problem slam hand
Adam Lea
Playing 5cm, 3 weak twos, 15-17NT:
North South
864 AQ953
AKQT96 7
J AKQT6
AK3 J7
1H 1S
3H 4D
4S 5D
6S P
4D was natural
5D was a cue bid
As you see partner and I fell for the "weak trump suit" trap, where 6S
suffers from two spade losers (KTx offside) but grand slams are available in
7D, 7H and 7NT. This was a large swing out. How could we have avoided this?
Thanks
Adam
Bob Lipton
It happens. You might have used blackwood and made your way to 7NT at
MP or even 6NT, but it's hard to switch out of of a major. However, you
you were going to bid 6S, why not continue cuebidding with 5H?
Bob
Alex Martelli
7H is NOT a good contract -- the HJ is unlikely to fall in 3 rounds.
7NT isn't all much better (there are some squeeze chances if Hs don't
split but the S finesse works, but they still fall short of the overall
chance you should have for a grand).
I believe that 7D, while probably the best grand (if D are 4-3 you get
the chance to ruff out a H and make even with H 4-2 and the HJ with the
4-cards length, as is more likely), is still a little below the chances
you'd LIKE to have before you bid a grand. I wouldn't lose any sleep
about missing any of these grands -- the only one that MAY be worth
bidding (7D) depends on the DJ (or from N's viewpoint the DT) which are
almost impossible to ascertain in the bidding.
Focus should rather be on avoiding 6S with the too-weak trump suit, and
even that is hard enough. Basically, over 5D N _might_ have bid 5NT to
mean "choice of slams", i.e. "let's play a slam, but which one's best is
not yet clear to me" -- if that's what 5N bids mean in your partnership,
and I do think that's what they SHOULD mean. On 5N, S would bid 6D
(very happy to play there if that's what N had in mind), and N would bid
6H -- this implies totally yecchy spades and incredibly good hearts,
since this ratio of suit quality is the only one that would leave N in
doubt between H and S slams. Now S would have to show good judgment by
passing 6H despite his H single -- N's hearts, for this sequence, should
be strong enough to play in front of a singleton, and spades should
surely be xxx, offering discarding opportunities on diamonds.
But it's definitely a difficult hand to bid in a natural system!-)
Alex
frisbie...@yahoo.com
South has shown a weakness in clubs and doesn't seem excited about
spades, North knows something funny is going on but not what, so in
hindsight North should bid 5NT to show clubs and minimal spades and let
partner take over.
Sandy E. Barnes
news:n8-dnYBVgbUMCHLZ...@bt.com...
***
1H - 1S!
3H - 4D
4S - 5D
6H - PASS
! = either leas than 3 hearts, or less than a game force, or both.
Partner can convert from 6H 2 6S if he chooses, since you have already
offered support for spades..
Sandy Barnes
***
ewleo...@hotmail.com
I don't think it is clear how to bid to a grand without taking a view.
However, the final bid of 6S was ridiculous. With three small spades,
you would want to present other alternative slam contracts to partner.
The hand screams for bidding 6H with a near solid six card heart suit.
Eric Leong
Charles Brenner
Alex Martelli wrote:
> Adam Lea <asr...@btinternet.com> wrote:
>
> > Swiss Teams, N/S Vuln
> >
> > Playing 5cm, 3 weak twos, 15-17NT:
> >
> > North South
> > 864 AQ953
> > AKQT96 7
> > J AKQT6
> > AK3 J7
> 7H is NOT a good contract -- the HJ is unlikely to fall in 3 rounds.
"Unlikely" sounds like under 50%, whereas it is 54% (using my
well-known rule). Agreed that's not good enough -- especially if the
other table is not even making a slam as the actual result shows is a
possibility.
> I believe that 7D, while probably the best grand (if D are 4-3 you get
> the chance to ruff out a H and make even with H 4-2 and the HJ with the
> 4-cards length, as is more likely), is still a little below the chances
> you'd LIKE to have before you bid a grand.
unclear. It's 74%.
I'm not sure how a capitalized LIKE differs from lower case. At rubber
bridge, once I bid a grand I suppose I'd LIKE it to turn out to be
100%. At teams and before bidding it I want the mathematical odds that
give me a positive expectation taking everything into account -- but
that's just the ordinary, uncapitalized meaning of "like".
All very picky I admit. Your English is so expert and idiomatic that I
cut you no slack, maybe unfairly.
