| Player 1 |
|
#character-encoding UTF-8
#player1 Noah Noah
#player2 Carl Carl
>Noah: ADFINPS 8G DIF +14 14
#note This is very close on the sim with FIND. I also considered FINDS, which sims way worse. I analyzed the first few turns of this game (~3 per side) with NWL20, even though we played the game using NWL18.
>Carl: DEEHLRT 7H HE +17 17
>Noah: AHMNNPS 9F NAM +18 32
#note I also considered HM J6 and NAP/NAPS 9F. I didn't like setting up the S hook on column E, but maybe it's not so bad. These plays are all simming within a few points of each other, with HM behind the other plays.
>Carl: ADEGLRT 10E DAG +16 33
>Noah: AHILNPS J1 PLANISH +75 107
>Carl: EILORTT 2D TORTEL.I +60 93
>Noah: CEIORUU 4G COE.URI +20 127
#note I thought for a bit here.
>Carl: BEEKOSW 1A WOKE +39 132
>Noah: AIOPRUV 1J .AVIOR +47 174
>Carl: BEEOOST 11F BOOTEES +79 211
#note I didn't notice in-game that Carl picked a lower-scoring, but perhaps more defensive, placement of BOOTEES here.
>Noah: EEELPSU 1G UP +26 200
#note 3E PEE is simming a little better here.
>Carl: ?AADFQU 3M QUA +46 257
>Noah: EEELMNS 5I L.MES +38 238
#note This wins the sim by a bit. I had trouble deciding between this and EM 10I. I also considered LIME, which sims second best. MENSE seems decent, and sims third best.
>Carl: ?ADFLRV 12H FARD +28 285
>Noah: CEEIINO O3 .EONIC +27 265
>Carl: ?LNNVYZ J10 N..VY +21 306
#note Carl also considered NONYL. NONYL keeps some (minor?) Z potential on rows 4 and 10, whereas the Z is perhaps worse after NERVY, though LNZ? seems nicer for bingos down the line than VZ? does. NONYL sims a bit better than NERVY, with the Details numbers for my 4-ply sim mostly favoring NERVY except for the post-rollout numbers, which favor NONYL, so that its ultimate sim win seems to be a result of those numbers along with NONYL's greater score. We might trust the sim more if the post-rollout numbers are more due to Carl more gradually playing off bad tiles after NONYL than after NERVY. However, given that the other details favor NERVY over NONYL, it's possible that the board is somehow better adjusted to LNZ? than VZ?, in which case we should trust the post-rollout numbers less. And there is, of course, the consistent possibility of error due to the use of Quackle's static evaluator in the rollouts.
>Noah: AEEITTX 2A EX +41 306
>Carl: ?BLNOTZ 14F ZLOT. +37 343
#note I think Carl also considered NOB and liked that over BOT to keep the T for OFT.
>Noah: AEIITTY 15G AYE +37 343
>Carl: ??BNNSU L9 BU.S +30 373
#note iirc, after the game Carl seemed a little sheepish about having made this play, but I agreed that it looked good to score well and keep bingo tiles here. Thinking about it more after the game, I wonder if BUNNieS and the like have merit - my options look really weak after those bingos, but then again, they look weak even after BUSS. Generating choices from a pool of AADEEGGIIIJRRTTW post-BUSS shows no bingos, and a high score of 41 for JEW/JAW 6F. Maybe those plays give me some winning chances, though? BUNNIES seems to as well - BEWIGGED yields at least one endgame which Champ says Carl loses. BUSS might be better for spread, because it gives me more hope than BUNNieS in many positions that carry little. And perhaps I could infer a weaker leave, like EIRT, from Carl's previous plays. Still, if J?W plays are strong enough, they could be a problem for BUSS - I'm not sure how strong they are here.
>Noah: AAIIRTT M8 AI +8 351
#note Going for ROTATI?? words and, of course stuff on column N. As it happened, I was drawing dead. A sim much prefers AIT, both for my winning percentage and my bingo percentage. I don't know if the sim just doesn't have Carl blocking column N often enough. If AIT even does a better job of hitting bingos on the left side of the board, it looks like the play - maybe even if it doesn't.
>Carl: ??DENNR E1 N.RmaNDE +76 449
>Noah: AEGIRTT 6D E.G +8 359
#note I hadn't noticed WHIM yet when I played this. Champ likes this and NAG best. I didn't do much calculation here, just kinda felt this was good or felt like playing it. I think I often don't know how to calculate in a way that's very helpful - the calculations I do mostly seem fast and / or intuitive / opaque to me exactly what they are doing.
>Carl: GIJW 9L ..G +6 455
#note Champ likes GI best, by 7 over BIG. TRAIT doesn't seem to be a concern, since I can score a lot more by going slower.
>Noah: AIRTT 3L A... +16 375
>Carl: IJW H6 W... +12 467
>Noah: IRTT 8A TITR. +18 393
>Noah: (IJ) +18 411
|
