[blackshoe]

I had originally intended to post this in "laws and rulings", because I was thinking about how RAs handle this now, but it's really a question about how they should handle it, so I've put it here in "changing laws and regulations", even though I'm not sure anything needs to be changed.

In the thread about alerting partner's T/O double showing a balanced opening hand in Simple Rulings, the point was made that

Vampyr, on 2012-August-31, 18:34, said:

In the EBU, the systemic possibility of a singleton is announced. I think this is a good idea.

This brought an interesting (to me, anyway) question to mind: how do we handle "systemic possibilities" that arise on the fly? If a player opens 1NT with a singleton, having not seen the situation before, not thought about it, and not discussed it with his partner, is it a "systemic possibility" at all? If not, can partner be expected to disclose it? What if partner is much more experienced than the bidder, so that he might be aware the situation might come up, even though his partner wasn't? We accept, I think, that a rare deviation from system need not be alerted or included in explanations, but how rare is "rare"? Clearly somebody has decided that a 1NT opening with a singleton is not rare enough, or at least it's "not rare enough" that it will come up often enough to trigger the "implicit agreement" bit in the law. Can we put an objective frequency on this? Should we? If probability says the situation is likely to come up once a year, and actually it comes up thrice in a week, is the "implicit agreement" triggered? If it's come up once in the three years the pair have been playing together, presumably that would not trigger an implicit agreement, but what if it comes up again in a week, or two? I daresay some would argue, particularly if they're the other side at the table, that it does trigger, but I'm not so sure it should. How should this be handled?

[paulg]

Hidden away on the ACBL site is the suggestion that 1% of 1NT openings containing a singleton should be the threshold.

http://www.acbl.org/...aSingleton.html

[blackshoe]

Hm. And given an agreement that 1NT shows 15 to 17 HCP and a balanced hand, how often does a 4441 hand come up that otherwise meets the criteria? If it's less than a 1% probability, then the reference to 1% is redundant. If it's more, then there seems to be a problem.

[MickyB]

blackshoe, on 2012-September-01, 07:37, said:

Hm. And given an agreement that 1NT shows 15 to 17 HCP and a balanced hand, how often does a 4441 hand come up that otherwise meets the criteria?

5.9%, I believe. It'll be less for most pairs in practice because that doesn't include any 5422s/6322s, but still well clear of 1%.

[blackshoe]

Not sure I understand your reference to semi-balanced hands (two doubletons) since we are discussing unbalanced (one singleton) hands.

For myself, I would only open 1NT if the singleton were the ace or king. I imagine that narrows the odds a bit.

If the frequency of appearance of these hands is significantly greater than 1%, then how does a player keep track? What does he do? "I play 100 hands of bridge a week, and it came up on Monday and I opened 1NT, so I cannot open 1NT on these hands for the rest of the week"? That seems not only unlikely to be considered by most players, but also a little silly.

[campboy]

There are 30897212184 balanced hands in range (15-17), and 1931443704 (4441) hands in range, which is where MickyB's 5.9% comes from: it's the proportion of 1NT openers which have a singleton if you open all balanced or (4441) hands in range.

If we restrict our singletons to be ace or king (and believe that I didn't make a mistake when changing the code) there are only 588783384 suitable hands with singletons, so the chance of a 1NT opener having a singleton drops to
588783384/(30897212184+588783384) which is about 1.87%.

[paulg]

blackshoe, on 2012-September-01, 07:52, said:

For myself, I would only open 1NT if the singleton were the ace or king. I imagine that narrows the odds a bit.

If the frequency of appearance of these hands is significantly greater than 1%, then how does a player keep track? What does he do? "I play 100 hands of bridge a week, and it came up on Monday and I opened 1NT, so I cannot open 1NT on these hands for the rest of the week"? That seems not only unlikely to be considered by most players, but also a little silly.

I have no idea what the ACBL meant by 1%, but I presumed that it meant 1% of 1NT openers rather than 1% of all hands.

[Vampyr]

blackshoe, on 2012-August-31, 21:13, said:

What if partner is much more experienced than the bidder, so that he might be aware the situation might come up,

What "situation" is it that might come up? Is it, say, 4=4=4=1 inside the NT range? If the partnership has decided to handle these hands by opening 1NT, then it should be disclosed. Similarly if before discussing it they notice that they handle this holding by always opening 1NT; now they should make it explicit.

