Earlier today, I wrote "I'd like to know (since
I can't figure it out for myself) what sequence
of plays, however unlikely, would have led me to
a win."
I don't enjoy being in situations where I can't
figure things out for myself, so I poked around
to see what I *could* figure out, using Quackle's
List function. Additional work with Simulation
gave me a Win% = 3.56 with the play H1 CRE(M)E
for 27 points. That would take 4 of the 5 tiles
in the bag and close the deficit to 79 points,
with a C at H1 and an R at H2 -- nice letters to
bingo to/through.
I put my leave ('INO') and the 12 unseen tiles
into Quackle's Lister function, Ctrl+L, checked
'Build', and clicked 'A-Z?* Query' to build a
list, nicely sorted by length -- thank you! --
to see what words might play on the top row.
Keeping 'INO' and drawing 'ADRS' from the unseen
tile pool of 'ACDDDFHORSSU' would give me the
possibility of 1A SARDONI(C) for 89 points, so
if my opponent were to empty the bag and fail to
block the top row -- possible -- I could bingo
out and potentially catch him with enough points
on his rack to let me win the game. That gives
a bit of insight into how to use Quackle to find
the "miracle sequence" that salvages the game,
but doesn't answer the question of how to query
the program for best-case/worst-case scenarios.
Is there a way to get a list of possible winning
sequences? A way to get best-case/worst-case
scenarios in addition to the means & standard
deviations given in the 'Simulation Details'
panel and its 'Explanation' dialog?
Thanks, Quackle team.
John Hart
--- In [email protected], "John Hart" <[EMAIL PROTECTED]> wrote:
>
>
> Hello again, Quackle team and all,
>
> This is a recent game that went really wrong,
> mainly because I allowed a high-scoring phony.
> Still, late in the game, Quackle thought I had
> a slim 2% chance of salvaging the win. Further
> details after the first 10 moves...
>
> #player1 John John
> #player2 opp opp
> >John: ?EGOUVY - +0 0
> >opp: AWW 8G WAW +18 18
> >John: AEMMUUY -AYMMUU +0 0
> >opp: EKP 9F KEP +20 38
> >opp: EKP -- -20 18
> >John: EENNTXZ H8 .NNEXE +21 21
> >opp: ACEKLMP 13G K.P +17 35
> >John: EIORTTZ 11B EROTIZ. +32 53
> >opp: ?BEEEST C8 BEE.iEST +70 105
> >John: AILMNOT 10B M.TA +26 79
> >opp: ?EIRRTV 14H RIVeTER +102 207
> >John: IJ 8A JI. +36 115
> >opp: EIY 15A YE.I +24 231
> >John: DLLNNOP M10 POLL.N +16 131
> >opp: AAEGT D4 AGATE +28 259
> >John: ADHINOO L6 DAHOON +20 151
> >opp: AGTU 8K G.AUT +27 286
> >John: IILNRRV 5A VIR.IN +20 171
> >opp: IOQU O4 QUOI. +24 310
> >John: ABFLNRS 15F FAS +63 234
> >opp: ELMOOSY N6 YO. +29 339
> >John: BEGILNR K2 BILGE +19 253
> >opp: ALMU 4H MUL.A +20 359
>
> So here's my situation: I'm way behind, and
> there are darn few tiles (5) left in the bag.
> The move I played, tantamount to resignation,
> was K10 COIN(e)R for 27 points. That play
> was not listed when I did 'Generate Moves'.
> The Sim that I ran included that move and 2
> of Quackle's suggestions, and the result (on
> 2 plies, 1,000+ iterations) was:
>
> Move Score Leave Win % Valuation
> 9E CRO(N)E 26 EIN 2.18 30.0
> K10 COIN(e)R 27 EE 0.23 30.3
> H1 ICE(M)EN 30 OR 0.16 29.7
>
> Hmm. It seems that there's roughly a 1-in-50
> chance of salvaging the win if I play CRO(N)E
> instead, but I can't figure out how to get the
> program to reveal to me just what the magical
> sequence of plays & draws might be that would
> lead me to that miracle come-from-behind win.
>
> So I asked 'Championship Player' what to do,
> and got these 2 suggestions:
>
> Move Score Leave Win % Valuation
> 9E CRO(N)E 26 EIN 1.21 5.7
> K10 (Z)E(K) 25 CEINOR 0.76 7.7
>
> Is there a way to get Quackle to show me how I
> might have won the game after 9E CRO(N)E ? Even
> if it's highly improbable, I like to hold out
> some hope, even in desperate circumstances. The
> 'Simulation Details Explanation' dialog refers to
> average score, standard deviation, and average
> differential. What I think I'm looking for is
> best-case or worst-case scenario. In particular,
> I'd like to know (since I can't figure it out for
> myself) what sequence of plays, however unlikely,
> would have led me to a win.
>
> For the sake of completeness, the game ended:
> >John: CEEINOR K10 COIN.R +27 280
> >opp: CHOS O12 COSH +46 405
> >John: DEEFRSU B2 SER.F +32 312
> #rack2 ADD
> ... and he played ADD(ED) rather than A1 DAD.
>
> Thanks for any insight on querying Quackle as
> to how I might have pulled this one out.
>
> John Hart
>
