| Player 1 |
|
#player1 Dave Dave
#player2 Nigel_Richards Nigel Richards
>Dave: EINOOPW 8D WOOPIE +30 30
>Nigel_Richards: AEEIR 9C AERIE +21 21
>Dave: LMNOSTU 7F MOLT +24 54
#note I was torn between this and OLM for the same score. In the end I chose this mainly for board shape reasons. I figured that there was some value in taking away the double-double lane on column E, because Nigel has had a knack for double-doubling through twos against me.
>Nigel_Richards: AKM 10B KAM +29 50
>Dave: ?CFNSUZ 9I FUN +20 74
#note I don't care what the computer says. This is definitely the right play. (Simulation has it neck-and-neck with CUZ in the same spot, but I would choose this play every time.)
>Nigel_Richards: ENORW 8K OWNER +38 88
>Dave: ?CDIOSZ M6 ZI.CODeS +96 170
>Nigel_Richards: AAILNST K4 NATI..ALS +86 174
#note What did I say?
>Dave: EEGNRYY L10 GEY +32 202
#note Quackle seems to like GEYER in the same spot significantly more. I guess it scores 6 more points and there are still lots of E's and R's unseen, but this surprises me. With the board starting to get tight and me having tempo, dangling the open R on row 14 seems unnecessary. I guess there's also the factor that if the board does tighten up even more, the one pointers become less valuable than scoring tiles. But in general I'm happy to keep a flexible leave like this and keep pressing the tempo on a tight board while letting the opponent create lanes for my bingos.
>Nigel_Richards: BDI 11D BID +23 197
>Dave: AENRTUY N2 YURTA +33 235
>Nigel_Richards: GJOS 12A JOGS +64 261
>Dave: DEENSTX 4G EXTE.D +44 279
>Nigel_Richards: DFOR A11 F.ORD +48 309
>Dave: AHINNNS 3G HAIN +35 314
>Nigel_Richards: CHIL 5D LICH +27 336
>Dave: AEINNPS O1 NAPE +33 347
#note To be frank, I had lost focus on the EXTENDER spot which I had created a few turns ago and didn't think about M1 PANE for four more points here. That being said, Quackle thinks my play is better, because the other play leaves a lot more scoring opportunities for him with the O1 spot and even the 1H-M line. Although the score is still close, my tempo lead is valuable here, and it may be hard for Nigel to win without hitting a bingo or something like it.
>Nigel_Richards: RR C12 .RR +8 344
>Dave: ABEEINS 6B BEIN +16 363
#note I first thought of playing the new BAE at 5J, which is among my stronger plays available. But I eventually came to prefer BEIN, and I am still happy with that choice. Interestingly, neither Quackle's static evaluator nor Championship Player come up with BEIN at all, but when I added it to a simulation to the end of the game, it did very well, comparable to Quackle's preferred plays. There is great value in killing off the 4D lever by making LI, which takes no front hook. This makes it both harder for Nigel to score with the V's and cuts down the possibilities of him making setups.
>Nigel_Richards: ETT 11H ETT... +7 351
#note Super interesting. At the time, I inferred that Nigel was likely holding at least the Q and the blank, and maybe the U too, and trying to set up EQUAL or EQUID for a bingo-like score. In retrospect, I'm not sure this inference was completely right, as I will explain later. This is a very difficult endgame, and Quackle analyzes it incorrectly at this point. Figuring out what to do over the board was well beyond my ability, and I was also running low on time. After the next play, I'll summarize what I have figured out with the aid of the computer. The big problem here is that how to play best depends greatly on what is in the bag, which is far from random, given the play Nigel just made. In retrospect, I am stymied about what I can correctly infer, and I am very curious to know what Nigel was holding. But I know that he will never tell in a million years.
>Dave: AEEOSUV 12F OVA +13 376
#note So here's a summary of the main tries: EAVES H13 wins outright in all variations *except* when I draw the Q and he plays AVULSING through the S. I thought that was a worthwhile gamble, but I thought that I might be able to win without giving any possible bingos, since there are none right now if I don't create an opening. The main other tries are AVE 12F and OVA 12F. I want to shed the V because holding it and pulling a second V could be very bad for me in the endgame. Furthermore, I chose OVA specifically so that a V draw would give me an out of REVUES in two places. It turns out that most draws after OVA win, but drawing the Q loses by exactly one point if Nigel plays a perfect endgame starting with GURL O6, threatening outs of DIVA N12 and VIFDA L2. (I did not calculate any of this during the game, of course.) On the other hand, playing AVE provides insurance against the Q draw by giving ROQUES to go out in two places, but it loses to the V draw. But there's more to the story! It turns out that both OVA and AVE lose to the draw of U! Oddly, the rack Nigel would most prefer to have is with the Q and no U. That is because he will have 15L LAVING hooking GRRL and threatening to dump the Q for 31 points. This will force me to block the QI spot, but then he will just dump QIN up top and win narrowly. My best guess is that this is what Nigel was intending when he played ETTLED. So, by the percentages, EAVES is the best play, because it loses only to a Q draw, whereas OVA loses to Q or U, and AVE loses to U or V. However, this doesn't take into account that beating OVA after the Q draw requires perfect play by Nigel.
>Nigel_Richards: ?GILQUV L1 QUI. +24 375
>Dave: AEESU 14C .ESEAU +8 384
#note The A draw was a great one for me, guaranteeing me an out in two spots. Quackle points out that I could play out slowly to net one more point, but that is completely impractical in a game with the flag hanging. I was just glad to pull a winning tile.
>Dave: (?GLV) +14 398
|
