PatternOS v4 - Hypothesis & Combination Prediction Lab

Draws analyzed: 888

Important: This module ranks full combinations and challenges hypotheses. It is still experimental and cannot defeat a fair random lottery.
New in v4: whole-column scoring, why-this-column explanations, hypothesis discovery, challenge/backtest table and engine competition summary.

Prediction Lab - Top Candidate Columns

rank numbers total_score confidence why joker note
1 8-13-29-36-44 63.36 LOW fast rank blend; sum 130; odd/even 2/3 7 Fast v10.1 Full Analysis ranking. Use Backtesting/Hypothesis for deeper testing.
2 4-18-29-36-44 59.92 LOW fast rank blend; sum 131; odd/even 1/4 18 Fast v10.1 Full Analysis ranking. Use Backtesting/Hypothesis for deeper testing.
3 5-8-13-31-42 61.47 LOW fast rank blend; sum 99; odd/even 3/2 8 Fast v10.1 Full Analysis ranking. Use Backtesting/Hypothesis for deeper testing.
4 3-12-29-37-42 61.99 LOW fast rank blend; sum 123; odd/even 3/2 1 Fast v10.1 Full Analysis ranking. Use Backtesting/Hypothesis for deeper testing.
5 1-4-11-24-44 48.51 LOW fast rank blend; sum 84; odd/even 2/3 20 Fast v10.1 Full Analysis ranking. Use Backtesting/Hypothesis for deeper testing.
6 9-13-18-33-34 60.51 LOW fast rank blend; sum 107; odd/even 3/2 14 Fast v10.1 Full Analysis ranking. Use Backtesting/Hypothesis for deeper testing.
7 5-16-26-27-36 60.70 LOW fast rank blend; sum 110; odd/even 2/3 3 Fast v10.1 Full Analysis ranking. Use Backtesting/Hypothesis for deeper testing.
8 14-17-28-41-42 47.83 LOW fast rank blend; sum 142; odd/even 2/3 15 Fast v10.1 Full Analysis ranking. Use Backtesting/Hypothesis for deeper testing.
9 1-27-29-30-31 59.13 LOW fast rank blend; sum 118; odd/even 4/1 10 Fast v10.1 Full Analysis ranking. Use Backtesting/Hypothesis for deeper testing.
10 8-18-28-37-40 53.87 LOW fast rank blend; sum 131; odd/even 1/4 19 Fast v10.1 Full Analysis ranking. Use Backtesting/Hypothesis for deeper testing.
11 5-9-11-23-44 49.39 LOW fast rank blend; sum 92; odd/even 4/1 13 Fast v10.1 Full Analysis ranking. Use Backtesting/Hypothesis for deeper testing.
12 6-12-13-14-26 35.91 LOW fast rank blend; sum 71; odd/even 1/4 4 Fast v10.1 Full Analysis ranking. Use Backtesting/Hypothesis for deeper testing.
13 20-22-24-33-36 49.85 LOW fast rank blend; sum 135; odd/even 1/4 16 Fast v10.1 Full Analysis ranking. Use Backtesting/Hypothesis for deeper testing.
14 7-15-16-41-42 57.61 LOW fast rank blend; sum 121; odd/even 3/2 6 Fast v10.1 Full Analysis ranking. Use Backtesting/Hypothesis for deeper testing.
15 3-4-17-34-43 54.25 LOW fast rank blend; sum 101; odd/even 3/2 9 Fast v10.1 Full Analysis ranking. Use Backtesting/Hypothesis for deeper testing.
16 1-27-30-31-32 56.51 LOW fast rank blend; sum 121; odd/even 3/2 5 Fast v10.1 Full Analysis ranking. Use Backtesting/Hypothesis for deeper testing.
17 2-18-28-37-40 52.88 LOW fast rank blend; sum 125; odd/even 1/4 12 Fast v10.1 Full Analysis ranking. Use Backtesting/Hypothesis for deeper testing.
18 5-9-11-21-23 29.87 LOW fast rank blend; sum 69; odd/even 5/0 17 Fast v10.1 Full Analysis ranking. Use Backtesting/Hypothesis for deeper testing.
19 6-12-14-26-45 51.18 LOW fast rank blend; sum 103; odd/even 1/4 11 Fast v10.1 Full Analysis ranking. Use Backtesting/Hypothesis for deeper testing.
20 10-20-22-24-33 52.64 LOW fast rank blend; sum 109; odd/even 1/4 2 Fast v10.1 Full Analysis ranking. Use Backtesting/Hypothesis for deeper testing.

