inv+ jacoby?

[wank]

the variant of Gf jacoby which seems to be FOTM as far as i know is:-

 

3C = any min >>> 3D asks for opener's shortage if he's got one.....3H/3S/3NT show responder's shortage

3D = above min, no shortage

3H = club shortage

3S = diamond shortage

3NT = major shortage

4m = 5-5

4H = 5-5 majors 5-6 if it was a heart opener

 

if you're palying it as inv+ you could obviously just edit this and play 3C - 3M as to play with responder's higher shortages having to move up 1 space, but that takes you to 4C to show a major singleton which is getting awkward.

 

is there a better way to do it specifically for inv+ jacoby?

[inquiry]

the variant of Gf jacoby which seems to be FOTM as far as i know is:-

 

3C = any min >>> 3D asks for opener's shortage if he's got one.....3H/3S/3NT show responder's shortage

3D = above min, no shortage

3H = club shortage

3S = diamond shortage

3NT = major shortage

4m = 5-5

4H = 5-5 majors 5-6 if it was a heart opener

 

if you're palying it as inv+ you could obviously just edit this and play 3C - 3M as to play with responder's higher shortages having to move up 1 space, but that takes you to 4C to show a major singleton which is getting awkward.

 

is there a better way to do it specifically for inv+ jacoby?

 

 

If memory serves me well, I got this from Glenn Ashton's Bridgematters.com website. Go there and look for Jacoby 2NT Plus or something like that.

 

 

Over 2NT (limit raise or better Jacoby, opener rebids:

 

--------> 3C, no game interest or monster slam interest (can have shortness)

--------> 3D, Game interest, no slam interest, but better than 4M

---------------> responder now bids 3H to ask opener to show any shortness.

--------> 3H, (regardless of what major was opened) ask for 2NT bidder to show shortness if you have (yes, opener can ask responder for shortness)

-------> 3S short in "other major" (regardless of what suit was opened)

-------> 4C/4C short in bid suit (if not bid before naturally)

-------> 4M Just enough for game, no more (responder of course can bid again. Generally denies shortness

 

This has some advantage of hiding Opener's distribution if Game is the limit from EITHER side of the table (the concept of camouflage or hiding our hand pattern/stregth from opponents and only sharing when necessary to improve OUR bidding). . After the 3♣ bid if responder jump to 4M, opener with strong slam interest will, of course, be bidding slam or moving that way.

[gnasher]

if you're palying it as inv+ you could obviously just edit this and play 3C - 3M as to play with responder's higher shortages having to move up 1 space, but that takes you to 4C to show a major singleton which is getting awkward.

 

After 1♠-2NT;3♣, when responder has a major-suit shortage he can bid any of 4♣, 4♦ and 4♥. That ought to allow you to define responder's hand more closely, eg by playing 4♣ and 4♦ as cue-bids with a singleton heart, and 4♥ as a void. There's one less bid available after a 1♥ opening.

[JLOGIC]

Why does everyones structure include showing shape immediately? This is especially bad when partner is just inv+, you will have a slam even less frequently, and the information leakage will lead to them beating your game even more frequently. You can still show a lot of information when there is slam interest while never/almost never giving away information for free to begin with.

[JLOGIC]

Also, having only "min" and "non min" seems poor opposite a limit+. Does partner have to cooperate when you show non min and he has just a suitable limit raise since you might have slam? You need quite a lot extra for slam to make, and most of the time you will just have a little bit extra, so again you give away lots of info as well as having wide ranges in all auctions. Non serious/serious can help this but it still will hurt your accuracy on slam bidding. Having 3 or even 4 ranges seems much better.

 

Something very simple like:

 

step 1 = extras but not the nuts

 

step 2 = going to game opposite a LR but not really much extra (5431 12 count, 5422 13, etc)

 

3 of your major = rejecting a limit raise (you will almost never have slam over this, but if you do you still have room to cuebid. Obv you almost never have shortness).

 

Others = the nuts, can make slam opposite a suitable limit raise.

 

So now you have 4 ranges, partner can jump to game a fair amount over step 1 (like with all limit raises and min GFs), and you still have a ton of room to sort out shortness/5422s/cuebid when you both have extras.

 

You even have a significant amount of room over the common but not often slammish step 2 type of hand. Over all of those responder can show his shortness if he wants also rather than ask.

 

And most importantly, you rarely give away unnecessary information! That should be a goal when designing your system that has a higher priority than being able to show specific 6322s and stuff lol.

