The Lab ran 14 experiments across 5 research threads this cycle (2026-09-06 → 2026-09-14) on the idle GPUs — every verdict below is published as the data said it, wins, losses and coin-flips alike. No cherry-picking.
Which coins are actually good for grids?
- run of 2026-09-06 — 'bottom (cull?)' leads on mean_grid_pnl: +1.72
Is a simple trend filter better than the grid?
- run of 2026-09-06 — 'trend_follow_50' leads on mean_return_pct: +109
Would pausing grids in a BTC downtrend stop the bleed?
- run of 2026-09-07 — 'BTC-downtrend' leads on total_pnl: -15.80
Confidence-gated trading sim
- majors pack, 5m bars, 3-bar horizon (fresh-data re-run) — NO tradeable subset after 25bps/side fees — best cell (long @ p>0.65) still nets 15.0 bps/trade
- alts pack, 5m bars, 3-bar horizon (fresh-data re-run) — NO tradeable subset after 25bps/side fees — best cell (short @ p>0.65) still nets -37.2 bps/trade
- meme pack, 5m bars, 3-bar horizon (fresh-data re-run) — NO tradeable subset after 25bps/side fees — best cell (long @ p>0.7) still nets -16.0 bps/trade
- broad pack, 5m bars, 3-bar horizon (fresh-data re-run) — NO tradeable subset after 25bps/side fees — best cell (long @ p>0.65) still nets -26.8 bps/trade
- majors pack, 15m bars, 1-bar horizon (fresh-data re-run) — NO tradeable subset after 25bps/side fees — best cell (long @ p>0.65) still nets -21.7 bps/trade
- alts pack, 15m bars, 1-bar horizon (fresh-data re-run) — NO tradeable subset after 25bps/side fees — best cell (short @ p>0.65) still nets -33.3 bps/trade
- meme pack, 15m bars, 1-bar horizon (fresh-data re-run) — NO tradeable subset after 25bps/side fees — best cell (long @ p>0.6) still nets -27.5 bps/trade
- broad pack, 15m bars, 1-bar horizon (fresh-data re-run) — NO tradeable subset after 25bps/side fees — best cell (long @ p>0.65) still nets -34.1 bps/trade
- majors pack, 15m bars, 3-bar horizon (fresh-data re-run) — NO tradeable subset after 25bps/side fees — best cell (long @ p>0.7) still nets -22.4 bps/trade
- alts pack, 15m bars, 3-bar horizon (fresh-data re-run) — NO tradeable subset after 25bps/side fees — best cell (long @ p>0.7) still nets -27.7 bps/trade
Which crypto coins can a simple ML model actually predict?
- run of 2026-09-08 — 'ONDO' leads on auc: +0.529
Auto-published weekly by Gordon Lab. Each entry is a real backtest or model run on real market data; methods and full result tables live in the lab archive. This is research, not financial advice.