Dhruv Verma
← Writing

Oct 17, 2026

Same Board, Different Game: How a Chess Question Became a Research Paper

Six weeks of turning a broad question about women in chess into a focused literature review, including the wrong turns, the corrections, and what I'd do the same way next time.

The Backstory

I play chess recreationally, and one question has bothered me for a while: why are there so few women at the top of the game? Chess has no physical component. Nobody needs to be taller or stronger to calculate a rook endgame. So the gap had to come from somewhere else.

When my writing course at SJSU (CMPE 294) asked for a semester-long research topic, I picked this one. My starting questions were broad. Did the gap come from social realities of earlier eras? Could any physical difference explain it at all?

This post is how that broad curiosity narrowed, over about six weeks, into a focused literature review, and what I learned about research along the way.


Part 1: Finding a Case

The first assignment asked for one scholarly source and one popular source. The scholarly one came quickly: Arnold et al. (2024), a study showing parents and mentors judge girls to have less chess potential than boys, even at the same rating.

The popular source changed the whole project. In August 2026 the BBC reported on the She Plays To Win Lionesses, the first all-female team to win promotion to the top division of the Four Nations Chess League (4NCL). Their reward was a rule. Division 1 teams must include both male and female players, so the Lionesses could not play in the division they had earned without adding a man.

A rule designed to bring women into chess was keeping a women's team out. That contradiction gave the paper a case to stand on.

Lesson 1: Read the rule, not the headline. The BBC headline said the team "must include a male player." The actual rule requires one player of each sex. Later I caught a second version of the same mistake: the 4NCL rulebook never states why the rule exists. That claim came from the league's statement to the BBC. Small distinctions, but a paper built on a misread rule is built on sand.


Part 2: The Question That Answered Itself

My rough-draft research proposal got one piece of peer feedback that stung because it was right: my research question implied its own conclusion.

I had personal views about evolutionary and societal factors behind the gap, and they had leaked into the wording. The fix was not to delete them but to name them. They became a positionality paragraph, framed as hypotheses I would test, not conclusions I would defend.

The final proposal asked how three kinds of rules (mixed-team quotas, women-only events, and women-only titles) shape who plays and what they communicate about women's ability. I also added primary documents: the 4NCL rulebook and the FIDE title regulations, where the Woman Grandmaster title requires a 2300 rating against 2500 for the open Grandmaster title.


Part 3: Mapping Before Writing

The annotated bibliography came next, and my first draft ran three to five times the target length. What finally worked was a two-pass approach: write long to think, then cut each section down to one job. Summary covers question, finding, and conclusion. Credibility leads with the biggest weakness. Value says what the source gives me and how it complicates my thinking.

Then I mapped the sources visually: a time-and-place table, a "constellation" showing how sources connect, and a matrix plotting causes against each source's stance on separate structures.

The map showed something I didn't want to see. My sources answered only half my question. They explained what the rules communicate. None of them measured who plays after a rule takes effect. Nobody had tracked dropout or retention after a quota or a title.

So the question narrowed one last time:

What do the rules built to increase women's participation communicate about women's ability, and does that message match what research says causes the gender gap?

"Who plays" moved into the limitations section, where it honestly belonged.

The mapping also exposed a second problem. Some arrows on the map gave sources stances they never took. I started drawing my own inferences as dashed boxes, separate from what each paper actually claimed. Going back to the PDF of one esports study (Nolla et al.) showed that anxiety was not elevated in women-only tournaments, and that "disengagement" was the authors' inference, not a measurement.

That same study handed me my sharpest argument. Nolla et al. say women-only spaces should recruit and build skill, not permanently segregate. Applied to titles, that standard is hard to meet: a title adds no practice and is held for life.


Part 4: Writing the Review

Reverse-outlining my draft showed where I was summarizing sources one by one instead of synthesizing them. The fix was restoring the debates I had cut for length:

  • Vishkin vs. Napp and Breda on why women's participation varies across countries
  • Stafford vs. Smerdon et al. on whether women play worse against men because of stereotype threat

The debates were the synthesis. Cutting them had made the review shorter and weaker at the same time.

I also added plain-language explanations of ratings, titles, and board order, because a reader who doesn't play chess shouldn't have to. The final draft landed at 2,392 words.

What the research says, briefly: much of the gap at the top follows from how few women play at all (Bilalić et al.). Adult expectations shape girls early (Arnold et al.). National institutions matter, though researchers dispute exactly how. The finding I didn't expect was a gap in the literature itself: studies measure distributions, beliefs, and games, but almost never a rule. Read against that research, some rules, like a lower rating bar for a women's title, communicate an ability difference the evidence attributes mostly to participation and expectations.

And the physical-difference question I started with? I never answered it directly. It turned out not to matter for the question I ended up asking, which is its own kind of answer.


Part 5: Same Research, Different Shapes

Once the review was done, the same material got rebuilt several times:

  • An interactive explainer site, also called Same Board, Different Game, with a raffle simulator showing how pool size alone produces fewer women at the top, title ladders side by side, and a short quiz
  • A figure titled "Rules vs. Research," comparing each rule against what the research says causes the gap
  • A genre translation pairing an Onion-style satirical news story as the hook with the explainer site as the depth

Each version forced me to decide what actually mattered. Satire, it turns out, is a brutal test of whether you understand your own argument.


Lessons Learned

  1. Start with a case, not a topic. "Gender bias in chess" is a topic. An all-female team blocked by a rule meant to help women is a case. Cases give papers a spine.

  2. Read the primary document. Headlines compress. Rules, regulations, and PDFs don't.

  3. A question that implies its answer isn't a question. Name your biases as hypotheses, then let the sources test them.

  4. Map your sources before drafting. The map tells you which question your evidence can actually answer.

  5. Separate what a source says from what you infer. Dashed boxes for your inferences, solid boxes for their claims.

  6. Debates are the synthesis. If every source agrees in your review, you probably cut the interesting parts.

  7. Write long, then cut each section to one job. Concision is a second pass, not a first-draft skill.

  8. AI tools give leads, not sources. I used Claude and NotebookLM heavily for drafting, mapping, and visuals. NotebookLM's infographic got the Woman Grandmaster rating threshold wrong, described a league rule as "legally forbidden," and treated Elo win probabilities as literal win rates. Every correction that mattered came from going back to the original paper.


Closing Notes

The paper I wrote is not the paper I set out to write. I started by asking why women are scarce at the top of chess. I ended up asking what the rules built to fix that say about women, and whether the research agrees. The second question is narrower, but it's one my sources could actually answer.

If you're starting your own research paper: find your case early, read the rule yourself, and let the map of your sources tell you what you can honestly claim.