Ken Eichenbaum described to me a cue concept that makes some sense. I recently saw a post on the BBF that illustrates where this would make a lot of sense. The idea of a WAC is that an "impossibly negative cue" is actually an "absolutely positive cue," showing stuff everywhere with uncertainty about passing the game level.
Here's an example.
2C-P-2D(waiting GF)-P-
2H-P-3H-P-
4D?!?!?
How on earth can Opener have a 4D cue? No black suit controls at all means at best (in HCP) QJ-QJ. To even have a 2C opening, Opener would almost need solid diamonds and solid hearts. Sure -- possible. Kind of a picture jump without the jump. But, really? Why not simply bid 4H with that four-loser hand? Or, bid 3NT and then 4D if partner bids 4C?
Well, maybe this call (one under game), when impossible (or practically impossible) should rather than DENY any club or spade control actually PROMISE control in every suit (with slam interest) but express uncertainty about the five-level. In other words, if a cue denies "too much," then it instead promises "all."
The problem presented was after a 3S cue, a 4D cue from Responder, and a 4H "signoff" by Opener, could Opener have a club control but not enough cause to enter the five-level. If so, then Opener can never check on a club control and yet assuredly stay at game if this is denied. If not, then Opener is in a pickle with controls everywhere but not enough general strength to justify entry to the five-level.
The WAC grabs all and solves this problem (or possible/perceived problem at least).
Funny thing is that this concept first was presented to me as a result of a humorous analysis of a bizarre cuebidding decision. Someone (who shall remain nameless) made a cue in a specific sequence that seemed impossible. A technical hand was possible, even if that seemed really unlikely. So, as we are prone to do, we decided that this might make sense if the person thought up and launched without discussion (as we sometimes do ourselves, admittedly) a "new concept" that "just had to be obvious." In reality, it was just a bizarre call, a decision to cue for the first call and then jump into an undiscussed pattern bid for who knows what reason. But, every once in a while an absurd call out of the blue can serve as a discussion point for some actually relevant idea.
Of course, a good idea is to discuss (if you like the WAC idea) when this should apply. For, a call that is "impossible" is rare. More often, the call would be "highly improbable" or for obscure reasons actually redundant. This specific sequence of a one-under-game-initial-cue after a 2C opening and a 2-3 raise of a major seems fairly "impossible." But, whipping this out mid-stream, unless you are into that sort of thing, would be somewhat unwise, I suppose.
Thursday, February 4, 2010
Wednesday, February 3, 2010
Patterns?
Just thinking out loud...
MANY years ago, I remember using mental "flash cards" to learn hand patterns at bridge. This is a common learning tool for getting used to the numbers at bridge. 4432, 5431, 6322. Stuff like that.
The next time this ever entered my mind was in learning parity signals.
Recently, however, I started thinking as to whether there are other ways to categorize patterns. Whether this would add any benefit or not is unclear. But, just getting a grasp on pattern type thinking might lead itself somewhere.
Here's an example of what I mean.
What do hands like 4432, 4441, 5440, and even 4333 have in common with some hand patterns like 6421, 6430, 5422, and 5431?
Well, nothing obvious. However, let's assume that the first group has a common theme, namely that the shortest suit is in one of the majors. Then, let's assume that the second group has a common theme, namely that the shortest suit is the same major as in the first group, the longest suit is diamonds, and the second-longest suit the other major. Now any similarities?
I'll pick one short major, the short spades example:
Group One: 2443, 2434, 2344, 1444, 0445, 0454, 0544, 3433
Group Two: 1462, 2461, 3460, 0463, 2452, 3451, 1453
Now, do you see any pattern similarities?
There is nothing numerically similar, at least not that I can see, other than the similarities forced into the example. However, from a bridge perspective, these two pattern groups are "related" in that all are appropriate for a takeout double, if an equal-level-conversion call is systemically allowed.
