As originally posted to the crossword-games-pro (CGP) mailing list.
From: jac@doe.carleton.ca (James A. Cherry)
Date: Tue, 25 Nov 1997 15:49:44 -0500 (EST)
Subject: WSC97 final, game 4
Message-id: <199711252049.UAA05260@paraguay.doe.carleton.ca>
Joel: Turn 1 a b c d e f g h i j k l m n o -> Joel EHORRRU 0 ----------------------------- Matt AAEEGIT 0 1|= ' = ' =| 2| - " " - | 8g HOUR +14 14 3| - ' ' - | 4|' - ' - '| The only other choice might be RHO, but 5| - - | here's what 2000 1-move sims say: 6| " " " " | 7| ' ' ' ' | 17.2 8g HOUR (14) 8|= ' - ' =| -0.2 8e HOUR (14) 9| ' ' ' ' | -0.4 8f HOUR (14) 10| " " " " | -4.6 8h HOUR (14) 11| - - | -4.8 8f RHO (12) 12|' - ' - '| -6.8 8g RHO (12) 13| - ' ' - | 14| - " " - | This lunkhead joined at the tips with his 15|= ' = ' =| computer says, "Nice play, Joel." ----------------------------- Matt: Turn 1 a b c d e f g h i j k l m n o Joel AEPRRUY 14 ----------------------------- -> Matt AAEEGIT 0 1|= ' = ' =| 2| - " " - | 7f TAIGA +24 24 3| - ' ' - | 4|' - ' - '| Best. 5| - - | 6| " " " " | 7| ' ' ' ' | 8|= ' H O U R ' =| 9| ' ' ' ' | 10| " " " " | 11| - - | 12|' - ' - '| 13| - ' ' - | 14| - " " - | 15|= ' = ' =| ----------------------------- Joel: Turn 2 a b c d e f g h i j k l m n o -> Joel AEPRRUY 14 ----------------------------- Matt AEEEFIJ 24 1|= ' = ' =| 2| - " " - | 6b YAUPER$ +31 45 3| - ' ' - | 4|' - ' - '| Another interesting position. Block the 5| - - | hot spot or turn over and balance? 2000 6| " " " " | 1-move sims say: 7| ' T A I G A ' | 8|= ' H O U R ' =| 36.8 6j YUP (31) 9| ' ' ' ' | -1.6 6j YAUP (32) 10| " " " " | -2.2 6j PARRY (35) 11| - - | -2.8 6b YAUPER$ (31) 12|' - ' - '| -5.9 6j YAP (31) 13| - ' ' - | -7.9 6j YAUPER$ (36) 14| - " " - | 15|= ' = ' =| I still prefer Joel's play to YUP, though ----------------------------- PARRY seems OK to me as well. Matt: Turn 2 a b c d e f g h i j k l m n o Joel ABGOORT 45 ----------------------------- -> Matt AEEEFIJ 24 1|= ' = ' =| 2| - " " - | 6j JEFE +56 80 3| - ' ' - | 4|' - ' - '| B'oh! Shoulda blocked the hot spot. 5| - - | Matt's play is correct. 6| Y A U P E R " " | 7| ' T A I G A ' | 8|= ' H O U R ' =| 9| ' ' ' ' | 10| " " " " | 11| - - | 12|' - ' - '| 13| - ' ' - | 14| - " " - | 15|= ' = ' =| ----------------------------- Joel: Turn 3 a b c d e f g h i j k l m n o -> Joel ABGOORT 45 ----------------------------- Matt AEEEGIZ 80 1|= ' = ' =| 2| - " " - | 5i TABOO +40 85 3| - ' ' - | 4|' - ' - '| BOTARGO# and its plural don't fit. Which 5| - - | move is best of TABOO 5i (40), TABOR 5i 6| Y A U P E R J E F E " | (40), and ABORT 5j (40)? The answer 7| ' T A I G A ' | surprised me. I thought GR would be a 8|= ' H O U R ' =| better keep than GO, but sim says not: 9| ' ' ' ' | TABOR sims 0.3 ahead of TABOO. ABORT is 10| " " " " | only 1.3 back, perhaps surprising given 11| - - | that it opens column O. 12|' - ' - '| 13| - ' ' - | Roger Phillips defined BORTAGO# online as 14| - " " - | "a relish made of mullet or tunny roe". 