Every major structure should be publishable as a reusable structure card plus model positions and transformations.
Structure card schema
Structure Card
- Structure name
- House aliases / external aliases
- Canonical pawn skeleton
- Side orientation
- Common opening origins
- Static strengths — Side A
- Static weaknesses — Side A
- Static strengths — Side B
- Static weaknesses — Side B
- Thematic breaks — Side A
- Thematic breaks — Side B
- Typical plans — Side A
- Typical plans — Side B
- Typical piece squares
- Common transformations
- Endgame tendencies
- Cross-domain links
- Source notes
Canonical pawn-only diagram
For each family:
- show only pawns;
- use consistent orientation;
- state exact coordinates;
- mark thematic break squares textually;
- avoid implying non-essential wing pawns are mandatory.
Full-piece model position
Required metadata:
- FEN;
- side to move;
- structure family;
- defining pawn coordinates;
- key tension;
- candidate breaks;
- intended plan lesson;
- opening provenance if real;
- legality verification.
Transformation sequence
Use 2–4 boards:
- Before — current skeleton.
- Break — pawn move/contact.
- Exchange — if needed.
- After — new skeleton.
Caption must state:
- removed pawns;
- newly open/semi-open files;
- new passer/weakness;
- whether the original structure name still applies.
Model-game selection
For every major family, prefer a small balanced corpus:
- one ideal plan for Side A;
- one ideal plan for Side B;
- one thematic break;
- one instructive transformation/endgame if not already covered.
Do not use only wins by the side associated with the famous plan.
Historical game metadata
For a real game verify:
- players;
- event;
- date/year;
- exact PGN;
- position FEN;
- move number;
- source.
Do not invent player intention.
Break verification
A thematic break should be labeled at three levels:
Thematic
Known recurring structural operation.
Available
Legal/currently possible.
Correct now
Concrete evaluation proves/recommends it in the example.
Thematic does not mean immediately playable.
Weakness example verification
If teaching weak pawn:
- identify structural reason;
- show whether target is actually accessible;
- show possible defense/advance;
- include anti-dogma counterexample where same feature is useful.
Passed-pawn verification
Programmatically verify pawn-pawn geometry for:
- passed;
- protected;
- connected;
- candidate examples.
candidate passer remains an editorial/strategic judgment and should state the intended exchange mechanism.
Island-count verification
Island exercises should compute groups by occupied adjacent files.
Do not count:
- every pawn;
- doubled pawns as separate islands on same file.
Structure-recognition exercises
Learner-facing prompt should often hide:
- opening name;
- structure name.
Ask:
- describe skeleton;
- identify features;
- identify breaks;
- then match family.
This tests transfer.
Cross-opening exercises
Show same/similar skeleton from two opening move orders.
Required:
- endpoint FENs;
- note whether positions are exactly identical or only structurally similar.
Do not call structurally similar positions a true transposition unless full state matches.
Plan matching
When asking for a typical plan:
- allow multiple plausible plans when structure permits;
- distinguish
thematicfrombest move; - do not mark one slogan as universally correct.
Engine use
Use engine to verify:
- tactical feasibility of illustrative break;
- example move legality;
- no hidden tactical refutation.
Do not use engine evaluation as:
- definition of structure;
- proof that a structural feature is universally good/bad.
Database use
Use opening/game databases for:
- provenance;
- model-game discovery;
- recurring opening origins.
Do not infer:
- structural value directly from frequency.
Source anchors
Core strategic literature should include established pawn-structure references such as:
- Mauricio Flores Rios, Chess Structures: A Grandmaster Guide;
- Andrew Soltis, Pawn Structure Chess;
- Hans Kmoch, Pawn Power in Chess (terminology should be modernized/explained where dated).
Instructional cross-checks can include current structure courses and curated master-game material.
Current instructional cross-check
Chess.com's GM Johan Hellsten course currently presents 30 essential structures, including IQP, hanging pawns, Carlsbad, Exchange Slav, Scheveningen, Dragon, Maróczy, Botvinnik, Sveshnikov, Hedgehog, French, Closed Spanish, Benoni, Benko, King's Indian and Stonewall. This supports the broad recurring-structure corpus without making that course the canonical taxonomy.
Reference: https://www.chess.com/lessons/every-pawn-structure-explained
Accessibility
- break arrows need coordinate text;
- base/head labels in captions;
- do not rely only on color;
- transformation boards numbered;
- pawn coordinates available in alt text.
Final verification checklist
- [ ] Structure skeleton coordinates explicit?
- [ ] House terminology disclosed?
- [ ] Strengths and weaknesses both represented?
- [ ] Plans for both sides?
- [ ] Thematic breaks distinguished from immediate best moves?
- [ ] Transformation endpoint verified?
- [ ] Structure name still valid after transformation?
- [ ] Model game metadata verified?
- [ ] Anti-dogma counterexample included?
- [ ] the evaluation guides evaluation boundary preserved?
- [ ] the strategy guides plan-selection boundary preserved?
- [ ] Does the explanation distinguish structure-specific piece roles from general piece theory?
- [ ] Does the explanation distinguish structural endgame tendencies from exact endgame technique?
Teach the skeleton, then the break, then the transformation—and only then reconnect the pieces.
Weak pawn versus weak square
A structural weakness is sometimes important mainly because of the square it creates.
A backward pawn can be tolerable if the opponent cannot occupy the square in front. A doubled pawn can be acceptable if it controls key entry squares.
Always evaluate the pawn and the surrounding square complex together.
Space versus overextension
Advanced pawns create space when they restrict the opponent while remaining supportable.
They become overextended when the territory cannot be maintained and the weaknesses behind the pawns become exploitable.
The difference is support + counterbreaks, not how far the pawns have moved.