Ledger tables (Appendix D V.D; Chapter 18 "Whether the alignment is design or coincidence")
letter-group, 13 cases (0,9,4,0)                               p = 0.0014   (1 in 715)
letter-group + 3 alif-free openings, 16 cases (0,11,4,1)       p = 0.00275   (1 in 364)
letter-group, yaa pair as holds, 15 cases (0,9,6,0)            p = 0.0002   (1 in 5,005)
letter-group + alif-free, pair as holds, 18 cases (0,11,6,1)   p = 0.000377   (1 in 2,652)
formula-only, 13 cases (0,3,4,6)                               p = 0.294   (1 in 3.4)
formula-only, pair as holds (0,3,6,6)                          p = 0.185   (1 in 5.4)
formula-only + 2 sign controls (0,5,4,6)                       p = 0.154   (1 in 6.5)
formula-only + controls, pair as holds (0,5,6,6)               p = 0.0747   (1 in 13.4)

Frozen frame (Appendix D III.c, V.E): 14 openings, rate 1/19
five of fourteen hold (pending yaa pair counted)               p = 0.000541   (1 in 1,848)
three of fourteen hold (counted strictly)                      p = 0.0343   (1 in 29)
alignment, strict: sign 0/6 vs non-sign 3/8 (letter-group)     p = 0.154   (1 in 6.5)
alignment, strict: sign 0/5 vs non-sign 3/9 (formula-only)     p = 0.231   (1 in 4.3)
alignment, pair counted: sign 0/6 vs non-sign 5/3 (letter-group) p = 0.028   (1 in 36)
alignment, pair counted: sign 0/5 vs non-sign 5/4 (formula-only) p = 0.0629   (1 in 15.9)

Historical frame, V.A trials (calibrated per-trial probabilities as stated in Appendix D V.A)
three hard-tier trials (headline)                              p = 0.000124   (1 in 8,091)
three hard-tier trials, flat 1/19                              p = 0.000146   (1 in 6,859)
four trials incl. Sad as recited (sensitivity)                 p = 6.55e-06   (1 in 152,658)
four trials, flat 1/19                                         p = 7.67e-06   (1 in 130,321)
five trials with the pending yaa pair, flat 1/19               p = 4.04e-07   (1 in 2,476,099)