GFWS21 · Quantitative research
Statistical Edge Intelligence
Compare strategy-episode outcomes, market conditions and adverse price excursions using only saved GFWS21 data.
Extended closed-candle research
Strategy Intelligence 2.0
Existing Performance history is retained. Long path studies will accumulate from archived closed candles as new market sessions complete.
Historical strategy coverage · 5 closed bars
Legacy measured extrema (1% favorable-only definition). Rankings by sample size, never by assumed trading profitability.
| Strategy | Measured / episodes | Favorable only | P90 adverse | Evidence |
|---|
Research observations are not executable trade results. Signals without archived forward bars remain unmeasured; optimizing risk/target settings requires independent holdout or walk-forward validation, fees, spread, slippage, gap risk and sufficiently large comparable cohorts. This module never sends orders.
Thresholds are research barriers, not advised take-profit/stop-loss orders. One instrument/timeframe/strategy/episode is counted once.
Evidence assessment
No research result loaded.
A statistically proven tradable edge is not established by favorable price excursions alone. A trustworthy claim needs target-before-invalidation ordering, trade costs, an explicit benchmark, an out-of-sample period and realistic executions.
Outcomes: favorable only 0 · adverse only 0 · both crossed 0 · neither crossed 0 · pending 0. The favorable-only rate divides by all measured episodes, including ambiguous and unresolved ones.
Chronological 70/30 descriptive split: older — · newer — (0 measured recent episodes). This is not a pre-registered forward test.
Strategy evidence ranking
Ranked by sample size, not by a misleading win-rate leaderboard. Each strategy requires its own independently validated reference case.
| Strategy | Measured / episodes | Favorable only | Wilson 95% interval | P90 adverse | Assessment |
|---|
No captured strategies match these filters.
Performance by market regime
| Condition | Measured / episodes | Target only | Both crossed | P90 adverse |
|---|
No saved observations for this selection.
Performance by market
| Condition | Measured / episodes | Target only | Both crossed | P90 adverse |
|---|
No saved observations for this selection.
Performance by timeframe
| Condition | Measured / episodes | Target only | Both crossed | P90 adverse |
|---|
No saved observations for this selection.
Performance by volume condition
| Condition | Measured / episodes | Target only | Both crossed | P90 adverse |
|---|
No saved observations for this selection.
Dataset coverage and measurement limits
Recorded signals in original database: unavailable. Existing Performance captures: —. Independent strategy episodes in this module: —; repeated alerts consolidated: —.
Existing measured Performance outcomes are currently available only for qualified PREDICTION+ / BINGO signals captured by the existing collector. Older signals without saved forward candles cannot be reliably evaluated without historical market data; this module does not download or reconstruct it.
Each horizon measures price extremes in subsequent completed candles after research capture; favorable and adverse extrema may occur in either order. Incomplete horizons remain pending. A counted strategy episode is not a filled trade, independent portfolio return or proof of profitability.
Source: existing Supabase Performance snapshots, excursion measurements and stored signal quality context. No scanner, strategy, provider, prediction score or existing Performance calculation is modified. Model —.