UK49s results are random lottery outcomes. Research models rank historical patterns and do not establish a reliable ability to predict future winning numbers. Experimental signals may disappear out of sample. 48 tests run. 0 validated edges found above baseline.
V8 Research · Prospective audit
V3 Engine Validation Trail
Frozen Candidate Engine v3 architecture and its pre-draw records scored after each draw.
Out-of-sample audit
Pre-draw records only, scored after the draw. No post-draw fitting. Rows are written before each draw and scored afterwards. Nothing is refitted after the fact.
Candidate Engine v3 — Frozen 2026-09-06
Signal, conditioning source and weight per draw type
| Draw | Signal | Source | Weight | Status |
|---|---|---|---|---|
| 🍳 Brunchtime | Trend + Hot balls | None | 0.00 | ✅ Confirmed |
| ☀️ Lunchtime | Pair→N lag=1 | BT same day | 0.40 | ✅ Confirmed |
| 🚗 Drivetime | Pair→N lag=1 | LT same day | 0.35 | ⚠️ Provisional (122 draws) |
| 🍵 Teatime | N→N lag=2 | LT same day | 0.35 | ✅ Confirmed |
Research closed. No architecture changes until 90+ prospective draws per draw type.
Historical walk-forward baseline (reference)
Fixed reference values from the completed walk-forward study — never recomputed here
| Draw | Baseline | Best signal | Backtest lift |
|---|---|---|---|
| BT | 61.3% | Trend-only | — |
| LT | 55.4% | Pair→N lag=1 | +3.8pp |
| DT | 64.8% | Pair→N lag=1 | +3.3pp |
| TT | 56.3% | N→N lag=2 | +3.3pp (last 250) |
V3 prospective validation table
One row per draw per date, most recent first
Running v3 performance
Live Any6 rate from resolved prospective rows only, against the fixed historical baseline
Rates stay marked unvalidated until 90+ prospective draws per draw type are recorded. Baseline percentages are fixed historical references, not live recomputations.
UK49s results are random lottery outcomes. Research models rank historical patterns and do not establish a reliable ability to predict future winning numbers. Experimental signals may disappear out of sample. 48 tests run. 0 validated edges found above baseline.
Prospective rows are written before each draw and scored afterwards. A rate above baseline in a small sample is not evidence of an edge.