[ASkolnick]

We are playing your normal, Jacoby 2N response opposite a 1 of a major precision opener and we are discussing the merits of bidding a 4 card suit versus a singleton at the 3 level. I have come up with a compromise which lets you do both, but what would be the pivot point of a 5-4-3-1, 6-4-2-1 hand?

How good should your 4 card suit be? I know what I have proposed, but I don't know if it is right. I can do simulations, but I am not sure what I would be testing for anyway.

[helene_t]

I have given some thoughts to this.

Consider the model

logit(p(WeCanMakeSlam)) = a + b*AdjustedLTC + c*NumberOfTrumps + AdditionalTerms

Depending on which AdditionalTerms are included, the model will become more or less accurate. One example of an additional term could be d*NumberOfShortnesses+e*NumberOfQueenPointsInPictureCardsOppositeShortnesses. Another example could be d*NumberOfBrokenLongSideSuits+e*NumberOfFillersOppositeBrokenSideSuits. Or something like that. The terms will have to be phrased such that they reflect the information that you could actually uncovcer using some kind of follow-up scheme.

Now generate some 100,000 DD solved hands and see which of the competing models makes the most accurate predictions.

[karlson]

I think it matters not only matters how likely slam is, but how well responder will evaluate.

In my opinion, it's generally pretty easy to reevaluate your hand when partner shows a singleton. It can be much harder when partner shows a side suit.

As such, I would suggest a very strict definition on what the 4-card suit is. Sure, it won't come up as often, but at least you'll know what to do when it does. Otherwise, you just show the shortness and there's no reason to suspect you'll have a bad auction.

[cherdano]

ASkolnick, on Nov 6 2008, 01:47 PM, said:

We are playing your normal, Jacoby 2N response opposite a 1 of a major precision opener and we are discussing the merits of bidding a 4 card suit versus a singleton at the 3 level. I have come up with a compromise which lets you do both, but what would be the pivot point of a 5-4-3-1, 6-4-2-1 hand?

How good should your 4 card suit be? I know what I have proposed, but I don't know if it is right. I can do simulations, but I am not sure what I would be testing for anyway.

I think its more important to show your shortness than to show your 4-card side suit. (It may be worthwhile to be able to do both.)
Fans of gaining IMPs by finding a 4-4 minor grand slam when 7M in the 5-3 fit is down will disagree.

[xcurt]

It's not enough to know what the key features are. You need to think about the optimal coding scheme given the bandwidth (number of steps) available. You might even need to think about the extend to which exchanging information benefits your side (higher chance of reaching the best contract or a reasonable alternative) versus the other side (letting them defend more accurately or double an artificial call to set up a defense or find a save). In other words, this is a very complex optimization problem.

This is very similar to an issue discussed in another thread, what to do holding 6M4m minimum after 1M-1NT (no opposition bidding). There, when designing your system, you need to weigh the extent to which showing more of your hand (2m -> 9 cards shown) benefits you versus the extend to which showing the high-value features of your hand (2M -> 6 cards shown, but the value of a sixth card in a suit is much higher than the value of a fourth card in a suit, ceteris paribus).

[ASkolnick]

Karlson: I agree you shouldn't show a random 4 card suit, which is why I asked how good the suit should be.

Cherdano: You can get the best of both worlds since I can show both

Xcurt: I agree that the complexity is there since you have more steps, there is more of a chance of accurate defense. Also, which thread are you talking about? I know the one recently talking about 6-4 hands, but there was no minimum suit quality.

My suggestion is to have a minimum suit quality of KJ9x for the suit. Would this be considered too good? too bad? or just right?

Didn't want to give what I was doing until we heard more explanation.

[FrancesHinden]

This may be considered to be changing the subject, but as an alternative, how about having opener relay out his entire shape, rather than having a choice of whether to show a singleton or a side 4-card suit?

And to address the original question: I think it depends on the hand types that bid Jacoby. How likely is it that the partnership may not want to play in its major suit fit? To take an extreme example, in one partnership we make our forcing raise on all hands with support and the requisite strength, including very unbalanced hands, and as a consequence opener has the chance to relay out complete shape and honour location as we might want to alight in a fit elsewhere. But if responder is (e.g.) known to be balanced, opener will only want to show a 4-card suit if it has, say, 2 of the top 3 honours, because it's only a potential 4-4 fit at the slame level.

