you know your luck is running bad when .., (part 2)

[Adam Lea]

This hand came up last night against a couple of LOL's:

QJT764
J95
Q75
9
85 A3
T2 87643
AT9863 KJ42
T53 K6
K92
AKQ
-
AQJ8742

North opens 2S, no alert or announcement from South, so I ask what it is
(in EBU land it has to be alerted or announced). South said something
along the lines of "I don't know, I can't remember what we agreed to
play" and giggles. The auction proceeds:

N E S W
2S P 3C P
3S P 4NT P
5C P 6C AP

Partner asks how many aces the 5C bid shows and we get the answer "I
don't know, can't remember if it is 0 or 1" and another giggle. Partner
leads the diamond ace, declarer leads the spade king, partner signals
count with the eight, now I have to think. If partner has two it doesn't
matter but if partner has four, I should take the ace and declarer is
cut off from dummy, so I take the ace. As you can see declarer does have
three spades so can always get to dummy to finesse my king and claim 12
tricks, cue the phrase "that was lucky partner" and another giggle.

This was worth 2 out of 14 matchpoints for us, only one pair found 6S,
the rest were in club or spade games.

[Thomas Dehn]

How exactly will declarer get to dummy when you duck the first spade?


Thomas

[Adam Lea]

Damn, I hadn't spotted that.

[sbt]

In article <H_edneZJdJ-rVzPS...@bt.com>, Adam Lea

<lea...@btinternet.com> wrote:

> On 12/05/12 20:36, Thomas Dehn wrote:
> > On 05/12/2012 09:34 PM, Adam Lea wrote:
> >> QJT764
> >> J95
> >> Q75
> >> 9
> >> 85 A3
> >> T2 87643
> >> AT9863 KJ42
> >> T53 K6
> >> K92
> >> AKQ
> >> -
> >> AQJ8742
> >>

<snipping part of original>

> >>
> >> Partner asks how many aces the 5C bid shows and we get the answer "I
> >> don't know, can't remember if it is 0 or 1" and another giggle. Partner
> >> leads the diamond ace, declarer leads the spade king, partner signals
> >> count with the eight, now I have to think. If partner has two it doesn't
> >> matter but if partner has four, I should take the ace and declarer is
> >> cut off from dummy, so I take the ace. As you can see declarer does have
> >> three spades so can always get to dummy to finesse my king and claim 12
> >> tricks, cue the phrase "that was lucky partner" and another giggle.
> >
> > How exactly will declarer get to dummy when you duck the first spade?
> >
> >
> > Thomas
>
> Damn, I hadn't spotted that.

However, you could easily recast the play to keep the story intact by
having declarer lead a small spade to the board instead of leading the
king, initially. Then club finesse and drop your king, pull last trump
and then lead SK.

--
Dennis Cohen

[Adam Lea]

Well in the actual scenario it comes down to guessing whether declarer
has one or three spades. If she had three, then why is she not in 6S
when her partner has shown at least six?

[Player]

Try a small S to the dummy
Ron

[Thomas Dehn]

If partner has 9852 in spades, partner might play the S9 rather than the
S8. Trust partner rather than a weak declarer's play.

Then, declarer had already ruffed the D opening lead.
If she also had a S singleton K, her distribution
would be 1-3-0-9 or 1-4-0-8. In case she has
K-AKQx-void-AQxxxxxx, whether you take the SA is
just an overtrick; she can always reach dummy with
the HJ and finesse your CK.


Thomas

[Jürgen R.]

"Adam Lea" <lea...@btinternet.com> schrieb im Newsbeitrag
news:S4idnXPf6q96dzPS...@bt.com...

Because, according to Reese, "Little old Ladies are tricky" while
some of their opponents are not.

[Will in New Haven]

With KXX in Spades, declarer can always get to dummy, just not by
leading the King. So one would never play a competent declarer for
that holding if she played that way. On the other tentacle, we have
plenty of evidence for playing this declarer for a bad play.

