| Player 1 |
|
#player1 IJB Izak J. Bulten
#player2 JJB John J. Bulten
>IJB: AGIIINT -AII +0 0
#note ~0:30 [~24:30] This first recorded game featuring a father-son team was a great test of IJB's latest year of club and tournament training, even though racks were usually more duplicative than bingo-prone. IJB starts with a trade, very accurately; the best is to AGINT, with GINT second by 5.2, although IJB noted later that AGINT should have been obvious. Atigi 14 rates 3 behind trading.
>JJB: ?BCNPRR 8G BRR +10 10
#note 0:14 [24:46] Obvious though humorous.
>IJB: DDGINOT 9C DINGO +15 15
#note 1:54 [22:36] IJB considers trodding* by judging the merits of trodder* and trotter, and rates dingo by recalling dingoes, determining each rightly. This play rates only behind odd 7i 9, by 2.5.
>JJB: ?CMMNOP 8B POM +20 30
#note 3:33 [21:13] JJB finds it hard to believe he has found the best play so keeps looking awhile; but there it is.
>IJB: DFIPRST 10G DRIFT +23 38
#note 1:42 [20:54] IJB assembles up-drifts and correctly rules it out by remembering updrafts, finding both best plays (j7 seemed less safe).
>JJB: ?CLMNTW -CLMW +0 30
#note 2:19 [18:54] JJB takes a little play-hunting time and more leave-planning time, choosing a leave 1.1 behind CN?, .6 behind CL?, and equal to C?. Milt and cwms are 7 points behind. Hold the C when studying the C!
>IJB: AEEIPSS 10B PE +14 52
#note 3:45 [17:09] IJB selects a leave worth +10.2 and makes one of the two best dumps with it, equal to cep d11 14. The obscure bingo missed is epitases 70, 45.8 points ahead in evaluation. Keeping the middling tile would have supplemented the rack, but IJB's directed hook is also worth noting.
>JJB: ?AENORT E4 NORTE.As +86 116
#note 0:42 [18:12] JJB has drawn 42 bingos here, including brominate/protamine 75 and overtrain 61, and calculates out the quadruple as the highest slot. It looks like wantoner defends better (not resonant), so it might not have been observed in-game.
>IJB: EGILMSS D2 SLIME +24 76
#note 3:02 [14:07] The other quadruple line is open and accommodates gimlets 40 (18.4 diff). Otherwise column D is indicated, and second place is a minor change, glime d2 26, 7.7 ahead solely on the S. Playing the S is not problematic here, and miles d1 26 also works.
>JJB: EEGIIOO I3 GOOIE. +10 126
#note 1:26 [16:46]
>IJB: AGHNST H1 THAN +30 106
#note 1:15 [12:52] One letter off: thang was available for 9 more (approximately 11.4 diff), but the G hangs on IJB's rack awhile.
>JJB: AEILLUU K5 ULULA.E +28 154
#note 1:09 [15:37] JJB is overwhelmed that the word exists and that it scores so highly (16.8 above the next, pilau 14). He has pulled the double quadruple. Honorable mention: pilule.
>IJB: GISX L11 XI +27 133
#note 2:21 [10:31] Best play from the known tiles!
>JJB: EEEINOW C1 EEEW +26 180
#note 1:19 [14:18] The better parallel is eew g1 28 (up by 1.5) but JJB prefers the novelty, not even directing a hook.
>IJB: AEGHIIS 8K .EASH +27 160
#note 1:06 [10:31] Right hotspot. This vocab test calls for finding a five that dumps an I: laigh 30, preferred by 21.6 over using ES. Geisha/xis 34, reishi 30, and sage 31 (4-overlap) also score comparably.
>JJB: AEINOSY G2 OY +25 205
#note 2:28 [11:50] JJB is recalling his one good rack so far. At least a bingo dump finally avails.
>IJB: AAGILOY B1 AY +24 184
#note 0:49 [8:36] The top plays all cash winy, so the F column is indicated, starting with yoga f4 37 (14.0 ahead). IJB is starting to conserve time and takes an acceptable overlap that opens riskily. For turnover if not for evaluation, he also had laogais n2 20, having studied the 4-voweled 6s regularly.
>JJB: AEFINRS N2 FAIRNE.S +76 281
#note 1:03 [10:47] The players like to recite simple lists to each other, like arisen, so this is a welcome JJB draw. Naif a1 49 rates just above sanserif n5 65.
>IJB: AGILO A1 GAL +31 215
#note 0:33 [8:03] IJB gets to use his own setup. The only improvement on the known tiles is to extend it to gaol/goal 37 (approximately 12.9 diff).
>JJB: ACDDEJR O1 JAR +37 318
#note 4:10 [6:37] Quackle prefers card 39 by 2.8 for undoubling, and certainly columns A, H, L, and M are J-accommodating. JJB has now gotten into time tightness himself.
>IJB: EINOTTV M3 VOE +28 243
#note 0:48 [7:15] Excellent overlap, within .4 of vet/yage j2 25.
>JJB: ?CDDENW J2 WED +27 345
#note 1:47 [4:50] JJB now has the blank but has been hampered all game from standout racks. He has another easy decision (wed/dew lead the pack by 8.9).
>IJB: ABINOTT 1J OBIT +15 258
#note 0:50 [6:25] IJB is alert to all the hooks. Over the board he was not 100% certain if he could carry off obtain/ope a10 32 (the better play by 20.9) or if that were merely crosswordese, although he often gets through the full 3-list cold. He chose the tamer hook and excellent leave. Oba f4 26 also is the column F cash in second place.