15|= ' = ' =| ----------------------------- Matt: Turn 3 a b c d e f g h i j k l m n o Joel EEGLNRS 85 ----------------------------- -> Matt AEEEGIZ 80 1|= ' = ' =| 2| - " " - | e4 ZIP +28 108 3| - ' ' - | 4|' - ' - '| Adam Logan saw Matt's best play, saying 5| - T A B O O | David Boys had played it against him at 6| Y A U P E R J E F E " | the 1995 WSC: AGUIZE# d4 (52). Second 7| ' T A I G A ' | best is GAUZE d4 (32) simming 24 points 8|= ' H O U R ' =| behind, and Matt's play is 34 points 9| ' ' ' ' | behind. 10| " " " " | 11| - - | 12|' - ' - '| 13| - ' ' - | 14| - " " - | 15|= ' = ' =| ----------------------------- Joel: Turn 4 a b c d e f g h i j k l m n o -> Joel EEGLNRS 85 ----------------------------- Matt AEEEGOQ 108 1|= ' = ' =| 2| - " " - | c3 GLEANERS +72 157 3| - ' ' - | 4|' - Z ' - '| Both ENLARGES c3 (72) and GENERALS c1 5| I T A B O O | (70) play; obviously, GENERALS is 6| Y A U P E R J E F E " | inferior because Joel shouldn't be 7| ' T A I G A ' | opening triple-triple lines when ahead. 8|= ' H O U R ' =| Between the other two moves, ENLARGES 9| ' ' ' ' | simulates about 3 points better. This 10| " " " " | is likely because it places the G at 8c 11| - - | instead of an E, reducing the size of 12|' - ' - '| Matt's potential comebacks at 8a. 13| - ' ' - | 14| - " " - | 15|= ' = ' =| ----------------------------- Matt: Turn 4 a b c d e f g h i j k l m n o Joel ??ALRTW 157 ----------------------------- -> Matt AEEEGOQ 108 1|= ' = ' =| 2| - " " - | 4b GLOZE +30 138 3| G ' ' - | 4|' L - Z ' - '| Using the same tiles, Matt could play 5| E I T A B O O | GOE# d2 (35) for five more points. It 6| Y A U P E R J E F E " | simulates only 3.4 better than GLOZE, 7| N T A I G A ' | however. 8|= E ' H O U R ' =| 9| R ' ' ' | 10| " S " " " | 11| - - | 12|' - ' - '| 13| - ' ' - | 14| - " " - | 15|= ' = ' =| ----------------------------- Joel: Turn 5 a b c d e f g h i j k l m n o -> Joel ??ALRTW 157 ----------------------------- Matt ACEEIMQ 138 1|= ' = ' =| 2| - " " - | 10c StALWaRT +63 220 3| G ' ' - | 4|' G L O Z E ' - '| Adam commented that all of TREW#, ALEW#, 5| E I T A B O O | and AREW# 8a (33) are reasonable 6| Y A U P E R J E F E " | alternatives. In fact, they simulate 7| N T A I G A ' | more than 20 points ahead of StALWaRT. 8|= E ' H O U R ' =| TREW# seems to be the best of the three, 9| R ' ' ' | presumably because of its strong ??AL 10| " S " " " | keep. Online, Marlene Milkent said 11| - - | whenever she chooses not to bingo in 12|' - ' - '| favor of high play, she draws junk; Zev 13| - ' ' - | Kaufman responded, "How bad can ??AL be?" 14| - " " - | 15|= ' = ' =| ----------------------------- Matt: Turn 5 a b c d e f g h i j k l m n o Joel IIINNUY 220 ----------------------------- -> Matt ACEEIMQ 138 1|= ' = ' =| 2| - " " - | d8 METIC# +31 169 3| G ' ' - | 4|' G L O Z E ' - '| I kept being impressed at the OSW-only 5| E I T A B O O | words both players were playing. Matt 6| Y A U P E R J E F E " | makes an excellent play here, beating 7| N T A I G A ' | out QI# 4h (23) and QAT i3 (22) in sim by 8|= E ' H O U R ' =| 7.5 and 10.4 points, respectively. I 9| R ' ' ' | learned from a newspaper article how much 10| " S t A L W a