| Player 1 |
|
#player1 Dave_Koenig Dave Koenig
#player2 Ian_Weinstein Ian Weinstein
>Dave_Koenig: ABEEMOS 8H OBE +10 10
#note Sim and I agree that this is the best opener. Both the 8F and 8H placements fared approximately equally in the sim, but I like this placement for a very simple reason: I have a back hook tile, but no front hook.
>Ian_Weinstein: ?DEISTV K6 VISiTED +76 76
>Dave_Koenig: ACEMOSS L9 SOME +25 35
#note The clear sim winner: MO L11. In retrospect it makes perfect sense, but during the game I failed to appreciate how strong the ACESS leave was. MOC in the same spot is about a half point behind in the sim, despite scoring six more. (Shows you how much more potent ACESS is than AESS.) SOME and several other plays are around 4 points behind.
>Ian_Weinstein: RRR I8 .RRR +7 83
>Dave_Koenig: ACIOPST H11 OPTIC +38 73
#note Doh! I realized I had missed PISCATOR *immediately* after pressing my clock.
>Ian_Weinstein: DIYZ 13F DI.ZY +29 112
>Dave_Koenig: AAFMRSX M12 FAX +33 106
>Ian_Weinstein: OO 14L O.O +20 132
>Dave_Koenig: ?ACHMRS 14A MARCHeS. +82 188
#note The second best bingo. If you can find the best one, you will impress people and draw them to you. I saw it during the game, but doubted myself on its validity.
>Ian_Weinstein: AEGLNRS G3 ANGLERS +68 200
>Dave_Koenig: ADEPUUW A12 PU.A +33 221
>Ian_Weinstein: HITW D11 WIT.H +34 234
>Dave_Koenig: DEFINUW E8 WIFED +37 258
>Ian_Weinstein: AAEL H1 ALAE +20 254
>Dave_Koenig: GJKLNNU K4 UN....... +24 282
#note I was relieved to have this play. Not only did it score well, but it created a spot for me to clear some of the JUNK off my rack. Ian held it for a while, but the challenge computer remained unvisited.
>Ian_Weinstein: AAL C11 ALA. +22 276
>Dave_Koenig: EGIJKLN 4J J.NK +30 312
#note This was clearly a case of letting the situation on the previous play cloud my judgement. I had set myself up for this exact word, but after my two tile draw it was no longer my best play. JUKING is about 5 points better in the sim, but JUKE keeping GILN is about 10 points better. That's a pretty hefty difference for two plays that score the same and keep a different one point tile. The lesson? GI is much more powerful when accompanied by an N. All the unseen E's probably tip the balance even more in favor of JUKE.
>Ian_Weinstein: INY N4 YIN +27 303
>Dave_Koenig: EEGILOT O6 OGEE +17 329
#note This is an interesting moment. Given the close score and the fact that column O is probably the most dangerous opening, it makes sense to play long and score as much as possible. I immediately thought of ELEGIT, but it is just not playable because of the Q, which is by far the biggest scoring opportunity left unseen. Of all the plays on column O, my play sims best. It's worth taking the three extra points over EGO both because the points are really important here and there are a lot of E's left to draw. There are a couple of strong simming plays that go elsewhere, namely GONE 5I and TEE F9. However, in this game situation, leaving the O column open feels unnecessarily risky to me.
>Ian_Weinstein: AINOQU D4 QUINOA +40 343
>Dave_Koenig: EILRTTU E3 TIT +21 350
#note This is where I think either I can't trust the sim results, or I just don't understand Scrabble well enough to interpret them. Sim to the end of the game likes two plays TUIS G11, and UT C7, predicting that they both win about 57% of the time, even though the second scores about 9 lower equity than the first. UT is definitely the bingo at all costs play as it not only keeps the best bingo leave, but also opens up another line. The problem is that we could be opening the line for a bingo or big play by the opponent too, which might explain the equity loss. But it also interferes with ENDOW/INDOW, which is a winning idea for Ian in some close endgames, so maybe that's how it makes back up the winning percentage. TUIS scares me. It sacrifices a lot of points to go all out for a bingo in only one spot. If opponent hits the E column himself with something like BINT and I don't hit the bingo, I could be screwed. Also, I worry that a human opponent will read this play as a sign to block row 1 at least partially, which sim is not able to take into account. One key fact about this position is that it is very hard to block the top row while still scoring. Since the game is so close, both sides may just have to leave it open while they pursue the points. To me, playing off of the Q made perfect sense because I still left myself with bingo possibilities (granted, not as good with the U in hand) while grabbing as many points as I could, to optimize my chance of winning when I don't hit a bingo. But the Duck thinks IT E4 wins only 49.6% of the time and TIT only 46.9% of the time. It's definitely clear that LIT in the same spot is worse, because there is another T unseen but no more L's. In a close endgame with bingo possibilities like this one, minimizing the chance of having duplicates can be crucial. Maybe my judgement is off and Quackle knows best. But the way things went, my choice was rewarded.
>Ian_Weinstein: BDEEINO C6 BED +21 364
#note Ian really doesn't have a good play that blocks the bingo line, so he scores and leaves himself the ENDOW/INDOW possibility. Sim suggests that this play, or the immediate ENDOW, or DEE F9 optimize his winning chances.
>Dave_Koenig: EEGLRTU 1C REGUL.TE +60 410
>Dave_Koenig: (EINOV) +16 426
|
