Match 1, Game 2, Move 5, Forum Double

[blitzxz]

Cube?

Score:
Black(forum) 1, White 0, match to 5

Pips:
Black 135 White 149

GnuID:
4HPKAwjGnoMBKA:cAmgAAAACAAA

[diane]

Spoiler

I think he will pass this, so that would push me towards not cubing, but on the other hand - it would be an easy point.  My instincts would be to play this a bit longer and see if it turns into something more lucrative.  It doesn't look to me like we will lose the option to cube at any time - so I would wait and see how this one goes.  If in  couple of rolls we haven't taken advantage of all his blots and our stronger board, then we can cube.

[close]

[NIHILIST]

Spoiler

The only reason to not double is that you think its too good. Otherwise not doubling allows him to play for free. The THREE Ps favor us, cash it.

[close]

stogaddedspoiler:)

Bob

[ah_clem]

As I said yesterday...

Spoiler

Double.
PRAT.  Position, Race and threats.  If you're ahead in two out of the three, it's a usually a double.

We've got a three point board, plus active builders, vs a one point board for factorum. We clearly have a better position.
We're ahead by 15 pips, so we're ahead in the race.
And we've got three blots to choose from, with no real threats against us.

That's three for three. If you start ticking off the possible rolls, only 66  plays poorly, and that's are not too bad.  55, 53, 33 and 22 are devastating, so there are market losers here.  64 63 54 and 53 probably lose market too, depending on the return roll.

Too good?  I'm not seeing that we're that far ahead.  But we might be next roll if we sit on the cube this time.
So, I say turn the cube.

[close]

[diane]

The only reason to not double is that you think its too good.

Spoiler

That is the thing with this game just now, I think there IS a soupcon of 'too good' with it

[close]

[socksey]

Spoiler

I say get it while the getting is good.  Too many things can go wrong.  Why wait to find out and possibly lose the game?

[close]

socksey



"The difference between genius and stupidity is genius has its limits." - Albert Einstein

[stog]

Spoiler

double/take!

[close]

[Zorba]

Spoiler

A very strong double, even though factotum is not on the bar. We have several advantages in this game, and they do add up. In no particular order:

• race (on roll with 14 pips advantage =~ 10%, just a double in itself if it were a race)
• homeboard/prime: 6,5 and 2 made and the eightpoint for some extra blocking. no prime yet but well set up to create one. factotum has nothing.
• backcheckers: although we still have two backcheckers versus factotum's one, they are nicely split and don't face any threats or a blockade right now
• threats: many indirect shots at two outfield blots, a hit would also start escaping a backchecker, many loose hits at factotum's backchecker, some pointing numbers that make the fourth innerboard point, several great double rolls

Anybody who's not completely sure that this is a take, should double! (The famous Woolsey doubling rule that can't be repeated often enough)

It looks like a clear pass to me, because factotum will lose a considerable amount of gammons as well and basically just has nothing going for him.

If you're absolutely sure this is a pass, you might consider playing on for the gammon. Is it safe here, to just take a roll? No, it's not at all. Consider a roll of 6-4, for instance. Not bad, we hit the outfield blot, escape the backman, get 10 extra pips with 24/14*. But what if factotum rolls 3-2 from the bar? He anchors and hits back. Can we still cash a point then? It seems clear that we can't, in fact it doesn't look like a double anymore, the 22pt anchor stabilizes a lot, and with us on the bar we don't threaten nearly as much as we do now.

Similar things happen if we hit loose 8/3* or 6/3*, and factotum hits back. In this case, factotum doesn't get the anchor, but he gains a lot of pips. Maybe we still have a double after some of these, but not cashes anymore. That means we're worth less than the full point we might get now by doubling.

Another good sequence for factotum might be when he rolls good doubles next turn, and gains back some of the initiative.

So, it's far from safe to play on for the gammon here, there are many sequences where factotum has a good take next turn which is costly, or where we don't even have a double anymore: very costly!

If we'd win an awful lot of gammons otherwise, it might still be correct to play on for the gammon, as a sort of gamble for two points instead of one, with the risk of losing one or more points occasionally. Here, we still need quite a lot to happen before we can really claim a gammon, so this doens't look like a good gamble.

Last but not least: suppose factotum will take our cube, while it's actually too good. Yummy!!!

So, as a slight modification of the Woolsey Rule, you could say: if you're not absolutely sure it's a stonecold pass, DOUBLE and even if you are, if you're not absolutely sure it's too good to double: DOUBLE.

[close]

[ah_clem]

Spoiler

...  My instincts would be to play this a bit longer and see if it turns into something more lucrative.  It doesn't look to me like we will lose the option to cube at any time - so I would wait and see how this one goes.

