Match 1, Game 3, Move 11, Forum

[blitzxz]

4-4?

Score:
Green(forum) 2, White(factotum) 0, match to 5

Pips:
Green 135 White 148

GnuID:
sN3BBAMcn8EMIA:UgmyAAAAEAAA

[diane]

I knew we should have let lews roll...

Spoiler

I havent voted yet, but the two moves I am torn between are:
15-7 [2] - which leaves the mid point for that back checker to aim for, and some protection for moving it out, but a point on 7 is too close to factos point to be able to make use of it.
15/11 (2) 13/9 (2 - this gives a block in our outer board, which might work for us to give us a chance to move that back checker, but I dont like breaking the mid point just now, it is too early.
I shall await some persuasion from others, or a completely different view  

[close]

[blitzxz]

Spoiler

Doubles are hard to move as usual. Just too many choices. Everything looks good and I have hard time to figure out what exatly are the differences in these choices. I voted for 15/7 (2) because it looks like the safest way to get checkers home and is line with my safe play strategy which I only do in these holding/one prime games.   However it leaves the runner complitely disconnected and doesn't block factotum at all... Safe is still my choice.

[close]

[lewscannon]

Spoiler

This makes for an interesting choice. I don't know if I'm making the best one with 13-9 (2) & 6-2(2), but this makes the most sense to me. I don't want to move off the 15 point right now, because you concede the outfield, even if moving the 15s blocks the 6 from his 5 point. I also prefer the 6-2 to to the 8-4 because getting a roll with a 6 in it afterwards is going to open up a lot of hitting possibilities for factotum, though the more I think about it, I'm not as certain that the 8-4 isn't a better move, though I am pretty positive about the 13-9.

[close]

[ah_clem]

Spoiler

Our #1 priority is escaping the runner.  Unfortunately, we can't do that with this roll, so what are our other priorities?

• building our board
• blocking factorum's men on the five point
• unstacking the candlestick on the 6 pt
• clearing the 15 point

15/11 (2) 6/2 (2) scratches all these itches, so that's why I voted that way.

15/11 (2) 13/9 (2) does the best job at blocking, but does nothing for the other objectives.  And it leaves no landing spot for the runner.

15/11 (2) 13/9 (2)  or 15/7 (2) don't help unstack.

13/9 (2) 6/2 (2) is similar in theme to my choice, but I'm wary of it because it abandons the midpoint and blocks from four points away instead of six.  However, the 15 may be a better "midpoint" here since the runner could reach it with a 63, and maybe we don't want to get into a holding game with both sides holding points 6 away.

If this was a chouette, I'd go along with that play. 

[close]

[diane]

Spoiler

Eeek, you people did nothing to help clarify my thoughts, but just gave me more to stress over!!     

The unstacking is compelling, but they look too far forward, the 13-9 is equally compelling, but opens the midpoint and leaves the 15 point all alone...
The 15-11 also is tempting, but then they look all at sea too    I am gonna wait a little longer, at this point I am mostly inclined to follow blitz to the 7 point...

[close]

[ah_clem]

Eeek, you people did nothing to help clarify my thoughts, but just gave me more to stress over!!     

If it's any help at all, allow me to add that I have no confidence that my preferred move is best.  (c:

[Zorba]

This will clear it all up for diane 

Spoiler

Hmm, this is the kind of tough decision that made me think it might've been a pass, practically speaking  A lot of options, and nothing seems to stand out or looks awful right away. To create some order in all the possible moves, it's important to look at our goals and how to prevent factotum best from achieving his goals.

With this roll, we put ourselves firmly in the lead in the race, but unfortunately our backchecker can't move. The result is that our timing is starting to get tricky. We really need to run our backchecker very soon, or it will become virtually impossible to get it around safely, as factotum brings in his builders and starts dominating the outfield.

Therefore, our 15pt is an asset here, not a liabilty! It's much easier to reach for the backman than the midpoint; a 6-3 does it right away and if we move up, more rolls will go there, including perhaps a direct 6. Also, that 15pt is one of the few assets we have here that makes factotum's life a bit difficult. Look at what he'd like to do: extend that 4-prime from the back and bring the builders in as diversified as possible. If we give up the 15pt, that becomes very easy for factotum. These two things combined make our 15pt much more important than our midpoint here. So, 13/9(2) beats 15/11(2) IMO.

