| Player 1 |
|
#player1 Mack Mack
#player2 Quackle Quackle
>Mack: DINOORV H8 OVOID +22 22
>Quackle: CEILORU 12D CLOU.IER +72 72
>Mack: ANNORRS 11J NOR +12 34
>Quackle: EFIIILO M10 FILO +18 90
>Mack: AEEFNRS 10L E.F +19 53
#note This is aggressive, but EFFS is really hard to block without an S so at my early deficit and with a great leave it seems warranted.
>Quackle: AEGIIRX N13 XI +27 117
>Mack: AEMNRST O4 MARTENS +96 149
>Quackle: AEEGIRY N4 AYE +21 138
>Mack: ?AALMNP 15H NAPALMs +92 241
>Quackle: EGINRSY 14B SYRINGE +78 216
>Mack: CDEELNW B10 CLEW. +16 257
#note Quackle says I should take the extra 6 with M1 WELD. I thought about it but it didn't seem quite worth it to open the floating E and the row 1 scoring spot. #questionable
>Quackle: AGPQSTU 10F QU.TA +41 257
>Mack: ADDEENW A12 WE +23 280
#note Could do AWE for the extra 3 points and turnover, but the pool was starting to look somewhat consonant heavy and I wanted the extra bingo % here.
>Quackle: AGIPSTU L13 GU. +15 272
>Mack: ADDEINV I2 INVADED +67 347
#note Basically a coinflip for which INVADED to play. I went with this one since the I is a bit harder to overlap on the TWS and it restricts the placement of the floaters a bit better for potential bingos.
>Quackle: AEIKPST M2 KIP +25 297
>Mack: ADHJORZ 5H H.DJ +30 377
#note Looks good - blocks lanes, scores, and cleans up my rack.
>Quackle: ?AEHOST 8A STAtEHO.. +86 383
#note Still confused why he didn't switch the blanks for 3 more. But the 3 points made the pre-endgame tremendously more interesting: stay tuned!
>Mack: AGOORUZ H1 ZOO +40 417
>Quackle: ABEEITT 7C BET +18 401
>Mack: ABGINRU A4 GUAR. +8 425
#note This play is the whole reason I posted the game. It's one of the most complex one in the bag pre-endgames I've encountered in a long time. Quackle is incapable of solving it, claiming the best I can do is win 25% of the time (2/8).
First let's dispose of the possibility of me winning more than 75% (6/8): if I draw an E, I can't block both STRIATES on column A and STATI(ON)ER on row 3. I'd need a 6 ending in -AT on column B or an 8 ending in -ON on row 3, neither of which exist (as I correctly determined in game).
So the question becomes whether I can win 6/8. My play, as I calculated in game, wins 4.5/8 - better than Quackle finds, but as shown below, not optimal. Here's what can happen with each draw:
A draw: he plays L4 SIREE, I go out with BINA(TE) and tie.
E draw: he bingos out with STATIONER and I lose.
I draw: I can't go out, so he outruns me with the sequence 1K RESET + 12M LAT, and I lose.
R draw: he plays L4 SITE, and I go out with B2 BRIN and win.
S draw: I win easily having both outs and the huge L4 SIB available.
T draw: I win by 1 after L4 SIREE and I go out with B3 BINT.
The reason I went for GUARS is because it set up the BRIN and BINT plays mentioned above. I ended up drawing the R and going out with BRIN and winning. Unfortunately, I missed an opportunity to win 5.5/8. The key is to get rid of the I so that I can preserve outs if I draw the I. I recognized this in game, but due to the delicious tacos that became ready as I was contemplating this position, did not spend enough time to find the star play of BRUINS/BURINS (both accomplish the same).
After BRUINS/BURINS, here are the possibilities:
A draw: We tie after he plays L4 SIREE and I go out with 6C AGA.
E draw: He bingos out and I lose.
I draw: I win by going out with 4A RAGI after his best play of 1K RESET.
R draw: I win after he plays L4 SITE and I go out with 6F GRAD.
S draw: I win easily.
T draw: I win after he plays L4 SIREE and I go out with 9C GAT.
So BRUINS/BURINS loses with an E draw, ties with an A draw, and wins otherwise. Another notable option is UNRIGS, which wins with an A draw but ties with a T draw - this is of course worse than BRUINS/BURINS since there are 2 Ts to draw but just 1 A. Hence UNRIGS wins 5/8 as opposed to BRUINS/BURINS' 5.5/8.
I'm yet to find a play that wins 6/8, or even another 5.5/8 for that matter. Any play that wins at least 5.5/8 will need to accomplish the following:
(1) Block column A bingos.
(2) Preserve sufficiently high scoring outs with an A, R, or T draw.
(3) Preserve either an out or a high-enough scoring sequence to win or tie with an I.
Criterion (1) is obvious. (2) follows from the fact that I don't care what happens if I draw an E since I'm losing to an outbingo regardless. Drawing an S will always win since he has neither column L or row 1 plays and will never outrun me. Drawing an A, R, or T will require me to outrun big L4 plays of either SITE or SIREE, which will require me to draw a high enough scoring out as he will always be going out with whatever he has left. Drawing an I is the interesting case, and brings us to criterion (3). The good news is that he won't have big HADJI overlap plays, so I'll certainly win if I draw an out and may be able to outrun him without going out. The bad news is that another I does not go well with my current tiles for drawing an out, and any non-out winning sequence will still have to score substantially to win after 1K RESET. An example of this is after UNRIGS, an I draw results in a 1-point win for me after 1K RESET by him, 12M LAB for me, and 9C TA for him to go out.
We can be a bit more specific in the case of an A/R/T draw. An A or T draw allows L4 SIREE, against which I will need to score a combined 24 points on my two turns to tie (25 to win). An R draw allows L4 SITE, against which I will need to score a combined 18 points on my two turns to tie (19 to win).
Any candidate play must be along column A, ending in -AT on column B, or along row 6 above BET. I've done an exhaustive subanagram search and, guided by the above criteria, have not found anything that equals or outperforms BRUINS/BURINS. A lot of possibilities, like GUARS, have losing with an I draw as a fatal flaw. B5 GNAT is the notable best option down column B, winning 5/8 as it saves TRIBU(TE), BRUI(TE)R, and A1 URBIA as possible outs (losing by just 1 with an I draw to the 1K RESET + 12M LAT sequence -- hence my comment earlier about why the 3 points lost from switching the blanks in STAtEHOOD made all the difference!). All row 6 plays seem to do even worse than 5/8.
In sum, my conclusion is that A3 BRUINS/BURINS is optimal, winning 5.5/8. A handful of other plays, including UNRIGS and GNAT, win 5/8, and other plays win even less.
If anyone finds another 5.5/8 or perhaps a 6/8 I've missed, please do let me know!
>Quackle: AEEISTT L4 SITE +40 441
>Mack: BINR B2 BRIN +17 442
#note In the end I still won and got to eat my tacos while they were still hot. Yum!
>Mack: (TAE) +6 448
|
