| Player 1 |
|
#player1 Joel Joel
#player2 Gibson Gibson
>Joel: FNOPTTV -FNOPTV +0 0
#note This is Round 25 of 2016 NASC in Ft. Wayne, Indiana. Notes by Joel Sherman
20k 4-ply sim ranks top 4 keeps here as NT, PONT, NOT, and T; but Duck draws so insufficiently randomly that this order is the complete opposite of correct. Keeping any more than one tile here defeats the purpose of trading -- it decreases the likelihood of fixing the imbalance, and of drawing ?, S or other power tiles. There is no multi-tile leave from this rack that bingoes in real life drawing often enough to justify eschewing the primary motives.
>Gibson: CINNUUY 8E UNCI +12 12
#note You've gotta be kidding me! Trade, keep N, IN or CN are about equal, toss a coin.
>Joel: EORRTTU H3 TROUT.ER +59 59
>Gibson: HNQTUWY 5G T.WNY +22 34
#note Best.
>Joel: AEELMRS E5 SER.MAL +36 95
#note 10K 4-ply sim claims AM i8 21 beats ACME 24 by 44.8-44.5 equity margin, but ACME is in fact the best play if I want to go for a bingo, as it sets up a place for S ender. TOWNY screams David has no vowels, so I can be less wary of spending an extra vowel myself -- same logic applies to SERUMAL, which I chose to go for blanks first with a larger lead on quieter board in case I miss them, tho sim claims this is 7 points worse in equity (Duck really likes to fish, I don't because the Duck is a lot better drawer.) I also considered AMEER i8, but passed it up because of the X and Z setup on 10j, and generally more volatile board when I have a moderate lead.
>Gibson: AEHINQU F8 .AH +31 65
#note Also best.
>Joel: BCEEESU D4 CUBE +24 119
#note Saw possible BERCEUSE here, no such luck on the board.
>Gibson: EIINQSU J2 QUI.INES +89 154
#note After opening UNCI with NUY leave instead of trading, David is the only opponent I know who winds up with this result only 3 turns later, regularly.
>Joel: EEEGLSS D10 LEG +16 135
#note Saw eDgeless and elegIses. 18k 4-ply sim inexplicably rates adding ES to my play for only 4 more raw points as being nearly 3 equity points better. Having a place to start 7's with S and holding S's for both ends of words is surely pointing to about a 10 point error in sim.
>Gibson: ADDENOV K8 DOVEN +23 177
#note Here David gives up 13 points (vs. EVADED 8j) for what he thinks is a tighter board, one that is now crawling with S-hooks, and he's not holding one, and he could easily have inferred from my LEG play that I am. Why does this work for him?
>Joel: EEMSSXZ L11 MEZE +38 173
#note According to sim, I'm going to hit AX 14i for 61 often enough to make MEZES 13g 39 very nearly as good.
>Gibson: AABDEEV M12 BADE +44 221
#note Slightly annoying that he got more out of my Z than I did, but "luckily", I have not 1 S but 2, so I can pluralize MEZE *and* NEB or LEG, cuz I was not dumb enough to play LEGES. Had I played MEZES 13g, he could stuff the AX threat with BEAD h12, leaving my remaining S totally useless as well.
>Joel: ?AASSTX 15L S.XT +53 226
#note It's hard to complain at this point, I'm getting plenty of power tiles, and scoring well enough to be not worse than even; and I've made 7 out of 7 best moves, not as tho they were difficult however.
>Gibson: ?AEEFRV 14K R..E +28 249
#note This position is so complicated because of David's blank and the limitations of the board that sim ran verrry slowly. Duck greatly favors fishing for bingos here, with FA or FE c7 or FEED 14j. REDE is several points behind those, but is certainly stylistically fitting with David's game, taking out the TW bingos ending -NT and 9's into ED, and in the bargain, it nets 14 to 1 for raw value of tiles used, which makes it a move I might even consider, if I was already ahead by at least 40. However, from behind by 5, that clunky leave makes CRAVE g8 for 26, and setting up his F for 40+ plays next turn that are unlikely to be blocked with only 1 F and 1 H left in unseen 40 tiles much the best; and points in the direction of my own upcoming error.
