| Player 1 |
|
#character-encoding UTF-8
#player1 LM Leslie Millard
#player2 JJB John J. Bulten
>LM: AGQU 8F QUAG +28 28
#note 0:50 [24:10] In the final game of the swiss, none of the top three seeds (each at 7-3) can play each other, so each gets an opponent with fewer wins and is seeking to maximize spread for the king-of-the-hill final. LM has 3.5 and gets to play JJB after she has been so helpful getting him board photographs. She has an easy decision at the start.
>JJB: AEILOPR 9E LIPA +29 29
#note 2:30 [22:30] Though it takes time, this rack resolves well by selecting best leaves and seeing the yielded plays that overlap best. Qi 11 rates almost equally well and wins almost as many games, for bingo hunters and fishers.
>LM: AHOT E9 .OATH +16 44
#note 2:46 [21:24] (oh 10g 16 +.4) Same score and better tiles is to use parallels rather than doubles; but it depends on LM's rack and she is doing fine with that.
>JJB: BEENORU 12C BU.EO +14 43
#note 2:52 [19:38] (boreen 14b 29 +3.3) Also better than doubling is using a hook to advantage, which JJB is not immediately alert to. Options include boreen, rouen, and unbe. JJB likes extending with upborne but it's way behind on score, leave, and placement. He considered lube until it was blocked, and unrobe doesn't play.
>LM: ADDEINR 11G DRAINED +71 115
#note 0:56 [20:28] (drained 13g 79 +8) LM jumps ahead in a hurry, without inverting the play to get the superior placement in row 13.
>JJB: AEENRRS J2 REEARNS +76 119
#note 1:13 [18:25] (earners 14a 77 +1) JJB also calculates too little, as he is proud to work out that j2 scores more than 13g, but that is still to neglect the hook available to earners 14a; opening up multiple triples at a time is usually not a drag on spread.
>LM: ADJMM L7 JAMM.D +38 153
#note 2:08 [18:20] Good management.
>JJB: CILNRST 4H CR.TIN +22 141
#note 2:33 [15:52] JJB word-builds, going from citrin 16 to cretin 22, keeping a good leave. While this could be a good bingo rack, fishes simply don't get much higher than lac/cal 10 and it's better to use up the points and redraw with the S.
>LM: AADGGO M3 A.AGOG +31 184
>LM: AADGGO -- -31 153
#note 3:26 [14:54] (dagoba c8 22 +30.4) LM misremembers analog versus anagoge; yet, if she had the E, anagoge 40 would be the best play! From recalled tiles, however, dagoba 22 ranks above the best parallel, gama 23, and there is no killer play. This means JJB is likely to pull ahead of LM's J play.
>JJB: EEILSTU 14D LEU +14 155
#note 1:21 [14:31] (jeu 7l 11 +5.3; leud m8 17 +3) JJB takes 32 seconds to hold and challenge, desiring to be extra certain, but the correct words here are just anagoge/s, anagogy, anagogic. He then proceeds to use the hook in rather wimpy fashion because his synergy is much improved. Leu 14 is better than lieu 15, but keeping the L with jeu 11 is worth even more, and he also misses the safe overlap of leud m8 17.
>LM: AADGGO 15C GAD +20 173
#note 1:132 [13:41] (gaga 15a 24 +9.7) Out of protectiveness, LM uses her D hook so as to neutralize both triples. But scoring with gaga 24 is not bad either, and dagoba is still available, and the triples are not too significant on an open board like this.
>JJB: AEEIIST M10 I.EA +17 172
#note 0:56 [13:35] (aide m9 18 +1) JJB could have used that L, having drawn all vowels; EILST with subdraws AE or AI would have yielded common bingos through R. As it is he must dump, and a little calculation would show that aide is 1 above idea despite not doubling.
>LM: AFGIO M2 GO.IF +18 191
#note 2:02 [11:39] Sometimes the lone double really is the best play!
>JJB: EIOOSTX N9 OX +53 225
#note 0:33 [13:02]
>LM: ?AEESVW O4 WEAVES +50 241
#note 1:20 [10:19] (weavers o3 91 +15.4) LM resolves to go into column O and lays down waves. She then takes some time to consider and can insert the E for weaves 50, hitting the premiums correctly for the second-best play. After clocking, she is consternated to find that she could have also inserted the blank she retained to get weavers 91! However, this omission is not the worst case, since a blank alone is still worth 25.6, meaning she has only sacrificed 15 points and can still hope to bingo. Weaves 50 and weave/oe 35 both rank some points ahead of waverers 71! The option should be based on game tempo and opponent expectation. Cashing both power tiles now is worthwhile, but both weaves and weavers top the simulation with 65-70% win chances, not knowing JJB's bingo-prone tiles.
