| Player 1 |
|
#player1 Rafi Rafi
#player2 Mike Mike
>Rafi: AFIOSTU 8G FOU +12 12
>Mike: CGHLQTY -GHQTY +0 0
#note It's worth playing through with LYCH J8. Just didn't see it.
>Rafi: ?AGINST 9F cASTING +70 82
#note The long-term positional advantages of CASTING seemed pretty clear here, as a play hooking UT without a front hook will wreck half the board pretty thoroughly. The only problem is the immediate S hook, but I think that's a pretty small risk considering the pressure I put on Mike if he doesn't bingo pretty quickly.
>Mike: CILNOOW L9 .LOW +16 16
#note Though the sims don't hammer this too hard (to the tune of 1.5-2.5% below the best play of COWLING L3), this is probably a game-losing play. It's clear from Rafi's choice of bingos that he is deliberately pursuing a closed board. While it's true that GLOW forks the board a little bit (on the theory that it's hard to use the 8K-8O spot without opening something else up), COWLING does this much better by making the S-hook on CASTING nearly impossible to block. COWLING is absolutely the play here.
>Rafi: AAEIITU 8K EAU +16 98
>Mike: ?BCINOY K11 YOB +31 47
>Rafi: ADIINTU M10 AUNT +22 120
#note Thought about UNAI, but I especially like keeping it closed against Mike, who has strong word knowledge. A bingo on the triple completely turns the tide, while a bingo hooking YOB scores much less and I'm likely to hit a decent play off the bingo. Only real liability is the X, which Mike didn't have (BOXY). 4-ply prefers UNAI, many-ply prefers AUNT.
>Mike: ?CEEIJN I7 J..E +19 66
>Rafi: BDIIKOU F9 .UBOID +14 134
#note UDO/JO is interesting, setting up my K. Didn't think of that possibility over the board though. My reasoning on CUBOID was that it blocks the S-hook on JUTE, and while it does open up an S hook in the triple, a bingo ending in S will no longer fit. If Mike uses an S for 30-40 points, I'm still strongly in the driver's seat. Quack prefers UDO by a tad.
>Mike: ?ACEEIN 14B EChI.NAE +70 136
#note Normally, when two bingos score equally but one opens a triple-triple line, the other bingo will sim 3-4% better. Here, it's EnNEADIC 14A that wins the sim by that margin. I think this is less about the triple-triple and more about the 15H hotspot being made much harder to use. I probably do want to play ENNEADIC, but I'll listen to arguments on both sides.
>Rafi: EIIKNSW 15F SEWN +49 183
#note Played SEWN over SEW because of a 20v 36c pool. While that decision was right, I flat out missed the far superior WEAK.
>Mike: DDEGHOV B10 HEDG.D +20 156
#note I can understand the rationale behind the sim winner, HOOVED 12D. But is it really 8-12% worse? I know the lines around the H are easy to shut down (any play forming EH does it), but still, that's an awfully large margin.
>Rafi: DHIIKMP N9 HM +24 207
#note Can't do much about the triple spot... but HM sets up PIKI (and KIP, which I didn't see at the time) and PIKA/PIKED fairly often with almost a quarter of the bag being A's and E's. My other thought was KID setting up PHI, but with 5 A's left I didn't think it would stay available. Thinking about it more, I probably should have stayed with the closed theme... HIMP is pretty flexible so even if PHI is blocked I can still score pretty well on this board. DIM might be the play though... KAPH and KOPH for 55 can happen!
>Mike: OOQRTVV -OOQVV +0 156
#note Take your pick of exchanges. Elise is suggesting TORO A7, which makes no sense whatsoever.
>Rafi: DIIKNPT O8 KIP +40 247
>Mike: IIILRRT 11D LI.RI +14 170
#note 3-ply winprob simulation (1210), -115.38 / 5.3% [31.42s]
>Rafi: DEINNOT A5 INTONED +82 329
>Mike: IOPRSTT B2 PORT +14 184
#note At this point, tournament considerations come into play. Going into this game, I was in third place, one game and 240 spread points behind Rafi (the leader). I'm not playing him in round 18, so I need to make sure that Rafi doesn't get so far ahead on spread that, if I'm playing him in round 19 and I'm two games behind, the spread difference is manageable. I'll take this one last shot to try to win the game, but after this, it's about spread preservation.
>Rafi: AAEEMOV 1A EAVE +31 360
>Mike: AGIQSTT E13 Q. +22 206
#note 3-ply winprob simulation (2540), -192.02 / 0.1% [47.29s]
>Rafi: AIMORRR C1 .AR +20 380
>Mike: AGISTTZ D3 TAG +11 217
#note 3-ply spread simulation (880), -176.99 / 0.1% [93.72s]
>Rafi: ILMORRS E4 MOR +20 400
>Mike: EEISTXZ D8 EXI.E +26 243
#note Flat-out missed XIS 7M. This spread-preservation plan isn't working out so well.
>Rafi: AILRRSY 14M YA +18 418
#note Maybe not best, but certainly mistake inducing.
>Mike: ACOSTVZ K5 COV.. +20 263
>Rafi: EFILRRS 6J L.RE +6 424
>Mike: ALSTZ M3 LAZ. +26 289
>Rafi: FIRS O12 FIRS +39 463
>Rafi: (ST) +4 467 |