This is no great observation, in its own right. However, thinking about these groupings in a sort of brainstorming effort leads to some thoughts. (Might work better with pot.) Like, I notice that hands with primary clubs and secondary other-major do not fit into this grouping, which means that they are "left out," possibly needing a "solution." I think out loud about the "physical" qualities of this situation, where "weight" seems easily spread out or grouped at the top BUT with primary weight just below the top. It seems that the interests of bidding goals, coupled with practicalities and safeties, force certain "weight" tendencies to patterns suitable for certain approaches.
One might, then, plausibly devise a methodology where a certain call handles "convenient pattern holdings" but an alternative handles "inconvenient pattern groupings."
Imagine, for simple example, an overcalling structure over 1NT where immediate calls show "convenient grouping holdings," such as with, say, Cappelletti, whereas a double might simply show "inconvenient pattern holdings." After a 2C relay from Advancer, the Doubler might indicate some core feature of some "inconvenient pattern group." For instance, 2D after the relay might show some hand with diamonds plus a shorter major. Or, 2D might also cover some other diamond-focused "inconvenient pattern group" as well, if there is a way to describe the inconvenient pattern group because we know all of the diamond-focused entries into that pattern group. Of course, I don't know that group right now -- hence the thought experiment.
I mean, we all have these hands that pop up where we look at the hand and gasp at the complete lack of options. If we could somehow find that these "ugh!!!" hands have some "mechanics" causing the difficulties, we might somehow figure out a cure along the lines of some artificial capture technique.
Consider an example. Maybe a 2NT response to a major opening might actually serve better (than other ideas typically used) as a "flag" of a trouble pattern hand, with 3C asking to explain, and Responder identifying the difficult pattern.
Just a thought, perhaps.
MANY years ago, I remember using mental "flash cards" to learn hand patterns at bridge. This is a common learning tool for getting used to the numbers at bridge. 4432, 5431, 6322. Stuff like that.
The next time this ever entered my mind was in learning parity signals.
Recently, however, I started thinking as to whether there are other ways to categorize patterns. Whether this would add any benefit or not is unclear. But, just getting a grasp on pattern type thinking might lead itself somewhere.
Here's an example of what I mean.
What do hands like 4432, 4441, 5440, and even 4333 have in common with some hand patterns like 6421, 6430, 5422, and 5431?
Well, nothing obvious. However, let's assume that the first group has a common theme, namely that the shortest suit is in one of the majors. Then, let's assume that the second group has a common theme, namely that the shortest suit is the same major as in the first group, the longest suit is diamonds, and the second-longest suit the other major. Now any similarities?
I'll pick one short major, the short spades example:
Group One: 2443, 2434, 2344, 1444, 0445, 0454, 0544, 3433
Group Two: 1462, 2461, 3460, 0463, 2452, 3451, 1453
Now, do you see any pattern similarities?
There is nothing numerically similar, at least not that I can see, other than the similarities forced into the example. However, from a bridge perspective, these two pattern groups are "related" in that all are appropriate for a takeout double, if an equal-level-conversion call is systemically allowed.
This is no great observation, in its own right. However, thinking about these groupings in a sort of brainstorming effort leads to some thoughts. (Might work better with pot.) Like, I notice that hands with primary clubs and secondary other-major do not fit into this grouping, which means that they are "left out," possibly needing a "solution." I think out loud about the "physical" qualities of this situation, where "weight" seems easily spread out or grouped at the top BUT with primary weight just below the top. It seems that the interests of bidding goals, coupled with practicalities and safeties, force certain "weight" tendencies to patterns suitable for certain approaches.
One might, then, plausibly devise a methodology where a certain call handles "convenient pattern holdings" but an alternative handles "inconvenient pattern groupings."
