Public strategy ranking

…

Ranked by the Janorium score: how good the typical backtest of the current version is, discounted by how much evidence stands behind it. One lucky run does not top this table.

#	Strategy	Market	Score	Return / yr	Avg max DD	Sharpe	Profitable	Backtests	Coverage

How the score works

Score = 100 × quality × evidence. Both factors are 0–1, so neither can carry the other.

Quality — how good is the typical run, after its drawdown?

Return / yr: Median annualised return, weight 40; full marks at 50% a year.
Sharpe: Median Sharpe ratio, weight 35; full marks at 2.
Consistency: Share of runs that ended in profit, weight 25.
Drawdown: Multiplies the three above by 1 − average max drawdown: 20% drawdown keeps 80 of the quality.

Evidence — how much stands behind it?

Runs: Distinct setups (symbol × timeframe × period), log-scaled; full at 20. Re-running one setup counts once.
Coverage: Symbols up to 5, timeframes up to 3, years of history up to 4.
Trades: Median trades per run, square-root scaled; full at 100.

Only completed backtests of the strategy's current version that paid trading fees are scored — an earlier version is a different strategy, and a fee-free run is a profit no exchange pays out. Medians, not best runs: the single best return is still shown on the strategy page, but it does not rank.