Charles
Alex Martelli
> Alex Martelli wrote:
> > Adam Lea <asr...@btinternet.com> wrote:
> >
> > > Swiss Teams, N/S Vuln
> > >
> > > Playing 5cm, 3 weak twos, 15-17NT:
> > >
> > > North South
> > > 864 AQ953
> > > AKQT96 7
> > > J AKQT6
> > > AK3 J7
>
> > 7H is NOT a good contract -- the HJ is unlikely to fall in 3 rounds.
>
> "Unlikely" sounds like under 50%, whereas it is 54% (using my
> well-known rule). Agreed that's not good enough -- especially if the
I apologize for not knowing your well-known rule -- URL please? Anyway,
54% is definitely correct, yes.
> other table is not even making a slam as the actual result shows is a
> possibility.
Right. The conditions are "vulnerable at IMPs" (and in a short match
too -- Swiss teams -- though the effect of _that_ is harder to gauge).
7H making, vs 6S-1, is worth 2310, 20 IMPs; 6H+1, vs the same 6S-1, is
1560, 17 IMPs (same IMPs as 1530 for 6H just made).
So, if you bid the 54% grand, and the other table is going down 1 in 6S,
your IMP expectation is 20*0.54 = 10.80; while if you bid 6H, your
expectation is much larger: 17. Bidding 7H "costs" 6.2 IMPs (by
reducing your expected gain) in this case.
If the other table is playing 6H, then in 7H you will be +750 when it
makes (13 IMPs), -1530 when it goes down (-17 IMPs); your expectation
for bidding a 54% 7H is now 13*0.54-17*0.46 = -0.80 IMPs, not quite as
costly but still a losing proposition. While "unlikely" may carry a
subtly wrong nuance, "less likely than it would have to be to make the
grand worth bidding" (14 times as many words;-) might be precise enough
to gladden any pedant's heart!-)
> > I believe that 7D, while probably the best grand (if D are 4-3 you get
> > the chance to ruff out a H and make even with H 4-2 and the HJ with the
> > 4-cards length, as is more likely), is still a little below the chances
> > you'd LIKE to have before you bid a grand.
>
> unclear. It's 74%.
I may be off-base, but it appears to me that your number is wrong -- are
you considering the risks of a first-round ruff of the lead, which even
thinking just of spades must be something like 2%? So it seems to me it
should be below your stated 74 and closer to 71 instead. But if you're
kind enough to show us your computations in detail we'll all be able to
agree on the theoretically right number (of course at the table one may
always hope to get a little bit more -- opps may mislead [or misdefend],
opps' silence may sometimes be taken as a slight indication that their
hands don't have wild shapes -- but that's a very iffy basis anyway).
> I'm not sure how a capitalized LIKE differs from lower case. At rubber
> bridge, once I bid a grand I suppose I'd LIKE it to turn out to be
Ah, but you're addressing an importantly different viewpoint here: what
you would like to be the case AFTER you've bid the grand -- I was very
specifically referring to what one would like the probability to be
BEFORE bidding it (as below explained, caps == emphasis).
> 100%. At teams and before bidding it I want the mathematical odds that
> give me a positive expectation taking everything into account -- but
> that's just the ordinary, uncapitalized meaning of "like".
>
> All very picky I admit. Your English is so expert and idiomatic that I
> cut you no slack, maybe unfairly.
NP -- as a confirmed pedant myself, there are few things I like better
than debating with another. Uppercasing a word is a common Usenet
typographical convention for emphasis (which in other typographical
realms might be rendered with italics, in ordinary conversation with
emphatic tone of voice and/or [particularly to an Italian like me] with
appropriate hand gestures). So, for example, one might choose to write
"I like bridge" to indicate ordinary and mundane liking, vs "I LIKE
bridge" to indicate emphatic, enthusiastic, intense liking. If you had
to rephrase it without any support from tipography, tone of voice, or
body language, you might rewrite "LIKE" as "really, truly, intensely,
emphatically like", if you were into colorful, verbose redundancy.