 

BTW this is similar to what meckwell plays though theirs is different because they play strong club. IMO this is why rodwell is the best theorist, because he is a practical one, and recognizes one goal in designing your structure is to not give away info so that meck can take a ton of tricks after the dummy comes down!

[gnasher]

Why does everyones structure include showing shape immediately? This is especially bad when partner is just inv+, you will have a slam even less frequently, and the information leakage will lead to them beating your game even more frequently. You can still show a lot of information when there is slam interest while never/almost never giving away information for free to begin with.

They don't. In the structure that Wank posted, opener shows his shape only with a non-minimum. I'd define that as something like 15+ in high cards. It would be pretty rare to go down in game with that much and a singleton, opposite a limit raise.

 

With a hand that's weaker than that but still accepting the invitation, you bid 3♣ and then raise partner's signoff to game.

[TWO4BRIDGE]

They don't. In the structure that Wank posted, opener shows his shape only with a non-minimum. I'd define that as something like 15+ in high cards. It would be pretty rare to go down in game with that much and a singleton, opposite a limit raise.

 

With a hand that's weaker than that but still accepting the invitation, you bid 3♣ and then raise partner's signoff to game.

Correctamundo.

 

I think it was the Swedes who came up with the idea of 2NT! = limit+ w/4.

And folks, including myself... and Bill Higgins, have been improving/modifying it -- for better or worse -- ever since.

 

For example, Responder can still use Bergen Raises and a different splinter structure:

1M - 3C! = lower Bergen

 

but now 3D is not needed for Upper Bergen, so:

1M - ??

3D! =

♣

-splinter

3M! = normal preemptive Bergen raise

3oM! =

♦

splinter

3NT! = oM splinter

 

This frees-up 4-level splinters to show VOIDs.

[JLOGIC]

They don't. In the structure that Wank posted, opener shows his shape only with a non-minimum. I'd define that as something like 15+ in high cards. It would be pretty rare to go down in game with that much and a singleton, opposite a limit raise.

 

With a hand that's weaker than that but still accepting the invitation, you bid 3♣ and then raise partner's signoff to game.

 

Then when you show a minimum partner has to ask for your shape/try for slam more often if you can have a 5332 11 count or a 5431 14 count. He then might have to further cuebid/slam try if his hand is good opposite a 5431 14 count but not a 5431 11 count. You cannot really have it both ways. Yes it would suck to go down after all of that but I wouldn't be surprised when they find the killing defense the small # of times they can.

 

Do you play different structures at matchpoints than imps? Giving away information that leads to overtricks sucks.

[Phil]

I haven't played 2N as inv + for awhile, and I only played this in a strong club setting since opener had a smaller spectrum of hands.

 

If I were to go back, I don't like the idea of having three classes of "non-nuts" hands since it seems like a waste of useful low-level space. It seems Jlall's #2 and #3 could be compressed into one call without any loss, since responder can bid a NF 3♠ over a minimum 3♣ and opener can either pass or bid 4♠.

 

There's a fine line between looking for a reasonable (and not perfecto) slam with opener showing extras with a 14 count and responder cooperating with a 12 versus the quest for overtricks.

[Phil]

He then might have to further cuebid/slam try if his hand is good opposite a 5431 14 count but not a 5431 11 count.

 

Seems a Non-Serious 3N would work nicely here.

[hrothgar]

Why does everyones structure include showing shape immediately? This is especially bad when partner is just inv+, you will have a slam even less frequently, and the information leakage will lead to them beating your game even more frequently. You can still show a lot of information when there is slam interest while never/almost never giving away information for free to begin with.

 

Here's my guess

 

1. Response structures for Inv+ Jacoby are optimized towards game exploration rather than slam exploration

2. Most players feel that the gains from being able to make an accurate decision regarding game outweigh the losses from information leakage

 

Consider the following example:

 

The auction starts

 

1M - (P) - 2M - (P)

???

 

Many players use fairly elaborate structures that allow opener to do all of the following

 

1. Show length in a suit

2. Show shortness in a suit

3. Make a general purpose range ask

 

When folks bump one of these options off the stack, the general range ask is often the first to go...

Information about opener's shape is often the most useful piece of information to make an informed decision.

 

I don't find this example too far removed from the invite+ raise with 4+ card trump support.

[gnasher]

Then when you show a minimum partner has to ask for your shape/try for slam more often if you can have a 5332 11 count or a 5431 14 count. He then might have to further cuebid/slam try if his hand is good opposite a 5431 14 count but not a 5431 11 count. You cannot really have it both ways. Yes it would suck to go down after all of that but I wouldn't be surprised when they find the killing defense the small # of times they can.