Imagine, for simple example, an overcalling structure over 1NT where immediate calls show "convenient grouping holdings," such as with, say, Cappelletti, whereas a double might simply show "inconvenient pattern holdings." After a 2C relay from Advancer, the Doubler might indicate some core feature of some "inconvenient pattern group." For instance, 2D after the relay might show some hand with diamonds plus a shorter major. Or, 2D might also cover some other diamond-focused "inconvenient pattern group" as well, if there is a way to describe the inconvenient pattern group because we know all of the diamond-focused entries into that pattern group. Of course, I don't know that group right now -- hence the thought experiment.
I mean, we all have these hands that pop up where we look at the hand and gasp at the complete lack of options. If we could somehow find that these "ugh!!!" hands have some "mechanics" causing the difficulties, we might somehow figure out a cure along the lines of some artificial capture technique.
Consider an example. Maybe a 2NT response to a major opening might actually serve better (than other ideas typically used) as a "flag" of a trouble pattern hand, with 3C asking to explain, and Responder identifying the difficult pattern.
Just a thought, perhaps.
Monday, January 11, 2010
4NT in the Context of a Double Closeout
So, now we get to the big idea.
Suppose that two suits have been closed out.
An example.
1S-P-2C-P-
2D-P-2S
So far, GF, spades agreed.
1S-P-2C-P-
2D-P-2S-P-
3C-P-3D-P-
3S-P-4C
In this auction, Opener's bypass of 2NT (which would deny two of the top three spades) inferentially promised two of the top three spades. His cue of 3S later showed the third top spade, "closing out" the spade suit. We now know about the Ace, King, and Queen of spades.
Opener's cue of 3C showed one of the top three clubs. Responder's cue of 4C showed the remaining two top clubs. Thus, the side suit clubs is also "closed out."
In this situation, then, the closing out of clubs (one of the two side suits) converts diamond and heart cues to straight control bids.
So, a few things happen, and we know that the reds are covered before someone bids 4NT.
What is that?
Well, at this point, the only unknowns seem to be the nature of the red suit controls. Are these aces or kings? Shortness?
It seems to me that closing out both the trump suit and a side suit should make 4NT a rather precise bid. If the nature of the red controls in this example is unknown, then the responses to 4NT should clarify that issue. To explain, consider the example (black suits closed out, red suit control type unknown). What could be done now?
4NT = please explain red suit control types:
5C = One Ace
5D = No Aces
5H = Two Aces
If the person answering has one Ace and a void in the obvious suit, he bids 5S.
If the person answering has one Ace and a void, but there is no obvious location for that void, then 5S shows the lower void and 5NT shows the higher void.
If the person bids 5C (one Ace), then 5D asks for a possible void. Without a void, sign off at the lowest level. With a void in the obvious suit, cuebid resumes. If the location of the void would not be obvious, show the location of the void by steps (lowest non-signoff shows lowest void; highest non-signoff shows highest void).
If a void is not possible, then 5H+ show two Aces and are immediate cuebid resuming bids (including 5H). In that event, as well, there is no "asking bid" for the void, but cuebidding resumes.
If a triple closeout occurs, if that is possible, then 4NT should simply ask for clarification of the remaining suit. I'll let others come up with those rules, if anyone wants to go that far.
Suppose that two suits have been closed out.
An example.
1S-P-2C-P-
2D-P-2S
So far, GF, spades agreed.
1S-P-2C-P-
2D-P-2S-P-
3C-P-3D-P-
3S-P-4C
In this auction, Opener's bypass of 2NT (which would deny two of the top three spades) inferentially promised two of the top three spades. His cue of 3S later showed the third top spade, "closing out" the spade suit. We now know about the Ace, King, and Queen of spades.
Opener's cue of 3C showed one of the top three clubs. Responder's cue of 4C showed the remaining two top clubs. Thus, the side suit clubs is also "closed out."
In this situation, then, the closing out of clubs (one of the two side suits) converts diamond and heart cues to straight control bids.
So, a few things happen, and we know that the reds are covered before someone bids 4NT.
What is that?
Well, at this point, the only unknowns seem to be the nature of the red suit controls. Are these aces or kings? Shortness?