What about blocking factotum? He's way behind in the race, so factotum is no hurry to run from his anchor. His goal is not to run, but to block us upfront and maybe later, attack our backman. So blocking is not much of a priority here; most of the time, factotum will simply not care at all that his backmen are blocked. Also, we don't have much time to keep the blockade in place anyhow.

What about board building? This goes behind factotum's anchor, which is usually not what you want to do. It does help to maintain a good outfield presence though, if we move some fours upfront 8/4(2) or 6/2(2), and that might be more important here, where the race is by far our best bet, and we thus need outfield points to hop home. Also, a stronger board is always helpful if some blot-hitting will start.

Bottom-line? I'm still not sure, but I'm pretty sure I want to keep the 15pt. Two points in factotum's outfield looks like a bit too much here, so I'd play 13/9(2). Then either 8/4(2) or 6/2(2); I picked the latter to create a more compact position that seems less likely to give us awkward rolls later, but of course the 4pt is a better homeboard point.

Tough call, 13/9(2) 6/2(2) is my choice.

[close]

[stog]

Spoiler

well explained. my choice too was the 13/9 6/2 x2 - i think the 15 point is v important for us.

[close]

[diane]

Spoiler

Yup, after several more hours considering this, I kinda like that move too - the 6-3 is the only real saviour for that back checker, so it seems best to leave it that option. Then the only other real choice is the 2 point, yukky as it looks.  So, I will do what Zorba said [not at all what Lews said of course    ]

[close]

[ah_clem]

Zorba's comments make a lot of sense, so I'm changing my vote. 

[lewscannon]

What's Zorba got that I ain't got?

[diane]

What's Zorba got that I ain't got?

How would I know...perhaps if I did, we wouldn't be in this pickle     

[blitzxz]

Sorry, I'm away this weekend and can't do the posting from here so diane, stog or somebody could do the posting if the match is in hurry. 

[ah_clem]

Sorry, I'm away this weekend and can't do the posting from here so diane, stog or somebody could do the posting if the match is in hurry. 

I nominate lews to be in charge of rolling the dice. (c:

Here's a 0-ply gnubg rollout:

Spoiler

Boy did we screw the pooch on this one. Once again, nobody voted for the best move.  In retrospect, it appears that building our board is the right thing to concentrate on here - for one thing it  effectively prevents factorum from hitting loose, so it can be seen as protecting the runner. 

Doubles like this are often hard decisions when several different approaches look good and the problem is figuring out which one is best.  At least we're still a 49% chance to win the match.  Hang tough and hope for the best...