>JJB: ?CDIKNV D11 ICK +19 364
#note 3:18 [1:32] JJB's first real multiple-choice. All leaves look ugly even though carrying the blank. He did not spot the clear kin f4 30, 10.6 ahead, or dick/dink 12k 20/18, with ululated. All the same this was a good opportunity to keep looking, rather than the previous plays where he consumed unnecessary time.
>IJB: AINT B4 TIN +12 270
#note ~1:20 [~5:15] As the board closes and bingos get unlikely, less is recommendable. Here the plays are in F and 13, starting with overlap ani f5 17 (approximately 5.5 ahead). IJB also empties the bag.
>JJB: ?DNQUUV A5 QUa +24 388
#note 1:11 [0:21] JJB laughs again at QUU being 3 of the last 4 tiles but is making time-trouble inaccuracies, neglecting the risk of V stick when Q is well-supplied. No QV play or pair exists, so three outplays are required, one a Q play and one a different V play. However, IJB has two out-in-two setups: za/pacts and zas/act. The analysis of out-in-three is mathematically interesting; four out-in-threes can be forced. (1) In all unblockables, the V must be played first, eruv 6d 9; the Z must be cashed, also blocking the best Q, zas/icks 14b 34; the second-best Q must be cashed, also setting up DN, xis/suq 31; code c7 10; dun 12k 15+2; an elegant net of 13 points. (If out-in-two is threatened, eruv 9, za 15, then pad 6 blocks to retain options (not pan 5), code 10, suq 40, tes 12+2, net of 16; zas/act also nets higher for JJB.) (2) Without continuing to compensate for IJB's out-in-two threats: the other sequence unblockable against ordinary racks is vau 9, zas 34, dun 11, code 10, qi (2 ways) 11+2, net of -11. (3) Theoretically blockable, but the UU cannot be blocked without locking out the Z: nav 9, zas 34, qi 11, code 10, kudu 10+2 (or swap JJB's last two), net of -12. (4) Theoretically blockable, but o6 block locks out the Z: vodun 13, zas 34, qi 11, code 10, nu 2+2, net of -16. (5) That exhausts unblockability and leaves blockable situations. First we try suq 14d 40, zas g12 25, and 13g 6 (blocks act), code 10, eruv 9+2, which happily goes out with 22 net! But suq 40, mot 5 (sticking V), un/noh 12, za/ape 16, da/de 10, suqs 12, pass!, code 10+8, will net only 11, which is 2 behind eruv! (Suq 13l 31 falls 9 points behind for the same reason, and every other non-V play fails, even the setup mon 5.) Therefore the V must always be played first to do better. Assume IJB wants three plays in some order (zas 34, code 10, te/ta 6) to get 50+; then JJB is required either to play suq after eruv for points (already discussed), or to block immediately (and no V plays block those 3 plays). This completes proof that JJB's best play is eruv and nets him 13. (6) In the 1:11 actually available (!), JJB played qua 24. IJB could now stick him with the V and get 35 net, so JJB's net is -11 and is 24 behind optimality.
>IJB: ACSTZ 13J ZAS +22 292
#note ~1:25 [~3:50] IJB now has an easier task: merely blocking eruv and sticking Dad with the V (or else getting more points going out in two, or some other such trap). We already established that he can simply use mot 5; it leads to un 12, zas 34, de 10, code 10+8, net of 35. Blocking zas with dunk leads to cuz, which is worse. What if he threatens out in two instead? Za 15, eruv 9, pacts 20+6 is net of 32; za 15, pan 5 (blocking pacts, which is preferable), mot 5, doh 11, code 10, nu 2, si 10+8, net of 30. The other setup is IJB's actual play: zas 22, eruv 9, act 5+6, net of only 24; zas 22, dunk 18 (scoring and self-sticking), code 10, pass, te 6+8, net of 28; zas 22, ands/unde 14 (scoring without V), mot 5 (blocking eruv), nu 2, code 10+8, net of 29. So, counterintuitively, mot is IJB's best play and nets 35 as a classic V stick. IJB's play is 11 behind at net 24; despite containing an out in two, it doesn't go in for the kill when the V is susceptible.
>JJB: DNUV 6D ..UV +9 397
#note ~0:11 [~0:10] Forced, and obvious even under duress.
>IJB: CT C7 C... +10 302
#note ~0:40 [~3:10] It appears JJB has spotted act and has run down to about 10 seconds because IJB can go out for 5+4. But IJB still has one more trick up his sleeve, a psychological one, an area of game theory that JJB is learning from him. IJB declines act and goes for the mathematically inferior code 10, which should be prey to ands/unde 14+2, net -6, 15 behind. It has its effect.
>JJB: DN C7 ....D +2 399
#note ~0:20 [-0:10] JJB, catching up paperwork, is unnerved with the need to play again and made no preparation for ands/unde, nor even spotted and, where act had just been spotted. He throws down coded/dis 12, allowing te/ta 6+2, net of 4, compared to ands 14+2=16, a mistake of 12 net points on turn. More important, he loses 10 time points (included in score here), so that IJB's cleverness results in 10-12+6+2-(-10), 16 net, instead of the 9 net from act, and IJB becomes the only player to run JJB over on time all day.
>IJB: T 7M T. +6 308
#note ~0:40 [~2:30]
>IJB: (N) +2 310
#note More important than the game result is the amazing number of imbalanced racks that required careful management at all game stages, providing great instructiveness especially in one of the most complex full-knowledge endgames ever analyzed. Also unique: the bottom two rows were never used. JJB in particular had many easy decisions jump out of hard racks throughout, and kept sufficient endgame control, while IJB often made slight nuances to the best-rated play. Points improvable based on tiles played: IJB 28, JJB 36. Points improvable based on values: JJB 52.0, IJB approximately 184.6.
|