Theoretically, at IMPs over a _long_ match, you should "like" to bid
vulnerable grand slams with odds of about 57% or better, if you were
certain that the other table bids at least a small slam (one that's
certain to make, ignoring for now the "5 or 7" hands). If you knew the
other table had gone 1 off in a wrong contract, so that you have a +17
available for the taking by just stopping at a certain small slam, your
IMP expectations would be higher if you only bid grand slams with an 85%
or better chance of making. If you knew the other table had stayed in
game, then you have +13 available in a small slam -- if you bid the
grand you will be +17 if it makes, -13 if it goes down; so your IMP
expectation would be higher if you only bid grand slams with about 87%
or better chance.
The latter two probability thresholds are close enough that we can
usefully conflate them in an 86% threshold for bidding a grand slam if
the other table "bids badly" (stops in game or reaches the wrong
contract) vs a 57% threshold if the other table "bids well" (reaches the
right small or grand slam). So, calling p the probability (as a
fraction) that the other table bids well, your threshold would be 57%
times p plus 86% times (1-p). For example, if p is 0.5, your threshold
should be around 72; so, you'd like to bid a 74% grand, but it shouldn't
be a particularly intense liking -- the difference in expected gain
between bidding that grand and stopping in a small slam is too small to
get excited about it, and it depends on your guess about p, anyway.
By contrast, you'd really truly emphatically LIKE to bid a 90% grand
slam -- THAT is where the emphasis and excitement should properly be,
because it's a choice that may offer very substantial gains in the long
run, particularly against the stronger teams (p closer to 1) where you
especially need those gains (it should be clear that, strategically and
under win-loss conditions, gains are more important against teams close
or superior in strength to yours -- against markedly inferior teams you
can expect to win the match based on their errors by just avoiding
serious mistakes yourself).
A shorter match complicates the strategic analysis, in ways that are are
somewhat hard to take into account: in a sufficiently short match there
may be very little practical difference between winning 17 and winning
20 IMPs on one hand, compared to the difference that would make in a
sufficiently long match, for example. But translating these qualitative
conditions into arithmetic is another matter;-).
I hope I have clarified the role of the uppercasing "LIKE" in my
original post and sufficiently justified my use of that device;-).
Alex
Andrew
The obvious route would have been for responder to bid RKC. A 4NT call
appears to violate one of the common Blackwood rules--don't bid
Blackwood when lacking control in a side suit--however, I think it is
excusable on this hand. Responder holds 16 prime HCP opposite a hand
that was strong enough to make a jump rebid. It is very unlikely that
opener can hold a hand that is strong enough to rebid 3H yet lacking a
club control.
An RKC auction might go like this:
1H 1S
3H 4D
4S 4NT
5H 6D
6H 6NT
Andrew
Charles Brenner
Alex Martelli wrote:
> Charles Brenner <cbre...@berkeley.edu> wrote:
>
> > Alex Martelli wrote:
> > > Adam Lea <asr...@btinternet.com> wrote:
> > >
> > > > Swiss Teams, N/S Vuln
> > > >
> > > > Playing 5cm, 3 weak twos, 15-17NT:
> > > >
> > > > North South
> > > > 864 AQ953
> > > > AKQT96 7
> > > > J AKQT6
> > > > AK3 J7
> >
> > > 7H is NOT a good contract -- the HJ is unlikely to fall in 3 rounds.
> >
> > "Unlikely" sounds like under 50%, whereas it is 54% (using my
> > well-known rule). Agreed that's not good enough -- especially if the
>
> I apologize for not knowing your well-known rule -- URL please? Anyway,
> 54% is definitely correct, yes.
I mentioned it a couple of times ("well-known" is facetious) a year
ago in r.g.b., but finding the message and obtaining the URL of a
message are beyond my skill level. However:
http://dna-view.com/suitbreaks.htm.
> > > I believe that 7D, while probably the best grand (if D are 4-3 you get
> > > the chance to ruff out a H and make even with H 4-2 and the HJ with the
> > > 4-cards length, as is more likely), is still a little below the chances
> > > you'd LIKE to have before you bid a grand.
> >
> > unclear. It's 74%.
>
> I may be off-base, but it appears to me that your number is wrong -- are
> you considering the risks of a first-round ruff of the lead, which even
> thinking just of spades must be something like 2%? So it seems to me it
> should be below your stated 74 and closer to 71 instead.
I didn't consider the spade void. Incidentally, I will go down if
either opponent has a spade void, regardless of the lead.
So assume --
No spade void, plus either
diamonds 5-2 and hearts run, or
diamonds 4-3 and hearts run or Jxxx or J-fifth onside.