    1. Rollout          8/4(2) 6/2(2)                MWC:  50.67%
       0.394 0.127 0.003 - 0.606 0.158 0.005 CL  50.67% CF  50.67%
      [0.001 0.001 0.000 - 0.001 0.001 0.000 CL   0.11% CF   0.11%]
        Truncated cubeful rollout (depth 11) with var.redn.
        1296 games, Mersenne Twister dice gen. with seed 859267361 and quasi-random dice
        Play: 0-ply cubeful prune [expert]
        Cube: 0-ply cubeful prune [expert]
    2. Rollout          13/9(2) 8/4(2)               MWC:  50.33% ( -0.34%)
       0.382 0.109 0.003 - 0.618 0.138 0.004 CL  50.33% CF  50.33%
      [0.001 0.001 0.000 - 0.001 0.001 0.000 CL   0.11% CF   0.11%]
        Truncated cubeful rollout (depth 11) with var.redn.
        1296 games, Mersenne Twister dice gen. with seed 859267361 and quasi-random dice
        Play: 0-ply cubeful prune [expert]
        Cube: 0-ply cubeful prune [expert]
    3. Rollout          13/9(2) 6/2(2)               MWC:  49.02% ( -1.65%)
       0.371 0.098 0.003 - 0.629 0.155 0.005 CL  49.02% CF  49.02%
      [0.001 0.001 0.000 - 0.001 0.001 0.000 CL   0.11% CF   0.11%]
        Truncated cubeful rollout (depth 11) with var.redn.
        1296 games, Mersenne Twister dice gen. with seed 859267361 and quasi-random dice
        Play: 0-ply cubeful prune [expert]
        Cube: 0-ply cubeful prune [expert]
    4. Rollout          15/11(2) 8/4(2)              MWC:  47.58% ( -3.09%)
       0.351 0.107 0.003 - 0.649 0.155 0.005 CL  47.58% CF  47.58%
      [0.001 0.001 0.000 - 0.001 0.001 0.000 CL   0.10% CF   0.10%]
        Truncated cubeful rollout (depth 11) with var.redn.
        1296 games, Mersenne Twister dice gen. with seed 859267361 and quasi-random dice
        Play: 0-ply cubeful prune [expert]
        Cube: 0-ply cubeful prune [expert]
    5. Rollout          15/7(2)                      MWC:  47.31% ( -3.36%)
       0.338 0.099 0.003 - 0.662 0.124 0.003 CL  47.31% CF  47.31%
      [0.001 0.001 0.000 - 0.001 0.001 0.000 CL   0.10% CF   0.10%]
        Truncated cubeful rollout (depth 11) with var.redn.
        1296 games, Mersenne Twister dice gen. with seed 859267361 and quasi-random dice
        Play: 0-ply cubeful prune [expert]
        Cube: 0-ply cubeful prune [expert]
    6. Rollout          15/11(2) 6/2(2)              MWC:  46.48% ( -4.19%)
       0.341 0.097 0.003 - 0.659 0.169 0.005 CL  46.48% CF  46.48%
      [0.001 0.001 0.000 - 0.001 0.001 0.000 CL   0.09% CF   0.09%]
        Truncated cubeful rollout (depth 11) with var.redn.
        1296 games, Mersenne Twister dice gen. with seed 859267361 and quasi-random dice
        Play: 0-ply cubeful prune [expert]
        Cube: 0-ply cubeful prune [expert]

[close]

[stog]

ive just messaged zorba on fibs to this effect
"blitzxz is away this weekend and has asked whether diane or myself could post the new forum play, i am unfortunately away this week so i wonder if u or diane can manage this - i thought u might like to post the next play as practice for the next match if you have time. np if u haven't as blitzxz is only away for the weekend....but cld u ask diane if u see her if u can't.. did any of this make any sense regards and good dice -- stog

oh and have a great weekend mr blitzxz!

[Zorba]

GNUBG 2-ply rollout:

Spoiler

Medium sized error, but at least we avoided the whoppers

With the cube on 4 in a 5pt match, obviously there's quite a lot of MWC at stake, as ah_clem's rollout shows.

I'm quite surprised by the best move and probably not the only one; still trying to find a good explanation for it. The strong board will help against factotum playing too loose, and if we get to hit factotum. The follow-up after 4(2) 2(2) looks tricky though, with the stripped and stretched out position. Perhaps keeping both the 15pt and 13pt in the outfield is a strong factor, as it makes factotum's clean up of his blot harder and might also make his future moves into the outfield riskier (also because of our stronger board of course).

