[straube]

I've understood that 1C (strong)-1D (0-7) is...

1) too frequent an occurrence (for us it would be 65% or so), bogging down continuations and...
2) leaves the partnership vulnerable to RHO action.

So there seem to be two alternate families of responses. One is used by Moscito and Scream

1D-GF or most GFs
1H-semipositive or most semipositives
1S-DN

and the second by awm's sytem (and others?)

1D-GF or DN
other-semipositive/light GF

and I'm starting to wonder whether either the classic 0-7 or awm's structure is better than the Moscito/Scream family.

I've never liked that after 1C-1S, responder is not in a position to know whether opener has extra values or not. 1C-1S, 1N shows something like 15-20 or 16-21 depending. It's just a very large range and responder may have a hand with lots of jacks and spot cards or perhaps a very distributional hand where he would want to invite or bid game but can't....because he is unaware that opener has extra.

I also don't like that opener has to retreat to 1N with various 5m431s. Not so bad the 4441s, but the 5m431s are collectively pretty common. So now responder is transferring into opener's shortness, etc. Yuck.

But the real unfortunate thing is that opener can't even play 2m now.

And this is just what I dislike about the 1S response. There are similar problems after an 1H semipositive response.

We've looked at structures over this and concluded that it makes more sense for...

1C-1H,
.....1S-GF relay
.....1N-16-18, possible 5m431 or 4441
.....2C-stayman
.....2D-transfer
.....2H-transfer

than to use 2m as a place to play. Maybe that's the wrong conclusion (I don't think it is), but even so, the point is that the 1H has taken away space for part score bidding. Yes, on the one hand, the 1H bid has empowered opener to bid more, but OTOH it gives him less room to show his hand.

Another subject, but after 1C-1D (1H) we have found that we don't have a very good use for double here. We reverse relay here, so we can use Pass as S1 and dbl as S2, but then we have to count steps and find that it's not really worth the confusion. So our double is penalty and that's pretty useless. If the 1D response were negative, then 1C-1D (1H) dbl would be takeout and useful.

So my question is probably aimed for Moscito players, but what sorts of results do you get after 1C-1S and is the immediacy of knowing that partner has a semipositive reponse (particularly the 1H response) worth the cramped or somewhat cramped auction that follows?

[PrecisionL]

straube, on 2012-January-04, 12:31, said:

So my question is probably aimed for Moscito players, but what sorts of results do you get after 1C-1S and is the immediacy of knowing that partner has a semipositive reponse (particularly the 1H response) worth the cramped or somewhat cramped auction that follows?

I play our version of Millennium Club (15+ hcp) and 1♣ - 1♠ = 0-7(8) hcp and No 4-card major.

This works OK, but 75% of our 1♣ openings are balanced hands, so opener was going to rebid 1NT anyway.

21% of 1♣ openers are 1 or 2-suited with clubs primary, and

4% of 1♣ openers are balanced & 22+ HCP or G.F.

[Free]

As an ex-MOSCITO player, I don't really remember disasters after a double negative. I've played 2 different continuation schemes, both with their (dis)advantages.

One was:
1NT = around 15-20 (semi)balanced (basically less than 15 AKQ-points)
2♣ = GF relay (basically 15+ AKQ-points)
2♦/M/3♣ = canapé transfer (either signoff in the transferred suit, or planning to bid a longer suit)
2NT = strong NT

The other one was based on multi-landy, which was more geared towards better partscores:
1NT = around 15-20
2♣ = both Majors or some strong hands
2♦ = 6+M or 4M-5m
2M = 5M, 4+m
2NT = both minors

The large 1NT range isn't a big problem. Responder showed 0-2 AKQ-points, so he can hardly have an invite. If opener wants to invite, he can always bid stronger. Bidding 1NT means you'll only go to distributional games, for example after a transfer which gets super accepted.

Similar after 1♣-1♥ showing any semi-positive without 5M, but the 1NT range is smaller because opener has a lot more GF hands. Also, responder will invite much more frequent after 1♣-1♥-1NT. Btw, I never played stayman and transfers after this semi positive response, I find the scheme based on multi-landy much better.

[awm]

Dunno what you mean by unsound, but comparing my method to Moscito...

