The 75% Figure and the Empty Room Behind It: What the FIDE Survey Didn't Say

In November 2026, ten tables were set up on the first floor of the National Library in Kolkata. Three days, nine rapid rounds of Tata Steel Chess India. Viswanathan Anand, Hikaru Nakamura, Wesley So, Levon Aronian—ten chairs, and a hall where the sound of pieces on wood was the loudest thing in the room. I was there shooting footage for a four-part football documentary commissioned by a Kolkata club's title sponsor, paying roughly Rs 4.2 crore a season. The sponsor wanted football; the city gave me chess. I shot eleven hours across those three days. One scene never made the camera.

A young woman player stood by a board, adjusting pieces. A spectator asked, “Whose student are you?” She said, “I am a player.” The spectator laughed. The laugh did not reach the microphone, but it stayed in the air. Seven years later, a secondary report of a FIDE survey carried a number: three in four female chess players have faced gender bias or harassment. The figure turns that laugh into a percentage. The question is what the percentage tells us, and what it leaves out.

FIDE is the global governing body of chess. It commissioned the survey. The report carries the 75% figure. It carries no sample size, no geography, no questionnaire, no fielding date, no named player, no named national federation, no link to the primary FIDE document. The package that delivered the headline also included a website privacy notice among its “key information points.” That does not make the headline false. It tells us the extraction was noisy, and the package was incomplete. In my thirty-seven years on this beat, a survey published without methodology usually reads like the first draft of a policy document, not the final word of journalism.

The 75% is a prevalence metric, not a performance metric. It tells us how many women experienced something; it does not tell us how that experience changed their chess. The first problem is the question itself: “bias or harassment” was aggregated. Bias is a wide category—being called a “women's player,” having your preparation questioned, being assumed to be someone's student. Harassment is a different category; it can include criminal conduct. Merging both into one percentage is rhetorically strong and analytically weak. It cannot tell an organizer whether the problem is culture or safety. Those require different interventions.

The missing denominator is the second problem. If n ≈ 400, the 95% confidence interval on 75% is roughly ±4 percentage points; if n ≈ 100, roughly ±8.5 points. The figure is directionally robust and precision-unknown. The third problem is selection bias. Surveys distributed through federation and platform channels over-sample the most engaged, tournament-active, online-present players—precisely the group most exposed. Prevalence estimates from self-selected online samples typically run high. The fourth problem runs the other way: in a small, tightly-networked professional community, the most severe conduct is the least reported. A complaint can end a career. The 75% may be a floor, not a ceiling.

The structural backdrop matters. Top women's ratings sit around 2550–2600; the world No. 1 sits around 2830+. The gap is 230–280 Elo. Fewer than 50 women have ever held the full GM title—under 2% of all GMs. These are approximate benchmarks, pending verification against primary sources. They point to a pipeline problem. If the pipeline is narrow, the 75% is the experience of a narrow, high-achieving group, not the whole population. That group is also the most audible.

In chess, the most measurable effect of harassment may not show up in moves. It shows up in the tournament entry list. A player knows a known harasser is in the field. She declines the invitation. That withdrawal does not appear in her rating. It does not appear in any ACPL. It is still a performance decision. This is the “event selection” channel—and it is exactly where the survey is thinnest. The report offers no withdrawal rates by gender, no participation continuity across a cycle, no error rates under time pressure, no drop-out rates by age band. Without those, the 75% remains a sentiment metric, not a performance metric.

There is another layer. The survey almost certainly included online experience, because a large share of harassment in chess now occurs via platform chat, direct messages, and social media. If so, FIDE is publishing data on conduct largely outside its own jurisdiction. That is not a weakness of the institution; it is a limit of governance. You can survey a problem you cannot regulate.

Collective memory will remember the 75% as a scandal. My contrarian reading is different: the number is not the story; the empty room after it is. FIDE published a survey. What happened next? Was a reporting channel created? Did any national federation form an independent complaints committee? Will the next survey separate online from over-the-board? Or will 75% become a static fact, quoted in 2030 as if it were measured in 2026, while prevalence and reporting norms both drift? The survey reads like the first draft of a policy package. The most dangerous part of a draft is the part left unwritten.

I wait six weeks for the script the headlines forgot. That habit dates to 2026, when I took a voluntary separation package from a Kolkata daily's sports desk and flew to Chennai to record Manuel Aaron for eleven days. He was India's first International Master, from 2026, and the founder of the Tal Chess Club. At 82, he spoke about Moscow 2026, India's Olympiad debut. The 26-minute script, The First Move, was rejected by four broadcasters and never filmed. The rejection taught me not to open on results. Every treatment now begins with one dated image—a room, a board, a hand—and I keep a silence ledger of three moments per interview I deliberately leave unnarrated.

The same discipline applies here. Silence is not empty; it is the scene where the truth rehearses. The silence behind the 75% sits in three places. The first is the sample size. The second is the online/over-the-board split. The third is the women who did not fill out the survey because they have already left the game. That third group matters most. A player who has exited the system will not appear in any prevalence sample. Her absence is the largest data point of all.

Chess has an old habit of outsourcing every problem to the next generation. Girls will come, ratings will rise, titles will follow, and bias will fade on its own. That habit is born of convenience, not evidence. The 75% says bias will not fade on its own. It needs reporting structures, training, and consequences for organizers who treat complaints as “atmosphere.”

Three things to watch over the next 12–18 months. Whether FIDE publishes its methodology—sample size, questionnaire, geography, date. Whether national federations create independent reporting channels that do not sit under the organizer's authority. Whether the next survey separates online from over-the-board. Without those, the 75% becomes an arrow fired every year at a target no one has drawn.

The tables are still there in that Kolkata hall. In one corner, a girl is setting up pieces. A spectator asks, “Whose student are you?” She says, “I am a player.” The exchange repeats every season. The question is not how large 75% is; the question is which room the other 25% are sitting in—and whether its door is open.