| Player 1 |
|
#character-encoding UTF-8
#player1 Mack_Meller Mack Meller
#player2 David_Weisberg David Weisberg
>Mack_Meller: KLOORTV H7 VOLK +22 22
#note Actual racks courtesy of Rack Teller
>David_Weisberg: FIILUXY 8G F.XILY +23 23
#note Actually rather poor. I missed 10H KUFI which sims well ahead of anything else. I8 XI 22 comes next at almost 6 points behind. The elegant 7H VILIFY 20 is about 8.5 behind KUFI, though it doesn't look too bad to me. Even 10E FLUKY 17, at 9 behind KUFI, does better than my play, which comes 13 behind, perhaps using the X too cheaply and keeping the dreadful IU combination.
>Mack_Meller: IOOORST 9K OOT +11 33
>David_Weisberg: ACIINRU M9 .UNIC +14 37
#note I forgot to write down my rack but I think the leave was AIR. I remember considering UNCIAL. Since I'm not sure of the whole rack though, I'll forgo analysis for this turn.
>Mack_Meller: AINORSV 7F VI.A +28 61
>David_Weisberg: AIIJMQR 12L Q. +22 59
#note I really hoped to draw an O or U...
>Mack_Meller: INNORSW 6I WORN +22 83
>David_Weisberg: AIIJMRU 14J JUMAR +68 127
>Mack_Meller: ?GILNNS O8 LiGNINS +89 172
>David_Weisberg: ABIIOST 15H BIOTA +40 167
>Mack_Meller: DDENRTW 10L D.N. +15 187
>David_Weisberg: EEINSUZ 10F ZE. +36 203
#note Sim prefers N6 ZEE by 4 points, maybe because ZEK interferes with my S hook to VOLK. He did just play off only two tiles though, so I'll stick with ZEK.
>Mack_Meller: ADEMRTW 11E DAW +30 217
>David_Weisberg: EGINSSU 12F GENUS +26 229
#note This turn felt difficult. Some very strong tiles and at least a couple I need to shed, but no good way to open the board for quick bingos. The computer likes hanging onto both S's with 12F GENU 25, with the opposite kind of play, using both S's for maximum points and turnover, 12A GUISES, coming a point behind. That play feels like it's asking a lot of my next draw and giving him a chance to score equally in response as I've just blown up a promising rack. It's a bad board for bingos, bit that could change.
My play sims 2 points behind GENU, maybe because there's value in the two S's when I'll likely use one for a non-bingo play with ZAGS. It advertises a second S, but he's powerless to do much about it unless he has his own S, or an A to play above BIOTA, likely scoring very little. I10 SUE 20, the most directly bingo-opening play, sims about 5 behind GENU. I suppose it suffers from being bingoed on by Msck first, or blocked with all the hooks SUE takes, or even just plays with P that form ZAP. A human would also probably play more defense against SUE. I think I should have withheld the information about my second S and played GENU, and cashed one of the S's next turn, since bingos will be hard to come by now without the blank.
>Mack_Meller: ?EIMRRT N2 TRIREMe +71 288
>David_Weisberg: EEEIRST M1 TEE +12 241
#note REE or TEE? Though it may be easier to end a bingo in R then T, I had EIRS already, and there were 2 T's unseen vs. 1 R. It's much easier for him to hit the triple with a word ending in R to block many of my bingos than to end in T (or just score a lot), and if I recall, that was the deciding factor for TEE. TEE sims 5 points ahead of REE, but 2M ET 2 or K5 ER 4 both perform better than REE in simulation, with ER a tiny bit ahead of TEE. I disagree with this and believe a human would handily block everything after ET or ER, and TEE is as close as I can get to forking the board (threatening possible 9s through ER, for example). I need a miracle here, but TEE may be my best bet for one.
>Mack_Meller: ADEEEOR 4J ROAD.E +21 309
>David_Weisberg: CEIPRSS 1F CRISPES. +89 330
#note !!!
>Mack_Meller: AAEEHOT O1 THA. +30 339
#note 3I HETAERA 22 is great here. It sims about 3 ahead of THAE, not scoring much but doing just about everything one could hope for on a rack like this. Mack told me he didn't see it at the time and found it in analysis the night after that day's session.
