| Player 1 |
|
#player1 JJB John J. Bulten
#player2 TB Tim Bottorff
>JJB: EEIKLOR 8G ORELIKE +82 82
>JJB: EEIKLOR -- -82 0
#note 0:35 [24:25] The first game of the day is the last of the round robin, and the next has been paired with the same two players, as the first of a swiss that works like king of the hill; so they anticipate several games together as the top two. TB leads by a half-game, and the top two seeds also still have the best odds for the top two places. JJB decides to pay the expert opponent the respect of showing him his first rack, with a probable phony (making it harder to play an anagram hook tile: he sees T while TB admits not seeing it, and he hopes to calculate the others using -like, namely PS). With oke 8h 14 rating at 21.2, this phony has an excellent reward of 61.3 and risk of 22.2.
>TB: IQ 8G QI +22 22
#note 0:53 [24:07] TB is having none of it, challenges after 37 seconds, and puts out maybe the least helpful word possible.
>JJB: EEIKLOR 9F KI +19 19
#note 0:27 [23:58] It's unclear what keeps the O on JJB's rack, when it adds 1 point to the play and 5.9 in obvious static value to the leave (koi 9e 20).
>TB: GORY 7H GORY +12 34
#note 1:42 [22:25] Not anagramming to gyro has an apparent cost of 3 points, with some defensive value achieved in column K and row 8. Sims flipflop on which has the better win ratio; gory might hold it by a percent.
>JJB: EELORRT K4 REL. +14 33
#note 2:09 [21:49] Oddly, toyer 16 and rotely 18 tie, 1.4 ahead of rely. JJB could take either, but preferably rotely. Instead, the O stays put; JJB is even considering girl 12 but not giro 12.
>TB: AEP 6I PA.E +16 50
#note 3:04 [19:21] Slightly better: aper j4 18.
>JJB: AEFOORT 4J F.O +12 45
#note 2:30 [19:19] JJB sees the 8s with BC (not those with GW), knowledge that will prove useful later. He considers that kief beats kif or kef, but then takes the best play assuming AEORT. Again one O holds steady, even when foo/or 16 is an obvious leader in retrospect (10.3 diff).
>TB: ATU I2 UTA +11 61
#note 1:46 [17:35] For this dump uta/ka gets 13, but the hook plays into JJB's O either way, and there is no real value in shutting down the fro hook. Uta/palet 11 is worth considering.
>JJB: AEMORTW 1H WOMERA +40 85
#note 1:07 [18:12]
>TB: ELLTY 2K TELLY +22 83
#note 1:15 [16:20] If TB has the blank and no bingo (such as artillery j6), palely 6i 19 is demonstrably better, rating from 1.0 to 4.1 higher, but it is not best in every scenario.
>JJB: AEJRSTV E5 TAJES +31 116
#note 4:45 [13:37] JJB weighs another well-balanced rack and decides to take the S points (although vaster 33 is somewhat higher for that). Playing from 5k is tempting but goes nowhere. The hook he didn't see was taj/palea m5 25, up by 11.8, a significant oversight.
>TB: ?ABGINR M5 BRAvING +68 151
#note 1:04 [15:16] Now of the 29 bingos the best is qis/sabring 72.
>JJB: BELORRV 8A ROBL. +24 140
#note 1:39 [11:58] In another big miss, breve 42 is strongly indicated by 16.7, but does not suggest itself when the spot is first analyzed. So far JJB's play is not up to yesterday's standard, and TB's is.
>TB: DDE O1 D.ED +33 184
#note 1:27 [13:49] TB coolly seizes the triple JJB could not advantage.
>JJB: AEIRSVW F4 VAW +32 172
#note 0:52 [11:06] JJB walks backward into another bingo rack, and the correct dump to convert it is only a beginner-level problem.
>TB: NNOU D3 NOUN +17 201
#note 1:16 [12:33] TB is thankful for the added points to his low-value rack.
>JJB: ?DEIMRS B7 M.RtISED +70 242
#note 5:27 [5:39] JJB faces a crucial choice among 26 bingos because he wants to preserve the lead after playing the last R. The primary choice is playing upward (closing) or downward (opening). He misses the best two plays, misorder/disarmer 80, which would definitively settle it upward, but the other slots he seeks, moidores 78 and bromides 76, would be acceptable downward. It's just not on the study list for now. He does not see better than mortised downward or admirers upward (marrieds and emeroids are too poor), and after much seeking makes a gut decision to open (10 under max) and hope to bingo again (above or below) given the bag.
>TB: AGIO A10 AGIO +15 216
#note 3:13 [9:20] Ago c3 19 gets TB 1.9 better partial value, but one often cannot tinker with vowel dumps. Opponent is holding a second bingo through the S and wondering what possible play might make sense to block it: the answer comes, a vowel dump could easily.
>JJB: ADINOTU N9 DUO +16 258
#note 1:15 [4:24] Since the bingo through the R doesn't fit and the S is now blocked too, JJB decides it's fishing time again. He misses column C (udo 19), and the fact that duo rates second does not help his misgivings about the opening holding his lead now. Udo would probably improve his endgame.
>TB: EINPS O11 SPINE +34 250
#note 1:36 [7:44] JJB has just left two good S hooks open without holding either S, and so TB moves swiftly to decommission the more dangerous one. Quackle assigns another 1.8 to pie/alp/li 22.
