about a medal, one about a low win rate — forms a skewed impression. Sponsors have no obligation to understand wushu rules. Writers do. And it blurs what failure actually is. In combat sports, failure is losing to a specific person and can be dissected round by round. In performance, failure is an unfinished difficulty movement, a fall, a 0.03 gap on panel A. Collapse them and you erase the ability to analyse either.

What exists in taolu is not a win rate, it is variance

I built my own metric because nobody sells one.

I call it score variance — the standard deviation of an athlete's score across their three most recent outings in the same event. I pulled published score sheets from national championships and regional Games, entered them into a spreadsheet, and added by hand.

Three months of tracking produced a few observations. In domestic men's and women's taolu events, the gold-to-silver gap usually sits under 0.1 points; in many events it is 0.02 to 0.05. At that margin, a single small panel A error — a foot not quite planted, a stance not quite low enough — changes the colour of a medal.

In a discipline where gold and silver sit 0.03 apart, the concept of a win rate is technically wrong and it also destroys any ability to understand why anyone won.

From that spreadsheet I extracted a usable signal: a variance under 0.05 suggests a stable technical base. A variance above 0.15 across outings usually means a failed difficulty movement, or a fall. I call that trace the fall mark, and it tells me more than any win-rate bar I have ever seen on a graphic.

To be explicit: this is data I collected and calculated myself, not official federation data. I publish the method so that anyone who disagrees can disagree with numbers rather than with feelings.

In sanda, win rate says less than people assume

Sanda looks like the easy family to analyse, because there are wins and losses. Not quite.

Domestically, an athlete's season is short: a draw, a bracket, done. Five bouts in a year is a lot. With a sample that small, a 7-3 record is close to statistically meaningless — change one draw and the whole picture changes.

What carries meaning sits at round level. I record four things: average points per round, third-round win rate, knockdowns conceded, and penalty faults. In my data, the knockout rate in domestic sanda is far lower than audiences imagine; most bouts end on points. Look only at the win column and you are reading the last line of a process you skipped entirely.

Third-round win rate is my favourite metric because it measures conditioning and tactical nerve at once. An athlete who wins round three is usually one who adjusts after two rounds. That belongs on a graphic. A meaningless percentage bar does not.

When the frame is wrong, every comparison downstream is worthless

A wrong analytical frame makes every comparison it produces worthless, even when each individual number inside it is correct.

You can take an athlete's true win count, true loss count, a true percentage accurate to two decimals, and place it beside a performance athlete's name. Every number is real. The conclusion is still entirely wrong, because it answers a question the discipline never asks.

There is a linguistic habit worth flagging too: Vietnamese martial arts coverage borrows boxing vocabulary for taolu — duelling, knockout, counter-attack. The words sound spectacular. They also erase the specificity of a discipline whose entire value lies in movement precision and repeatable stability.

Where I could be wrong

People call me a contrarian; I call it how I stay sober. But contrarianism has traps, and I should name mine.

I may be underestimating the audience. Vietnamese martial arts viewers are not naive. The same person can watch sanda on Saturday night and taolu on Sunday morning, and can tell them apart. If I assume that any win-rate graphic is swallowed whole, I am insulting the people I write for.

Then there is convergence. Sanda absorbs technique from Muay Thai and kickboxing. Kickboxing absorbs from boxing. Some promotions run mixed rulesets where an athlete competes in both contact and performance formats on the same card. When reality blends, my demand for a hard classification system can become rigidity.

And here is what I fear most in myself: the wrong-frame argument can be used as a shield. If every judgement can be excused with a charge of using the wrong ruler, nobody is accountable for a bad call. The contrarian trap is not meant to fight the player, it is meant to fight prejudice — but a trap built to fight prejudice and then used to cover carelessness has been used for the wrong job.

I once underrated a sanda athlete because his win-loss record looked ugly. Looking closer, I found the opposite: he chose a safe style, avoided injury, lasted longer, and banked points in later rounds. What I called a low win rate was a career-survival strategy. I misread a person by reading one metric correctly.

What to measure instead

If I could rewrite one martial arts bulletin, I would start with a small label in the corner of the screen naming the family: contact, performance, or mixed. That single label removes most downstream errors.

Then the metrics. For taolu, drop win rate and use split A, B and C scores, score variance across the three most recent outings, and difficulty-movement completion rate. For sanda, replace overall records with points per round, third-round win rate, knockdowns conceded and penalty faults. For Muay Thai, state which ruleset is in play, because the same fight can produce two different winners under two scoring traditions.

The hardest part is admitting limits. If a metric does not measure the thing being discussed, it should not be on screen.

The question I leave behind, and I genuinely want the answer: if tonight's graphic read average score 9.62 with a variance of 0.04, instead of a 62 percent win rate, would you see that athlete differently?

I believe you would. Viewers deserve to see what the athlete actually trained for, not a percentage bar born of confusion.