| Player 1 |
|
#character-encoding UTF-8
#player1 Noah Noah
#player2 Josh Josh
>Noah: ?EENORS 8G pEREONS +64 64
#note 2022 Word Cup R6. Played this pretty fast. OpENERS 8F seems sounder to me now.
>Josh: FOO 7F OOF +14 14
>Noah: AEILNRS E6 ALINERS +71 135
>Josh: IITZ 12A ZITI. +48 62
>Noah: EEEIKMN A12 .INE +39 174
#note Josh might've had a slight perplexed reaction to this play, or maybe I'm making that up. Not sure I thought of ZEIN, which seems reasonable since it has fewer and weaker extensions. Rows 14 and 15 are maybe a bit easier to block after this, but they are also a bit better. (Added after writing the previous: Josh says he was annoyed when I played ZINE because he thought I was not fishing for an extension, but just being reckless.)
>Josh: ACFR 8A FARC. +39 101
>Noah: AEEIKMO 7A OKE +22 196
#note Didn't see SMOKIE M8. Still kind of attached to OKE, but SMOKIE seems good.
>Josh: EQTU B10 QU.TE +74 175
>Noah: AADEEIM I7 A.E +14 210
#note Mostly considered AMIDASE as an alternative, I think, but MEDIA 9H looks like the real other consideration here.
>Josh: ELNOPSY 6I POLEYNS +72 247
>Noah: ADEGIJM O4 MA.JID +57 267
>Josh: EGHINW M8 .HEWING +36 283
>Noah: ABEGTVX N9 AX +57 324
#note A little unsure of this and the next two racks.
>Josh: GIRV C3 VIRG.. +20 303
>Noah: BEEGTTV L10 VEG +38 362
>Josh: ABDW M2 BAWD. +28 331
>Noah: ABERTTU 2J BAR.UT +32 394
>Josh: ADRT 1H TRAD +22 353
>Noah: EHMOTUY 6B O.H.M +38 432
#note Not sure I considered MOUTHY, which might be better, here. It blocks probably the best bingo lines on the board, leaving ones that open me scoring options in the endgame and going more for the blank.
>Josh: NOPS 14J PON.S +28 381
>Noah: ?AETUUY 11K Y.. +18 450
#note Roughly my in-game analysis: "only bingos I can find are COLLIED/COLLIDE and IDIOLECT. A sufficiently good row 15 play beats everything, but the best actual play I can see there is cUTEY, which IDIOLECT edges out. IDIOLECT is 4/36, COLLIDE/COLLIED is 2/36. YEW beats IDIOLECT it I draw the L (UlULATE), or I can block COLLIED/COLLIDE. That's 33/36 for YEW and 32/36 for blocking on row 15. So play YEW." I might have had an inkling that this would also beat IDIOLECT if I drew the I, but I'm not sure. It sort of feels like ... it never occurred to me to block IDIOLECT?? Which seems kind of silly in retrospect. What this may really come down to is inference, which I barely or didn't consider in-game. PONGS seems likely to have kept a three-consonant leave, since the pool from Josh's perspective contains a good number of vowels, and P?NGO plays were possible if he had a vowel. Most three-consonant leaves are possible, with DLT the only one I clearly feel like I can weight lower, due to DIPLONTS. Since DLT can result in IDIOLECT but not COLLIDE, COLLIDE's value should be more likely than it is under an assumption of randomness, but just what are the likelihoods? Well, we'll assume Josh can have CDL, CDT, CLT, CLL, DLL, or LLT, with the first three weighted at 2/9 and the last three at 1/9 due to the duplicate Ls. His chance of drawing a given bingo, with different instances of a tile counting as separate tiles, is 1/6C2 = 1/15. Summing up gives 2/9*2/15 + 1/9*(2/15+2/15) = 8/135 for COLLIDE and 2/9*(2/15+4/15+2/15) = 16/135 for IDIOLECT. Including DLT would've added another 4 in the numerator for IDIOLECT (changing the denominator, but in the same way for both probabilities), so somehow my calculations show that Josh having 3 consonants increases the relative probability of IDIOLECT to COLLIDE ... I don't really understand why. In any case, handling IDIOLECT over COLLIDE still seems like the call. There are other factors I've left out of the analysis, but the final judgment looks pretty safe.
>Josh: L 9I .L +4 385
>Noah: ?AEITUU 3G UvEA +10 460
#note Managed to make a probably good play with not much time left - Champ likes this, UrEA, and Ad(I)EU best.
>Josh: CDEILOT 15F COILED +34 419
>Noah: ITU 10B .UI.T +16 476
>Noah: (T) +2 478
|
