Game 4, move 4 : roadkillbooks 5-2

[diane]

Ok, rkb in white, to move 5-2

4LsNAwLgu00AAw:MIGqACAAEAAA

[ah_clem]

I take back all my complaints about you rolling the dice.  (c:

Welcome back.

[diane]

I take back all my complaints about you rolling the dice.  (c:

Welcome back.

what does he want? 

oh well, it is nice to feel wanted 

[dorbel]

I doubt if he was doubling, but perhaps you should have asked. He's a favourite with some gammon chances and it's 3-away, 3-away, where aggressive doubling pays.

[stog]

how did our last move lead to us having 2 back? me no understand where we are?
before moving last we were (blk going a'clockwise) then we chose 24/20 8/7

[lewscannon]

I doubt if he was doubling, but perhaps you should have asked. He's a favourite with some gammon chances and it's 3-away, 3-away, where aggressive doubling pays.

Yeah, I would've considered cubing here very strongly. What does gnu say on this? That being said, good roll diane!

[vikingblood80]

how did our last move lead to us having 2 back? me no understand where we are?
before moving last we were (blk going a'clockwise) then we chose 24/20 8/7

these are rkb's men...its exactly the same position as yours, only directions are swapped

[stog]

ok thx -- couldnt see it b4

[ah_clem]

Yeah, I would've considered cubing here very strongly. What does gnu say on this? That being said, good roll diane!

I punched it into gnubg this morning before leaving for work.  The analysis had our chosen move as best giving us about a 45% chance of winning.  That's not a double for rkb.  It wasn't even a double for my choice of move, which was about a .8 error.

I'll post the rollout as soon as I return home.  There seemed to be 8 10 candidate moves, so the rollout took longer than usual....

UPDATE:  Rollout.  Cubing is a double whopper.


Cube analysis
Rollout cubeless equity  +0.286 (Money:  +0.246)

Cubeful equities:
1. No double            +0.414
2. Double, pass         +1.000  ( +0.586)
3. Double, take         +0.165  ( -0.249)
Proper cube action: No double, take (29.8%)
Rollout details:
Centered 1-cube:
  0.578 0.146 0.007 - 0.422 0.060 0.003 CL  +0.286 CF  +0.414
[0.002 0.002 0.001 - 0.002 0.003 0.001 CL   0.004 CF   0.013]
Player You owns 2-cube:
  0.580 0.155 0.008 - 0.420 0.089 0.006 CL  +0.723 CF  +0.165
[0.002 0.003 0.001 - 0.002 0.004 0.001 CL   0.014 CF   0.019]
Full cubeful rollout with var.redn.
1296 games, Mersenne Twister dice gen. with seed 865816185 and quasi-random dice
Play: 0-ply cubeful prune [expert]
Cube: 0-ply cubeful prune [expert]

[roadkillbooks]

I was not going to double since I knew a 5 was coming.  If I could have know for sure that I wasnt going to get a five.. Im sure doubling would have been correct .

Well here I see 3 main candidate plays: a- I could play safe with something like 11/4
b- I can play hit an stack with 11/6, 7/5* hit
c- I could hit and unstack/double blot with 7/5* hit, 6/1

a- I don't like the safe play here.  Forum is only a 5 or 6 away from putting me into a 5 point holding game and I think my position deserves more than that. Though on the positive side if forum doesn't get a 5 or 6 Id be strong again.  One tactic passive but possibly correct. 

Hitting on the 5 point when I control forum's 5 point looks like a must. Although many times errors are made because your expecting to hit and point... then a bad number comes up and the game plan stays the same.

Here the two main hitting plays (excluding a premature break of my anchor are b and c.  Sometimes I like the hit, unstack, double blot type plays.  I think that if forum had a 3 point prime with like the one point made instead, that play might be right.  Here I think it seems to lead to too many bad numbers.

So Id go for the aesthetically unpleasing 7/5*, 13/6
Even 5/1 would have been better for me playing it 11/5*

[diane]

So Id go for the aesthetically unpleasing 7/5*, 13/6
Even 5/1 would have been better for me playing it 11/5*

I am gonna go out on a limb and assume you mean 7/5, 11/6   

[dorbel]

RKB finds the best play here, even with that horrible distribution! I would have tried 11/4, a small error, while Snowie 3-ply will play 7/5*, 6/1! However, a short rollout supports the match play, showing that hitting is still correct even when it leaves a position that is almost unplayable.