| Player 1 |
|
#player1 Zev_Kaufman Zev Kaufman
#player2 Jackson_Smylie Jackson Smylie
>Zev_Kaufman: ADELPRT 8D PLATED +24 24
#note Zev Kaufman vs. Jackson Smylie. Decembre 7th 2011. This game was the third game of the night at the Toronto Scrabble Club. The endgame was also featured in the Last Word Newsletter. I apologize if any of the percentages are incorrect.
Exch. L, Exch. D, and Exch. P simulate best, winning 55.10%, 54.19%, and 53.23% respectively (after 1873 iterations). You draw a 7 47/93 times after exchanging L (50.54%). You draw a 7 57/93 times after exchanging D (61.29%). You draw a 7 52/93 times after exchanging P (56.99%). The fact that exchanging L simulates best indicates there are many 8s using the letters ADEPRT. Exchanging L bingoes 73.04% of the time. Exchanging D bingoes 74.69% of the time. Exchanging P bingoes 71.92% of the time.
PLATED did not place vowels adjacent to any double letter scores. It scored well and kept a decent one-tile leave of R. Zev thought that exchanging one tile would not be worth the major point sacrifice.
>Jackson_Smylie: EEISTTW D8 .EWIT +20 20
#note I considered (P)EWIT, (P)EEWIT, WET H7, TEW F7, and WITE F9. Speedy player prefers WITE. I played (P)EWIT because it forked the board, making it impossible for my opponent to close all the bingo lanes. WITE set up A10-G10, I10-O10, and J9-J15. WITE closed 5 lanes (lanes through the LATE and D). I felt the 3-point sacrifice was correct.
>Zev_Kaufman: ADLRTUW F7 W.UL +9 33
#note WAT simulates best, at 53.18%. Though W(A)UL's leave is the best in the rack, it scores too few points to choose over WAT. Although WAT is asking for a 35-40 point play from the opponent, the 35 point play would give Zev the triple word score to work with.
>Jackson_Smylie: AEEELST 7E E.E +12 32
#note I didn't know TEASEL or TEASELED. I haven't yet come to AEELSST or ADEEELST in probability. Out of the nonbingo plays, E(W)E is superior because of the leave and the face that it sets up another bingo lane alont row 6.
>Zev_Kaufman: ADEGRRT H2 GARTER.D +68 101
#note GARR(E)TED scores one more point. Nice play though.
>Jackson_Smylie: AEILRST 5E TER.IALS +82 114
#note Nobrainer.
>Zev_Kaufman: ADEGINY K3 DE.AYING +63 164
#note Bingo bonanza!
>Jackson_Smylie: AAINNOR C9 NAOI +17 131
#note I missed A(E)ON J4. I looked for ways to get rid of my ANO, but didn't see a place. NAOI simulates a hair behind after 1482 iterations.
>Zev_Kaufman: GMMNNSV -GMMNNV +0 164
#note N(A)M J4 simulates 3.5% above any other play. V(A)N G3 simulates 1% higher than all exchanges. Keeping MS simulates 0.4% better than keeping S. I would consider the synergy of GNS or NS to be good enough to surpass keeping S.
>Jackson_Smylie: AEINORT 12C ..ERATION +60 191
#note The only bingo.
>Zev_Kaufman: AEIOOOS -IOOO +0 164
#note Surprisingly, OS(S)IA and OES both simulate above all exchanges. Due to a high ratio of consonants to vowels in the bag, keeping AES is the best exchange. 2813 iterations were completed.
>Jackson_Smylie: ?BDHILN J4 B.H +33 224
#note Far superior to any other play.
>Zev_Kaufman: ABEKORS 2C BROKA.ES +77 241
#note The only bingo. It is worthy to note SOAK L12 for 46. It kills the ITERATION hook. I would not consider it to be close to BROKA(G)ES by any means.
>Jackson_Smylie: ?DEILNR 1I REDLINe +78 302
#note I think REDLINe is superior to RELINeD. I believe more 8s begin with D than with E. Putting the N in the N column also prevents most bingoes (very few begin with N). REDLINe wins approximately 81.02% of the time (after 735 iterations).
>Zev_Kaufman: ?AAMNOO 3A MOO +18 259
#note with 30 unseen tiles, I think MOO was a tad desperate. Unless I have a bingo elsewhere, it is very unlikely I will not block his opening along the A column. The leave of AAN? doesn't guarantee a bingo the following turn. Quackle prefers (L)OON(S) or MOA A1.
>Jackson_Smylie: FGIOOST L8 FOOTS +58 360
#note Best.
>Zev_Kaufman: ?AAHNVX J11 H.AX +30 289
#note AX(M)AN. Zev thought he might be able to hit an X bingo after H(O)AX. (RE)X is a consideration because of the points scored.
>Jackson_Smylie: FGIMNPU A1 FU.ING +39 399
#note I saw MU(M)PING but turned it down out of fear that it was invalid.
>Zev_Kaufman: ?AESUVZ H12 .AV +6 295
#note A very good setup by Zev. The only play that supersedes it is S(I)ZE I11 for 58 points, keeping AUV?. Not only does it put Zev down by 52, it sets up the DRS hooks in case Zev draws a 40 point play or bingo. It simulates 6% higher than (T)AV after 3670 iterations.
>Jackson_Smylie: CCIJMPV 10B J.. +29 428
#note A pivotal move in the game. J(AW) rids me of my hefty J, and scores 29 points. I reasoned that I could withstand a bingo from Zev (which was a small chance anyways, given the QUUYZ's presence in the bag). J(AW) put me up 133 points, so I figured it was the correct play. I was wrong. (E)PIC and (E)MIC were much superior because they block Zev's aggressive setup. Horrible leave notwithstanding, one must block in a situation like this.
>Zev_Kaufman: ?ENSUUZ 9K ..UN +5 300
#note Speedy player does not even register (NO)UN. When inputted and simulated, (NO)UN wins over 6% more often than ZEiNS (after 10 000 iterations). ZEiNS blocks the only large scoring spot left. Zev would be down 43 with UU and 5 other tiles from the disgusting CCEIIMOPQVY pool. Zev was very smart to turn down the compelling 90 points. (NO)UN forks the board, made possible by my lack of blocking. By forking the board, just like with the play of (P)EWIT, I am unable to block both TAV -> TAVS and NOUN -> NOUNS.
>Jackson_Smylie: CCEIMPV I14 EM +17 445
#note Emptying the bag was ripping the win to shreds. The bag contains the bingos ISOZYME, OBSEQUY, OUTSIZE, QUIZZES, SEQUOIA, and SOYUZES, all of which are playable. As I am typing this paragraph, I see MI(N)CE, which simulates very well, winning 90% of the time (after 20 000 iterations). Drawing one of the 6 possible bingos is unlikely, but drawing QUIZZES or SOYUZES is unlikelier (yes, unlikelier is good). After playing MI(N)CE, there are only 2 potential bingos. At this point in the game, if Zev's rack is EIQSUZ? or EOSUYZ?, the game is over. EM wins approximately 34% of the time, (A)M wins approximately 69% of the time, and MI(N)CE wins approximately 90% of the time (after 20 000 iterations).
>Zev_Kaufman: ?EOSUYZ O9 SOYUZEs +112 412
#note Wow. A great ending to a great game. Well played Zev. EM, my play last turn, was definitely a mistake. But as it turns out, the only play that could have hypothetically won (It would have won 50% of the time) was dishing off the M. If I had drawn the I and Zev the Q, he would have eaten it and I would have won. Wp.
>Zev_Kaufman: (CCIIPQV) +50 462
|