Discovered Hypotheses

hypothesis evidence challenge status details
Number 8 is in an active return window after a relatively long gap. 0.990 Small sample / fair draws remain independent; gap pressure can be random noise. Candidate current_gap=32, avg_gap=8.163, gap_pressure=3.92
Pair 4-27 has a possible cycle/return pressure signal. 0.990 Pair cycles are sparse; repeated pairs in 78 draws can easily be coincidental. Candidate frequency=20, avg_cycle=41.42, last_seen_gap=83, pressure=4.01
Pair 16-22 has a possible cycle/return pressure signal. 0.990 Pair cycles are sparse; repeated pairs in 78 draws can easily be coincidental. Candidate frequency=18, avg_cycle=37.77, last_seen_gap=155, pressure=8.21
Pair 13-36 has a possible cycle/return pressure signal. 0.990 Pair cycles are sparse; repeated pairs in 78 draws can easily be coincidental. Candidate frequency=18, avg_cycle=44.94, last_seen_gap=92, pressure=4.09
Number 43 is in an active return window after a relatively long gap. 0.879 Small sample / fair draws remain independent; gap pressure can be random noise. Candidate current_gap=29, avg_gap=11.0, gap_pressure=2.636
Pair 33-44 has a possible cycle/return pressure signal. 0.771 Pair cycles are sparse; repeated pairs in 78 draws can easily be coincidental. Candidate frequency=19, avg_cycle=43.56, last_seen_gap=42, pressure=1.93
Number 29 is in an active return window after a relatively long gap. 0.769 Small sample / fair draws remain independent; gap pressure can be random noise. Candidate current_gap=17, avg_gap=7.364, gap_pressure=2.308
Pair 13-44 has a possible cycle/return pressure signal. 0.758 Pair cycles are sparse; repeated pairs in 78 draws can easily be coincidental. Candidate frequency=19, avg_cycle=40.11, last_seen_gap=38, pressure=1.89
Number 18 is in an active return window after a relatively long gap. 0.748 Small sample / fair draws remain independent; gap pressure can be random noise. Candidate current_gap=20, avg_gap=8.907, gap_pressure=2.245
Number 13 is in an active return window after a relatively long gap. 0.644 Small sample / fair draws remain independent; gap pressure can be random noise. Candidate current_gap=16, avg_gap=8.279, gap_pressure=1.933
Number 1 is in an active return window after a relatively long gap. 0.564 Small sample / fair draws remain independent; gap pressure can be random noise. Candidate current_gap=16, avg_gap=9.451, gap_pressure=1.693
Number 12 is in an active return window after a relatively long gap. 0.560 Small sample / fair draws remain independent; gap pressure can be random noise. Candidate current_gap=15, avg_gap=8.928, gap_pressure=1.68
After 4, number 13 appears unusually often after lag 5. 0.442 Follow relations are vulnerable to multiple-testing false positives. Candidate cases=111, hits=29, hit_rate=26.13%, lift=2.21
After 18, number 6 appears unusually often after lag 2. 0.428 Follow relations are vulnerable to multiple-testing false positives. Candidate cases=98, hits=21, hit_rate=21.43%, lift=2.138
After 39, number 25 appears unusually often after lag 5. 0.423 Follow relations are vulnerable to multiple-testing false positives. Candidate cases=84, hits=17, hit_rate=20.24%, lift=2.114
After 32, number 13 appears unusually often after lag 4. 0.418 Follow relations are vulnerable to multiple-testing false positives. Candidate cases=93, hits=23, hit_rate=24.73%, lift=2.092
After 31, number 2 appears unusually often after lag 2. 0.408 Follow relations are vulnerable to multiple-testing false positives. Candidate cases=103, hits=22, hit_rate=21.36%, lift=2.039
Number 31 is in an active return window after a relatively long gap. 0.549 Small sample / fair draws remain independent; gap pressure can be random noise. Weak current_gap=14, avg_gap=8.5, gap_pressure=1.647
After 43, number 26 appears unusually often after lag 5. 0.394 Follow relations are vulnerable to multiple-testing false positives. Weak cases=79, hits=17, hit_rate=21.52%, lift=1.97
After 40, number 4 appears unusually often after lag 5. 0.393 Follow relations are vulnerable to multiple-testing false positives. Weak cases=96, hits=24, hit_rate=25.0%, lift=1.965
After 38, number 9 appears unusually often after lag 4. 0.385 Follow relations are vulnerable to multiple-testing false positives. Weak cases=83, hits=18, hit_rate=21.69%, lift=1.926
Pair 34-42 has a possible cycle/return pressure signal. 0.187 Pair cycles are sparse; repeated pairs in 78 draws can easily be coincidental. Weak frequency=19, avg_cycle=47.17, last_seen_gap=11, pressure=0.47
Pair 24-28 has a possible cycle/return pressure signal. 0.124 Pair cycles are sparse; repeated pairs in 78 draws can easily be coincidental. Weak frequency=18, avg_cycle=51.53, last_seen_gap=8, pressure=0.31
Pair 21-26 has a possible cycle/return pressure signal. 0.098 Pair cycles are sparse; repeated pairs in 78 draws can easily be coincidental. Weak frequency=18, avg_cycle=48.88, last_seen_gap=6, pressure=0.25

Challenge Backtest

test_draw actual best_hits avg_hits_top best_column random_expected_hits result_note
3114 15-16-17-33-37 1 0.667 4-13-29-34-37 0.556 Fast bounded smoke test
3115 4-11-23-28-40 1 0.500 4-18-19-29-37 0.556 Fast bounded smoke test
3116 4-7-9-19-40 2 0.667 4-13-16-19-37 0.556 Fast bounded smoke test
3117 3-11-16-34-37 2 1.000 3-5-33-34-44 0.556 Fast bounded smoke test
3118 10-22-24-30-35 1 0.333 8-13-18-24-42 0.556 Fast bounded smoke test

Engine Competition Summary

metric value
tests 5
avg_hits_top_columns 0.633
random_expected_hits_per_column 0.556
best_hit_2plus_draws 2
confidence LOW - fast bounded v10.1 smoke test