R T " | work Matt had been doing to learn 11| - - | SOWPODS, and I recall Joel saying he 12|' - ' - '| hadn't done much studying for 1993 and 13| - ' ' - | only a little more for 1995. Both 14| - " " - | players seemed quite comfortable with 15|= ' = ' =| their nonnative$ lexicon in these games. ----------------------------- Joel: Turn 6 a b c d e f g h i j k l m n o -> Joel IIINNUY 220 ----------------------------- Matt AEIOQSX 169 1|= ' = ' =| 2| - " " - | 12b UNCINI +8 228 3| G ' ' - | 4|' G L O Z E ' - '| I was pleased to get this one right 5| E I T A B O O | myself as I watched on MarlDOoM. This is 6| Y A U P E R J E F E " | easily the best play, beating LINNY# f10 7| N T A I G A ' | (16) by 5 points, no duh. IIU is an 8|= E M H O U R ' =| 'orrible keep. 9| R E ' ' ' | 10| " S t A L W a R T " | 11| I - - | 12|' C ' - '| 13| - ' ' - | 14| - " " - | 15|= ' = ' =| ----------------------------- Matt: Turn 6 a b c d e f g h i j k l m n o Joel ADILSUY 228 ----------------------------- -> Matt AEIOQSX 169 1|= ' = ' =| 2| - " " - | b11 QUAI$ +26 195 3| G ' ' - | 4|' G L O Z E ' - '| Sim misses the point here, in my 5| E I T A B O O | opinion: QUAI$ is better than QI# 4h 6| Y A U P E R J E F E " | (23) by 0.3 or so, but Matt really needs 7| N T A I G A ' | to (a) ditch the Q and (b) open a big 8|= E M H O U R ' =| spot for himself. Hooks for QUAI are 9| R E ' ' ' | LR#S$, and Matt holds an S with LRRSS 10| " S t A L W a R T " | unseen (or UNCINI, as the case may be). 11| I - - | 12|' U N C I N I ' - '| 13| - ' ' - | 14| - " " - | 15|= ' = ' =| ----------------------------- Joel: Turn 7 a b c d e f g h i j k l m n o -> Joel ADILSUY 228 ----------------------------- Matt ELMOSWX 195 1|= ' = ' =| 2| - " " - | 15a SLAY +47 275 3| G ' ' - | 4|' G L O Z E ' - '| Clearly correct. Maven's second choice, 5| E I T A B O O | SLAID# 15a (35), sims 10 points back. 6| Y A U P E R J E F E " | 7| N T A I G A ' | 8|= E M H O U R ' =| 9| R E ' ' ' | 10| " S t A L W a R T " | 11| Q I - - | 12|' U N C I N I ' - '| 13| A - ' ' - | 14| I " " - | 15|= ' = ' =| ----------------------------- Matt: Turn 7 a b c d e f g h i j k l m n o Joel BDEINOU 275 ----------------------------- -> Matt ELMOSWX 195 1|= ' = ' =| 2| - " " - | 14d MOW +23 218 3| G ' ' - | 4|' G L O Z E ' - '| Matt keeps hammering away, trying to 5| E I T A B O O | stay in it. OWL 4l (30) sims the best, 6| Y A U P E R J E F E " | presumably because Matt's S gets used 7| N T A I G A ' | well the next turn. Check it out: 8|= E M H O U R ' =| 9| R E ' ' ' | 46.7 4l OWL (30) 10| " S t A L W a R T " | -5.1 4l OW (28) 11| Q I - - | -6.6 4m WOX (27) 12|' U N C I N I ' - '| -7.5 4l OWLS (34) 13| A - ' ' - | -8.9 14d MOW (23) 14| I " " - | 15|S L A Y = ' =| OWL gets Matt to within 33, while MOW ----------------------------- keeps him back 42, on average. Joel: Turn 8 a b c d e f g h i j k l m n o -> Joel BDEINOU 275 ----------------------------- Matt EEFLRSX 218 1|= ' = ' =| 2| - " " - | 15f OBIED# +29 304 3| G ' ' - | 4|' G L O Z E ' - '| MarlDOoM watchers were stunned when Joel 5| E I T A B O O | didn't spot BEDOUIN 3f (66). We 6| Y A U P E R J E F E " | conjectured it was a miss under what must 7| N T A I G A ' | have been hideous amounts of pressure. 