Open research infrastructure
Research Notebook Lab
Reproduce the published analyses with downloadable Python notebooks, public inputs, and validation checks. Each notebook records its sources and interpretation limits.
Notebook execution route
Market evidence: public rows become reviewable codeLatest share from StatCounter Global Stats and linked public records- Notebooks
- 7
- executable packets
- API routes
- 14
- public JSON inputs
- Default rows
- 2,193
- bounded observations
- Artifacts
- .ipynb
- downloadable source
Reproduce the analysis, not a conclusion.Each notebook keeps its declared dataset, entity universe, period, source custody, and interpretation boundary. Recipes never combine unlike metrics into a composite market-power score.
Download active notebook4 public endpoints / Python 3
Curated market evidence packet
What does the latest public record show about one curated market, and which source rows support it?- Default rows
- 34
- 4 years / 1 source
- Code cells
- 7
- parameterized + bounded
- Source review
- Jul 17, 2026
- StatCounter Global Stats and linked public records
- Output
- Ranking
- Latest participant shares and evidence checks
Executed result contract
Latest participant shares and evidence checks
Input4 public endpoints
Validate4 assertions
AnalyzeLatest share
Publish.ipynb + figure
Parameter ledger
Defaults remain editable in the notebook- Market
- us-search-engine-usage
- Endpoints
- markets + observations + participants + sources
- Default output
- latest participant-share ranking
Research handoff
From public rows to reviewable evidence.
Every notebook keeps the endpoint grammar visible, checks the fields it consumes, produces one bounded figure, and routes the result back to live profiles and methods.
Open public briefings