--
Will in New Haven

[Charles Brenner]

On May 13, 1:19 am, Thomas Dehn <thomas-use...@arcor.de> wrote:
> On 05/13/2012 12:58 AM, Adam Lea wrote:
>
>
>
>
>
>
>
>
>
> > On 12/05/12 23:12, sbt wrote:

> >> In article<H_edneZJdJ-rVzPSnZ2dnUVZ8r6dn...@bt.com>, Adam Lea

Why? I might play top from four small in an exceptional circumstance
but the normal count card from four small as I have learned and play
is 2nd best. If you have the agreement instead to play top from four
that would coincidentally have helped given the reported layout, but
it has no theoretical advantage and would have confronted the
corresponding ambiguity if West held instead 9x. If you have no
understanding about the play from four small then there is always a
guess.

Charles

[Thomas Dehn]

If you have an explicit agreement to play the second highest from
four small in this situation, that is fine.

I am making a restricted choice type argument here.
If you do not have such an agreement, partner
might play either the 8 or the 9 from 98xx.
From 8x, partner will always play the 8.

Thus, when partner plays the 8, and you have to "guess"
whether he has 98xx or 8x, per restricted choice you have to
half the a-priori probability you assign to partner
holding 98xx, as something like half the time he
actually holds 98xx, he will play the 9.

If you have other partnership understanding that impact which
card partner might play from 98xx, then of course
percentages might be different.



Thomas

[David Goldfarb]

In article <9456d302-c44d-46cd...@vy9g2000pbc.googlegroups.com>,

Player <ron...@msn.com> wrote:
>> On 05/12/2012 09:34 PM, Adam Lea wrote:
>> > QJT764
>> > J95
>> > Q75
>> > 9
>> > 85 A3
>> > T2 87643
>> > AT9863 KJ42
>> > T53 K6
>> >
>> > K92
>> > AKQ
>> > -
>> > AQJ8742
>>
>>

>> How exactly will declarer get to dummy when you duck the first spade?
>

>Try a small S to the dummy

As the cards lie, west ruffs that, and declarer is down 2 since she
has no way to finesse in trumps and must lose to the CK also.

--
David Goldfarb |"I see more than you, child. I see an end to hell.
goldf...@gmail.com | What do you see?"
gold...@ocf.berkeley.edu | "I see a man in a lot of pain."
|"Pain? Yes. Consider it a preview." -- _Zot!_ #18

[Charles Brenner]

According to that argument, won't you always figure partner is more
likely to have 2 cards than 4? If so, there is something fishy. If
not, why not?

[Frances]

> Charles- Hide quoted text -
>
> - Show quoted text -

I agree with two of your statements:
- it is better to have an agreement on what to play from 4 low, and
stick to it, than not to have one
- whatever your agreement, you will sometimes not be able to tell the
difference between 2 and 4

However, I disagree that there is no theoretical advantage either
way. I always play top from 4 low.

The disadvantage is that sometimes you can't afford to play the top
card (e.g. dummy has 10853, you have 9642, partner has KQx, declarer
has AJx, playing the 9 gives 3 tricks)

The advantage is that partner can read it: if you play second highest
from 4 then you can't always tell the difference between bottom of 3
and second highest from 4. Even if you can't differentiate between 2
and 4 you do at least want to be sure if it is an odd or an even
number.

[Charles Brenner]

I can't figure out what the x's represent (teasing), but of course the
top can cost a trick and that seems to me the argument in favor of
second best. I may have overstated in assuming that it is standard.
Many years ago I learned it from Jeremy Flint and never considered
that it might be controversial.

> The advantage is that partner can read it: if you play second highest

> bfrom 4 then you can't always tell the difference between bottom of 3

> and second highest from 4. Even if you can't differentiate between 2
> and 4 you do at least want to be sure if it is an odd or an even
> number.

Interesting point. True that top is more likely to disambiguate
between 3 and 4, though it seems to me an exaggeration if you claim
that partner can always read it.
Although what I said was wrong in the two ways you have mentioned,
perhaps the idea that I had in mind was right -- that for purposes of
clarifying 4 vs 2, the two methods are equally good. If so, then it
comes down to whether the extra clarification of 4 vs 3 card length
outweighs the danger of dropping a trick by playing the highest card.
(Of course one can say that responder should settle for 2nd best when
there is a danger of dropping a trick by signalling highest, but such
a more complicated policy has downsides of its own.) Your conclusion
surprises me though I always assume in these matters that you've
thought it through more than I have.

Charles

[Thomas Dehn]

On the actual hand, there still is declarer's play of the SK,
and there is the fact that declarer already showed a D void.
In this particular hand, I'd play declarer for holding three spades,
but that is easily said when seeing all four hands.

