| Player 1 |
|
#player1 AR Anita Rackham
#player2 JJB John J. Bulten
>AR: APPU 8G PUPA +16 16
#note 0:35 [24:25] Straightforward opening to a game that remains suspensefully neck-and-neck the entire way.
>JJB: ?ABJORT J7 J.B +12 12
#note 3:31 [21:29] With lots to think about, JJB goes for leave (score 12, but leave +35.3) rather than best combo of leave and points (jota/jato 25+29.7, difference of 7.4 in estimated available points). Jabot 9d 22, jot 18, and jo 15 are close to best.
>AR: AEFRT K9 AFTER +20 36
#note 0:29 [23:56]
>JJB: ?ADORRT 12D pROTRAD. +68 80
#note 2:34 [18:55] JJB sees many bingos, knowing the 5 through E+P, and eschews toreador 70 (2 points higher, also roadster) for one that might tempt a phony S hook. S hooks prove important later!
>AR: AELNNRT L3 LANTERN +71 107
#note 0:30 [23:26] AR has also been rack-balancing and, unfazed, plays the only bingo.
>JJB: AAGIIOV H10 VI.AGO +30 110
#note 0:47 [18:08] JJB starts making up for lost time.
>AR: AEIRWY 3G WEARI.Y +36 143
#note 0:55 [22:31] AR turns over 6 tiles in reply. Both draw poorly.
>JJB: AIIKMNW H1 KN.W +45 155
#note 1:07 [17:01] Wiki k2 32 has a much better leave and rates 3.5 higher statically. Points are fine, but dumping I would be nice, as rack imbalance continues to persist after. JJB looks for forms of manikin (5f 26) but does not find that play, and it counts as inferior anyway.
>AR: Q -4 +0 143
#note 0:13 [22:18] Rack unknown.
>JJB: AAIILMZ E10 ZI.AM +32 187
#note 2:31 [14:30] Difficult choice eats more time. Ziram/mirza are 2.6 behind azan 2e 38, but also in the cluster of premiums are zig, tiz, zamia, and banzai.
>AR: EHISSST M7 HEIST +31 174
#note 1:53 [20:35] Playing off EST is unnecessary, given ahi i12 22 and hijab. This is probably where AR later recalls drawing three S's, which means she has shiest 15a 47 for 12.7 better value.
>JJB: AAEIILN K3 .LIA +16 203
#note 2:01 [12:29] Too much bingo-hunting and vowel-planning. At least JJB finds the highest-value slot in column F, with anole 32 or anoa 23, but correctly recognizes the leaves are horrid (anole is only .2 above ilia). But the better advice is just to find the real I dump, animi 14b 18 (3.8 diff).
>AR: EOSS 15D OSE +21 195
#note 1:40 [18:55] Quackle generally agrees with burning another S here, but rates os 20 at 1.4 better with partial rack.
>JJB: AEEGINU 14G A.UE +7 210
#note 1:30 [10:59] Now the vowel imbalance is hurting because eugenia(s) is the only bingo set and JJB has blocked the cogent row 9 with his weak Z play. Top combo is gauze 21 or enigma 14a 29, with ague 6.8 behind, all on points (JJB turned down eau 9 as too blocky).
>AR: EGS I12 .G.E +14 209
#note 2:28 [16:27] Copycatting seems the best play.
>JJB: EEGILNX 13K .EX +20 230
#note 1:43 [9:16] Good choice; nothing better with the X on this board. Heisting 24 is way behind.
>AR: EIMS 1F MI.E +10 219
#note 0:52 [15:35] Mi 14l is 9.4 ahead in partial rating and mise 25 also rates better if risk is acceptable.
>JJB: EGILNOT 4G O.L +13 243
#note 1:57 [7:19] JJB makes a key error and misses lentigo 65 after having put it on two fast-track lists (39.8 diff). Mentally, the rack looks inviting, and in fact also plays through the letters of MAW (not through K of course), but K gets blocked anyway and the invitation does not get answered. While he thinks opening with owl is rack-conducive, ow 10 is better and second behind lentigo. (After ow, 18 of 29 draws will bingo in row 5, although it's very block-tempting.)
>AR: CEITV 5C EVICT +26 245
#note 1:45 [13:50] AR declines to put out an anagram hook for JJB's new rack (gin-tone: drank).
>JJB: EGINNOT D3 GI.EN +18 261
#note 2:40 [4:39] Quackle still counsels using Z for ginzo 21 (JJB's play is 7.6 behind and too S-conducive). Sims also put tigon, tignon, or toeing on top at about 50%, perhaps on turnover for the blank more than anything else.
>AR: OORS 8A ROOS +26 271
#note ~3:00 [~10:50]
>JJB: DDNOQTU B6 QU.D +34 295
#note 0:15 [4:24] Best response is obvious and keeps nodose in reserve.
