| Player 1 |
|
#character-encoding UTF-8
#player1 JJB John J. Bulten
#player2 AA Asif Ali
>JJB: AEHILNQ H7 QI +22 22
#note 0:37 [24:23] Now second and third seeds face off in the swiss. The opening rack does not take much time; inhale is 13 points behind.
>AA: DINR I6 RIND +22 22
#note 0:44 [24:16] Also easily indicated for an I holder, even if it invites a large hook later.
>JJB: AAEHLNU G8 HAUL +14 36
#note 3:39 [20:44] (aha j9 17 +1.0) With a deliberately closed board, leave is essential, so JJB prefers haul/hula over aha 17 or nah 20 using j10. He tries to remember what's weird about inhaul: the answer is that it's a noun but it takes an -er as a synonym (but not an -ed). This tempting extension setup, while he's not sure which is the extension, doesn't score enough to commend itself above the hook.
>AA: AFIRT H11 AFRIT +38 60
#note 0:27 [23:49] AA has the tiles to take the first initiative in lead.
>JJB: AABEINT 13D ATAB.INE +64 100
#note 1:05 [19:39] And this gives JJB all the tiles of retina, to make the only bingo easy.
>AA: DERW 14C DREW +33 93
#note 1:23 [22:26] (dew 14d 33 +1.1) AA is not one to let go of a 30+, so he makes a tempting triple opening. Keeping the R would score the same and be harder to exploit.
>JJB: ?AEISWX L8 WAXIESt +101 201
#note 0:25 [19:14] (wax f8 66 +1.8) In one of those rare cases, the obvious bingo for 101 is not the best move in static value: if JJB notes the wide-open f10, and builds on it to get to wax/whin/aa/xu, he gets 66 and keeps tiles well worth 36.8 in reserve. Further, these plays are very close in simulation, over 85% in win ratio, with wax 66 ahead of waxiest (l8 101 or j2 98) by a couple points. What expert has the guts to play wax here with confidence? Certainly the setpoint of IES? +36.8 is worth memorizing in a list of current static values.
>AA: FOY 12B FOY +30 123
#note 1:24 [21:02] (foy m7 31 +1) AA continues to take large steps to keep up, missing only the 1-point diff in m7 placement.
>JJB: ABEGHLL K9 BAH +39 240
#note 2:43 [19:14] JJB can also demonstrate good parallel synergy, while keeping the first G in reserve given row 5.
>AA: EIV 8L .IVE +30 153
#note 0:30 [20:32] Seal up the triple safely. If AA is struggling building bingo racks, or alternatively struggling with overloaded racks, he certainly isn't showing it.
>JJB: ADEEGLL N6 GA.ELLED +69 309
#note 0:31 [16:00] (alleged 5e 93 +24) Finding the 8 and missing the 7! The difference in win percentage is about 99% alleged, 98% gavelled. Statistically, if the spread were much lower then the win percentage gap for giving up 24 points would be much higher.
>AA: ALMU O12 MAUL +37 190
#note 1:36 [18:56] AA holds for a minute, during which time JJB spots his omission and writes "93!" on his scoresheet. Ah well. AA takes full advantage of another premium and is still in striking distance with good tiles and good judgment.
>JJB: AEGIMOO 5H AGIO +17 326
#note 3:21 [12:39] (main j10 36 +7.2) Starting to peel off time, JJB considers trading, but makes the best of the missed G hook, not noticing his synergy at j10 that's worth a little rack retardation at this lead. About to cash the hook, he is comfortable on turn disarmingly describing his missed 93-point opportunity to AA.
>AA: GIRT F5 GRIT +14 204
#note 1:22 [17:34] (git f6 15 +2.1) AA does not need to play the R in this line, as he has a slight point and leave benefit for not doing so.
>JJB: EKMOPSY 4J SMOKEY +44 370
#note 2:04 [13:05] (mike m7 51 +13.7) Faced with an embarrassment of premiums, JJB is comfortable scoring the S, and notes kemps/agios 42 but rightly prefers smokey 44. He didn't see the key premiums indicating mike/pike 51, nor the S retainers pokey/mopey in his row 4, nor the hook of kyar allowing pesky 46, all of which have better static value. Of these, pike (given another P in bag) seems to hold the win in virtually all cases, namely about 99.8%.
>AA: OOT 3L TOO +22 226
#note 1:16 [16:18] Now AA is compelled to discharge the two Os on his rack, but he still has very few bingo lines. Possibly took 16 should be considered since AA probably holds an S and/or blank while JJB is drawing a random rack unlikely to fit into the nonuple lane.
>JJB: DEJPRTT M7 T.RE +27 397
#note 3:17 [7:18] (tide m7 31 +7.6) Having finally seen the valuable spot, JJB has unaccountable myopia about playing off D instead of R.
>AA: EIV 15A VIE +21 247
#note 0:26 [15:52] AA must continue to balance his rack, as he is very near to a bingo and has little other win chance. However, opening up the nonuple with veto 11 (the simulation winner), or a double with veg 14, seem more hopeful choices (0.1%) than giving up the final I and shutting down a line that might be used more effectively later. AA has unconsciously shifted from seeking the win to seeking to minimize spread loss.