There exists one holding for partner where he
has all missing four small spades, and there are five doubletons
headed by either the 9 or the 8. The latter combined are way
more likely. So it boils down to the question whether your
confidence in declarer's ability to not make the dumb play
of the SK rather than a small spade is sufficient
to outweigh those odds.


Thomas

[Charles Brenner]

Deciding on the prior is an irrelevant digression. Look back to what
you wrote above, which is about the logic after a prior has been
decided.

> There exists one holding for partner where he
> has all missing four small spades, and there are five doubletons
> headed by either the 9 or the 8.

That too is just a way of saying that in a sense (namely the dealing
probability for a single suit) the prior probability favors partner
having a doubleton rather than four cards. So this too doesn't help
resolve the apparent paradox in what you claimed.

To understand whether partner's spot can be helpful, suppose a 50/50
prior whether partner has two or four. You put forth a line of
reasoning which says that under that assumption (and assuming no
partnership agreement about signalling with four cards), partner's
card can be helpful. Now it's easy to see through the paradox. Partner
has all 4 with probability 1/2, and has any of the three doubletons
with total probability 1/2. Any of the 3 non-smallest cards is 1/3 to
appear from four, and equally appears from just one of the 3
doubletons. That's why there is no information from it.

Charles

[Thomas Dehn]

Ok, I'll word it differently.

Denote as x the probability that partner has 98xx,
and as y the probability that partner has 9x (including 98), and
as z the probability that partner has 8x (but not 98).
3z=2y, as here there are two 8x doubletons and three 9x doubletons.
I am ignoring the possibility of 98x.

I then denote p the probability that declarer plays the
K from S Kxx.


When you have an agreement to alway play the 9 from 98xx,
o when partner plays the 9, you have all of x and p of y (but no z)
o when partner plays the 8, you have z

When you have no agreement on whether partner will
play the 8 or the 9 from 98xx
o when partner plays the 9, you have half of x, and p of y (but no z)
o when partner plays the 8, you have half of x, and p of z (but no y)


So, lets take fake numbers. K vs. xxxx a priori is one
of ten 4-1 splits, 0.02826. Each of those doubletons
a priori is one of 20 3-2 splits, each 0.0339.
Then, lets say you judge that from Kxx in spades, this
particular declarer is very clueless, and
would play the K from S Kxx one third of the time, p = 1/3.

When you have an agreement to alway play the 9 from 98xx
o when partner plays the 9, 0.02826 vs. 1/3*3x0.0339; 46.01% for 98xx
o when partner plays the 8, 0% for 98xx

When you have no agreement on whether partner will
play the 8 or the 9 from 98xx
o when partner plays the 9,
0.5*0.02826 vs. 1/3*3x0.0339, 29.42% for 98xx
o when partner plays the 8,
0.5*0.02826 vs. 1/3*2x0.0339, 38.47% for 98xx


Thomas

[Charles Brenner]

I wish you hadn't. Better for purposes of communication would have
been to reply to what I wrote and say what is wrong with it or compare
it with what you believe.

Instead, you provide a post lacking either introduction or conclusion,
and littered with private language (e.g. "you have ..." means what?)
The net is that you decline to take the effort to understand what I
wrote and instead impose great effort on your readers to divine what
you have in mind but have not expressed in writing.

At a guess, your point is to contradict the assertion in my previous
post that neither spot (9 or 8) provides any information compared to
the other. Probably true and a consequence of the asymmetry I
overlooked from the fact that 98 doubleton always plays the 9. But was
that the logic in your mind when you claimed there is a restricted
choice argument? As I said, that argument as stated seems to say that
whichever spot the Ax defender sees the defender can take as evidence
for a doubleton. ("Evidence for" means increasing the probability of
the event compared to the prior probability.)

Charles

[Frances]

When the decision is close or ambiguous I do what my partner prefers
so I haven't personally done a detailed analysis on which is better.
Most of our arguments about defensive carding in this type of position
come when one of us doesn't play the highest card because they think
they can't afford it, and partner then objects that the card was
ambiguous. However, this is far more common in a (reverse) attitude
position than a count position.

I've been trying to construct a hand where 3 vs 4 actually matters and
it's not easy. (It's easy to come up with situations where there is
ambiguity, but less so where it affects your choice of plays)