>AR: BI 10C BI. +14 285
#note ~5:05 [5:45] AR has time to search for bingos and still has hopes. Bi also plays simply at 2m for 15, and the bag is growing worse for JJB so risk is minimal.
>JJB: DFNOTUY 15A NOD... +21 316
#note 0:57 [3:27] JJB sees his nominal row-14 duty (e.g., duty 14l 28), and has no comfort from the added bingo risk in response (e.g., secondly o7; no triple-triples available though, thankfully). Fud 28 rates best statically (13.1 ahead), then fund 30, but they have much greater spread risk (e.g., cheloids 176). He believes his real duty is to block column A; but rechosen, reschool, roundels bingo in only 3/36 of cases, so foy 16 and nodose 21 (pretty except for its leave) are both way behind in winning likelihood. When an endgame is about even, blocking low-risk potential bingos may strain the gnat and swallow the camel; instead, keep scoring, and setting up scores. It is unclear what play would win the most of the 36 possible endgames. Fudgy 3a wins about 90% of simulations, but sims omit some tactical plays, and AR's actual rack can defeat fudgy with unfed 27, con 14, hos 29+4.
>AR: ?DEHNOS F9 NE +21 306
#note 2:15 [3:30] AR plays the endgame well enough despite her concern about them. She could get all 7 tiles decently into columns F and N, with the sequence redon 29, fly 16, hos 29+10, which nets her 52 and wins by 21. (JJB loses value to block with tutty 8 because of ohs 17+16.) She sees the slots, albeit not the blank potential; yet JJB's last draw is so poor that her play still nets 42 (diff of 10).
>JJB: CFLTUY 11A FLY +16 332
#note 2:27 [1:00] JJB has cruft, almost literally, and mutters, "Now what am I going to do?" seeing that hod is an easy 30 for opponent and intuiting correctly that the game is mathematically lost. He believes himself compelled to play fly/by to block reshod 12, not suspecting that options are so limited that it is also the best rated (fluty/fluyt go nowhere unless column N should get used). He searches for an out in two that does not exist, only because other theoretical boards would support it. Only a partial out in two can be found, and no blocking play actually beats the default fly/cud (T) for points (e.g., fun/rely (holding CT) is 7 behind); so JJB's instinctive guess of best play is validated, even as he is about to make a different endgame error.
>AR: ?DHOS N9 SHOD +38 344
#note 2:04 [1:26] AR admittedly believes she has only one way to win now: cashing as many tiles as necessary. It appears that the highest score is 30 for hod n10 or holm 14b, and the latter also blocks cud 12 for a slight edge, making JJB's reply cod 12, followed by reds 15+4, net of 37 to AR, winning by 11. (Hod 30, cud 12, banish 15+2 wins by 9, and other wins exist.) But AR throws in a monkey wrench by gambling on mis-pluralizing a plural to get 38 points, with shod/banis* (leading to cud 12, agues 13+2, and netting 41 if accepted). JJB should simply challenge off, leading to the calmly visible cud 12, hod 30, att/theist 12+2, net 4 to AR, enough for him to win by 22. This means that, if challenged, she would be 33 behind best play (holm), and, if not challenged, she is only 4 ahead of best play. (The metric of unclaimed points available is designed to yield higher numbers when players make weaker plays and leave more points "on the table"; when a player successfully phonies, this can be rewarded by subtracting the opportunity points gained from the unclaimed points metric, because they were psychologically claimed even when they were mathematically unavailable. By this standard AR's play cancels out 4 unclaimed points with 4 claimed opportunity points, otherwise unavailable.)
>JJB: CTU C13 CU. +12 344
#note 0:52.7 [0:07.3] Flustered by not even having considered "shod plus plural", JJB takes none of his last minute wondering if the plural is phony; he knows the definition of bani, and that the set is just basin/nabis/sabin, but the play was carried off so calmly and his idealist outs are so unforthcoming that he doesn't even test the perpendiculars for validity, nor recall that he had ruled out row-9 hooks wordlessly. The only path to win is invisible to him. Instead of net -4 (shod 0, cud 12, hod 30, att 12+2), he allows shod 38 without even holding, grudgingly plays cud 12 in the last seconds, to which agues/afters 13+2 would net -41 to him.
>AR: ? 1F ....s +10 354
#note 0:12 [1:14] AR in turn does not take the minute to find agues for 3 extra points; but any play wins after JJB could only tie the current score.
>AR: (T) +2 356
#note After exchanging early bingos, both players are tested by a series of vowel-heavy and imbalanced racks. JJB at least misses one bingo, and then is more hurt by draws and miscalculations in the end than AR; she has a won game once JJB empties the bag, then yields the possibility by playing a phony, then gets the win back by JJB being too distracted to think about challenging. This is not a "swindle" game, because she has the win without a phony and it happens before the bag empties. Rather, this is called a "double-blind" ending, where both erred. So, does Rackham mean Scrabble connoisseur? Known points available: AR 10, JJB 41. Overall points available: AR 33.5+, JJB 123.6.
|