>David_Weisberg: AEFHIOP 2H FOE +21 351
#note I felt pretty lost here. I think I played FOE because I figured I was likely to redraw an E, but I didn't notice how the remaining scoring tiles like B and Y could fit neatly under it for a lot of points. FIE might have been better to at least prevent plays using SOY. I did consider that I could play 5J HA 22 next turn if it wasn't blocked, but FIE allows that too. On the other hand FIE allows plays with SIG (and SIB), and G is a very bad tile to have on this board.
More generally this seemed like the area to play in if only to block the bingo line above ROADIE, which I'll have no chance to use myself with this rack, and that also ruled out plays like 2J IF or EF 18 (PIER or PEER), though OF remains viable. By this point I recall only having a few minutes left.
I narrowed down the bewildering set of weak options to just 4 to sim. OF came out best, followed by 5J HA 22 a point behind, FIE 2 more points behind, and FOE a very slight bit behind FIE. At this stage, when win percentage in simulation becomes more relevant, OF and HA win about 49% (with OF at 0.5% better), and FIE and FOE both win around 46%, with FIE having a nearly 0.5% edge. I'll have to agree with the sim here about OF, because there's just so little he can do on this board afterward.
>Mack_Meller: ABEEEOY 3I BEY +30 369
>David_Weisberg: ADGHIPT I12 .P +7 358
#note Another fiendishly difficult turn with little time left. I thought my play cleverly threatened HIP, HAG, DIG, etc., with him needing to play for only 6 to block it, and potentially having no consonants left, leaving me free to set up things elsewhere.
The computer thinks this is misguided, and that I should play F1 CHIT 17, winning nearly 63%, or 13H AH 20, winning nearly 62% (AH does slightly better on equity, but less than half a point). At this point I was kind of flying by the seat of my pants, so at least setting up something and not emptying the bag seemed better than scoring more and giving Mack full information, as he was in time trouble as well. It could get close enough that whoever goes overtime loses, so this was important.
>Mack_Meller: AEEILOU 11J OI +6 375
#note In simulation 11J AE 6 wins 51% vs 42% for OI, and 40% for OE. He may have been considering that I cannot hook OS for HOS, which is only valid in NWL2018 (and soon to be NWL2023), and those 14 points could mean a lot. However, the H plays for 15 at 2L for HETH, which I didn't see and I'm not sure he saw at the time either. Retaining an A is also of value in case I have the last two, for hooks above BIOTA. I might have liked OE, though the I could turn out to be a worse liability than 2 E's. It's looking very possible one of us could get stuck with tiles on this board before the other goes out.
>David_Weisberg: AADGHIT 14E AGHA +16 374
#note Quackle thinks I have no win here and can only tie with F1 CAD 8, but Mack asked Cesar Del Solar to run this endgame through his Macondo engine, which powers HastyBot and BestBot.
In Mack's words, "It found that you could've actually won by 1 point (!) had you played CIG or SH/HETH instead of AGHA. Quackle does not come up with this. It's a crazy sequence -- the idea is my best play next turn is YOUR regardless (stuff like ALEE/AG after CIG gives you way too much counterplay), and then you play whichever of CIG and HETH you didn't play before. Then you're threatening DATA/AB, ALEE doesn't work again since it gives you DATA for even more. I can't outrun you otherwise, so I have to block with LEA/AB. Then here's the truly amazing part -- I'm E-stuck!! You play AM for 5, I only can play RE for 2 on the bottom right then I have nowhere for my last E, and you play 1 at a time and win easily. So going back, my best play instead of LEA is to just play LEV for 8 and let you go out with DATA and win 398-397. Crazy stuff."
Needless to say, neither of us were in a position to see all this in the moment.
>Mack_Meller: AEEELU 15C LEA +6 381
#note Probably the only win. If I then block K3 YOUR 14 with AIN, he can still play his U with 7M UM 5, and I can't stop his out play of F1 CEE 7.
>David_Weisberg: DIT B14 TI +8 382
>Mack_Meller: EEU K3 ..U. +14 395
>David_Weisberg: D J10 D.. +8 390
#note One for the ages! Well done, Mack!
>David_Weisberg: (EE) +4 394
|
