The complete series · 20 essays
Chess, scaffolded.
Follow the project in research order: surprise, diagnosis, instrument design, and the operating discipline required to make a result real.
01The anomaly02Diagnosis03The instrument04Research discipline05Laboratory operations
01
The anomaly
Two Moves Are Better Than Ten
Frozen evidence8 min
↗02The anomaly
Full Games, Not Chess Puzzles
Frozen evidence7 min
↗03The anomaly
Elo Is a Property of a Protocol
Frozen evidence9 min
↗04The anomaly
The Candidate-Set Paradox
Frozen evidence9 min
↗05Diagnosis
Is the Model a Searcher, a Critic, or a Policy?
Protocol / design9 min
↗06Diagnosis
Where Does the Game Fall Apart?
Frozen evidence8 min
↗07Diagnosis
The Price of Thinking
Frozen evidence12 min
↗08The instrument
Building a Chess-Native Reasoning Architecture
Protocol / design21 min
↗09The instrument
Candidate Curves as Model Fingerprints
Frozen evidence11 min
↗10The instrument
An Open Benchmark for Tool-Mediated Chess
Protocol / design16 min
↗11Research discipline
The Experiment We Refused to Run
Protocol / design8 min
↗12Research discipline
Why Not Skip Straight to Higher Elo?
Protocol / design9 min
↗13Research discipline
Ideas or Vision? Why the Score Alone Cannot Tell
Protocol / design10 min
↗14Laboratory operations
Five Workers, One Clock: Why LLM Chess Runs Feel Slow
Provisional9 min
↗15Laboratory operations
The Result That Refused to Exist
Provisional11 min
↗16Laboratory operations
The Tool Call That Looked Like Causality
Frozen evidence12 min
↗17Laboratory operations
The Control Arm Was Full of Chess
Protocol / design8 min
↗18Laboratory operations
The Analysis That Existed Before the Results
Protocol / design8 min
↗19Laboratory operations
The Bytes We Could Prove
Protocol / design6 min
↗20Laboratory operations
The Request That Failed Before the Network
Protocol / design7 min
↗