I see lot's of market losing sequences - pretty much any hit/dance sequence loses the market.  It might be a pass already, in which case cashing it now before we find out how wrong this one can go (and as Zorba points out, there are plenty of sequences that drop us below the double point)

[close]

[lewscannon]

Spoiler

cube. i doubt he'll take.

[close]

[diane]

Spoiler

It looks like a clear pass to me, because factotum will lose a considerable amount of gammons as well and basically just has nothing going for him.

So, it's far from safe to play on for the gammon here, there are many sequences where factotum has a good take next turn which is costly, or where we don't even have a double anymore: very costly!

If we'd win an awful lot of gammons otherwise, it might still be correct to play on for the gammon, as a sort of gamble for two points instead of one, with the risk of losing one or more points occasionally. Here, we still need quite a lot to happen before we can really claim a gammon, so this doens't look like a good gamble.

Last but not least: suppose factotum will take our cube, while it's actually too good. Yummy!!!

This might be why I still have confusion and get a lot of cube decisions wrong...there is an awful lot of contradiction there...

Why will he pass because he has the chance to lose a lot of gammons, but it isn't worth us seeing if one of those chances happens?

Why do cube to take his drop, but secretly hope he will take, because we like our chances here a lot?

[close]

[ah_clem]

Spoiler

Why will he pass because he has the chance to lose a lot of gammons, but it isn't worth us seeing if one of those chances happens?

A good rule of thumb is that you need to have twice as many gammon wins as (ordinary) losses to make it worth playing on for the gammon. The reason: If you can turn the cube and cash, you've gained 1.  If you play on, you're risking the sure point to gain a second point.  If you lose, you've gone from +1 to -1, or a loss of two. So, you're risking two points to gain one.  Thus, you need the gammon to happen twice as often as the loss to make it worth the risk.

I don't see twice as many gammons as wins in this position, so I don't think it's too-good-to-double.

Why do cube to take his drop, but secretly hope he will take, because we like our chances here a lot?

Because if it's actually a pass, taking is an error and his equity goes down.  It's  good for you when your opponent makes cube errors because your equity goes up.  (although dice being what they are, the cube mistake sometimes works in his favor)

[close]

[

[playBunny]

Spoiler

This might be why I still have confusion and get a lot of cube decisions wrong...there is an awful lot of contradiction there...

Why will he pass because he has the chance to lose a lot of gammons, but it isn't worth us seeing if one of those chances happens?

Why do cube to take his drop, but secretly hope he will take, because we like our chances here a lot?

It's a question of balance. Gammon chances without risk are great but not that ommon. In a situation like this, where we have decent blitzing chances but definitely no guarantee, and two backrunners with a minefield to traverse if our attack goes wrong, it's better to make our opponent pay the price for risk rather than us. You mentioned the "easy point" and that's what we go for.

We'd go for the gammon if we had good winning rate as well as a good gammon rate but those backrunners drop the wins right down. In these positions the high gammons give the opponent pause for thought and it's an optimistic spirit that will think more about escaping the blitz and going on to win than losing 4 points in a gammon.

Heh heh, I've been that optimistic spirit so many times it's uncountable. That's why I hate that cheating bot at Vog!  

[close]

[NIHILIST]

Spoiler

We'd love him to take this.

[close]

Bob

[ah_clem]

Spoiler

We'd love him to take this.

“You've got to ask yourself one question: 'Do I feel lucky?' Well, do ya punk?”

[close]

[factotum]

If the forum doubles, I pass.   But with a good roll, you might be on your way to a gammon.  You'll probably be able to double me out next roll anyway.  (Is there any provision for changing one's vote?)

[diane]

Spoiler

A good rule of thumb is that you need to have twice as many gammon wins as (ordinary) losses to make it worth playing on for the gammon. The reason: If you can turn the cube and cash, you've gained 1.  If you play on, you're risking the sure point to gain a second point.  If you lose, you've gone from +1 to -1, or a loss of two. So, you're risking two points to gain one.  Thus, you need the gammon to happen twice as often as the loss to make it worth the risk.

I don't see twice as many gammons as wins in this position, so I don't think it's too-good-to-double.

Because if it's actually a pass, taking is an error and his equity goes down.  It's  good for you when your opponent makes cube errors because your equity goes up.  (although dice being what they are, the cube mistake sometimes works in his favor)

Ok - you got me with you on the first point, I can see that now  

But still confused with the next one, I don't give a hoot where the equity is if I still lose, despite being the favourite to win.  I will stomp if that happens    So, I don't get any happiness from him taking what is a pass and then 'pulling an adrian'   and winning anyway.  So, if there are things to go wrong for us, then letting him get an extra bunch of points for a stroke of luck aren't my idea of fun.
[I have a thought I am struggling to pin down here].  So with an opponent who takes anything I would definitely not cube this until it looked like I was winning and had no hope of gammoning, with an opponent who is going to drop it, you have convinced me it is worth cashing the single because we don't have sufficient gammons to make it worthwhile.  I think Factotum will drop it - so that would mean you succeeded in changing my mind...here at least  

[close]

[socksey]

Very nice explaination, ah_clem!  Very informative.  Thank you!

socksey



"There's so much comedy on television. Does that cause comedy in the streets?" - Dick Cavett

[lewscannon]

Ok - you got me with you on the first point, I can see that now  

So, I don't get any happiness from him taking what is a pass and then 'pulling an adrian'

I thought that pulling adrians was mookie's specialty.

