FIBS Board backgammon forum

If the board image does not appear in the question above, you may view it in the attached PNG file below.

one vote per user .. results of the poll are shown after you have voted

Sometimes the cube can be a most intimidating weapon, but it is a double edged sword. In this situation, RED's cube, if declined, gives RED an imposing 3-0 lead. If accepted, WHITE with chances of escaping and maintaining the lead in pip count can tie the match with a win or with a gammon or a redouble have a chance to take the lead into the Crawford game. GNUbg says the correct actions are

Spoiler

In this match, the finals in adz's "cheap and nastyl" - 3/5 tourney, with some hesitation I accepted johwayne's agressive cube, escaped and tied the match with a declined redouble a few moves later. Frankly, I thought it was a good cube and considered passing. GNUbg said it was an awful cube making my considering to pass equally awful.

Cube analysis

4-ply cubeless equity  +0.131 (Money:  +0.098)
  0.518 0.138 0.002 - 0.482 0.077 0.002

Cubeful equities:
1. No double            +0.060
2. Double, pass         +1.000  ( +0.940)
3. Double, take         -0.557  ( -0.617)

Proper cube action: No double, take (39.6%)

0-ply cubeless equity  +0.128 (Money:  +0.091)
  0.518 0.144 0.003 - 0.482 0.088 0.002

Cubeful equities:
1. No double            +0.043
2. Double, pass         +1.000  ( +0.957)
3. Double, take         -0.501  ( -0.544)

Proper cube action: No double, take (36.2%)

[close]

you may enter the ID's below into GNUbg or Snowie to evaluate your answer:

Position ID: OM/BAyDGTuGAAw Match ID: MAGgACAAAAAA


There comes a time when a man must spit on his hands, hoist the black flag and begin slitting throats.
-- H. L. Mencken US editor (1880 - 1956)               

What's a "staggler"? (I had to ask before dorbel does). johwayne is far from the only fibster I know whom's extremely aggressive with the cube - I suspect surfing alters one's perception of risk.

a "staggler" is what you get when spell check doesn't work in the poll question  thanks for the correction

Spoiler

The first option in poll should have been no double, take instead of no double. Like gnu says: Proper cube action: No double, take (39.6%)

[close]

Ok , I am picky , i know.

adrian, the context in which i ask the question of "What are the proper cube decisions?" is slightly and subtlely different from the way GNUbg presents its cube analysis. So, a decision of "No cube" stands alone as an independent choice as there is no response when no cube is offered. If and only if a cube is offered, do the take/pass responses make sense. Too good to cube is another category still. IMO, these four choices for answers fit almost all cube problems in which I ask the question as above. A differently worded question might use the style you suggest

Quote from: sixty_something on November 27, 2008, 12:31:53 AM
IMO, these four choices for answers fit almost all cube problems in which I ask the question as above. A differently worded question might use the style you suggest

Sometimes it's Too good to double/Take. I saw the related position in a BG board a few months ago, if my memory serves me right.

recently, one of my cube problems (117 or 118) drew a comment from spielberg which caused me to go back and add a Too good to double/Take answer .. i thought about that afterwards .. in hindsight, i didn't like that answer, but in the context of that problem and spielberg's comment i left it in

in general, i don't ever recall GNUbg suggesting a Too good to double/Take in its evaluations .. for me, in the context  or GNUbg evaluations, it doesn't even make sense as a possibility .. while it certainly may be the case where a given match score when a cube is offered, even when too good to double, a take and redouble might be in order or optional .. however, in that case it is not likely a cube thrown from what GNUbg describes as a too good to double situation .. so, for the context of my cube problems and the manner in which i ask for "proper cube decisions" such a case would more likely simply fall into the category Double/Take .. i use Too good to double/Pass on almost all cube problems .. adding Too good to double/Take to all such problems may be worth considering, but at present i can see doing that on a case-by-case basis at best

I'm not entirely certain but I think it's something like this: Too good to double/Take means that the gammon chances are good for the cube holder, as usual, however the double not only wastes that potential but is also a gift to the trailer because they are so weak in position and match score that they are pretty dead. The cube is a boon and they should take and redouble with gratitude - and a prayer to the Dice Gods for the hit that they need.

Too good to double/take is a nonsense. See my reply to Spielberg at the end of the thread to pos 118. In matches there are only four possible cube actions. They are no double/take, double/take, double/pass and too good to double/pass.
A position that is too good to double means that the player on roll has an equity of greater than 1 point if he plays on, so clearly if he incorrectly offers the cube then the player being doubled passes to limit his loss to 1 point.

Stranger things have happened. In a hypergammon match I got, and I swear this is true, Proper cube action: Eat it!

