32 strategies. Top-50 coins. 7-day window. Here’s what happened.
Data as of 2026-08-24 02:57 UTC (the ranking run this post was generated from) — source data (JSON); the strategy counts, table figures and 30-day line below are recomputed from it by scripts/weekly-snapshot.py.
Status column added (2026-10-08). The Status column in both tables was added on 2026-10-08 and shows each strategy’s verdict on that day, not on the day this post went out.
Top 3
Some links in these tables pointed at strategies that were already retired or shelved. This post went out on 2026-08-24, and HV Squeeze shelved 2026-02-15 (190 days before) · Donchian Breakout killed 2026-07-12 (43 days before) · MACD Cross killed 2026-07-12 (43 days before) · Stochastic RSI killed 2026-07-12 (43 days before) · Keltner Squeeze killed 2026-08-15 (9 days before). The weekly tally ran without reading the verdicts. The numbers stay — they are what that week produced and this post owns that record. What changed is where those links go: to each strategy’s journal, where the kill is explained, not to a simulator run. We do not swap in a different strategy; that would erase the week’s record. Dates are each journal’s
dateKilled.
⛔ 3 of this week’s top 3 are retired or shelved. They appear because the backtest engine still scores them on fresh data, but each has been audited and found unsuitable for trading. Links go to their journal pages — read the verdict before running them.
| Rank | Strategy | Status | WR | PF | Return | Trades | Move |
|---|---|---|---|---|---|---|---|
| 1 | MACD Cross LONG | ⛔ Retired (◐ conditional on current data) | 93.2% | ≥99.99‡ | +5.7% | 59† | — |
| 2 | Stochastic RSI LONG | ⛔ Retired (◐ conditional on current data) | 89.8% | 19.32 | +9.7% | 118 | ▲4 |
| 3 | Keltner Squeeze LONG | ⛔ Retired (◐ conditional on current data) | 84.2% | 9.65 | +4.3% | 38† | ▲2 |
This table is the weekly tally as published on 2026-08-24 — and all three of the Top 3 had already been killed by then. MACD Cross killed 2026-07-12 (its 2026-08-14 preset-default re-measure read PF 0.83 over 2,656 trades) · Stochastic RSI killed 2026-07-12 · Keltner Squeeze killed 2026-08-15 — that is 43, 43 and 9 days before this post went out. The weekly tally ran without reading the verdicts. The numbers stay (they are what that week actually produced, and this post owns that record). What changes is where the links go: to each strategy’s journal, where the kill is explained, rather than to a simulator run. If you click a row that says “rank 1”, the place you should land is why we killed it. We do not swap in a different strategy — that would erase the week’s record.
18 out of 32 strategies posted a win rate of 50% or higher this week.
Bottom 3
| Rank | Strategy | Status | WR | PF | Return | Trades | Move |
|---|---|---|---|---|---|---|---|
| 1 | Donchian Breakout | ⛔ Retired (◐ conditional on current data) | 4.2% | 0.02 | -6.4% | 48† | — |
| 2 | HV Squeeze | ⛔ Shelved | 4.0% | 0.04 | -6.2% | 50† | — |
| 3 | Stochastic RSI | ⛔ Retired (◐ conditional on current data) | 12.1% | 0.12 | -10.1% | 116 | — |
The Bottom 3 is the same story — those three (Donchian Breakout, HV Squeeze, Stochastic RSI) are killed or shelved today too. All three carry a date in their journals: Stochastic RSI killed 2026-07-12 · Donchian Breakout killed 2026-07-12 · HV Squeeze shelved 2026-02-15 — 43, 43 and 190 days before this post (each journal owns its verdict). The links point at the journals for the same reason as the table above.
† = fewer than 100 trades in this 7-day window — low sample, interpret with caution (n≥100 recommended). ‡ = profit factor is at the display cap, not a measured value — the underlying figure was higher (possibly the no-loss sentinel). Treat it as “no losing trades in this window,” not as a score.
Monthly Context
Over the past 30 days, Keltner Squeeze LONG 6H holds the top spot with +4.2% return and a 10.82 profit factor. (Keltner Squeeze was retired on 2026-08-15, nine days before this post went out — the journal explains why.)
So What
Rankings shift every week. Past performance doesn’t predict future results — that’s exactly why we re-run every strategy on fresh data. This ranking is a research snapshot: read each strategy’s verdict before using a result — retired and shelved strategies are not trading candidates.