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 · Experimental
V8 Backtesting
Weights optimised on the training split, frozen, then evaluated once on an unseen holdout test split.
Experimental model — unvalidated
In testing. No result here is confirmed out-of-sample. Holdout evaluation is retrospective research, not a pre-draw audited result.
V8 Weighted Ensemble — Holdout Evaluation
GET /api/v8/results — train / validation / held-out test split, 9 optimised signal weights
Interpretation
V8 result: 0/4 significant on holdout. BT overfit confirmed — val 44.4% collapsed to test 32.6%.
A validation-to-test collapse of this size is the expected signature of weight optimisation fitting noise. No draw type reached p<0.05, so no engine is promoted.