| Player 1 |
|
#player1 kacattac kacattac
#player2 kyan kyan
>kacattac: AAFGIST 8G FAG +14 14
#note ISC game. Rando opponent, so I entered myself as both players in order to progress past the upload screen here. (Help?) Late in the day when my brain tends not to be firing at its fastest. My offensive style has been getting me utterly clobbered lately, so I decided to try playing more defensively. I have just beaten a much higher-rated opponent by playing defense and drawing well, so I am feeling good about this scheme. Present opponent is very familiar to me: plays open, finds his bingos, not as good on in-between plays. So basically it's the perfect occasion to experiment with a more defensive style. Experimentation will have to wait for another play, though.
>kyan: AILMOOV 7C MOVIOLA +79 79
#note Nice word, nice placement. How come I feel like I'm already in a sticky spot even with TIRADES sitting on my second rack?
>kacattac: ADEIRST C3 READ.ITS +74 88
#note I spent forever on this. There is no truly "defensive" option here. Why even bother to have a Game Plan in a variance game? Quackle likes this or (M)ISRATED. D2 ASTER(O)ID scores 2 less and sims 3.5 worse. The other options are all poor. My immediate thought was to play ASTEROID because it blocks the hotspot at 8D. However, I mistakenly visualized this floating the A in row 1 rather than row 2, and this scared me off. I never visualized the whole word, just the D in hotspot, which is not actually where it would have gone! (I wish I could say I don't do this kind of thing in real tournaments...) I think that in my offensive style I would have played ASTEROID very quickly; I like the shape of it for offensive purposes. Also, after he has just turned over his entire rack and drawn from a nearly-full bag, isn't a clunky parallel play at 8A a massively greater danger than a bingo through any of the floaters I'm about to give him, no matter their placement? I tricked myself into making the play that Quackle likes, but I think Quackle might be wrong given a human opponent. I consider this a mistake.
>kyan: AFINQTT 8A NA.F +42 121
#note FML. He gets the hotspot and gets to use my I instead of his.
>kacattac: DEINOOR 6A ED.O +16 104
#note Again I spent a looong time grinding on this from the perspective of a self-consciously defensive style. Again it quickly becomes obvious that there is for the moment no point in having a Game Plan vis a vis offense or defense. In other words: there is no block to be made which doesn't leave or create an almost identical spot; and there is no opening to be made which is any more or less dangerous than the ones which already exist. The Game has dictated the Plan: get the next bingo down asafp. This is exactly what happens every time I make a Game Plan for a specific opponent, usually leading me to burn all my time early in the game just to settle on suboptimal plays. I decided I hated playing parallel in column B because it can give him another little slop play at A1 (more gratis rack-grooming services after his post-bingo draw), and that parallel plays in column D will just add score to his column B counterplays (ditto). Also, I noticed that I'm close to a bingo with this rack, hence (1) it's ok to sacrifice a bit of score now to maximize bingo chances next turn, and (2) the upper and lower column C floaters are my best bet for getting that bingo down, hence I don't want to take out one of those areas because it's possible that he will take out the other one. (Should I have inferred from the long time he took to play NAIF that there was a downside to that play for him, probably that the leave was bad and hence no bingo was forthcoming?) Eventually I came up with this. Quackle hates it. OR(D)O or DO(D)O in the same spot sim 2.5 better, scoring 2 or 3 less but keeping E instead of O and ostensibly not allowing as many big counterplays in column A. Just fishing an O at 6C sims around the same valuation as these with a higher win %. B4 ONO sims 7 better, and B2 RONDO wins the sim, scoring 28 points and simming 10 better than my play. Even though I'm both a musician and a basketball fan I might still have missed RONDO even in offense mode, or I might have balked at the 2/0 leave on this bingo-friendly board regardless of the far superior scoring. In hindsight I kind of like the fish from both stylistic perspectives. I reeeally hate ONO all around. I intensely hate it and I don't care what Quackle says. I get why it sims well, yet the mere thought of it triggers a deluge of traumatic memories: counterslop from A1; he bingos from the S and now I have blocked myself by taking out the R and E floaters up top; a 3/0 or 2/1 draw of 1-point tiles that don't go together. I am trying to learn to listen to my spidey senses on this stuff at the same time as I dip my toe into Quackle analysis. Even though I'm not good enough yet to justify my intuitions, it seems like intuition is ultimately a very important ingredient in really high-level play. So, I understand why EDDO sims so poorly, but I can't concede that it is quite as bad as Quackle says it is. Score, balance, and sychrony being the holy trinity of rack management, there are times, as here, when you can only have two of the three. Usually I can figure out which of these considerations Quackle (or a human analyst) is proritizing and subsequently relate that to the board and the bag. Here, given excellent floaters (I mean, I'm guaranteed either E or S! That's as good as it gets...), deep word knowledge, and a human opponent, I think it's okay to short (mildly!) on synchrony in favor of score and balance. Having spent the quarantine making and studying word lists, I think there needs to come a point where IONR is not actually 4.5 points worse for me than EINR. (Especially not when there is an E floater!) I have lost way too many club and tournament games clinging to Es while opponents with minimal word knowledge outrun+block with little slop plays. Both RONDO and the fish feel like plays that win against mediocre opponents but not against good ones. One final (I promise) point on why that might be true, if it is true: against a "good" opponent, I actually *like* floating the E instead of the O in column A. Considering the board, the score, and the sequence, I think this spot is more valuable to me than to him: for me it's a little bit of insurance against worst-case scenarios, whereas for him it's just another little scoring spot that's harder to hit than the ones he already had. For him, playing at A1 blows up his rack and floats me his starting tile; a good enough tile to reach the E will be just as good for me to bingo off of. What are the real chances that this is his optimal play here? I love plays like EDDO as a little bit of insurance against the nightmare scenario where he bingos off the S and I can't bingo back. A1 MASHI(E) or GROSZ(E) types of counterplays feel waaay less likely than a bingo through one of the excellent floaters. But most likely of all, after NA(I)F, isn't it quiiite likely that a "good" player will be looking to play short again rather than long, to further balance for a subsequent bingo? All of this considered, RONDO feels like a massive loss of control for a play that despite its superior score doesn't even gain tempo, whereas with EDDO I anticipate having good options next turn in literally any scenario, even the worst-case scenarios. I realize I am not nearly good enough to think this, but I do think it often in such positions.