Quote

If probability says the situation is likely to come up once a year, and actually it comes up thrice in a week, is the "implicit agreement" triggered? If it's come up once in the three years the pair have been playing together, presumably that would not trigger an implicit agreement, but what if it comes up again in a week, or two? I daresay some would argue, particularly if they're the other side at the table, that it does trigger, but I'm not so sure it should. How should this be handled?

Why is it suddenly coming up so often? Why is the player preferring 1NT to whatever the "systemic" bid is? Is there in fact no systemic bid for the hand? Maybe the player's preferences have changed sometime in the past three years; or... I don't know. I don't think it's good enough to hide behind "anti-systemic" if it happens... how often? Maybe a good threshold is, when it happens, partner is likely to say "I can't remember the last time she did that"

[blackshoe]

campboy, on 2012-September-01, 08:08, said:

There are 30897212184 balanced hands in range (15-17), and 1931443704 (4441) hands in range, which is where MickyB's 5.9% comes from: it's the proportion of 1NT openers which have a singleton if you open all balanced or (4441) hands in range.

If we restrict our singletons to be ace or king (and believe that I didn't make a mistake when changing the code) there are only 588783384 suitable hands with singletons, so the chance of a 1NT opener having a singleton drops to
588783384/(30897212184+588783384) which is about 1.87%.

paulg, on 2012-September-01, 08:47, said:

I have no idea what the ACBL meant by 1%, but I presumed that it meant 1% of 1NT openers rather than 1% of all hands.

I didn't think we were talking about all hands. If I understand campboy, the frequency with which suitable hands (15-17 HCP, 4441 distribution, singleton A or K) come up is 1.87% of hands which would open 1NT, not 1.87% of all hands.

[blackshoe]

Vampyr, on 2012-September-01, 09:20, said:

What "situation" is it that might come up? Is it, say, 4=4=4=1 inside the NT range? If the partnership has decided to handle these hands by opening 1NT, then it should be disclosed. Similarly if before discussing it they notice that they handle this holding by always opening 1NT; now they should make it explicit.

If a partnership is experienced enough to make these decisions, and if they've thought of the problem at all, sure. But many, many partnerships act on what they were taught when they learned the game (a 1NT opener shows a balanced — no singleton, no void, no more than one doubleton — hand of a defined point range) and have never given any thought to when it might be good to deviate from that. Granted many of them won't deviate from it ("I can't open 1NT, I have a singleton!") but some will sooner or later consider the possibility.

One of the problems here comes when the opponents complain — "he can't open 1NT with a singleton!" — because the TD is rarely going to have any clue how often this pair has done this before. We could say "one instance creates an implicit agreement", or we could tell the pair concerned that "you've had your one free shot, now you must either make this an explicit agreement, and disclose it properly, or never do it again". I'm certain the first of those is a bad way to handle it, and I'm not sure the second isn't also bad.

Vampyr, on 2012-September-01, 09:20, said:

Why is it suddenly coming up so often? Why is the player preferring 1NT to whatever the "systemic" bid is? Is there in fact no systemic bid for the hand? Maybe the player's preferences have changed sometime in the past three years; or... I don't know. I don't think it's good enough to hide behind "anti-systemic" if it happens... how often? Maybe a good threshold is, when it happens, partner is likely to say "I can't remember the last time she did that"

Let's noT go down the "hide behind" road. I'm not talking about trying to avoid disclosure obligations, or skirting the edges of cheating, I'm talking about figuring out what those obligations are, and avoiding any accusations of cheating by not cheating.

As I understand Standard American and 2/1, the systemic bid with 4441 is to open 1♣ with a singleton diamond, and 1♦ otherwise. The problem is "how do I convince partner I've got 16 points?" after I've done that. AFAIK there's no systemic way to do that*, and that's way people open 1NT with a stiff.