>JJB: DEEJPTU D12 ...D +20 417
#note 2:08 [5:10] (jug 5d 22 +11.4) JJB sees nothing better than to keep the board closed, despite the severe negative simulation value of holding the J. Tossing off joke 15 onto the board, or jefe 30 or judo 20, are closing moves too and do better at retaining the win in the squeeze cases.
>AA: NU 5D NU. +8 255
#note 1:37 [14:15] (nuthin 8d 10 +2) JJB's one-tile play has slight psychological effect, as AA stops the clock after 11 seconds to verify where the play was made. He could have a good line with the phonetic spelling nuthin 10, neglected by both players, but he is also directing a hook to his S/blank here.
>JJB: EEJOPTU 2N JO +38 455
#note 0:16 [4:54] (jupe 4a 46 +19.6) Low on time and having picked an O, JJB goes for the simple score rather than calculating out jupe/en 46 or jupon 44. However, it's true that those are a bit riskier and that a closing play makes the most sense, so jo stays ahead in win percentage. Joke 15 seems to win all simulations now.
>AA: ?CENRSS C3 RESCaNS +73 328
>AA: ?CENRSS -- -73 255
#note 1:33 [12:42] (cony o1 31 +35.0) Now comes the tragic flaw. AA decides the best-looking play is rescans* (not caserns, censers, censors, scanners, or scunners), and proceeds to claim his hook. But a bingo cannot win at this point anyway; the hope would be on a setup or score that somehow interacts with opponent to offer a higher return, including ers/sers in column O, a play ending with the S in snug from c2/c3, sensa/scena h1, or the play he also considered, cony/toon 31.
>JJB: EEOPTUZ D3 ZE. +24 479
#note 3:58.1 [0:55.9] (ponzu j11 44 +28.3) JJB remembers the ancress in the caserns (who is associated with scanners but not with rescans*), and can challenge confidently after a triple-check. Now it appears he must block all options, and uses all but his last minute to do so. Ponzu/phi 44 is tempting but fails this goal (also, zee 34 allows casernes, and puton 20 allows princess); ponzu d3 32 does the job, but cannot be verified quickly as it appears to keep a usable line open. JJB might have preferred zone to zen given more time. Ponzu 32 leads dynamic values, assuming one has time to trust it (ponzu 44 is oddly in second place, which is probably a Quackle endgame error because it still prefers cony 31 over bingos in reply even though that play cannot bingo).
>AA: ?CENRSS O1 C.N. +31 286
#note ~0:22 [~12:20] AA makes the indicated play, both in static and dynamic value, but no bingos are available for his power tiles given NNP in bag. From his point of view, he could still theoretically draw seizures, asperse, ospreys, netsurfs, ossetra/osetras, restuffs.
>JJB: ENOOPTU 3D .EP +14 493
#note 0:27.0 [0:28.9] (ponzu 3a 32 +29.0) Until he knows AA can't play something like princess, JJB will not volunteer ponzu 32 (even if he had seen it). Given that he controls the Os, he reasons in his limited time that opponent is unlikely to bingo in row 2 with the unseen tiles (which is correct).
>AA: ?ENPRSS O1 ....S +10 296
>AA: ?ENPRSS -- -10 286
#note ~4:28 [7:52] (superthin 8a 42 +57.7) AA goes on the wild goose chase for 4 minutes, but there is no anagram of persons, nor any fitting 8. He needed to get on the board by spotting the extensive superthin 42 here. Instead (whether misreading the empty bag or thinking he has time for more score) he misremembers a word, possibly molys, polys, or coadys, and assumes at great cost that conys* is correct.
>JJB: CNOOTU J11 CO.TO +17 510
#note 0:05.9 [0:23.0] (nuthin 10, posies 31, conto 17+6 +11) JJB challenges immediately and then goes with his prepared play, both in under 6 seconds total. (This also creates a configuration with 16 adjacent keyed letters on the right side of the board, a rarity; yet sers o7 would have made it 19.) However, if JJB had seen nuthin as one of the options for his leave, had realized the need to block superthin, and had inverted the order, he could still play conto safely afterward, or alternately toco 2e for one point less.
>AA: ?ENPRSS G1 PES +22 308
#note 1:24 [6:28] (superthin 42, nu 12+4 +25) Showing that you can always search even in a lost game, AA certainly had the time to search for superthin and risk it, or to take the more solid posies/pussies/pansies. Instead he adds 3 to opponent's response.
>JJB: NU 1G .UN +15 525
#note 0:02.9 [0:20.1] Scoring a bit more than nuthin; also beating the blocked nu 12.
>JJB: (NRS?) +6 531
#note JJB pulled ahead with three bingos, though missing a quadruple, and continued to score well to get his lead to 200, after which AA's cautious play was lost in misjudgments. But with JJB's need to play very conservatively and forgo scores like ponzu/phi 44, he could only extend his lead to 223. Both players had opportunities to extend thin to their credit (nuthin and superthin) but it didn't happen. Known points available: AA 28, JJB 35. Overall points available: AA 123.9+, JJB 154.6.
|
