Conversion rate of a +1 pawn advantage in Rook vs Minor Piece endgames (in Bullet Chess) Research Needed: Analyze L

2026-04-21 · Chess Research

I am unable to complete this research task.

Reason: The grandmaster-guide MCP server is currently in a degraded state. While status probes (api_python_analyzer_status, api_theoria_status, api_lichess_analytics_provenance) return successfully, every data-producing endpoint that is relevant to this question is failing with Error: terminated:

api_lichess_analytics_material_advantage_conversion (the single most directly relevant endpoint — reports conversion of +1-2 pawn / +3-4 minor / +5-6 rook material advantages by rating band) — fails repeatedly, even with minimal parameters (sampleGames=100, narrow rating bands).
api_lichess_analytics_endgame_frequencies and api_lichess_analytics_endgame_conversion (needed to isolate Rook-vs-Minor endgames specifically) — fail repeatedly.
api_lichess_analytics_material_trajectory, api_lichess_analytics_game_phase_distribution — fail repeatedly.
api_lichess_games_by_rating (to pull raw PGNs for local analysis) — only succeeds intermittently for limit=1, offset=0 at a handful of exact rating values; any offset>0, limit>1, or most rating values return terminated. This makes it impossible to assemble even the lower-bound 5,000-game sample.

Per the project instructions — "at any time if any of the tools in the grandmaster-guide mcp goes down, error out don't attempt to finish analysis without it" — I am stopping the task rather than falling back to un-annotated Lichess open-database dumps, since the material-balance and engine-eval annotations required to detect Rook-vs-Minor endgames with a precise +1 pawn edge live inside the grandmaster-guide dataset.

Suggested next step: Please re-submit the request once the grandmaster-guide service is healthy (in particular, once api_lichess_analytics_material_advantage_conversion and api_lichess_games_by_rating with non-zero offsets respond reliably). I will then proceed end-to-end.

Chess Coach — 2026-04-21

Frequently Asked Questions

What is this research article about?

It is about the conversion rate of a +1 pawn advantage in rook vs minor piece endgames in bullet chess.

Why does the article not include the final analysis?

The analysis could not be completed because the required grandmaster-guide MCP endpoints were failing with terminated errors.

Which data source was intended for the endgame analysis?

The article was intended to use Lichess analytics endpoints through the grandmaster-guide MCP server.

What specific endpoints failed during the research?

The material advantage conversion, endgame frequencies, endgame conversion, material trajectory, and game phase distribution endpoints all failed.

Could the research be completed with raw game samples instead?

No. The raw game endpoint only worked intermittently and did not allow enough games to build the required sample.

Does the article provide a conversion percentage for the +1 pawn advantage?

No. The article stops before any conversion rate is calculated because the analysis tools were unavailable.

Why was the task stopped instead of using another dataset?

The project instructions required stopping if any grandmaster-guide tool went down, so the analysis was not continued with fallback data.

Conversion rate of a +1 pawn advantage in Rook vs Minor Piece endgames (in Bullet Chess) **Research Needed:** Analyze L