Game 3, move 4 rkb 5-2 NEXT (m5) Forum 5-4 NEXT, rkb 6-1 NEXT (m6) Forum double?

[Zorba]

OK folks, hold on, here we go with some nexts first. You may want to view the board from the previous vote to see what's going on, or just trust me and look at the result below.  

Last vote, Forum(blue) decided to play the 5-3 B/20 24/21. Now we continue:

roadkillbooks(white) rolls 2-5: Obviously he hits with 23/16*

Forum rolls 5-4 from the bar: obviously we return hit AND make the anchor with B/20 13/9*.

roadkillbooks rolls 1-6 from the bar: after entering with B/24, I feel confident rkb would choose to play 24/18 with the six, unstacking the third backman and leaving a less harmful shot than after 13/7. So, B/18 for rkb.

That leaves us with this position, which I offer as a cube decision for the Forum (blue):

jGfwIDCMZ+GABQ:cAmgACAAAAAA

[ah_clem]

Nice job with the dice, Zorba.

Spoiler

Position: almost symmetric, but we have an anchor on his 5, while he has his "anchor" still sitting on our one.  We have a blot on his 4, he has a blot on our 7.  Slight to moderate advantage in position.

Race: 9 pip lead.  Not much, but something.

Threats: His blot on our 7 is under the gun (3 builders pointed at it) and with our advanced anchor we won't be shy about hitting it loose.  He has no builders and a stripped 8 point.  Another slight to moderate advantage.

So, we have an advantage in all three, but not all that big an advantage.  We're behind in the match, so we should double a bit earlier than at an even score (or for money).

Market Losers? 6-2 fan, 2-2 fan, 66, 33. Not that many, but enough.

double. Mainly due to match score.

[close]

[dorbel]

Spoiler

I wouldn't double this yet. The main feature of this position for me is that we have a high anchor and rkb does not. Whether he gets one or not may be decided in the next couple of turns. If he can get one, or even just exchange hits, we have no advantage at all. I would like to take a roll and look for a more clear cut advantage before offering a tougher cube.
The "Let's double earlier because we trail in the match" school of thought has a lot of merit. Owning the cube has very little value for rkb at this score, so we don't give up much by giving it away. However, a take here is trivially easy and the one thing that the match leader wants when he is a bit behind in the game is an easy take. Doubling aggressively and hoping to win a gammon is a good way to turn a match around, but the tough-but-takeable later cube often steals a point through gammon fear.

[close]

[diane]

Spoiler

I think I would like to cube this...I see a good chance to do attacking things, and if we wait he will drop.

[close]

[lewscannon]

Spoiler

I don't see this as a cube yet. It's an easy take for rkb if we do. Maybe if we were down 3-0, but at 2-0 it's not that crucial to cube here rather than force him to make a more difficult decision down the road.

[close]

[blitzxz]

Spoiler

Far from it. Even 1-point anchor gives a lot winning chances and there's a big chance that roadkillbooks get higher anchor effectively balancing the game complitely. Volatility is also quite low. I don't see very big changes for next exchange of rolls.

[close]

[Zorba]

Spoiler

At this matchscore, it's good to stop before rolling here and ask yourself whether it's a cube already. The take is easy: getting hit is bad for rkb, but usually not devastating. Also, there may be several return shots that are very strong, or rolls from te bar that make the 21 or 20pt anchor. And, there's still plenty of rolls that don't hit at all, and then rkb has good chances to improve the situation of his backmen; lots of rolls make the barpoint and more or less equalize the position.

So, easy take, but it could still be a double? There is a fair amount of volatility here (a lot could happen on the next exchange). Hit/dance is the obvious sequence you should consider here for market-losing potential. But there's also various doubles that play very well.

At this matchscore, several of these are indeed market-losers, but we usually don't blow rkb out of the water. He will always have his acepoint game, giving a fair amount of counter-play and even thought it's the lowest anchor, it will still keep gammon losses limited, compared to a situation where we could blitz and close rkb out.

If we don't hit rkb (triple shots are missed more often than we like...) our advantage of the high anchor may be countered right away, and we'd be quite unhappy with our double.

So overall, it's pretty close at this matchscore I think (as explained by others), but our advantage and threats don't seem big enough yet to justify turning the cube. One more serious improvement, and we'd have a strong cube.

[close]

[Zorba]

The vote is tied at 7-7 right now. As this is an important decision and I'll be away for most of the evening here anyway,  I'll extend the voting period a bit and see if there's any change. If it's still tied, I'll decide for NO DOUBLE eventually, as this was leading the polls earlier.

[blitzxz]

Spoiler

One thing is that if we double now we will make game very easy for rkb. Sure there are some market losers, like hit 'n' dance, but that's not very likely. I foresee   here that we will improve and have almost perfect double next roll. This will also be very hard take/pass desicion and we will give good chance for error. Other likely scenario is that rkb makes anchor and it also gives lot more error chances since doubling is difficult in 2-0 lead. I haven't checked bots but as a practical point of view I'm 100 % sure this is no double

[close]

[ah_clem]

Rollout.

Spoiler

A  close decision, but no-double is correct.



Cube analysis
Rollout cubeless MWC  37.79% (Money:  +0.346)

Cubeful equities:
1. No double            39.79%
2. Double, pass         42.80%  (  3.01%)
3. Double, take         39.67%  ( -0.11%)
Proper cube action: No double, take (3.7%)
Rollout details:
Centered 1-cube:
  0.609 0.203 0.011 - 0.391 0.084 0.003 CL  37.79% CF  39.79%
[0.001 0.001 0.000 - 0.001 0.000 0.000 CL   0.03% CF   0.03%]
Player rkb owns 2-cube:
  0.614 0.206 0.011 - 0.386 0.087 0.003 CL  40.78% CF  39.67%
[0.001 0.001 0.000 - 0.001 0.000 0.000 CL   0.04% CF   0.04%]
Truncated cubeful rollout (depth 11) with var.redn.
1296 games, Mersenne Twister dice gen. with seed 860457240 and quasi-random dice
Play: 0-ply cubeful prune [expert]
Cube: 0-ply cubeful prune [expert]

[close]

[roadkillbooks]

Was the match score entered into this equation?

[Zorba]

0/2-ply GnuBG rollout:

Spoiler

Perhaps with 2-ply play, the decision would be a bit closer, as it usually gains the favourite. But it looks like this is just not quite good enough for a double at this matchscore. Pretty close though.



Rollout details:
Centered 1-cube:
   60,10  19,07   1,11 -  39,90   9,33   0,94 CL  +0,3628 CF  +0,5912
[  0,05   0,05   0,02 -   0,05   0,06   0,05 CL   0,0017 CF   0,0035]
Player roadkillbooks owns 2-cube:
   60,38  18,77   1,03 -  39,62   9,93   1,33 CL  +0,6844 CF  +0,5581 (-0,0331)
[  0,06   0,06   0,03 -   0,06   0,07   0,06 CL   0,0038 CF   0,0043]
Full cubeful rollout with var.redn.
15552 games, Mersenne Twister dice gen. with seed 863347199 and quasi-random dice
Play: 0-ply cubeful prune [expert]
Cube: 2-ply cubeful prune [world class]

[close]