    1. Rollout          8/4(2) 6/2(2)                Eq.:  -0,3083
        39,39  16,01   0,83 -  60,61  14,46   2,27 CL  -0,3083
      [  0,28   0,56   0,25 -   0,28   0,32   0,43 CL   0,0059]
        Full cubeless rollout (trunc. at one-sided bearoff) with var.redn.
        648 games, Mersenne Twister dice gen. with seed 859211562 and quasi-random dice
        Play:  2-ply cubeless prune
        keep the first 0 0-ply moves and up to 8 more moves within equity 0,08
        Skip pruning for 1-ply moves.
        Cube: 0-ply cubeful prune
    2. Rollout          13/9(2) 8/4(2)               Eq.:  -0,3374 ( -0,0291)
        37,58  12,68   1,10 -  62,42  13,40   1,51 CL  -0,3374
      [  0,26   0,51   0,33 -   0,26   0,31   0,36 CL   0,0055]
        Full cubeless rollout (trunc. at one-sided bearoff) with var.redn.
        648 games, Mersenne Twister dice gen. with seed 859211562 and quasi-random dice
        Play:  2-ply cubeless prune
        keep the first 0 0-ply moves and up to 8 more moves within equity 0,08
        Skip pruning for 1-ply moves.
        Cube: 0-ply cubeful prune
    3. Rollout          13/9(2) 6/2(2)               Eq.:  -0,3632 ( -0,0549)
        36,50  12,61   0,96 -  63,50  14,03   1,43 CL  -0,3632
      [  0,26   0,54   0,34 -   0,26   0,30   0,32 CL   0,0054]
        Full cubeless rollout (trunc. at one-sided bearoff) with var.redn.
        648 games, Mersenne Twister dice gen. with seed 859211562 and quasi-random dice
        Play:  2-ply cubeless prune
        keep the first 0 0-ply moves and up to 8 more moves within equity 0,08
        Skip pruning for 1-ply moves.
        Cube: 0-ply cubeful prune
    4. Rollout          15/11(2) 8/4(2)              Eq.:  -0,3962 ( -0,0879)
        35,22  11,71   0,38 -  64,78  15,14   2,66 CL  -0,3962
      [  0,63   0,81   0,21 -   0,63   0,72   1,01 CL   0,0134]
        Full cubeless rollout (trunc. at one-sided bearoff) with var.redn.
        108 games, Mersenne Twister dice gen. with seed 859211562 and quasi-random dice
        Play:  2-ply cubeless prune
        keep the first 0 0-ply moves and up to 8 more moves within equity 0,08
        Skip pruning for 1-ply moves.
        Cube: 0-ply cubeful prune
    5. Rollout          15/7(2)                      Eq.:  -0,4043 ( -0,0960)
        33,69  10,49   2,95 -  66,31  11,76   0,00 CL  -0,4043
      [  0,76   0,99   1,48 -   0,76   0,82   2,62 CL   0,0155]
        Full cubeless rollout (trunc. at one-sided bearoff) with var.redn.
        108 games, Mersenne Twister dice gen. with seed 859211562 and quasi-random dice
        Play:  2-ply cubeless prune
        keep the first 0 0-ply moves and up to 8 more moves within equity 0,08
        Skip pruning for 1-ply moves.
        Cube: 0-ply cubeful prune
    6. Rollout          15/11(2) 6/2(2)              Eq.:  -0,4432 ( -0,1349)
        32,96  11,39   2,02 -  67,04  15,42   0,71 CL  -0,4432
      [  0,65   1,19   1,75 -   0,65   0,72   0,69 CL   0,0142]
        Full cubeless rollout (trunc. at one-sided bearoff) with var.redn.
        108 games, Mersenne Twister dice gen. with seed 859211562 and quasi-random dice
        Play:  2-ply cubeless prune
        keep the first 0 0-ply moves and up to 8 more moves within equity 0,08
        Skip pruning for 1-ply moves.
        Cube: 0-ply cubeful prune
    7. Rollout          15/11(2) 13/9(2)             Eq.:  -0,4860 ( -0,1777)
        30,33   8,77   1,16 -  69,67  13,95   1,90 CL  -0,4860
      [  0,61   0,87   0,80 -   0,61   0,83   0,80 CL   0,0131]
        Full cubeless rollout (trunc. at one-sided bearoff) with var.redn.
        108 games, Mersenne Twister dice gen. with seed 859211562 and quasi-random dice
        Play:  2-ply cubeless prune
        keep the first 0 0-ply moves and up to 8 more moves within equity 0,08
        Skip pruning for 1-ply moves.
        Cube: 0-ply cubeful prune

[close]

[Zorba]

Okay, I'll try to do the next roll. I just turned the "manual dice" off on GnuBG, however tempting it was to just pick a roll myself for factotum

Code Select Expand

GNU Backgammon will now use the Mersenne Twister generator.
Seed initialised by system clock.

Here we go, stay tuned