Research quality
Useful evidence. Visible limits.
Research-quality counts describe the publication process. They do not measure betting profitability.
Source completeness before confidence
Completed reports require sourced facts, unknowns, both paths to victory, an alternative script, evidence and a counterargument. Confidence describes research support, not a numerical chance of winning.
No manufactured precision
Missing statistics or prices stay unavailable. We do not label a last observed price as a closing price without a comparable timestamped close.
Separate research from results
A report is not an official selection. Simulations never enter the official record. Source completeness, stale-source incidents and update success will be reported only when those measurements are actually recorded.
Read the archive →