For the probability of all such distributions I get 71.8%, so yes,
closer to 71% than to 74%. (The spade void doesn't change the result as
much as we might have expected because spade voids tend to be hand with
bad red splits anyway. The same principle applies in the extreme with
respect to club voids -- if RHO can ruff the club lead the hand was
down anyway unless RHO is exactly 4540.)
> > I'm not sure how a capitalized LIKE differs from lower case. At rubber
> > bridge, once I bid a grand I suppose I'd LIKE it to turn out to be
>
> Ah, but you're addressing an importantly different viewpoint here: what
> you would like to be the case AFTER you've bid the grand -- I was very
> specifically referring to what one would like the probability to be
> BEFORE bidding it (as below explained, caps == emphasis).
Exactly. Had you meant AFTER, then I could understand the sense implied
in emphasizing "LIKE". But since you said "before" and I presumed you
meant it, I could not understand.
> > 100%. At teams and before bidding it I want the mathematical odds that
> > give me a positive expectation taking everything into account -- but
> > that's just the ordinary, uncapitalized meaning of "like".
> >
> > All very picky I admit. Your English is so expert and idiomatic that I
> > cut you no slack, maybe unfairly.
>
> NP -- as a confirmed pedant myself, there are few things I like better
> than debating with another. Uppercasing a word is a common Usenet
> typographical convention for emphasis ...
We agree that it indicates emphasis, indicating a distinction compared
to what one would mean without the emphaiss
> [snip excellent discussion of mathematical criteria]
> I hope I have clarified the role of the uppercasing "LIKE" in my
> original post and sufficiently justified my use of that device;-).
Not really. On the one hand you discuss the threshold for bidding a
grand in terms of various mathematical models. Making a judgement on
that basis, no justification for emphasis.
On the other hand you discuss the difference between a big expected
gain and a little one -- that's the context in which is implied by
"LIKE".
However, in your original post you gave a mathematical discussion
ending "is still a little below the chances you'd LIKE to have before
you bid a grand." The mathematical context and especially the phrase "a
little below" to me gainsaid the bonanza-vs-marginal distinction (my
"other hand") and suggested the "one hand". Hence my slight confusion
about the capitalization.
With friendly regards,
Charles
Alex Martelli
...
> > > > > North South
> > > > > 864 AQ953
> > > > > AKQT96 7
> > > > > J AKQT6
> > > > > AK3 J7
> > >
> > > > 7H is NOT a good contract -- the HJ is unlikely to fall in 3 rounds.
> > >
> > > "Unlikely" sounds like under 50%, whereas it is 54% (using my
> > > well-known rule). Agreed that's not good enough -- especially if the
> >
> > I apologize for not knowing your well-known rule -- URL please? Anyway,
> > 54% is definitely correct, yes.
>
> I mentioned it a couple of times ("well-known" is facetious) a year
> ago in r.g.b., but finding the message and obtaining the URL of a
> message are beyond my skill level. However:
> http://dna-view.com/suitbreaks.htm.
Thanks, very helpful mnemonic aid! Like the "72/rate" rule (to compute
how many years it takes to double money invested at compound interest
with rate% yearly) it allows useful instant computations, and I had not
heard of the rule.
> > > > I believe that 7D, while probably the best grand (if D are 4-3 you get
> > > > the chance to ruff out a H and make even with H 4-2 and the HJ with the
> > > > 4-cards length, as is more likely), is still a little below the chances
> > > > you'd LIKE to have before you bid a grand.
> > >
> > > unclear. It's 74%.
> >
> > I may be off-base, but it appears to me that your number is wrong -- are
> > you considering the risks of a first-round ruff of the lead, which even
> > thinking just of spades must be something like 2%? So it seems to me it
> > should be below your stated 74 and closer to 71 instead.
>
> I didn't consider the spade void. Incidentally, I will go down if
> either opponent has a spade void, regardless of the lead.
If you take the (say) trump lead with the J, and then (try to) use the
SA to come back to hand to draw trumps, yes, of course -- and I think
you're right in implying it's nevertheless the best line. Diamonds 5-2
is a solid 15.3% chance -- and under those circumstances the H jack
could still fall with about 48% conditional chance, if I'm computing
things right (from the extremely useful conditional probability of suit
divisions applet at <http://www.himbuv.com/cpe.htm>), so taking T1 with
the J in dummy is crucial to make the grand in 7.3% of the cases, while
S voids _in all_ are only 4%, and only in about half those cases could
you still make the grand by overtaking in hand at T1 (conditional
probability of D 4-3 either given S 5-0 being 51.6%).