8|= E M H O U R ' =| 9| R E ' ' ' | 10| " S t A L W a R T " | 11| Q I - - | 12|' U N C I N I ' - '| 13| A - ' ' - | 14| I M O W " - | 15|S L A Y = ' =| ----------------------------- Matt: Turn 8 a b c d e f g h i j k l m n o Joel AENNSTU 304 ----------------------------- -> Matt EEFLRSX 218 1|= ' = ' =| 2| - " " - | 13g FLEXERS* +103 218 3| G ' ' - | 4|' G L O Z E ' - '| The key play of this tournament. There 5| E I T A B O O | are many interesting questions: 6| Y A U P E R J E F E " | 7| N T A I G A ' | - Would you risk FLEXERS* at single- 8|= E M H O U R ' =| challenge? 9| R E ' ' ' | - Would you challenge FLEXERS* at double- 10| " S t A L W a R T " | challenge? 11| Q I - - | - If Matt plays REFLEX 14i (43) or FLEME# 12|' U N C I N I ' - '| 8a (30), how does the game come out? 13| A - ' ' - | 14| I M O W " - | I heard Joel told Matt at the table that 15|S L A Y O B I E D ' =| he'd have made the same play. FLEXORS is ----------------------------- good, but not FLEXERS; I hate -ER words! Joel: Turn 9 a b c d e f g h i j k l m n o -> Joel AENNSTU 304 ----------------------------- Matt EEFLRSX 218 1|= ' = ' =| 2| - " " - | 13g NUT +8 312 3| G ' ' - | 4|' G L O Z E ' - '| A great example of where computers need 5| E I T A B O O | improvement: both Maven and DDBot want 6| Y A U P E R J E F E " | UNOFTEN# l3 (22), but this gives up 7| N T A I G A ' | FLEXURES/REFLUXES 3h (92). Computers 8|= E M H O U R ' =| need to take into account the opponents' 9| R E ' ' ' | tiles when they challenge plays off. 10| " S t A L W a R T " | 11| Q I - - | Joel's play is good, not giving Matt much 12|' U N C I N I ' - '| to work with. 13| A - ' ' - | 14| I M O W " - | 15|S L A Y O B I E D ' =| ----------------------------- Matt: Turn 9 a b c d e f g h i j k l m n o Joel AEHIINS 312 ----------------------------- -> Matt EEFLRSX 218 1|= ' = ' =| 2| - " " - | 8a EXEME# +42 260 3| G ' ' - | 4|' G L O Z E ' - '| Probably FLEME# 8a (30) is better because 5| E I T A B O O | it keeps a vowel in a consonant-heavy 6| Y A U P E R J E F E " | unseen tile set. Adam, Maven, and DDBot 7| N T A I G A ' | all wanted FLEME# mildly over EXEME#. 8|= E M H O U R ' =| Though, down 94, maybe Matt wants points. 9| R E ' ' ' | 'Course, he can't mount a comeback with 10| " S t A L W a R T " | no vowels. Not an easy call. 11| Q I - - | 12|' U N C I N I ' - '| 13| A - N U T - | 14| I M O W " - | 15|S L A Y O B I E D ' =| ----------------------------- Joel: Turn 10 a b c d e f g h i j k l m n o -> Joel AEHIINS 312 ----------------------------- Matt EFLRRSV 260 1|= ' = ' =| 2| - " " - | 11g HI +16 328 3| G ' ' - | Unseen: AEOO CDDDFKLNPRRSTTVV 4|' G L O Z E ' - '| 5| E I T A B O O | Joel blocks the 11h/i bingos and keeps 6| Y A U P E R J E F E " | some vowels (and a nice rack, AEINS) for 7| N T A I G A ' | the consonant-heavy unseen letters. 