(1) If opponents are passing throughout, I am essentially always better placed. My auctions on double-negative hands will be much better with the extra space. I can devote many more calls to show semi-positive hands (because I don't need 1♠ for double-negative), and my relays are generally more efficient.
(2) If opponents are bidding over the response, then I'm better placed on semi-positives because opener has more shape information. I'm also better placed on some positive hands (because I don't bid 1♦ on shapely min positives ever). The only time I'm behind is when it goes 1♣-Pass-1♦-BID. However, when opener has length in the opposing suit he is safe to pass, since he doesn't want to be anywhere opposite a double-negative and the auction is of course forcing when responder has a positie. The only potential problem case is when opener is short in the opposing suit... but then he will often (okay not always) want to bid even opposite a potential double negative to compete for the partial. The vast majority of the time he can take some call and happily hear partner pass (double neg) or do anything else (positive). The thing is that opener doesn't have to distinguish between very minimum hands that still want to compete and hands with a bit extra because of responder's split range. I just don't think I lose much here.

[whereagles]

Here's an idea for you: based on the "shape first" principle, I suggest transfer responses:

1♣ ??

1♦ = 5+ hearts, 5+ hcp. Opener completes the transfer with 19+ GF or bids something else with 16-18.
1♥ = 5+ spades, 5+ hcp. As above.
1♠ = your DN.
1NT = clubs, 5+ hcp, yada yada.
2♣ = diamonds, 5+ hcp.
2♦ = balanced hand 5-7.
2♥+ = balanced hand 8+ GF.

[straube]

awm, on 2012-January-04, 19:42, said:

Dunno what you mean by unsound, but comparing my method to Moscito...

(1) If opponents are passing throughout, I am essentially always better placed. My auctions on double-negative hands will be much better with the extra space. I can devote many more calls to show semi-positive hands (because I don't need 1♠ for double-negative), and my relays are generally more efficient.
(2) If opponents are bidding over the response, then I'm better placed on semi-positives because opener has more shape information. I'm also better placed on some positive hands (because I don't bid 1♦ on shapely min positives ever). The only time I'm behind is when it goes 1♣-Pass-1♦-BID. However, when opener has length in the opposing suit he is safe to pass, since he doesn't want to be anywhere opposite a double-negative and the auction is of course forcing when responder has a positie. The only potential problem case is when opener is short in the opposing suit... but then he will often (okay not always) want to bid even opposite a potential double negative to compete for the partial. The vast majority of the time he can take some call and happily hear partner pass (double neg) or do anything else (positive). The thing is that opener doesn't have to distinguish between very minimum hands that still want to compete and hands with a bit extra because of responder's split range. I just don't think I lose much here.

I think your system is pretty impressive. I think it does have some tradeoffs though. After 1C-1D, 1N responder may have a shapely superpositive hand and he has much less room to show it. The 1D doesn't empower opener to act and I would think that would occasionally hurt you in competition. Lastly, you have the problem that all 2-way bids create....the further one branches from the relay bid (1D in this case) the more difficulty one has making a bid that is welcome to both hand types (weak and strong in this case).

One thing that is wonderful about your 2-way bid is that you've effectively created many more sequences. You might bid 1C-1D, 2C and opener has no idea (nor does he care) whether partner will pass or relay the hand out. For us, 1C-1D, 2C is forcing. We don't have the pass option. True, pass leads nowhere, but it obviates a problem hand....and we have many of those.

Another thing is that you have shorter auctions when slam is not in the making. Does your 2-7 range though also work against you? I'm wondering if opener sometimes doesn't have the strength to relay (for fear of 2) but if he had, he might have discovered a shapely pattern with 7 QPs that would have produced slam.

[straube]

whereagles, on 2012-January-04, 20:04, said:

Here's an idea for you: based on the "shape first" principle, I suggest transfer responses:

1♣ ??

1♦ = 5+ hearts, 5+ hcp. Opener completes the transfer with 19+ GF or bids something else with 16-18.
1♥ = 5+ spades, 5+ hcp. As above.
1♠ = your DN.
1NT = clubs, 5+ hcp, yada yada.
2♣ = diamonds, 5+ hcp.
2♦ = balanced hand 5-7.
2♥+ = balanced hand 8+ GF.

I think you are bidding too high with the balanced hands. 5-7 at 2D might just be too high for us to make anything, let alone get to the proper strain. 8+ at 2H prevents opener from showing a shapely hand. If you look at the transfer bids, you give very different amounts of room to (for instance) showing diamonds and hearts. That can't be right.

[straube]

Free, on 2012-January-04, 15:01, said:

As an ex-MOSCITO player, I don't really remember disasters after a double negative. I've played 2 different continuation schemes, both with their (dis)advantages.