>kyan: AGIIQTT F6 Q. +31 152
#note I really struggle with tilting in sequences like this, even if the game is meaningless.
>kacattac: IKNOORR A1 ROOKI. +45 149
#note Now I should be throwing a party. But I was proud of EDDO and expected it to work out one way or another, so this wasn't even exciting. Perhaps the secret to not tilting is to be as excited about my own luck as I am indignant about the opponent's?
>kyan: AGIISTT J3 GAITS +21 173
#note He played a strong game, but this was his least strong moment. It came down rather quickly following two previous quick plays which can't have left much time for consideration. There's no defensive value to this play, or any other kind of value that I can see. I could have thought much harder than I did about what all of that might indicate. On the other hand, this is totally an online, mid-quarantine type of move that isn't really going to happen in tournament play; and that's fine, I've made quite a few myself. B10 ATIGI (+9.5) wins a 2-ply sim rather decisively. 9H GAIT is close behind. A purely stylistic, mutually defensible choice! I want one of those.
>kacattac: CGINRUV 4H CR.VING +32 181
#note Careful what you wish for. This play wins the sim, unless he has another S, in which case it is not as good but still defensible. Admittedly, I tunneled on it and didn't think through it nearly enough. After a lot of grinding on READMITS and EDDO, I am sick of grinding and desperate for points. 4C (E)RUV keeping ICGN scores 18 less but sims only 2 less; giving him an S it sims 2 better, cowinning the sim with 41 C(A)RVING for 30, which instead of opening a huge S backhook merely floats a G in the TWS lane (ho hum). I'm amenable to any of these plays. I tend not even to look for plays like ERUV because it clogs the board without scoring much, and I am behind and need to bingo asafp; but it has strong merits here, and if I had the willpower to stick to my Game Plan I might have seen it. I definitely should have considered that it blocks slop plays in the D column: after CRAVING, he can score there *and* take out the dangerous row 1 lane, and then I am actually worse off than I was this turn.
>kyan: AEINRTU 9H TAURINE +67 240
#note Can't do anything about that. This is his only bingo and he has to play it, but it doesn't score much, it gives back a lot, and the game is barely half over. I need to be calmer in these situations.
>kacattac: JLNPUWY N1 JUN.LY +50 231
#note Instaplayed. Probably a mistake, definitely of process if not of result. 8L PUNY keeping JLW scores 3 less but sims 4.5 better. JUNGLY is very competitive with it in a 2-ply sim, but I switched to 5-ply when I considered what PUNY does to the board: I really hate that it kills all the new bingo lanes while I still trail decisively and there are sooo many little slop spots all over the board that he can pretty much coast on his current lead. I am willing to believe that creating the J hotspot is the greater long-term evil; that seems like just the kind of problem that a computer will be much better at solving than a human. I also failed to consider that he is very likely to play in the PUNY spot next turn: often it is best to simply concede that these lanes are going away, and that if they must go away it is best to jump on the hotspot yourself instead of just letting your opponent have it (the gratis rack-grooming problem again; post-bingo, that spot is probably very valuable to him). This comes up often enough that I have started to think of it as a general principle, and when I figure it out it usually proves to be a good play. It requires the player to assume an outlook which in moral and political thought is called consequentialism. Since I despise this way of thinking and those who engage in it, sometimes I play bad scrabble. Anyway, JUNGLY is too cool to pass up; that's good morality even if its bad scrabble. Max score, max lanes, max turnover into a bingo-friendly bag with both blanks unseen, and, in real scrabble though not here, the possibility of drawing a challenge. I think that makes JUNGLY defensible. For equity-loss purposes let's call it a tiny mistake.