8|E X E M E H O U R ' =| 9| R E ' ' ' | 10| " S t A L W a R T " | 11| Q I - - | 12|' U N C I N I ' - '| 13| A - N U T - | 14| I M O W " - | 15|S L A Y O B I E D ' =| ----------------------------- Matt: Turn 10 a b c d e f g h i j k l m n o Joel AEIKNOS 328 ----------------------------- -> Matt EFLRRSV 260 1|= ' = ' =| 2| - " " - | a6 REV +6 266 3| G ' ' - | Unseen: AAEIOO CDDDKNNPSTTV 4|' G L O Z E ' - '| 5| E I T A B O O | Only two vowels are left to draw in 11 6| Y A U P E R J E F E " | tiles, it turns out. I don't know what 7| N T A I G A ' | Matt should play to give himself a good 8|E X E M E H O U R ' =| shot. FROE m3 (14)? There's still a V 9| R E ' ' ' | unseen, and he shouldn't block 3f plays, 10| " S t A L W a R T " | so I can't fault REV. 11| Q I - H I - | 12|' U N C I N I ' - '| Given the bag, Matt has a 5/55 chance of 13| A - N U T - | drawing a bingo: AN gives SALFERN# n6 14| I M O W " - | (81), AO gives SAFROLE n6 (81), and DO 15|S L A Y O B I E D ' =| gives REFOLDS$ 3f (63). ----------------------------- Joel: Turn 11 a b c d e f g h i j k l m n o -> Joel AEIKNOS 328 ----------------------------- Matt DEFLRST 266 1|= ' = ' =| 2| - " " - | n6 SINK +27 355 3| G ' ' - | Unseen: AEO CDDDFLNPRSTTV Bag: AO CDDNPTV 4|' G L O Z E ' - '| 5| E I T A B O O | Suggested online were NAIK# 14h (21), 6| Y A U P E R J E F E " | OAKEN 14j (27), and EIKONS 14i (32). 7|R N T A I G A ' | It's reasonable to want to block n6 8|E X E M E H O U R ' =| bingos, but probably the best play for 9|V R E ' ' ' | that is SANKO# n6 (30). It doesn't 10| " S t A L W a R T " | matter a whole lot, however, since Joel 11| Q I - H I - | is up by 89. As I recall, he was getting 12|' U N C I N I ' - '| short on time; SINK or SANKO both give 13| A - N U T - | Matt, e.g., ADOPTER o9 (95)/(98), so I 14| I M O W " - | wonder if it's best to leave the spot 15|S L A Y O B I E D ' =| open after all. Hmm. ----------------------------- Matt: Turn 11 a b c d e f g h i j k l m n o Joel AACENOO 355 ----------------------------- -> Matt DEFLRST 266 1|= ' = ' =| 2| - " " - | 8l FEND +33 302 3| G ' ' - | Unseen: AAEOO CDDNPTV Bag: DDPTV 4|' G L O Z E ' - '| 5| E I T A B O O | Matt's play is reasonable, but with no 6| Y A U P E R J E F E S | vowels in the bag, FEND cannot win. DOFF 7|R N T A I G A ' I | l4 (22) also doesn't help given the bag, 8|E X E M E H O U R ' N =| and appears worse just looking at the 9|V R E ' ' ' K | unseen tiles because it blocks 3f bingos. 10| " S t A L W a R T " | 11| Q I - H I - | 12|' U N C I N I ' - '| 13| A - N U T - | 14| I M O W " - | 15|S L A Y O B I E D ' =| ----------------------------- Joel: Turn 12 a b c d e f g h i j k l m n o -> Joel AACENOO 355 ----------------------------- Matt DLPRSTV 302 1|= ' = ' =| 2| - " " - | 14j OCEAN +27 378 3| G ' ' - | Unseen: DDLRPSTTV Bag: DT 4|' G L O Z E ' - '| 5| E I T A B O O | Joel's got to be smiling given the 6| Y A U P E R J E F E S | tracking and his drawing the remaining 7|R N T A I G A ' I | vowels from the bag (AO). I can't check 8|E X E M E H O U R F E N D| it without my software crashing, but 9|V R E ' ' ' K | OCEAN looks correct to me: keep only two 10| " S t A L W a R T " | tiles, both