It seems to me that closing out both the trump suit and a side suit should make 4NT a rather precise bid. If the nature of the red controls in this example is unknown, then the responses to 4NT should clarify that issue. To explain, consider the example (black suits closed out, red suit control type unknown). What could be done now?
4NT = please explain red suit control types:
5C = One Ace
5D = No Aces
5H = Two Aces
If the person answering has one Ace and a void in the obvious suit, he bids 5S.
If the person answering has one Ace and a void, but there is no obvious location for that void, then 5S shows the lower void and 5NT shows the higher void.
If the person bids 5C (one Ace), then 5D asks for a possible void. Without a void, sign off at the lowest level. With a void in the obvious suit, cuebid resumes. If the location of the void would not be obvious, show the location of the void by steps (lowest non-signoff shows lowest void; highest non-signoff shows highest void).
If a void is not possible, then 5H+ show two Aces and are immediate cuebid resuming bids (including 5H). In that event, as well, there is no "asking bid" for the void, but cuebidding resumes.
If a triple closeout occurs, if that is possible, then 4NT should simply ask for clarification of the remaining suit. I'll let others come up with those rules, if anyone wants to go that far.
Closing out the suit?
What happens if a suit is "closed out" as far as controls?
Consider an example:
1S-P-2C-P-
2D-P-2S(spades agreed)
If Opener is able to cuebid 3C (two of the top three clubs) and Responder is then able to cuebid 4C (the missing club honor), or backwards (Responder cuebids 3C to show one and then Opener cuebids 4C to show the remaining two), then the suit is "closed out," meaning that this side suit is proven to be A-K-Q at the top,
A similar auction could occur after Golady:
1D-P-2C-P-
2D(hearts)-P-2H(hearts agreed)
Opener and/or Responder could "close out" the diamond suit.
If that occurs, then a 5-level call in a closed out suit cannot logically be Exclusion RKCB, as that makes no sense. Furthermore, it cannot be a "shifting" RKCB (show the king and queen of this suit as the secondary key cards) for the reason of redundancy -- the answer to this question is already known. (Same principle of why a call cannot be exclusion if partner has already denied a control in that suit anyway). 6KCB makes just as little sense.
This auction, therefore, involving a "close out," offers an opportunity for a surrogate asking bid at above game in that strain.
Suppose, further, that the close-out occurs below game sufficiently low to enable more than one cue between the close-out call and game. (A simple example is closing out clubs with a 4C call when spades is the agreed strain.) My position is that any cue after a side suit is closed out should be a straight control cue (honor or shortness allowed) if control is not yet established and even if the normal cue rules for that side suit would (a) allow a cue of a queen or (b) discourage or not allow a shortness control cue. This would further not require any "two honors" rule.
An example. One partner of mine likes a 3NT cue in the specific auction:
1H-P-2C-P-
2S-P-3H(agrees hearts)-P-
3S(two top spades)-P-3NT
...to be a responsive spade cue (I have the missing spade honor). Or, if you will, the "definition" of a "serious 3NT" in this specific auction is "I have the missing spade honor."
Fine.
But, once a 3NT call is made, closing out spades, it seems that Opener should be able to cue clubs with the Ace or King, but not the Queen, and that a cue of 4C could also be the call with a stiff or void in clubs. It just shows a club control. Similarly, a 4D cue by Opener should deny a diamond control but simply show a diamond control (Ace, King, stiff, or void).
Thus, closing out a side suit should convert other calls to straight control if control in that suit has not yet been established.
Consider an example:
1S-P-2C-P-
2D-P-2S(spades agreed)
If Opener is able to cuebid 3C (two of the top three clubs) and Responder is then able to cuebid 4C (the missing club honor), or backwards (Responder cuebids 3C to show one and then Opener cuebids 4C to show the remaining two), then the suit is "closed out," meaning that this side suit is proven to be A-K-Q at the top,
A similar auction could occur after Golady:
1D-P-2C-P-
2D(hearts)-P-2H(hearts agreed)
Opener and/or Responder could "close out" the diamond suit.