>Joel: ?AAJLSW 2J .AJAq +40 266
#note Sim has a large spread (9.5 equity points) between CLAW and JAWS 13a 33, and QAJAq another 2 points behind JAWS. JAWS does absolutely nothing that anyone should want to do in this position, I get less than 2.5 to 1 for two power tiles and limit my blank to being an S next turn, while David gets 11 to 1 for (A/I/O)G or 15 to 1 for IN in return; so that's another clear sim error. Quackle's suggestion list offered several other <15-point options with J or W; CLAW and JAWS were its only 2 choices that scored more than 20; and it ignored QAJAq entirely. To be clear, I failed to find CLAW during that turn, because I am thematically trained to not bother looking for it when I have the even more bingo-killing J to deal with, but 15 seconds after I clocked QAJAq, I noticed the LAHAR insertion and realized I had missed CLAW, and knew immediately that it was stronger than QAJAq and indeed the best play, and I wrote a note to that effect on my scoresheet. However, there is also no doubt in mind that QAJAq is easily the 2nd best play available to me, both for reasons that should be accounted for in sim, and reasons that Quack is not capable of accounting for. It is mathematically sound in that it definitely gains points, both this turn and next for scoring well immediately and offering David no comeback, except when he already has the other blank. The amount of points gained over plays that don't use the blank (that I had found before committing, IOW other than CLAW) is at least 20 this turn and 21 next, for a total of 41, easily in excess of the generally agreed value of a blank. And the intangible, incalculable factor is to try to make the already defense-oriented David's head explode trying to defend the secondary threat without losing ground on the scoreboard immediately, when he doesn't have the other blank; and if I were to make the assumption that he did, there'd be no reason to bother playing the game, and if there's no reason to bother playing that particular matchup, then there is really no reason for me to bother showing up to the tourney. Without a "nothing ventured, nothing gained" attitude, I would have quit this game and slit my wrists many years ago, so I work hard at holding onto that.
My estimate of equity here is CLAW should only beat QAJAq by about 3.5, not 11.5, but I did not find it because my mind was not open enough. JAWS is the move that should be 11.5 behind CLAW.
>Gibson: ?AEFINV M1 F.VA +20 269
#note So far, so good, it appears, but...
>Joel: IILOPSW C7 POI +17 283
#note ... seriously?!?!? The drawing monster returns, and my venture is dashed upon the rocks, time for an Elton John sad song. POI does rise to top of sim, beating POW c7, PSI 36 and SOW 31 on QAJAq, as it makes an effort to keep bingo lines on multiple fronts, and retains the chance to play CLAW after all, and confronts the vowel problem in the pool better than POW.
>Gibson: ?EHIKNP 4L H.IK +32 301
#note While ICK g7 33 is cute, the leave can get clunky and diminishing my threat should still be a higher priority. HAIK 32 is definitely better than KAPH 34 for that purpose.
>Joel: AILOOSW B7 ALOW +25 308
#note Without R for WOORALIS, this is just getting worse, but I can split the TW alley to allow bingos on top or bottom and score okay doing it, really need to draw the P...
>Gibson: ?EEGINP A1 kEEPING +94 395
#note ... except he already has it, and everything else useful! Never mind! (in Emily Litella voice)
>Joel: AAIIOST A10 OI +7 315
#note Nothing to see here folks, you can all head for the parking lot now.
>Gibson: DDGILOR 9J ..LDI +8 403
>Joel: AAFIOST O4 .OFTA +36 351
>Gibson: DGORRTY N8 Y.RD +15 418
>Joel: AINOS O1 OS +27 378
>Gibson: GORT 2A .RGOT +12 430
>Gibson: (AIN) +6 436
|