[jdonn]

FrancesHinden, on Nov 7 2008, 11:39 AM, said:

This may be considered to be changing the subject, but as an alternative, how about having opener relay out his entire shape, rather than having a choice of whether to show a singleton or a side 4-card suit?

I don't believe there is room to do so (at least not unless you totally give up describing anything about your strength)? I suppose you could do so with the more common shapes and just say something about your hand with wilder distribution.

[ASkolnick]

I think we will have most of the shapes covered since I didn't give the entire portion of the system. Part of the problem is we play so infrequently, we don't want to add too much complexity where the memory taxing will be much worse than the actual gain.

For the most part, 2N tends to be on a balanced hand (Yes, there are some exceptions but tends to be). The debate came up whether it was more important to show shortness or to show a 4-card suit? I was just trying to come up with the best way to do both without it being too taxing.

For the most part, it is pretty easy:
3C - I have a singleton
3D - I have a good 4-card suit (The discussion is what does good mean)
3H - I have a max (Don't need two steps since we are playing precision)
3S - I have a min
3N - I have a 6+ card suit no outside controls
4X 5-5 in appropriate suits.

It may be right to rearrange the 3H,3S,3N bids so you won't get doubled for leads with the appropriate hands.

[FrancesHinden]

jdonn, on Nov 7 2008, 05:59 PM, said:

FrancesHinden, on Nov 7 2008, 11:39 AM, said:

This may be considered to be changing the subject, but as an alternative, how about having opener relay out his entire shape, rather than having a choice of whether to show a singleton or a side 4-card suit?

I don't believe there is room to do so (at least not unless you totally give up describing anything about your strength)? I suppose you could do so with the more common shapes and just say something about your hand with wilder distribution.

This was suggested in the context of a Precision 1M opening, where strength is much more tightly defined

[jdonn]

FrancesHinden, on Nov 7 2008, 12:49 PM, said:

jdonn, on Nov 7 2008, 05:59 PM, said:

FrancesHinden, on Nov 7 2008, 11:39 AM, said:

This may be considered to be changing the subject, but as an alternative, how about having opener relay out his entire shape, rather than having a choice of whether to show a singleton or a side 4-card suit?

I don't believe there is room to do so (at least not unless you totally give up describing anything about your strength)? I suppose you could do so with the more common shapes and just say something about your hand with wilder distribution.

This was suggested in the context of a Precision 1M opening, where strength is much more tightly defined

It's easy to calculate how many shapes can be shown below 4M using Fibonacci.

With hearts which has less space than with spades, 1+1+2+3+5+8+13+21 = 54

There are this many shapes without including 8 in one suit or 12 in two suits.
5332: 3
5431: 6
6322: 3
6331: 3
6421: 6
6430: 6
5521: 4 (not counting 5-5 majors of course)
5530: 4
6511: 3
6520: 6
7222: 1
7321: 6
7330: 3
7420: 6
Total: 60

So if strength and slam suitability are totally ignorred (at least for the highest shapes that you show) it can't quite be done, although it could be by not disclosing full shape when very shapely. However even if you could do it, I do not believe it is very wise at all to ignore strength and slam suitability, even in the context of a limited opening bid system.

If the forcing raise is 2♠ then it's a whole different kettle of fish. I think that extra step actually helps quite a lot.

[han]

Agree with jdonn that it would be a poor idea to ignore slam suitability. I think it would be better to give up on describing exact shape and use some part of the available space to describe shape, some part to describe slam suitability and some space to describe location of honors. Of course these three phases don't have to be completely separate or in that order.

[skjaeran]

cherdano, on Nov 7 2008, 05:58 AM, said:

I think its more important to show your shortness than to show your 4-card side suit. (It may be worthwhile to be able to do both.)
Fans of gaining IMPs by finding a 4-4 minor grand slam when 7M in the 5-3 fit is down will disagree.

My experience is that it's good to show a side suit first at the 3-level with extras, to find a possible source of tricks, and then ask about singleton if appropriate. The usefulnes of showing a singleton first has always eluded me, probably partly because I'm unfamiliar with that method.

Recently I've used a more complex method that distinguish strength and shape more clearly.