>kyan: EEENRSZ O9 SNEEZER +109 349
#note FML. PUNY blocks this. But if I had played PUNY and said that blocking fresh-from-the-bag S-bingos in column O was the marginal consideration which swung the decision, wouldn't the probability police be issuing me a citation? As it happens he also has (S)NEEZERS at 10C, though it is easy to miss it there. But the really cruel thing about this is that a mere 80 point bingo in the same spot only puts him up 90 and blows the board wide open with plenty of game left. The challenge of that scenario is more psychological than strategic. I'm very capable of winning games that blow open late like this; in fact this probably is more favorable to me than the other likely scenarios. He is going to have big tempo regardless, and lightweight S-bingos in this spot create the lanes I need to come back. Isn't a PUNY-(S)NEEZERS sequence far more devasting? Isn't something cluttery like HUBS/H(I)/U(N)/B(E) for 54 much more devastating than, say, the massive volatility of something like sLATIER/(TAURINE)s for 78? Well, what actually happened was a #$%&ing black swan event. He gets super-max value from his Z and his bingo tiles in one lump sum payment, and he gets max turnover into a bag that is still looking incredible. It puts me down 118 with 27 tiles in the bag and a wide open board. Next turn's optimal play sims at 2.7% to win.
>kacattac: AILPTTW 13M WI. +30 261
#note Mistake. I end up getting away with this one, but even given my slim chances to win I think this is my worst mistake of the game. WALT(Z) in the same lane scores 4 more, keeps better, achieves the same positioning, and outsims my play by 8 pts and 0.6 win%. 14J WATTL(E) has high valuation but same win% as WI(Z). Really WATTLE is terrible, basically conceding the game. WIZ is better, but I played it too quickly and didn't look for better versions of it like WALTZ. No question row 13 is the place to play: it leaves me floaters in rows 1, 11 and 15 and creates hookability off either side of the W. The reason I played WIZ quickly is because it burns only two tiles; I need to bingo twice and the bag is likely to empty before that is possible to do. I am aware that some players strongly dislike this consideration. I think it is legit, but this time it did lead to a mistake because the lopsidedness of the score put me in the mood to sulk rather than grinding out the optimal score-leave-position play. Sometimes scrabble, like rebounding in basketball, is really just about effort.
>kyan: CINPRTX 8L XI +36 385
#note This sims second best to 2C PRINT/P(READMITS), which I think anyone can be forgiven for missing. Personally I think the third best play, N11 TR(I)P, is actually the best by far. If you had let me choose here whether he plays TRIP or literally anything else, I would have jumped at the latter. I don't see how I have any chance at all to win after TRIP, especially with him saving the XI. Also, keeping 0/5 here is brash, even given his lead; I've blown several club games against experts that way. The bag is only slightly vowel-heavy and there will be 21 tiles in the bag after he draws. Balance is still of huge importance, and turnover really can't get any more valuable than it is here.
>kacattac: AEELPTT 12G PALETTE +75 336
#note I draw exactly what I need, and my tile-conserving strategy seems to have been worth it.
>kyan: ACNPRTU 1L PU.A +48 433
#note FMFL. It's kind of insane that with three Os and two blanks in the bag there have been four plays made since JUNGLY came down without either of us even having the option play up here. Does anyone else ever wonder about ISC robodraws?
>kacattac: EHILMOY D10 HOMILY +41 377
#note Ok, I resummoned my commitment for this play because it's really my last chance to have any chance. My preference is for a short scoring play, but there are no good ones. Quackle's favorite short play is 10J HE, but it is objectively terrible compared to longer alternatives. I didn't even like HOMILY but I couldn't find anything I liked better. Now I have to hope that he has blown up his rack too badly to respond strongly to this; if he has any response at all I just can't win, which is why it's not a great play. The only alternative I almost talked myself into was 3L MO(N)Y; it sims very poorly. D10 HOMY is Quackle's best valued and second-best win%. I really strongly dislike it: few points, more clutter, again. D1 YLEM is valued a point worse than HOMILY but wins the sim decisively by win%. The leave is bad and it blocks a great lane, but when you're rearranging the proverbial deck chairs, I suppose it's possible for play like that to win more often than the others. Hey experts, how do you calculate your equity lost when (a) valuation, (b) win%, and (c) their sum, are triply conflated, as they are here?
>kyan: CDENRST 15C S.NCED +36 469
#note GGz
>kacattac: ADEEHOO 10J HE +28 405
#note I thought maybe I still had a swan's chance to win if I can score enough with a two-tile play. Looks like not. At least I got to 400 in ten turns with one S and no blanks. Do I get some orange slices for that?
>kyan: ??BERTU 1A .EBUTteR +77 546
#note FMFL scrabble GTH
>kyan: (ABDEOOW) +26 572
|