One was:
1NT = around 15-20 (semi)balanced (basically less than 15 AKQ-points)
2♣ = GF relay (basically 15+ AKQ-points)
2♦/M/3♣ = canapé transfer (either signoff in the transferred suit, or planning to bid a longer suit)
2NT = strong NT

The other one was based on multi-landy, which was more geared towards better partscores:
1NT = around 15-20
2♣ = both Majors or some strong hands
2♦ = 6+M or 4M-5m
2M = 5M, 4+m
2NT = both minors

The large 1NT range isn't a big problem. Responder showed 0-2 AKQ-points, so he can hardly have an invite. If opener wants to invite, he can always bid stronger. Bidding 1NT means you'll only go to distributional games, for example after a transfer which gets super accepted.

Similar after 1♣-1♥ showing any semi-positive without 5M, but the 1NT range is smaller because opener has a lot more GF hands. Also, responder will invite much more frequent after 1♣-1♥-1NT. Btw, I never played stayman and transfers after this semi positive response, I find the scheme based on multi-landy much better.

What I don't like about multi-Landy is that it threatens to "drop" responder prematurely. Say it goes 1C-1S, 2C-2H. That shows 4 hearts, right? So opener passes and responder has 6 hearts and the partnership has missed a 10-cd fit.

I also don't like if it goes 1C-1S, 2C-2H, 2N (which I assume shows 21-22 or so)-3H and opener has AKxx Q AKxx AKxx and responder has xx Jxxxx xx xxxx.

The "being dropped" issue is compounded when responder has a semipositive hand. We have occasional semipositive hands that always intend to game force...they are just lacking in the QP department.

Well, missing 2C with x AKxx AQx AJTxx is not the worse thing about 1S as DN, but I don't really like having to rebid 3m after 1C-1H. With a six-cd suit, I rate to make, but the bid puts a lot of pressure on responder when he wants to show a 5 or 6-cd major...especially when he wants to force to game with or without a club fit.

[awm]

straube, on 2012-January-04, 20:25, said:

I think your system is pretty impressive. I think it does have some tradeoffs though. After 1C-1D, 1N responder may have a shapely superpositive hand and he has much less room to show it. The 1D doesn't empower opener to act and I would think that would occasionally hurt you in competition. Lastly, you have the problem that all 2-way bids create....the further one branches from the relay bid (1D in this case) the more difficulty one has making a bid that is welcome to both hand types (weak and strong in this case).

One thing that is wonderful about your 2-way bid is that you've effectively created many more sequences. You might bid 1C-1D, 2C and opener has no idea (nor does he care) whether partner will pass or relay the hand out. For us, 1C-1D, 2C is forcing. We don't have the pass option. True, pass leads nowhere, but it obviates a problem hand....and we have many of those.

Another thing is that you have shorter auctions when slam is not in the making. Does your 2-7 range though also work against you? I'm wondering if opener sometimes doesn't have the strength to relay (for fear of 2) but if he had, he might have discovered a shapely pattern with 7 QPs that would have produced slam.

It's 2-6 QP... we haven't really had issues with it, especially because we can judge (at least partially) degree of fit.

What's interesting is, suppose after 1♣-Pass-1♦ (GF) - BID you play an inversion where Double is takeout, bids are natural, and pass asks partner to double if he would sit for a penalty double (or otherwise bid). You will find that opener extremely often makes the same call that he would over 1♣-Pass-1♠ (double-negative) - SAME BID. There are a small number of hands where opener wants to force game opposite the double-negative hand and can keep the auction lower opposite the GF (but it's pretty rare for partner to have a game force when you have enough for game opposite the double-negative, and even rarer for opponents to bid in that auction) and a few hands where the opponents bid at a high level and you would takeout double opposite the GF but pass opposite the double-negative... but by far the majority of hands you are making exactly the same call if for different reasons. This is why I don't think I lose much in these sequences.

[akhare]

I don't think it's unsound per se, but I am convinced that it's possible to do much better with awm's scheme.

For example, here's a sample scheme off the top of my head using the ambiguous 1♦ DN that combines some of the best elements of TOSR and Moscito.

Note how quickly this scheme unwinds some of the SP hands while right siding the positives:

1D: DN or some unbal hands with hearts
....1H asks:
.......1S: Confirms DN
.......1N: Majors or three suited with short minor
.......2C: H+C
.......2D: H
.......2H: H+D
....1S: Natural, 5+ spades
....1N: Strictly balanced hand shapes (opener can always bid 1H with bal hands)
....2C: Natural, single suited
....2D: Natural, single suited

1h: GF; Hands with spades, spades+minor, bal, three suited with short major
1S: SP; Mostly balanced, could be single suited with spades or hands with minors
1N: SP; 5+♥+m or single suited ♥
2C: SP; Majors
2D: GF; Clubs
2H; SP; ♠+minor
2S+: GF; diamonds

[akhare]

BTW, Rob F.'s modified TOSR method is analagous to awm's scheme and uses 1D as 0-7 and resolves shapes at symmetric - 1.