If that occurs, then a 5-level call in a closed out suit cannot logically be Exclusion RKCB, as that makes no sense. Furthermore, it cannot be a "shifting" RKCB (show the king and queen of this suit as the secondary key cards) for the reason of redundancy -- the answer to this question is already known. (Same principle of why a call cannot be exclusion if partner has already denied a control in that suit anyway). 6KCB makes just as little sense.
This auction, therefore, involving a "close out," offers an opportunity for a surrogate asking bid at above game in that strain.
Suppose, further, that the close-out occurs below game sufficiently low to enable more than one cue between the close-out call and game. (A simple example is closing out clubs with a 4C call when spades is the agreed strain.) My position is that any cue after a side suit is closed out should be a straight control cue (honor or shortness allowed) if control is not yet established and even if the normal cue rules for that side suit would (a) allow a cue of a queen or (b) discourage or not allow a shortness control cue. This would further not require any "two honors" rule.
An example. One partner of mine likes a 3NT cue in the specific auction:
1H-P-2C-P-
2S-P-3H(agrees hearts)-P-
3S(two top spades)-P-3NT
...to be a responsive spade cue (I have the missing spade honor). Or, if you will, the "definition" of a "serious 3NT" in this specific auction is "I have the missing spade honor."
Fine.
But, once a 3NT call is made, closing out spades, it seems that Opener should be able to cue clubs with the Ace or King, but not the Queen, and that a cue of 4C could also be the call with a stiff or void in clubs. It just shows a club control. Similarly, a 4D cue by Opener should deny a diamond control but simply show a diamond control (Ace, King, stiff, or void).
Thus, closing out a side suit should convert other calls to straight control if control in that suit has not yet been established.
Nice surrogate RKCB situation
From a sectional this weekend. Playing with a friend, whose cue structure is very similar to mine, but with a few small tweaks.
The actual hand:
KQxxx-KQJxxxx-x-V
1H-P-2C-P-
2S-P-3H(sets trumps)-P-
3S(two top spades)-P-4D(courtesy, not club Ace or KQ)-P-
?
So far, easy. Bid Exclusion RKCB, right? No. Partner has already denied the club Ace. With this specific structure, 4C would show the Ace, or the KQ. But, bypassing the 4C cue means no club Ace. Hence, Exclusion is off. This illustrates why. You don't need it. The 4NT answer will tell all.
Next, change the hand slightly.
AQxxx-KQJxxxx-x-V
A better hand, but more difficult. Same auction, up through 4D.
At this point, this partner likes 4S in a heart sequence to be a general slam move requesting partner to start asking questions. In other words, a desire to answer rather than ask.
However, "surrogate RKCB" works.
This is a method where a call that cannot be Exclusion inferentially (because of a denial or promise earlier that resolves all questions and therefore makes Exclusion redundant) takes on the "most useful meaning" by default rules.
In this sequence, the "most useful meaning" is obvious -- RKCB, but show the SPADE king and queen rather than the heart king and queen. Opener is looking at the heart king and queen and therefore knows that answer. But, the sequence is such that Responder could never cue spades. The 4S relay to answer won't work, because Responder won't guess 5-7-1-0.
So, 5C becomes RKCB, spade focus as to answering, heart focus as to strain.
The actual hand:
KQxxx-KQJxxxx-x-V
1H-P-2C-P-
2S-P-3H(sets trumps)-P-
3S(two top spades)-P-4D(courtesy, not club Ace or KQ)-P-
?
So far, easy. Bid Exclusion RKCB, right? No. Partner has already denied the club Ace. With this specific structure, 4C would show the Ace, or the KQ. But, bypassing the 4C cue means no club Ace. Hence, Exclusion is off. This illustrates why. You don't need it. The 4NT answer will tell all.
Next, change the hand slightly.
