| Player 1 |
|
#player1 Matt_Canik Matt Canik
#player2 Geoff_Thevenot Geoff Thevenot
>Matt_Canik: EEEILQS 8G QI +22 22
#note I'm analyzing this game to see how Quackle and Elise simulate things differently. Both programs have this play as the best, obviously, and both list it as 3.3 points better than QI in the other placement.
>Geoff_Thevenot: DDEEGRR 9H DREDGER +65 65
>Matt_Canik: EEEGLNS K7 LE.GE +14 36
#note I wasn't entirely sure DREDGER was valid. Geoff took over a minute to play it. So I hesitated to put the S on the end. Bad.
Q: -11.0 of GELEES
E: -13.5 of GELEES
>Geoff_Thevenot: OWW 8M WOW +34 99
>Matt_Canik: ADEENSV N2 ENDEAV..S +84 120
>Geoff_Thevenot: CIINSUX 3J UNCI.I +16 115
#note Geoff said this held SX. That's sweet, nice play Geoff.
Elise says that it's "less than likely" that my opponent has the X after this play, which already appears to be a limitation of its ability- detecting setups.
>Matt_Canik: AIJMOPR J6 JAP. +35 155
#note Elise's speedy player lists
JORAM 2F first,
JORAM 12I 2nd,
JAP(E) 3rd,
RIOJA 12K 4th, and
MAJOR L11 5th.
Quackle's simulation has this exact same order, however its speedy player switches JAPE and the inferior JORAM.
I just mised JORAM.
Q sims this 4.2 points down.
E sims this 4.3 points down.
>Geoff_Thevenot: EKX 2H KEX +55 170
>Matt_Canik: IIMORRU 12K RUMOR +24 179
#note "Current rack is better than approximately 2.1% of possible rack draws." -Elise
^that's going to become one of my favorite features.
Q sims this 4.2 down of MI(C)RO, which I should have seen.
E sims this 4.0 down.
These programs seem very similar for now. We'll see how they do on the closed board ahead.
>Geoff_Thevenot: ETY 13L TYE +30 200
>Matt_Canik: FIIMNRT L1 FI.IN +28 207
#note E puts this ahead by 10.4, Q says 12.2
>Geoff_Thevenot: AO 7M O.A +19 219
>Matt_Canik: BCDMORT 11G COMB.R +15 222
#note I'm having a lot of trouble getting Elise to routinely show me the tiles it expects the opponent to hold, and I can't get it to after this play. I myself had deduced that it was quite likely he was bingo-friendly and had either an S, a blank, or both.
The only options seem to be BR(I)O for 27, COMB(E)R for 15, or MORBID/BROMID for 14. I didn't want to draw into CDMT after my opponent is likely to bingo and open a good spot for me.
Elise says BRIO is better than COMBER by around 5.5 points, Q says 5.3 points.
Q says it's better by 5.3 points.
Interestingly, though, Elise gives BRIO a 5.6% higher chance of winning, whereas quackle puts it 2.8% higher. I don't know if this is because Elise realizes I'm about to be walled in, if it's just a relatively small sample size, or because Elise has decuded that Geoff is holding better than Quackle has.
>Geoff_Thevenot: ?AOPSTU 14F OUTPAsS +66 285
>Matt_Canik: ADOOTVZ H2 .AZOO +28 250
#note I'm really just missing a lot of obvious plays this game.
Elise has this 12 points down of TOAD, Q has it 10.7 down.
Again, though, Elise has the win% gap a good margin larger- 3.1% for Quackle, 5.4% for Elise. I'm not 100% sure how either program draws its win% numbers, though.
>Geoff_Thevenot: AAH 15H AAH +33 318
>Matt_Canik: DFHOSTV O1 TH.O +33 283
#note This is the first deviation I've seen between the programs. Q likes THIO by about a point, and has its win% 0.9% above SHAD or SHAT.
E likes SHAT by about a point over SHAD and THIO, and has SHAT winning 1% higher.
>Geoff_Thevenot: ?AINRTY 5B gYRATI.N +70 388
>Matt_Canik: ADFGSTV C2 DAV. +22 305
>Geoff_Thevenot: EEILLST E3 LE.LTIES +58 446
>Matt_Canik: FGLNNST 6L SL.T +28 333
>Geoff_Thevenot: BEEIIU D4 I.E +20 466
>Matt_Canik: FGNN 9D N.G +4 337
>Geoff_Thevenot: BEIU I11 .I +4 470
>Matt_Canik: FN F13 F.N +6 343
>Matt_Canik: (BEU) +10 353
|