Yet another variation on this scheme:

1D: DN or unbal with hearts
1H: Spades, bal, three suited
1S: SP
1N: Majors
2C: Minors
2D: Clubs
2H: Three suited
2S: Diamonds

[straube]

Does Rob's SP response handle all the semipositives? If so, that doesn't seem right. We would have a frequency distribution like (and I am guessing)...

1D-25%
1H-18%
1S-45%
etc

Here's awm's system...

http://www.cs.ucla.e...IMprecision.pdf

I don't feel like these others resemble awm's system very much. Adam is trying to show the distributional 2-6 RP hands immediately. The 7+ distributional hands (and perhaps the balanced 5-6+ RP hands) are biding their time with 1D. He does well when opener has a distributional hand and responder's balanced hand can relay it (which I'm sure is why he orders it that way), but imo loses when opener rebids 1N and responder has a shapely big hand. He also does well in part score situations and those hands where game is likely but slam out of reach. Fair assessment?

Our structure is very strong when it comes to slam bidding, but is a little vulnerable to preemption and more frequently misses the best part score. Lots of tradeoffs here.

[straube]

I think I'm misunderstanding Rob's structure. He must be using responses that show SP or better values.

[akhare]

straube, on 2012-January-04, 23:29, said:

Does Rob's SP response handle all the semipositives? If so, that doesn't seem right. We would have a frequency distribution like (and I am guessing)...

1D-25%
1H-18%
1S-45%
etc

Rob's scheme doesn't have any SPs, so 1♦ is 0-7 any or GF with ♥s

straube said:

I don't feel like these others resemble awm's system very much. Adam is trying to show the distributional 2-6 RP hands immediately.

The previous scheme attempted to do something very similar using a variation on Adam's idea. The point is that using 1♦ as a two way bid is (IMO) better than using 1♠ as DN and there are various possible permutations on the idea that can cater to various desired tradeoffs.

One possibility is to limit the number of GF hands in the 1♦ two way bid and utilize the extra space freed by 1♥/1♠ to describe more SPs immediately and / or resolve GF hands faster.

[straube]

Thanks for explaining. I think there's a big difference in what they're doing. Rob seems to be putting the GF heart hands (all of them?) into what is a fairly overloaded response. I think 0-7 accounts for something like 60+% of the hands. So is it like 70% for him? So it's great that he's -1 on most hands, but I wonder how he sorts out the 1D responses.

Adam is using the 2-way not as 0-7 or GF specific but into very weak or very strong. Either way tends to be very interested in opener's shape...very weak because a landing spot is needed very soon for a part score. Very strong because slam might be in the cards. In either case, it is opener's hand that is patterning out and to reverse relay (Rob's hearts for instance) would be cross-purposed. Again, it's not ideal (I think) when responder is very strong distributional and opener rebids 1N.

Have you looked at his notes? It really took some doing to sort out the auctions where responder starts to relay a semipositive/light GF hand and opener hasn't the strength to continue relays. In a sense, the semipositive/light GF responses are 2-way bids as well and opener can't know whether responder intended a GF when he first has the opportunity to simply relay the hand. The relay breaks are very strong, but I'm not sure I have the memory capacity for that.

[awm]

straube, on 2012-January-05, 00:24, said:

Again, it's not ideal (I think) when responder is very strong distributional and opener rebids 1N.

While this is true in principle, you're overstating the case somewhat. Opener's 1NT shows a balanced hand (a true one; we don't do off-shape 1NT rebids here) with a limited range (17-20). You can do a very good job on slam bidding over a natural and limited 1NT opening, and we've put in a lot of work to make our methods over 1NT openings as good as possible. In fact it is somewhat relay like, with responder able to show all common shapes below 3NT, or make single-suited slam tries effectively. Knowing that partner is balanced and limited really does help a lot. Certainly in some cases we'd be better off with opener running a true relay system, but there are other cases (such as where suit quality in responder's six-card suit becomes key, or where we want to make a straight quantitative invite to slam) that we actually do better by allowing responder captaincy. These hands also tend to be somewhat awkward in a more traditional relay structure because responder has to skip a lot of steps in showing strength/controls (and may have to remove an attempted sign-off by opener in the process).