But, regarding the Too Good to Double/Take, you might want to consider the following

http://groups.google.co.uk/group/rec.games.backgammon/browse_frm/thread/50c19d48e279049?hl=en&ie=UTF-8&q=%22Too+good+to+double/Take+%22

  very interesting, bunny .. thanks for that link .. i didn't know GNUbg ever used Too good to double, take, but that appears legit .. i don't have time to look at it right now, but will later .. on first glance it appears such a situation is dependent on the match score as we have suggested .. in that one, the cube is thrown by the side ahead 8-0 in an 11 point match .. i wonder if the correct next play is a redouble .. in that situation, facing a 0-10 Crawford game or even loss by gammon, i'd redouble immediately

I think that that post has to be a Gnu glitch.

Quote from: dorbel on November 29, 2008, 12:39:00 PM
I think that that post has to be a Gnu glitch.

You might be right.

Quote from: Douglas Zare in rec.games.backgammon on 12th Dec 2002
In match play, there are many positions in which you are too good to
double, but where your opponent can take if doubled. Of course, this
shouldn't be the case in money play... but is there a simple proof? Can
anyone cite a reference?

It comes out as a consequence of something that will be in my December
25th column in GammonVillage.com , but my proof is a bit convoluted.


Douglas Zare

And you might be wrong.

http://groups.google.co.uk/group/rec.games.backgammon/browse_frm/thread/42e272fe26001a85/7e23231a4eacc222?hl=en&lnk=gst&q=%22Too+good%20to+double%22+Take+#7e23231a4eacc222

If you understand zare's proof, expain it to me!

Quote from: dorbel on November 29, 2008, 11:44:53 PM
If you understand zare's proof, expain it to me!

Zare's proof doesn't pertain to what we're talking about. I quoted his entire first post as an appetiser. Methinks you didn't bother reading the thread, which is what you were being invited to do.

So there's GnuBg, Racoon and Douglas Zare all spouting "nonsense" and there's you declaring it to be this nonsense, yet you haven't read the threads. (Or if you have, your challenge indicates that you didn't understand it.)

Which doesn't further the discussion so I'll leave it at simply: "Experts' opinions are divided".

Actually I did read the thread. Zare's proof in his Dec 25th column will if I read him correctly pertain to this. Doug Zare's writing is fairly impenetrable to me at the best of times, so if he says that his proof will be "a bit convoluted", watch out. As I say, let me know if you understand it.
Perhaps in the end, as Hugh McNeill points out, it is all a matter of semantics. The important thing is to arrive at descriptions that enable us all to understand the position. For me, the theoretical possibility of "too good/take" isn't helpful, perhaps for DZ, who has a brain the size of a planet, it is helpful.
Note that all these descriptions are theoretical and assume correct cube actions by the opponent. Of course if you factor in the possibility of incorrect takes or passes, now and later, all sorts of cube actions become possibly correct.

Quote from: dorbel on November 30, 2008, 11:49:06 AM
Actually I did read the thread. Zare's proof in his Dec 25th column will if I read him correctly pertain to this. Doug Zare's writing is fairly impenetrable to me at the best of times, so if he says that his proof will be "a bit convoluted", watch out. As I say, let me know if you understand it.

As I said, his proof is irrelevant. It's a proof that ND>DP>DT (Too Good/Take) and ND>DT>DP (Double/Pass) cannot occur in money games.

lol, as I said, I can hardly ever understand what doug is on about! So where is the proof that Too good to double/take is a theoretical possibility in matches? Can we see a position that fulfils those requirements? The position we were shown in 2006 by fretmeister certainly isn't one.

Lol. No, my last post was only to correct your misunderstanding of the man whose output you cannot fathom. I'm quite happy to leave it as before: Experts' opinions are divided.

"Too good to double, take" is definitely an option in match play and not extremely rare or theoretical. Raccoon explained it very well.

Too good to double means your equity BEFORE doubling is over 1. It does NOT mean your equity AFTER a double/take is over 1. That's the catch.

It happens most often in situations where the player is LEADING at a 2-away score (versus 3-away or 4-away) in a highly gammonish position. Why? Because when the leader doesn't cube, gammons will win him the match, but when he does cube, his gammons have no value anymore and the trailer can rewhip to put the entire match on the line.

So if the LEADER has a position where he could win a lot of gammons, but at the same time he's not that big a favorite to win, the leader should play on for the gammon (he wins relatively many), but if the leader doubles anyhow, the trailer should TAKE, because now gammons have no value anymore, and the trailer can put the match on the line with his still considerable SINGLE wins.

In numbers, something like:
60% total wins, 40% gammon wins, 40% losses (i.e. a gammon rate of 67%, 2/3 of wins will be gammons)
will qualify for "too good, take" at a 2-away 4-away score