Published prediction scorecard · New York · Take 5 · draw 20260904:midday
Prediction 4 · 18 · 20 · 21 · 26 · result 8 · 9 · 32 · 37 · 39
This comparison uses the immutable snapshot created before this exact draw. The prediction is never rewritten after the result.
Pre-draw prediction
418202126
2026-09-04 · Midday · 2:30 PM · unordered-bayesian-recency-v1
Exact model miss
Verified result
89323739
0 of 5 main numbers matched · draw 20260904:midday · open result source ↗
Complete line rank
Not in locked field
The full official line was not present in this snapshot ranking.
Complete-line chance: 1 in 575757. No complete-line rank was recorded for this result; 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 Midday · 2:30 PM draws · input through 2026-09-02 · Midday · 2:30 PM · draw 20260902:midday.
Headline method: 55% same-draw smoothed number frequency + 30% all-draw frequency + 15% recency; unique combinations only
Complete comparison
Top 10 before this result
| Rank | Complete line | Score | Result |
|---|---|---|---|
| #1 | 4 · 18 · 20 · 21 · 26 | 0.033656 | |
| #2 | 4 · 17 · 18 · 20 · 21 | 0.033562 | |
| #3 | 4 · 17 · 18 · 20 · 26 | 0.033386 | |
| #4 | 4 · 18 · 20 · 21 · 30 | 0.033369 | |
| #5 | 17 · 18 · 20 · 21 · 26 | 0.033334 | |
| #6 | 4 · 18 · 20 · 21 · 29 | 0.033279 | |
| #7 | 4 · 18 · 20 · 26 · 30 | 0.033193 | |
| #8 | 4 · 18 · 20 · 21 · 32 | 0.033189 | |
| #9 | 4 · 15 · 18 · 20 · 21 | 0.033184 | |
| #10 | 18 · 20 · 21 · 26 · 30 | 0.033142 |
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 draw model 55% same-draw smoothed number frequency + 30% all-draw frequency + 15% recency; unique combinations only | Statistical blend | 4 · 18 · 20 · 21 · 26 | Not ranked | Miss | Published before draw |
| Same-draw frequency Smoothed whole-draw frequency for this drawing | Frequency | 9 · 17 · 21 · 26 · 38 | Not ranked | Miss | Published before draw |
| All-draw frequency Smoothed frequency across all earlier draws | Frequency | 13 · 17 · 20 · 21 · 26 | Not ranked | Miss | Published before draw |
| Recent momentum Weighted appearances in the 12 most recent earlier draws | Momentum | 4 · 12 · 18 · 20 · 29 | Not ranked | Miss | Published before draw |
| Draw-shape transition research Tests whether the next unordered field preserves a historically recurring shape after earlier fields with sum-family 9; white-ball shape and bonus-ball evidence stay separate | Structural pattern research | 4 · 9 · 20 · 23 · 35 Research comparison · 5 pre-declared structural features | Not ranked | Pattern miss · 2/5 features | Published before draw |
| Uniform control Deterministic pre-draw uniform comparison | Control baseline | 12 · 15 · 18 · 27 · 28 | Not ranked | Miss | Published before draw |