> So assume --
> No spade void, plus either
> diamonds 5-2 and hearts run, or
> diamonds 4-3 and hearts run or Jxxx or J-fifth onside.
>
> For the probability of all such distributions I get 71.8%, so yes,
> closer to 71% than to 74%. (The spade void doesn't change the result as
> much as we might have expected because spade voids tend to be hand with
> bad red splits anyway. The same principle applies in the extreme with
> respect to club voids -- if RHO can ruff the club lead the hand was
> down anyway unless RHO is exactly 4540.)
Yep, the interplay of conditional probabilities is very interesting here
-- as is your "qualitative" rationalization of why the numbers go a
certain way (at the table it's unlikely one can do the exact
computations, but having the right qualitative indicators can help gauge
the best line of play -- or in theory the right contract, though you'd
need to be playing a very subtle relay system to get enough detailed
info to actually do "bidding as mental play" at the table to this level
of accuracy:-).
> > > I'm not sure how a capitalized LIKE differs from lower case. At rubber
> > > bridge, once I bid a grand I suppose I'd LIKE it to turn out to be
> >
> > Ah, but you're addressing an importantly different viewpoint here: what
> > you would like to be the case AFTER you've bid the grand -- I was very
> > specifically referring to what one would like the probability to be
> > BEFORE bidding it (as below explained, caps == emphasis).
>
> Exactly. Had you meant AFTER, then I could understand the sense implied
> in emphasizing "LIKE". But since you said "before" and I presumed you
> meant it, I could not understand.
I'm sorry but I do not understand. E.g., at rubber: _after_ you've bid
the grand, you're committed, so you'd like (with or without emphasis)
its chance to be as high as possible, ideally 100%. But _before_ you
bid it -- when you're still not committed and have to decide -- there's
a mathematically computable threshold of unemphasized "like", at which
you have a positive "epsilon" small gain in expected result by bidding
the grand, and a higher fuzzier threshold of emphatic "LIKE", where the
expected gain from bidding the grand is "very substantial"; some
threshold depending on individual psychology, risk-aversion in
particular, where you'd feel bad (feeling you've missed a very likely
substantial gain you could have taken) if you *didn't* bid the grand, to
look at the same phenomenon from a specular viewpoint.
> > [snip excellent discussion of mathematical criteria]
Thanks.
> > I hope I have clarified the role of the uppercasing "LIKE" in my
> > original post and sufficiently justified my use of that device;-).
>
> Not really. On the one hand you discuss the threshold for bidding a
> grand in terms of various mathematical models. Making a judgement on
> that basis, no justification for emphasis.
>
> On the other hand you discuss the difference between a big expected
> gain and a little one -- that's the context in which is implied by
> "LIKE".
And that's obviously fuzzy, because risk aversion varies among
individuals and contexts.
> However, in your original post you gave a mathematical discussion
> ending "is still a little below the chances you'd LIKE to have before
> you bid a grand." The mathematical context and especially the phrase "a
> little below" to me gainsaid the bonanza-vs-marginal distinction (my
> "other hand") and suggested the "one hand". Hence my slight confusion
> about the capitalization.
The expression "a little below" is also fuzzy, just like "a bonanza".
What expected gain for the grand would overcome risk aversion (including
uncertainties about what goes on at the other table) by enough to really
cause enthusiasm (or conversely, angst at NOT having bid the grand)? I
guess each individual must answer for him or herself. Personally, I
know I'm quite cool about NOT bidding grands depending on a 3-2 division
of a suit, so about 68% on -- I'd "like" (no emphasis) to bid them, if I
could be sure that's all they need AND reasonably certain that the other
table bids well, but it's not as strong as "LIKE"... for me. 72% is
close enough to 68% to make minor psycological difference (again to me).
When we get into the 80s and 90s then at some point I start perceiving
the grand slam as "cold" (even though mathematically it isn't, it's
getting close;-), so I'd really *LIKE* to bid _those_ grands (even
though I'm perfectly resigned to missing some that are so good only
because of the happy presence of a key Ten-spot or thereabouts!-).
Thanks for offering ample occasion for clarification and opportunities
for interesting analysis on many planes,
Alex