>JJB: EINOSTV 6C INVESTO. +63 288
#note 1:02 [12:00] JJB finds the only bingo word and places it far away from the triples. Yet the wheels are already turning about C and T in column H.
>LM: BORW N12 BROW +26 267
#note 0:58 [9:21] (bow n1 27 +2.1) A tight choice, but it's probably better to keep the R and get the extra point with bow/go/ow.
>JJB: HIIOPSY 15K SHO.Y +54 342
#note 1:05 [10:55] JJB is glad the best play echoes his feelings about it, resolving not to sweat the leave of IIP -9.9.
>LM: EKU B13 KUE +20 287
#note 0:36 [8:45] (kue 14j 24 +4) Another effective hook, yet the K can be tripled at 14j.
>JJB: ?CEIIOP D3 PEI. +12 354
#note 2:39 [8:16] (epigonic 80 d1 +40.6; epic h1 24 +12) Now JJB's system fails him, as he has studied policier 63 and should have found it, while epigonic 80 is not out of his reach. Instead he opts to block as many bingo lines as possible given his lead and the unseen blank; albeit this should be done with the clever pionic 26, also available if he considers pion. Not time to use column H yet.
>LM: ENZ J10 Z.NE +33 320
#note 1:13 [7:32] (zin j10 32 +1.3) LM gives up the last E, the only real vowel on her rack, which would have helped her in the endgame, such as in column K or if she wants to hook zin later. Only three vowels are unseen out of 11 tiles, and it can be assumed JJB has a good balance of them (i.e. relying on drawing a vowel is an error because opponent has probably seen most of them in prior turns).
>JJB: ?CFIIOR I13 FIR +19 373
#note 5:18 [2:58] (fricot h1 33 +16.5; rif n1 23 +4) JJB sees the best static play of fricot 33, and is confident of the validity, but is still unjustly trigger-shy about LM's blank. He has tracked correctly but is conscious the player who empties the bag has a harder time going out. He busies himself considering multiple alternatives for each possible tile in the bag after the safe fir 19, convinced that it will be enough to guarantee the win; but he doesn't do the same for fricot 33 (or his other find, coif e1 18). But even if the worst case, Y, is in the bag, JJB can still play icily c6 or yince f2, which are close together but cannot both be blocked. He decides to stop looking with 3 minutes to spare, trusting that he has found enough options that his choice of fir will hold up well enough to win, though there's no strong reason to prefer CIO? over CI?. Simulation shows that if the blank is in the bag (actual odds closer to 1 in 35, not 1 in 8, since LM has likely seen the blank among all tiles she drew since her bingo), JJB can guarantee a bingo by playing the hookable ki/bi 20 (morbific, or frolics/fricots). Since there are 5 other chances to bingo besides the blank and they are very costly for opponent to block, ki might indeed be the best dynamic valuation, not because it makes a bingo usually, but because it threatens a bingo so effectively as to greatly handicap the opponent. Ki would be a masterful play.
>LM: ?LNTTTY C5 L.NTY +15 335
#note 1:37 [5:55] (nuttily c2 30, wilco a9 15+2 +17) Now finding nuttily is within LM's powers if she takes more time; it blocks opponent's best, colobi/ki 32, and clears all but one of her rack. If so, JJB has the choice of either wilco 15+2, or ocicat/chicot 27, because if he declines to go out then LM can do no better than debt 8+2 for the same spread. So with best play JJB scores the same whether he goes out or not.
>JJB: ?CILO F1 pOLIC. +9 382
#note 2:13.7 [0:44.3] (wilco a9 15+2 +6) JJB searches but does not find wilco, net of 17. He does see ocicat, one of his favorite words, but believes he needs to go out and takes police 9 (docile or collie would have been more fun, among other choices). Immediately afterward he wonders whether this had been a situation not to go out, because after ocicat 27, att/ay 8+2, he would have scored 4 more points! He is temporarily rueful not to have gotten to make the showier outplay, but in reality, after ocicat 27, LM can play att/debt/art 14+2 making it the lower spread for JJB. But recognizing wilco/ok 15 would have dispelled this dilemma.
>JJB: (TT?) +4 386
#note LM is already aware of her biggest opportunities this game, namely not phonying, and playing weavers for 91; her only other real miss was nuttily. JJB, however, failed to spot epigonic or policier, did not go for easy points with fricot, missed a setup opportunity well worth its power in the threat of s/ki, and gave up another 6 points for wilco. While tiles were about equal, LM can still win this matchup next time with slightly favorable tiles if level of play is not improved on either side. Known points available: JJB 27, LM 29. Overall points available: JJB 73.7, LM 88.3+.
|
