Brief: chess
Research brief for chess · how it was made.
Research brief — chess
| Field | Value |
|---|---|
| Slug | chess |
Primary query / seo_intent | The few checks that stop most club-game losses |
| Template (A–E) | E |
| Evidence tier (1–3) | 3, with named sources above the tier floor |
| Audience (one line) | Club and online players around 1000–1800 who already know the piece names and are tired of opening videos after they hang a knight |
| AI authors credited | ChatGPT-4o (original), Grok 4.7 (rewrite) |
| Reviewer or checker (optional) | none needed this pass; drills are standard and were not checked by a coach |
| Date | 2026-09-22 |
1. Concentration claim (one sentence)
In club chess, losses concentrate into a few position types - a hanging piece, a one-move tactic, a loud position played too fast, an unfamiliar first move, and a rook ending - while opening encyclopedias and equal ending tours fill the week.
2. Hard anchors (2–5)
- Ashton Anderson, Jon Kleinberg, and Sendhil Mullainathan, "Assessing Human Error Against a Benchmark of Perfection," ACM Transactions on Knowledge Discovery from Data 11(4), 2017. DOI 10.1145/3046947. Free text: https://arxiv.org/abs/1606.04956. On a balanced set of 600,000 decisions from solved positions, difficulty features predicted blunders at about 0.75 accuracy; skill features about 55%; time left about 53%. On FICS, for ratings 1200–1800, a 0.2 step in blunder potential moved the blunder rate more than 600 rating points. For players rated about 1500, 8–12 seconds in a safer position beat over a minute in a more dangerous one. Caveat: tablebase positions, mostly spare endgames, not middlegames.
- Genna Sosonko and Paul van der Sterren, New In Chess Yearbook 55, Interchess BV, 2000, p. 227, as reported on Wikipedia's first-move advantage page. Of 731,740 games, 349,855 began 1.e4 and 296,200 began 1.d4. Together that is 646,055 / 731,740, about 88.3%. https://en.wikipedia.org/wiki/First-move_advantage_in_chess
- Karsten Müller and Frank Lamprecht, Fundamental Chess Endings, Gambit, 2001, pp. 11–12 and 304, as reproduced with the table at https://thechessworld.com/articles/endgame/chess-statistics-most-often-occuring-endgames-in-chess/. Rook versus rook 8.45% of games; queen versus queen 1.87% on the same list.
- Endings with rooks and any number of pawns occur in about 8 to 10% of all games, including games that never reach an ending. https://en.wikipedia.org/wiki/Rook_and_pawn_versus_rook_endgame
2b. Field 80/20 examples (required — do not invent)
| Approx % claim | Field / context | Source (name, year, URL) or “adage — hedge” | Where it will sit in the article |
|---|---|---|---|
| 2 of 20 legal first moves (10%) covered about 88% of 731,740 games | Master-leaning database, year 2000, not a 2026 club pool | New In Chess Yearbook 55 (2000), via Wikipedia | Opening section, labeled 80/20 example, with the yearbook caveat |
| Difficulty features ~75% accuracy vs skill ~55% and time ~53% on a balanced blunder set | Solved chess positions | Anderson, Kleinberg, Mullainathan 2017 | Clock / danger section; not forced into 80/20 |
| Rook vs rook 8.45% of games; queen vs queen 1.87% | Endgame frequency list | Müller and Lamprecht 2001 | Endgame section; explicit that this is not an 80/20 |
3. Original observation (only-on-8020 seed)
A ten-game loss tag with five labels only: hanging piece, missed short tactic, rushed loud position, opening hole, rook ending. The next week drills only the winning tag.
How it becomes a section in the article:
"Tag ten games," with one labeled Illustrative row so the format is visible and not mistaken for a study.
4. Ignored majority (named)
Opening encyclopedias past move eight, puzzle streaks that are never reviewed, equal tours of forty ending types, blaming the whole clock, swapping repertoire after one loss.
5. Composite policy
| Scenario | Keep as Illustrative? | Cut instead? |
|---|---|---|
| A club player whose ten tags were 4 hanging, 3 rushed, 2 opening, 1 rook ending | Yes |
6. Vital few (draft list)
- One-second glance for a hanging piece
- Checks, captures, threats before a quiet move
- Extra minute only in a loud position
- One white first move and two black replies, to move eight
- Philidor, Lucena, and one rook-and-pawn idea
7. Device budget reminder
Template E: five bottlenecks, one drill each. One 80/20 example (openings). One 8020 move (the tag). No viz.
7b. Viz & misreads (optional — default is omit)
D3 viz? none
Misreads / edge Qs? Three: more clock, a bigger opening book, equal ending study.
8. Outbound links planned (Tier 1: ≥2 required)
- https://arxiv.org/abs/1606.04956
- https://en.wikipedia.org/wiki/First-move_advantage_in_chess
- https://thechessworld.com/articles/endgame/chess-statistics-most-often-occuring-endgames-in-chess/
- https://en.wikipedia.org/wiki/Rook_and_pawn_versus_rook_endgame
Internal: /decision-making, /concentration, /studying, /board-games
9. Sign-off
- ☑ Brief complete — ready to outline
- ☑ Reviewer/checker named if the topic needs one
- ☑ No fake citations planned