AQxxx-KQJxxxx-x-V
A better hand, but more difficult. Same auction, up through 4D.
At this point, this partner likes 4S in a heart sequence to be a general slam move requesting partner to start asking questions. In other words, a desire to answer rather than ask.
However, "surrogate RKCB" works.
This is a method where a call that cannot be Exclusion inferentially (because of a denial or promise earlier that resolves all questions and therefore makes Exclusion redundant) takes on the "most useful meaning" by default rules.
In this sequence, the "most useful meaning" is obvious -- RKCB, but show the SPADE king and queen rather than the heart king and queen. Opener is looking at the heart king and queen and therefore knows that answer. But, the sequence is such that Responder could never cue spades. The 4S relay to answer won't work, because Responder won't guess 5-7-1-0.
So, 5C becomes RKCB, spade focus as to answering, heart focus as to strain.
Wednesday, December 16, 2009
ebook available
Masterpoint Press has launched www.ebooksbridge.com, offering many titled, including Cuebidding at Bridge, a Modern Approach, in an ebook format. They are also considering publishing some books through this format only, perhaps for lower market books or lesser known authors. If you are interested, send me an email.
Thursday, December 3, 2009
Spoecific situation
A friend suggested an interesting treatment. Of course, I had to expand it even more.
The problem. Blah-blah-blah occurs. The opponents have shown spades. We have hearts. Partner's last bid was 4H. We want to explore slam, but we have not together worked out whether spades are controlled. What now?
My friend's idea makes sense. 4NT is RKCB. 4S is asking for a spade stopper and is RKCB if you have it. Thus, partner's replies to 4S will be:
4NT = 0/3 with a stopper
5C = 1/4 with a stopper
5D = 2 without the Q but with a stopper
5H = no stopper
5S = two with the Q and a stopper
I think this can be deepened even more, in a Meckwellian manner, if so inclined. You could also add in an immediate-answers with or without. For instance, Responder's 5C call could be 0/3 with a spade control. Opener could bid 4NT after 4S to ask or answer at 5C+. Something like that.
Or, a uni-directional. If the mover is weak, 4S asks, but 4NT+ answers with. If the mover is strong, 4S asks and is RKCB, 4NT asks and shows, 5C+ ask other questions.
In the end, though, I think the simple suggestion is easy and good. If 4S invites slam without a spade control, Opener answers RKCB with a spade control, skipping 5H as the decline bid. 4NT, instead, asks RKCB but shows a spade control. Simple, and effective.
BTW -- this solves a Bridge Bulletin problem (as noticed by my friend, Ken Eichenbaum).
The problem. Blah-blah-blah occurs. The opponents have shown spades. We have hearts. Partner's last bid was 4H. We want to explore slam, but we have not together worked out whether spades are controlled. What now?
My friend's idea makes sense. 4NT is RKCB. 4S is asking for a spade stopper and is RKCB if you have it. Thus, partner's replies to 4S will be:
4NT = 0/3 with a stopper
5C = 1/4 with a stopper
5D = 2 without the Q but with a stopper
5H = no stopper
5S = two with the Q and a stopper
I think this can be deepened even more, in a Meckwellian manner, if so inclined. You could also add in an immediate-answers with or without. For instance, Responder's 5C call could be 0/3 with a spade control. Opener could bid 4NT after 4S to ask or answer at 5C+. Something like that.
Or, a uni-directional. If the mover is weak, 4S asks, but 4NT+ answers with. If the mover is strong, 4S asks and is RKCB, 4NT asks and shows, 5C+ ask other questions.
In the end, though, I think the simple suggestion is easy and good. If 4S invites slam without a spade control, Opener answers RKCB with a spade control, skipping 5H as the decline bid. 4NT, instead, asks RKCB but shows a spade control. Simple, and effective.
BTW -- this solves a Bridge Bulletin problem (as noticed by my friend, Ken Eichenbaum).
Subscribe to:
Posts (Atom)
Posts