Another question: in jurisdictions where you're supposed to announce the NT range or where you say nothing about it, does an explicit or implicit agreement to open 1NT with a singleton make the bid alertable? I think the ACBL's position is that so long as you only do it 1% of the time (which effectively means you can only do it every other time the hand comes up) you don't alert. So by the rules, announcing the range is sufficient disclosure — but it doesn't feel like it. Of course you'll mention it if they ask, but they almost never do. And how does "he did it just last month" affect the alerting and disclosure requirements for the 1minor openings? Sometimes — when partner hasn't opened 1NT with a stiff in a while — you can be pretty sure (or can you?) that if he opens 1minor, he won't have a 4441 in the 1NT range, because he would have opened that 1NT. And once he's opened his one in a hundred times 1NT, you know that the 1minor opening might include such a hand. Keeping all this straight so you can 1) have an idea what the heck partner is doing and 2) properly disclose everything to your opponents seems like an awful lot of work.

*In some cases, I suppose you might reverse, but then you're distorting your shape.

[barmar]

It seems to me that this is "judge bridge". It's not generally a partnership agreement, it's just a player's stylistic judgement. I think it's pretty rare that partnerships have detailed agreements about it, except that some partnerships may have a rule prohibiting it at all ("If you do that one more time, we're never playing together again!"). I would judge that it's a "systemic possibility" for most experienced players unless they've made this kind of explicit prohibition.

As far as frequency goes, I've opened many NT hands like this, but I know I don't do it with all 15-17 4441's where the singleton is A or K. It probably depends on how the honors and high spots are distributed among the 4-card suits -- with aces and spaces I'd probably want the lead to go up to partner, with queens, kings, and tenaces I'd prefer to declare. This type of judgement probably gets it down to ACBL's 1% threshold.

However, with 20+ HCP and this type of hand, I almost always open 2NT (or open 2♣ planning on rebidding 2NT) with hands like this, since they're otherwise really difficult to describe.

Another stylistic issue like this is what gets opened with a minimum hand with 4♦ 5♣. Many players consider it "judge bridge" that you open 1♦ planning on rebidding 2♣, others would never think of doing it, and occasionally they may be partners. When I've started playing with a new partner, this rarely comes up in system discussion unless we have lots of time to plan.

[Vampyr]

barmar, on 2012-September-01, 13:06, said:

It seems to me that this is "judge bridge". It's not generally a partnership agreement, it's just a player's stylistic judgement. I think it's pretty rare that partnerships have detailed agreements about it,

I think that in England people give this issue a lot more thought, because it's part of the announcement regulations and is also listed on the convention card. As a result, a fair few people do announce "could have a singleton" or "could have a singleton <specific suit>".

[mycroft]

In the ACBL, it's also not a partnership agreement because the regulations bar you from having one. So partnership agreements are "they're balanced, and if you have a singleton A or K in a 4=4-(14) or (43)(1-weak 5) hand, I'm going to go out of my way not to notice what you did with it".

If you start opening with a small singleton, or if it could be in a major (especially if partner staymans with a bad (53)(xx) hand instead of transferring to the major), or it starts being wildish distributions, or if they can show that systemically, you *have to* open 1NT on that hand, they'll lower the boom. Otherwise, it's about right.

Basically, like a lot of convention regulation in the ACBL, "it works", and "if it ain't broke, don't fix it". The mathematicians and other people who expect precision in their regulations are driven nuts, and the other 99% of bridge players serenely go on playing bridge.

(note, I is one of the former category. I'm not denigrating them, just telling it like I see it happening. Also, I never speak online for any bridge organization)

[bluejak]

Vampyr, on 2012-September-01, 18:00, said:

I think that in England people give this issue a lot more thought, because it's part of the announcement regulations and is also listed on the convention card. As a result, a fair few people do announce "could have a singleton" or "could have a singleton <specific suit>".

My partner and I open [and therefore announce] hands with a singleton honour in a minor.

The WBF CTD used to play Precision and in his methods there were specific problems with 1-3 in the majors and 5-4 in the minors. He felt that these hands [and only these hands] should be opened with 1NT. Ok, pedants, have your fun.



To my mind frequency is the wrong criterion: it is expectedness. If you hold a 1=3=3=6 hand, partner opens 1NT, and you transfer into clubs because that is known to be an 8-card fit, then the fact that one day partner has a singleton club is irrelevant. If you wonder whether to transfer into clubs because partner might have a singleton then you have a disclosable agreement to allow singletons in your 1NT opening.

[paulg]

bluejak, on 2012-September-06, 14:08, said:

To my mind frequency is the wrong criterion: it is expectedness. If you hold a 1=3=3=6 hand, partner opens 1NT, and you transfer into clubs because that is known to be an 8-card fit, then the fact that one day partner has a singleton club is irrelevant. If you wonder whether to transfer into clubs because partner might have a singleton then you have a disclosable agreement to allow singletons in your 1NT opening.

