The Empty Spreadsheet in Brisbane: When a Swimming Analyst Has Nothing Left to Count

11:47 PM, one August night in Brisbane. On the laptop screen in my apartment overlooking the Brisbane River, I opened the Excel sheet I had spent three weeks preparing for Australia's national swimming trials. Column A: athlete names. Column B: qualifying times. Column C: 50m splits. Column D: stroke rate. Column E: distance per stroke. And column F — the most important one, where I record pool conditions, water temperature, grandstand pressure — was empty. Not a single row. I pressed Ctrl+Shift+End a fourth time. The spreadsheet returned me to cell A1.

That was the moment I recognized something five years in swimming analytics should have taught me earlier: a data pipeline can die in silence, and nobody notices until you need the number. That night I sat staring at the screen, not because there was nothing to write, but because I knew exactly what I had to do: write nothing at all.

Swimming differs from football at one fatal point. In football, you can rewatch the match and count passes. In swimming, everything happens underwater; in front of you there is only a lane and a touch of the hand. If you have no splits, no stroke data, no notes on pool conditions, you cannot analyze. You can only guess. And guessing, in my profession, is a form of professional crime.

I was born in Vietnam, work in Australia, and I learned this rule not in a classroom but in Kazan. The day Germany collapsed. The day I learned that a 99% probability can still die on the betting table. But Kazan taught me another lesson, less spoken of: when you have too much data, you can become arrogant. And when you have no data at all, you must stay silent.

That Brisbane night, I stayed silent. But silence does not mean there is nothing to say about how swimming data operates — and where it operates wrongly.

In swimming, splits are the entire story.

A swimmer swimming 200m freestyle in 1:46 tells you nothing unless you have their four 50m splits. A negative split — the second 50 faster than the first — is the signature of a swimmer in control of pacing. A positive split — slowing down — signals an over-fast start or a fitness problem. Same final time, two different split structures, two entirely different stories about potential in the next round.

I once analyzed a female swimmer in the 400m individual medley at a junior meet in Queensland. She swam 4:42 — nothing remarkable. But her splits revealed something odd: her breaststroke leg — her weakest — was 1.8 seconds faster than her previous personal best. Meanwhile, her final freestyle leg — supposedly her strongest — was 0.4 seconds slower. My conclusion was simple: she had swum breaststroke better than expected and freestyle below her ability. That meant 4:42 was not her ceiling. Her real ceiling, if she held the breaststroke leg and fixed the freestyle leg, was around 4:38.

Six months later, she swam 4:37.

Without splits, I would have overlooked her. And that is why I never trust a results table with only one time column.

But splits can deceive you too.

Stroke rate and distance per stroke (DPS) are two companion metrics I read together. A swimmer with high stroke rate but low DPS is "fanning water" — fast over 50m, broken over 200m. Conversely, low rate and high DPS is an energy-saving style, but may lack sprint speed. The technical problem is optimizing between the two, and that optimum shifts by swimmer, by distance, by pool conditions.

That is why my column F matters so much. Water temperature. Humidity. Starting block height. Crowd noise. All of these affect splits, and none appear in the official results table. A spreadsheet with only splits and no column F is a spreadsheet that lies by omission.

In swimming, starting equipment and underwater dolphin kick technique are often underrated variables. In short-course events, the underwater phase after the start and after each turn can account for up to 40% of race time for some swimmers. If you only look at 50m splits, you miss the underwater phase entirely. A coach once told me: "Splits don't measure the dive, so where do we measure the dive?" The answer is: in video, in observer notes, in data that does not sit in the scoreboard.

And here is where I have to say something hard to hear. Sports analytics has a disease: when data is missing, people do not stay silent. They fill the gap with models. Models predicting times. Models predicting potential. Models predicting medals. And those models, built on empty data foundations, are just numbers without gender, without pool, without people.

Numbers have no gender. But the people who read them do. And the people who write them, even more so.

That is why I sat still that night. I could easily have written a 1,500-word piece on Australia's national swimming trials based on memory of familiar athletes. I know who swims 100m freestyle fast. I know who has potential. But if I don't have splits, I have nothing. I only have my belief, and my belief is not data.

There is a probabilistic arrogance I have learned to recognize in myself. After Kazan, I once leaned toward the opposite extreme: dismissing results that fell within predictions, hunting only for exceptions. That too is an illusion. A 99% success still happens hundreds of times a year. The shadow of probability is not the only place worth living. But when data is absent, both extremes — arrogance and pessimism — are fabrication.

The Brisbane lesson is concrete. Before analyzing any swim meet, I check three data layers. Layer one: assertable data — splits, times, records, distances. Layer two: ambiguous data — visually measured stroke rate, manually noted pool conditions, subjectively judged form. Layer three: the sensory zone — where five years underwater, growing up in Vietnam and working in Australia give me an edge to speak without numbers. Those three layers must stay separate. Blending them is the fastest way to produce a beautiful but hollow analysis.

The empty Brisbane spreadsheet was not my failure. It was data. It was evidence that something had broken at the collection layer, and my job was to find the break, not fill it with inspiration.

The next morning, I called my data contact in Melbourne. It turned out a digitization error had wiped the whole team's field notes, and nobody noticed because everyone was using the summary sheet. It took forty minutes to restore from backup. Forty minutes. If I had written that night from memory, I would have published an analysis built on an empty spreadsheet, and none of my readers would have known.

What worries me is not the digitization error. What worries me is how many analysts are willing to keep writing when their column F is empty. When you are forced to deliver daily judgments, the pressure to produce data outweighs the pressure to seek truth. And in swimming — where everything important happens below the surface, where spectators see nothing but ripples — public opinion is easily steered by a number with no source.

I did not write that piece. I wrote this one. And I place my empty spreadsheet here as a testimony.

If you are reading a swimming analysis with no splits, no pool context, no source — ask yourself whether the author is counting, or guessing. When data is absent, honest silence is worth more than any model. And if an analyst cannot say "I don't have enough data," then what they are selling you is not analysis, but belief packaged as numbers.

Numbers have no gender. But an empty spreadsheet has a voice. And sometimes, to stay honest with the craft, you have to let it speak for you.