Published prediction scorecard · New York · NUMBERS · draw 20260903:evening
Prediction 4 · 4 · 8 · result 1 · 1 · 6
This comparison uses the immutable snapshot created before this exact draw. The prediction is never rewritten after the result.
Pre-draw prediction
448
2026-09-03 · Evening · 10:30 PM · slot-bayesian-recency-v1
Exact model miss
Verified result
116
0 of 3 main numbers matched · draw 20260903:evening · open result source ↗
Complete line rank
#291 of 1000
Score 0.105314 in the locked ranking.
Complete-line chance: 1 in 1000. Rank #291 is DrawPick's historical ordering; it does not change the lottery odds or guarantee an edge.
Locked evidence
What the model knew before the draw
322 verified earlier draws · 161 earlier Evening · 10:30 PM draws · input through 2026-09-02 · Midday · 2:30 PM · draw 20260902:midday.
Headline method: 55% same-slot smoothed frequency + 30% all-draw smoothed frequency + 15% recency weight
Complete comparison
Top 10 before this result
| Rank | Complete line | Score | Result |
|---|---|---|---|
| #1 | 4 · 4 · 8 | 0.131591 | |
| #2 | 4 · 9 · 8 | 0.129331 | |
| #3 | 4 · 1 · 8 | 0.12871 | |
| #4 | 4 · 8 · 8 | 0.128466 | |
| #5 | 4 · 6 · 8 | 0.127742 | |
| #6 | 6 · 4 · 8 | 0.125951 | |
| #7 | 6 · 9 · 8 | 0.123691 | |
| #8 | 6 · 1 · 8 | 0.12307 | |
| #9 | 4 · 7 · 8 | 0.123008 | |
| #10 | 6 · 8 · 8 | 0.122826 |
Winning line 1 · 1 · 6 ranked #291 before the draw.
Model field
Which model chose what?
Published rows were locked before the draw. Replay rows were calculated afterward from the same locked input and do not count as live predictions.
| Model | Type | Choice | Complete-line rank | Grade | Status |
|---|---|---|---|---|---|
| Balanced slot model 55% same-slot smoothed frequency + 30% all-draw smoothed frequency + 15% recency weight | Statistical blend | 4 · 4 · 8 | #291 | Miss | Published before draw |
| Same-time position frequency Smoothed digit frequency by position for this exact draw time | Frequency | 4 · 8 · 1 | #634 | Miss | Published before draw |
| All-draw position frequency Smoothed digit frequency by position across all stored draw times | Frequency | 0 · 8 · 5 | #638 | Miss | Published before draw |
| Recent position momentum Position-weighted appearances in the 12 most recent earlier draws | Momentum | 6 · 4 · 8 | #116 | Miss | Published before draw |
| Longest position gaps Longest time since each digit last appeared in its position; descriptive only, not a due-number claim | Gap analysis | 3 · 3 · 0 | #829 | Miss | Published before draw |
| Position transition pattern Smoothed position-by-position outcomes that followed the latest digit before this draw slot | Sequence history | 5 · 4 · 9 | #330 | Miss | Published before draw |
| Sequence-shape transition research Tests whether an ordered result shape recurs after earlier results with sum-family 4; exact-number scoring remains independent | Structural pattern research | 4 · 9 · 7 Research comparison · 5 pre-declared structural features | #91 | Pattern miss · 3/5 features | Published before draw |
| Uniform random control Deterministic pre-draw uniform ordering used only as a comparison baseline | Control baseline | 2 · 9 · 7 | #838 | Miss | Published before draw |