 

Your methods give away information too. When you bid 1M-2NT;3M-4M, they know you probably have 12-13 balanced; when you bid 1M-2NT;3D-4M, they know you're probably not balanced. With the methods that Wank posted, it will go 1M-2NT;3C-4M in both cases.

 

Do you play different structures at matchpoints than imps?

No, but this is an area where it might make sense to do so.

[gnasher]

He then might have to further cuebid/slam try if his hand is good opposite a 5431 14 count but not a 5431 11 count.

Seems a Non-Serious 3N would work nicely here.

Not in Wank's structure. It's gone

1M-2NT

3

♣

(any minimum)

Wank has defined responder's bids as:

3

♦

= asking for shortage

3M = non-forcing

3OM = club shortage

3NT = diamond shortage

4

♣

+ = shrotage in OM

You could play 3♦ as a serious ask and 3OM as a less serious ask, but then you push responder's shortage bids even higher.

[awm]

This may speak somewhat to the advantage of having 2NT = 3-card limit raise or 4+ game force raise.

 

One issue is that there are a lot of hands which are really quite minimum but want to accept a 4-card limit raise. So there is a wide range of hands where you are accepting game opposite the limit raise but have no real slam interest opposite the GF raise. Since a 3-card limit raise requires more to accept, a higher percentage of accepting hands will actually make slam (or at least make it worth exploring for slam) opposite the stronger option.

 

Adding the 3-card limit raise (instead of 4-card one) also makes it more dangerous for opponents to stick in a bid for the lead/sacrifice over the 2NT bid.

[bluecalm]

IMO this is why rodwell is the best theorist, because he is a practical one, and recognizes one goal in designing your structure is to not give away info so that meck can take a ton of tricks after the dummy comes down!

 

I wonder when he is going to realize puppet stayman after strong 2NT sucks. Meck will take even more tricks then :)

[whereagles]

Standard responses to J2NT give responder control and that is wrong. It is opener who should have control because 1. he has a wide ranging hand and 2. responder's hand is normally balanced.

 

If J2NT becomes inv+, that need is even more reinforced.

[bluecalm]

J2NT

 

Very strong hands with support shouldn't bid 2NT in the first place. There is a lot of place in 2/1. Start with 2♣ and get to know partner's hand exactly if you have SI to begin with.

Playing it as invite or invite and some weak GF hands (the ones which could only make slam opposite extremely good opener) makes much more sense.

This is what Italian and Polish pairs are doing for ages. I am surprised Americans haven't catch up on this one yet.

[inquiry]

Something very simple like:

 

step 1 = extras but not the nuts

 

step 2 = going to game opposite a LR but not really much extra (5431 12 count, 5422 13, etc)

 

3 of your major = rejecting a limit raise (you will almost never have slam over this, but if you do you still have room to cuebid. Obv you almost never have shortness).

 

Others = the nuts, can make slam opposite a suitable limit raise.

 

So now you have 4 ranges, partner can jump to game a fair amount over step 1 (like with all limit raises and min GFs), and you still have a ton of room to sort out shortness/5422s/cuebid when you both have extras.

 

 

 

This is basically how we play J2NT+ (as noted way above in this thread), but with the steps switched around a little. Hiding opener's hand pattern on the majority of the auctions is a key to the design. Reminder of the method I like to play.

 

3♣ = either super nuts, or no interest in game opposite limit raise

4M = just enough for game opposite limit raise nothing extra

3♦ = a bit better than just enough for game opposite limit raise, but not interested in slam opposite minimum GF raises

3♥ = (regardless of the opened suit) = GF, tell me YOUR short suits, big hand by opener

3♠/4m = short suit, BIG hand by opener, can envision slam opposite a limit raise if fit is working

 

The problem that i have had with this structure is that over the years, I have missed a perfect fit slam on 1M-2NT-4M a couple of times that maybe if I had shown my shortness we woujld have bid the slam. These were not bad boards, more along the lines of average minus to average for playing in four. The other auctions seem to work out ok.

[JLOGIC]

If I were to go back, I don't like the idea of having three classes of "non-nuts" hands since it seems like a waste of useful low-level space. It seems Jlall's #2 and #3 could be compressed into one call without any loss, since responder can bid a NF 3♠ over a minimum 3♣ and opener can either pass or bid 4♠.

 

"Without any loss" is going overboard, you are losing on hands where partner has some slam interest opposite #2 but not #1. I don't think these hands are that infrequent, because whenever partner might have some extras and a stiff in your weakest suit, you probably have some slam interest. I think many hands will investigate over #2 but not #1.