[akhare]

awm, on 2012-January-05, 00:31, said:

While this is true in principle, you're overstating the case somewhat. Opener's 1NT shows a balanced hand (a true one; we don't do off-shape 1NT rebids here) with a limited range (17-20). You can do a very good job on slam bidding over a natural and limited 1NT opening, and we've put in a lot of work to make our methods over 1NT openings as good as possible.

+1 -- I think that the balanced hand principle is overhyped as well.

[akhare]

straube, on 2012-January-05, 00:24, said:

Thanks for explaining. I think there's a big difference in what they're doing. Rob seems to be putting the GF heart hands (all of them?) into what is a fairly overloaded response. I think 0-7 accounts for something like 60+% of the hands. So is it like 70% for
him? So it's great that he's -1 on most hands, but I wonder how he sorts out the 1D responses.

Here's Rob's scheme:

1D H/H+m/4441 GF, or almost all negatives (excluding 1-suited 6+ semipositives)
........1S any negative, or GF 4441
........1N H+C
........2C H only
........2D+ H+D
1H S/S+m GF, or 1-suited S invite
........1N S+C
........2C S only
........2D+ S+D
1S GF balanced or C
........2C C only
........2D+ balanced
1N majors
2C D only
2D+ minors

Quote

Have you looked at his notes? It really took some doing to sort out the auctions where responder starts to relay a semipositive/light GF hand and opener hasn't the strength to continue relays.

Briefly, but as I see it, it's just a testimonial to what the two way 1♦ response makes possible. There are numerous ways of allocating responses to cater to showing immediate SP / GF hands.

The basic question is whether one buys into the two way 1♦ "meme" (I am sold)...

[benlessard]

I dont see why you should miss your fit when you are weak. IMO its much better to miss some of your fit when responder is SP, so at least when you endup in the wrong contract it should still make because you are in the 20-23 range and if opener is maximum you will still find the fit.

1D SP/H
and transfers for the rest is by far what is best IMO. I consider not playing transfer over a strong club to be borderline crazy.

[straube]

Generally speaking, I think 2-way bids are problematic because partner can't bid very high from that 2-way bid without getting in the way of one or the other type of hand. When 2-way bids are successful, they are usually accompanied by a great deal of complication. I'll give two examples from our own system.

1) We play that opening 2C to show an opening hand with 6 clubs followed by 2D is either a constructive hand with hearts OR any GF hand that wants to relay partner's hand. We had to devise a whole new relay structure to handle this. Opener's first rebid (2H) showed 0-1 heart, second response showed 2 hearts, and higher responses showed 3 hearts. Again, a whole new relay structure and relay breaks, too. So imo, it's successful but complicated.

2) We've added 1D-1H to be hearts OR GF balanced without a major. This works well for most of opener's rebids. When opener has hearts, we've had to order all of his 2D+ responses so as 1) to assist us getting to game or slam when responder has hearts and 2) assist us getting to the right game or slam when responder has the big balanced hand. We weren't entirely successful accomplishing this. For example, 1D-1H, 2N shows a void and now we can relay the hand but haven't room to find out the void prior to deciding whether to bid game in hearts. We have 3D as a general invite. We have to guess. I think that just illustrates that the higher one bids away from the 2-way bid, the more difficulty one has orchestrating that bid to work against both hand types.

I think it would be difficult to play 1C-1D as 0+ which it would be if it were say 0-7 OR 8+ with hearts or something else like that. For example, opener might try to show extra strength on some hands in case partner has 0-7, but that would be the wrong approach when partner turned out to have GF hearts. I mean, let's say we were playing a vanilla relay system and it went 1C-1H showing GF hearts...we wouldn't use a 1S rebid to show that opener has 20+ or so, right? That's the wrong hand showing. So that's probably not even how Rob continues, but I'm just trying to illustrate that it's very difficult to order opener's rebids against something that can be 0+.

I think I get what awm is doing though. It's very rare for opener to have a hand that wants to force game opposite a DN response. When he doesn't, he just starts to show his shape, and this is successful opposite weak or strong hands. When he does want to force game, he has forcing bids available in 1M rebids. 1C-1D, 1H is forcing and 1C-1D, 1S is forcing. He'll find out next round whether partner has the DN or not. He might have to jump then if he hasn't found a fit. I haven't sorted through that yet. The other difficulty is when opener has a GF hand with the minors and as he's pointed out, these are problematic for other systems as well.

To make his structure work, however, requires a lot of complication, at least compared to what we're doing. Lots and lots of relay breaks.

btw, I hope awm or Rob (if he's following this) don't mind our/my discussing their systems.