[lewscannon]

If the forum doubles, I pass.   But with a good roll, you might be on your way to a gammon.  You'll probably be able to double me out next roll anyway.  (Is there any provision for changing one's vote?)

It's early in the game, you're a good player and lots of bad things can happen, too. I'll take my point and run. Maybe at 2 away, 2 away it's worth seeing what happens, but not at this score. Besides, at 2 away I would've already cubed a couple of rolls ago, anyway.

[ah_clem]

But still confused with the next one, I don't give a hoot where the equity is if I still lose, despite being the favourite to win.  I will stomp if that happens    So, I don't get any happiness from him taking what is a pass and then 'pulling an adrian'  and winning anyway.  So, if there are things to go wrong for us, then letting him get an extra bunch of points for a stroke of luck aren't my idea of fun.

Well, nobody gets much happiness when their opponent takes an incorrect cube and comes back to win.   But you have to balance that out against the number of times he takes incorrectly and doesn't come back to win, and add in the number of times he not only doesn't come back but gets gammoned to boot.  Make sure you count the ones in the latter category twice.  Those double- and quadruple- wins outweigh the occasional double-loss.

Overall, you're ahead with your opponent's  incorrect take, but you should expect to be beaten occasionally. 

[Zorba]

If your opponent takes a pass, you should be very happy with that! And it's an extra incentive to double!

Sure, he might turn the game around occasionally, but he can also do that if you don't double now, so it's not as bad as it seems.

We are big favourites to win here though, with pretty high gammon chances too. If factotum takes, the most likely outcome is that we win 2 points instead of 1. Sometimes we even win 4 points (and the match) instead of 2, with a doubled gammon. And then also, sometimes, we lose; most likely 2 points. That's all in the game. What matters is the average outcome in the long run, not the few unlucky "adrians" you'll get  Since we're big favourites in this game for now, we'd like it a lot if we can raise the stakes.

In general, it's a very bad thing to let fear influence your doubling decisions. You'll do much better by taking the opportunistic view. When considering a double, look at the glass as half full, not half empty. It's all those wins and gammons you want for double points. The occasional losses or nasty recubes you'll get don't offset the big gains in the long run from doubling positions like these.

Quick example with numbers: suppose we win the game 2/3 of the time here. Suppose half of these wins are gammons, the other half just single wins. That means we win single 1/3 of the time and a gammon 1/3 of the time, and lose 1/3 of the time.

Th average or expected outcome without any cube (cubeless) would then be:

1/3 * +1 =  1/3 (single win, no gammon)
1/3 * +2 =  2/3 (gammon win)
1/3 * -1 = -1/3 (single loss)
---------------+
                2/3 points (this is also called cubeless equity)

A pass is 1 point, so we obviously gain from cube access here.

A double/take would double all values, and give us 4/3 points, or 1 1/3, so that's even better than getting a pass!

It's unfortunately not as easy as this, since if we don't double we can still double later, and if we double and factotum takes, he can redouble us later. The result is that the values are all closer to 1, but the idea is still the same: You win less than one point if you don't cube, you win exactly one point after double/pass, and you win more than one point after double/take, on average (i.e. in the long run).

P.S. About being too good here: you can't compare playing on for the gammon, with a double/take which of course also causes the play to continue. The difference is obviously that the cube will be on 2 after a double/take, which is very beneficial to us, as explained above.

Furthermore, as ah_clem explained, gammon wins over single wins are usually only about half as valuable as single wins over single losses.

[adrian]

So, I don't get any happiness from him taking what is a pass and then 'pulling an adrian'   and winning anyway.  So, if there are things to go wrong for us, then letting him get an extra bunch of points for a stroke of luck aren't my idea of fun.

I thought that pulling adrians was mookie's specialty.

That's all in the game. What matters is the average outcome in the long run, not the few unlucky "adrians" you'll get 

Do you speak about me here ? Am I already famous for my bad takes followed by incredibile luck?

[diane]

Do you speak about me here ? Am I already famous for my bad takes followed by incredible luck?

Hehehehe, I wondered when you find you are famous    It isn't so much the bad takes and incredible good luck, as maybe you have a good eye for when the tide is going your way 

[lewscannon]

Do you speak about me here ? Am I already famous for my bad takes followed by incredibile luck?

My comment had nothing to do with backgammon.