>JJB: AEEINTT N13 AIT +14 272
#note 1:29 [2:55] Search the board! Abet c7 18 yields the optimal EINT (+8.6), with EENT only +3.2 and the play 9.4 behind. Blocking is unnecessary as the bag gets heavier per capita.
>TB: ACEIOUV A3 VOICE. +16 266
#note 4:24 [3:20] TB struggles between cover and voicer, but the leave guides him correctly there. Quackle gives vau c1 16 the lead by 2.2; the opening risk is not severe and going out seems likelier. Second place is the Quackle trick of trading all with 8 in bag, worth 10.0 by default. TB is also correct that two case vowels are not a liability with 4/15 unseen vowels remaining.
>JJB: EEHINNT C3 EH +19 291
#note 0:44 [2:19] JJB makes the right static decision but endgame considerations now invade with both players low on time. He considers emptying with hen c2 18, but that directed hook can backfire. Only after his play does he realize that column O will eat up the rest of his lead, and no block is obvious. He is also not absorbing the fact that the board now strongly resists plays of 3 or more tiles. There is a low-score but thorough block of column O, ain 7m 3, but it does not seem to help in simulation. Eh 19 is still favored on points and leave at about 65%, if he can play out the rest effectively (not affectively!).
>TB: AHSTUXZ O7 ZAS +40 306
#note 2:36.7 [0:43.3] TB is blessed with SXZ together and sufficient synergy, and the high-score play is self-commending. Quackle ranks it along with thus/huts but the static value does not reflect the need to peel XZ quickly. Zas provides a mathematical win that only needs conversion into reality; thus/huts might not.
>JJB: CEFINNT 9E ...NT +14 305
#note ~2:09.0 [~0:10.0] JJB seeks too long for a lengthy play; the conflicting goals (catch up, go out, set up, block X) don't come together in findings, and enough cogent preparable plays do not exist. Under time pressure he hopes that skint 14 will help him find a killer followup. (Consider the never-seen knife f9 14, but note that its setup of cent is blockable with ex 39.) Neither player has out in two without someone contributing a setup. JJB can still control pace with a blockable setup similar to knife: centai 16, xu 20 (blocking fine 14, and setting up better than thee 14 with fit 6 in reply), if 19 (rejecting ineffective blocks), khet 19+2, which loses by 21 (net of -6). Another response also barely wins for TB: centai 16, hex 13, if 19, skite 14 (because going out with duet would tie), gin 4+2, which loses for JJB by 1. These limitations provide hope that the opponent might miss the only saving sequences. For alternate paths, first, eft 17 can be parried with uh 19, in 8 (blocking ex), text 12+8, losing by 29. The play of if 19 surprisingly sets up better than it allows: ex 36, centai 16, the 12, skin 12+2, losing by only 14; so the setup of centai needs blocking: ex/egos 23 (canceling the out in two), skint 14 (blocking skit), uh 20, exec 13, alt 6+2, losing by 20. Ef 19 does not do as well because of leave, and ficin 10 is parried by ex 23, exit 24 (blocking hut/eft), un 19, skin 12, skint 10+2, losing by 23. With this partial analysis, it appears that, algorithmically, the choice of if 19 is mathematically the best (net -5), but, heuristically, centai 16 (net nominal -6) forces opponent to choose xu 20 or else hex/skite to retain the win. Against an endgame-based computer, play if 19 to threaten centai; against a human, try centai 16 first to threaten fine. JJB is right to seek a long extension after all, but does not find centai and its entourage, or successfully play into that set by chance. Under time pressure he makes the wide opening skint 14; it has net of -15 according to the next calculations, or 10 behind if.
>TB: EHTUX 10I EX +27 333
#note ~0:13.3 [~0:30] Skint still has useful leave despite its setup nature, and JJB need not be so disconsolate about it. TB's ex 27 allows if/xi 19, uh 19, cete 9, alt 6+2, net of 26 (winning by 27; also fin/fen 11g 9, uh 19, cete/cite 9+2). Other responses to ex allow too much power in reply. The somewhat better response for TB is to take uh 19 first (setting up text 11 and ex 36), ice 13, ex l11 23 (blocking fen), fids 8, alt 6+2, net of 29, winning by 30. Or, in blitz terms, just note that uh sets up ex/xu and you can immediately keep the X safely.
>JJB: CEFIN 11J IF +9 314
#note ~0:45.0 [~-0:35] (Overtime 10-point loss included in score.) As before, if/xi 19 nets 1 to JJB; though other plays look tempting, they allow better replies.
>TB: HTU 2B UH +19 352
#note ~0:10 [~0:20] Now the right play becomes clear.
>JJB: CEN 1A EN +8 322
#note ~0:19.9 [-0:54.9] JJB misses cete (9 leaving 2) for en/nu (8 leaving 6), diff of 5.
>TB: T 14M T.. +3 355
#note ~0:07.9 [0:12.1] TB under the same pressure misses alt 6 for tin 3, diff of 3.
>TB: (C) +6 361
#note A heavily fought, mostly even game leads to a consonant-heavy deep endgame. TB, whose divergences from Quackle usually have good heuristics, consistently plays more accurately than JJB, who makes several intermediate-level errors; this is no time to play an opening phony, to miss palea or breve, or to hold suboptimal "sorta" bingo leaves. After the well-balanced voicer 16, TB is rewarded by SXZ together, making the endgame his to give away, and he holds on successfully while JJB mishandles analysis and time. TB raises his tournament lead to 1.5 games. Known points available: TB 17, JJB 38. Overall points available: TB 22.9+, JJB 116.7.
|