 

Similarly, a lot of your room will be used trying to figure out if opener does have some extras or not. Gnasher recommends "non min" being like 15+, that leaves a lot of 10-14 hands which is a big difference imo.

 

Seems a Non-Serious 3N would work nicely here.

 

Yes, non serious will work well, but only after partner has shown his shortness to the defenders first (where had it gone 1S-2N-3C-4S partner wouldn't have), and only if there is room (where if it had gone 1S-2N-3C-3D-3N, there is no longer room for non serious 3N). And there is still utility in having a smallish range and being able to later non serious or serious imo, there is a lot of difference in slam suitability within those hand types.

 

Your methods give away information too. When you bid 1M-2NT;3M-4M, they know you probably have 12-13 balanced; when you bid 1M-2NT;3D-4M, they know you're probably not balanced. With the methods that Wank posted, it will go 1M-2NT;3C-4M in both cases.

 

Yes obviously I will give away info sometimes by having 2 ranges instead of 1 when responder has a hand that would just sign off over both ranges. I think on average I will give away less information by a lot. Also, I don't think you can define my ranges as well as that, perhaps it depends on how light you are opening. If you play sound openings you almost always have a GF over a limit raise anyways I guess. I can have balanced or unbalanced hands in either category, though obviously the first is weighted towards balanced.

 

All in all I think there is easily enough space to handle the majority of issues, still have good slam auctions, and minimize information leakage as much as possible. You could go even further and have just 3C=non min, 3D asks, 3H=worst, then bal lmh, and over 3H you can ask. Now you are having even less information leakage but quite possibly hurting your constructive bidding too much (shortness in the other major and a minimum non min has no room to cuebid, etc). I guess the question is just where is the right balance.

 

I wonder when he is going to realize puppet stayman after strong 2NT sucks. Meck will take even more tricks then

 

They stopped playing puppet for a long time, they have only started again fairly recently.

[aguahombre]

If they knew Bluecalm's opinion that it sucks, they probably would stop again.

[bluecalm]

If they knew Bluecalm's opinion that it sucks, they probably would stop again.

 

Maybe they have another reason, like reducing variance and playing along with the other table.

No idea, I spent enough time on this one to be confident that at least at imps puppet is not profitable and it's better to just bid 3nt on all those 4-4-3-2, 4-3-3-3 and 5m-3-3-2 hands.

If someone very good had told me "I spent a lot of time analyzing this and my conclusion is that puppet is +EV instead of bashing 3NT" then I would probably reconsider but intuition alone even from someone very good isn't going to convince me as I saw tons of layouts and I don't have selective memory bias on this one (as I saw them consecutively with the goal of determining which is better).

 

One day when my program is ready which can make reasonable human-like lead on most NT biddings I will just simulate 10k hands and see. For now it have to wait :)

[JLOGIC]

Who bids puppet on 4333? Obviously it's good to be playing puppet when you are like (43)(15) or 33(16) or something, and on some 4432s I would definitely want to find my 5-3 fit if I could.

 

Again, what are you going to simulate? There is no doubt that finding your 5-3 fit is good on a certain subset of hands, the bigger question is how frequently do you lose when you look for a 5-3 fit and don't find one vs how often do you gain when you do find one, and how often do you lose by not being able to bid smolen/whatever you play 3C 3D 3M is normally. Good luck simulating that!

 

And again, there is no chance that meckwell are playing puppet to try and reduce variance, I cannot believe you would actually think that.

[aguahombre]

Jerry Helms, in the Feb Bulletin, attempted to clarify his previous statement about not liking puppet/2NT. Basically, the problem is the 4-5 and 5-4 hands, but an experienced pair, and/or a great pair, can make adjustments to other auctions and include puppet if they have the inclination to do so. It boils down to context. The convention sucks for those who have holes in their structure because of puppet. And is simply a matter of preference for the rest.

[bluecalm]

Again, what are you going to simulate?

 

I took some hands which people routinely bid puppet with and look at the possible layouts (100's of them because I am an addict).

My conclusion was that puppet causes more loses (because of better 1st lead) than it ever gains on 4-4-3-2 hands and 5-3-3-2 hands. Also system with puppet is weakier because there is less space for other hand types but obviously this one is difficult to quantify.

Obviously all the muppets and romexex which doesn't allow you to bid 2NT - 3NT aren't even worth considering.

[JLOGIC]

Then perhaps the results of your simulation show that people are bidding puppet on the wrong hand types, not that puppet is a bad convention (btw I do not play puppet over 2N).