vowels, and draw two from the 11| Q I - H I - | bag. 12|' U N C I N I ' - '| 13| A - N U T - | 14| I M O W " - | 15|S L A Y O B I E D ' =| ----------------------------- Matt: Turn 12 a b c d e f g h i j k l m n o Joel ADOT 378 ----------------------------- -> Matt DLPRSTV 302 1|= ' = ' =| 2| - " " - | f2 PRE# +11 313 3| G ' ' - | 4|' G L O Z E ' - '| Matt has to block TOAD g1 (23) and DATO$/ 5| E I T A B O O | DOAT/TOAD 3f (10). Turns out PRESET f2 6| Y A U P E R J E F E S | (16) is Matt's best endgame play, leading 7|R N T A I G A ' I | to a final spread of -87 after TA 13l 8|E X E M E H O U R F E N D| (10), GLOZED 4b (17), and OD 4l (22). 9|V R E ' ' ' K | PRE# leads to -93 after OD 4l (22), DEV$ 10| " S t A L W a R T " | l13 (11), and TAG b2 (11). 11| Q I - H I - | 12|' U N C I N I ' - '| 13| A - N U T - | 14| I M O W O C E A N | 15|S L A Y O B I E D ' =| ----------------------------- Joel: Turn 13 a b c d e f g h i j k l m n o -> Joel ADOT 378 ----------------------------- Matt DLSTV 313 1|= ' = ' =| 2| - P " - | 4l ODA# +24 402 3| G R ' ' - | 4|' G L O Z E ' - '| Ever pedantic, I'll note OD 3l (22) maxes 5| E I T A B O O | spread. Of course, when you can be world 6| Y A U P E R J E F E S | champion at Scrabble with either ODA# or 7|R N T A I G A ' I | OD, who the hell cares? 8|E X E M E H O U R F E N D| 9|V R E ' ' ' K | 10| " S t A L W a R T " | 11| Q I - H I - | 12|' U N C I N I ' - '| 13| A - N U T - | 14| I M O W O C E A N | 15|S L A Y O B I E D ' =| ----------------------------- Matt: Turn 13 a b c d e f g h i j k l m n o Joel T 402 ----------------------------- -> Matt DLSTV 313 1|= ' = ' =| 2| - P " - | l13 DEV$ +22 324 3| G R ' ' - | 4|' G L O Z E ' O D A '| Correct. 5| E I T A B O O | 6| Y A U P E R J E F E S | 7|R N T A I G A ' I | 8|E X E M E H O U R F E N D| 9|V R E ' ' ' K | 10| " S t A L W a R T " | 11| Q I - H I - | 12|' U N C I N I ' - '| 13| A - N U T - | 14| I M O W O C E A N | 15|S L A Y O B I E D ' =| ----------------------------- Joel: Turn 14 a b c d e f g h i j k l m n o -> Joel T 402 ----------------------------- Matt LST 324 1|= ' = ' =| 2| - P " - | n3 TA +2 404 3| G R ' ' - | (LST) +6 410 4|' G L O Z E ' O D A '| 5| E I T A B O O | For the record, ETA e8 (6) maximizes 6| Y A U P E R J E F E S | spread. 7|R N T A I G A ' I | 8|E X E M E H O U R F E N D| Final score: Joel 410, Matt 324 9|V R E ' ' ' K | 10| " S t A L W a R T " | Joel: 84.3%, Matt: 81.8%, says DDBot. 11| Q I - H I - | 12|' U N C I N I ' - '| Joel Sherman wins the 1997 World 13| A - N U T D - | (English Language) Scrabble Championship, 14| I M O W O C E A N | three games to one. Congratulations, 15|S L A Y O B I E D V =| Joel! -----------------------------
SCRABBLE® is a registered trademark of J.W. Spear & Sons Limited, Maidenhead SL6 4UB, England except in Canada and the United States, where it is a registered trademark of HASBRO. Used under license from Mattel Europa BV.
For technical assistance with this web site, please contact John Chew, webmaster of the (North American) National SCRABBLE® Association, at jjchew@math.utoronto.ca.