About

This site publishes two things: eval-scored strategies — a strategy the desk already tested, scored against a prop firm's published evaluation rules, with the rule set quoted and sourced — and death reports on strategies that never got that far. A death report documents a line that died before it ever had a profit-and-loss series to score, which means there is no eval result to show and the failure itself is the finding.

Who runs this. Dan Cobb — U.S. Army veteran, MBA, runs a solo research desk that tests trading strategies and publishes the ones that die.

What "tested" means here. Every post is generated from a row in a run ledger. The row records the spec that was frozen before the run, the commit the code was at, the data window, and the verdict. If a run was made without a pre-registered spec, or on a dirty working tree, the post says so above its own results — the generator prints that block automatically and a post cannot switch it off.

What this site is not. It is not a signal service, not a track record, and not advice. The strategies here have mostly been rejected by the desk. A rejection is a result, and a rejected strategy scored against a public rule set is a more useful worked example than a winner would be, because nobody is tempted to trade it.

Numbers are shown, not chosen. Where a rule has two defensible readings, both are reported. Where a venue does not publish something the result depends on, the unknown is named and the conservative branch leads.