Interestingly, at Brighton last year, the Director's first question was, "how often has he opened with a singleton in the last twelve months?".

But I think frequency and expectedness are the same, in essence.

[barmar]

bluejak, on 2012-September-06, 14:08, said:

To my mind frequency is the wrong criterion: it is expectedness.

They're closely related.

If something happens every few sessions, you would have to be an idiot not to expect it, and will presumably make allowances for it. The opponents are entitled to the same expectation.

If it only happens once a year, that's close enough to impossible that you shouldn't expect it on any particular occasion, and you probably don't need methods to find out about it.

[Zelandakh]

I think David is right here and here's a simple example to show it. Suppose I open 4NT Specific Ace Ask. I cannot remember the last time I did this but it is more than a year ago. But if I have the agreement with partner that this is the meaning then it is expected and should be disclosed. The same is true for any other rare tendencies partner might have - if I know partner's style well enough to expect them then they are surely disclosable.

An interesting point on this subject of 4441 hands and 1NT. My partner and I have the agreement that we can open 1NT on any 44(41) hand within range, thus making our 1NT opening alertable. The difficulty is that while my partner opens close to 100% of such hands with 1NT, I almost never do this, instead preferring to open the 4 card minor. In other words, we have adapted our system to cater to both styles. Indeed one might argue that we are effectively playing different systems for this class of hand. Is everyone ok with this?

[pran]

Zelandakh, on 2012-September-10, 02:02, said:

I think David is right here and here's a simple example to show it. Suppose I open 4NT Specific Ace Ask. I cannot remember the last time I did this but it is more than a year ago. But if I have the agreement with partner that this is the meaning then it is expected and should be disclosed. The same is true for any other rare tendencies partner might have - if I know partner's style well enough to expect them then they are surely disclosable.

An interesting point on this subject of 4441 hands and 1NT. My partner and I have the agreement that we can open 1NT on any 44(41) hand within range, thus making our 1NT opening alertable. The difficulty is that while my partner opens close to 100% of such hands with 1NT, I almost never do this, instead preferring to open the 4 card minor. In other words, we have adapted our system to cater to both styles. Indeed one might argue that we are effectively playing different systems for this class of hand. Is everyone ok with this?

I don't think so.
As far as I know most jurisdictions require that both players in a pair play the same system.

The two players may of course have different attitudes (like one player stretching a 15-17 NT towards 14 "good" points more often than the other), but IMHO your example is streching this variation in attitudes too far.

[paulg]

Zelandakh, on 2012-September-10, 02:02, said:

An interesting point on this subject of 4441 hands and 1NT. My partner and I have the agreement that we can open 1NT on any 44(41) hand within range, thus making our 1NT opening alertable. The difficulty is that while my partner opens close to 100% of such hands with 1NT, I almost never do this ...

So your partner alerts the 1NT, saying that 44(41) is possible by agreement when, in practice, it is not. Doesn't sound so good put like this, does it?

As pran said, style differences are generally permitted and should be disclosed, but I infer that you have reached the point of having different agreements.

[Zelandakh]

paulg, on 2012-September-10, 03:37, said:

So your partner alerts the 1NT, saying that 44(41) is possible by agreement when, in practice, it is not. Doesn't sound so good put like this, does it?

As pran said, style differences are generally permitted and should be disclosed, but I infer that you have reached the point of having different agreements.

Well yes and no. The methods over 1NT are essentially SAYC and pay no attention to the possivble 44(41) hand whatsoever. On the other hand the methods after 1♦ specifically cater to the 4441 hand by including a 2♦ rebid after 1♦ - 2♣ as a 1 round force which could have only 4 diamonds. The actual theoretical meaings of all bids are identical, it is only that there are 2 possible ways of bidding these hands. There are inferences from the radical difference in style which the opps are entitled to. What I was getting at is where a difference in style crosses over from that to two partners actually playing different systems. Note that the example could just as easily be the choice of opening the major or the minor in a 4432 hand in Acol, or of opening 1M or 1NT with a 5332 hand within range. If 2 options are available and one partner always chooses one while the other always chooses the other, then what is the criteria for this becoming 2 different systems?