[rbforster]

I was thinking some about the differences between 1♥-1NT (GF) and 1♠-1NT (GF) and the implications for the rest of your invitational hands. Wouldn't it make sense to play some sort of Kaplan-style inversion so that 1♥-1♠ became the GF and 1♥-1N was available as something else (spades?). Among other things, it would give the same level of relay resolution for both hearts and spades (whereas the 1♥-1N will be more cramped currently).

Along these lines...

Echognome, on Jun 16 2008, 05:17 PM, said:

I didn't put in our scheme after 1M - 2♣, although it is one of our more difficult sequences.

1♠ - 2♣ (not 3♥ or an IJS)

2♦ Any minimum (may have 4♥, 6♠), then:
--2♥ 4♥, NF, usually 1444, 14(35), or 04(54)
--2♠ Hx, NF
--2NT NF
--3♣ 6-card suit, not worth an IJS
--3♦ Same, but most would pass 2♦
2♥ 4+♥, Extras, GF
2♠ 6+♠, Extras, GF
2N 5♠, bal or semi-bal (may have 4m), Extras, GF
Higher Natural, GF

Similar after 1♥ - 2♣, except now 2♠ and higher are natural and GF.

Gnome - I notice that your 1♥-2♣ bid seems pretty narrowly defined, being an invite without

3♥ (else 2♦)
4♠ (else 1♠)
both minors (else 2N), or
a long good minor (else 3m)

This pretty much just leaves balanced hands - 3244 and 32(53), plus maybe some 6 card minors with poor suits and/or some xx(54)'s if you don't bid 2N with those. Do you think it would be too much of a stretch to include these hands in 1NT forcing and maybe leave a 2N rebid by responder as a natural invitation? I'm not sure what response structure you're playing over your 1M-1N GF relay, so I don't know if this would fit in well (for example if higher bids than 2N showed extras). If you did this, you could maybe include your IJS in clubs or both minors in 2♣ now (maybe even the IJS in diamonds?), and have more space to sort things out, as well as freeing up 3♣ and/or 2N for Bergen raises or something. I notice that your current scheme doesn't give you a mixed raise of opener and while 2♦ (3+♥s) is great for constructive bidding but 2♦ or 2♥ doesn't exactly shut out the spade-holding opponents when you've got a random 6 count with 1=4 majors as responder.

Of course over 1♠-2♣, I can see the point that you have hands with 4♥ that you want to be able to find the heart fit (basically all the 4 card 1♠ response shapes with swapped majors), so this covers a lot more ground than does the 1♥-2♣ bid.

[Echognome]

Rob F, on Jun 17 2008, 07:36 AM, said:

Gnome - I notice that your 1♥-2♣ bid seems pretty narrowly defined, being an invite without

3♥ (else 2♦)
4♠ (else 1♠)
both minors (else 2N), or
a long good minor (else 3m)

etc..

Rob,

I appreciate what you are saying. Note that including non GF hands into the relay makes it no longer midchart compliant. As far as I am aware, everything else I've described is midchart compliant. That means we can play the system in some F2F events. If I'm going to play a system that's not midchart, then I'd go for Echo Club instead of Gnome Club, which is my variant on Moscito (think something between TOSR and Moscito and you'll be close).

Also, I want to add that we're not all that excited that we have to pass non-fitting hands that are less than invitational. I consider this a loss in the system, but I just don't see a place to put them. Second, the 1M - 2♣ is pretty unwieldy unless narrowly defined. These are hands that in standard will often go 1M - 1N - 2m - ? So the room you lose might seem minimal (one step), but it really takes away opener's potential rebids. (You just have no place to show a minimum hand with 5M/4♣.) So we just alter things slightly and then tried to take as many hands out of the bid as possible. What may not be clear is that the 2NT showing the minors (and we pretty much reserve for 5-5) is only over 1♥ (since over ♠ it shows a mini-splinter). So over 1♥, the 2♣ bid is a LOT more well defined than over 1♠. However, we also can show 5♥ and a weak hand directly over 1♠ which is nice. Majors rule, minors drool... you know how it goes.

[joshs]

Echognome, on Jun 17 2008, 10:48 AM, said:

Rob F, on Jun 17 2008, 07:36 AM, said:

Gnome - I notice that your 1♥-2♣ bid seems pretty narrowly defined, being an invite without

3♥ (else 2♦)
4♠ (else 1♠)
both minors (else 2N), or
a long good minor (else 3m)

etc..

Rob,

I appreciate what you are saying. Note that including non GF hands into the relay makes it no longer midchart compliant. As far as I am aware, everything else I've described is midchart compliant. That means we can play the system in some F2F events. If I'm going to play a system that's not midchart, then I'd go for Echo Club instead of Gnome Club, which is my variant on Moscito (think something between TOSR and Moscito and you'll be close).

Also, I want to add that we're not all that excited that we have to pass non-fitting hands that are less than invitational. I consider this a loss in the system, but I just don't see a place to put them. Second, the 1M - 2♣ is pretty unwieldy unless narrowly defined. These are hands that in standard will often go 1M - 1N - 2m - ? So the room you lose might seem minimal (one step), but it really takes away opener's potential rebids. (You just have no place to show a minimum hand with 5M/4♣.) So we just alter things slightly and then tried to take as many hands out of the bid as possible. What may not be clear is that the 2NT showing the minors (and we pretty much reserve for 5-5) is only over 1♥ (since over ♠ it shows a mini-splinter). So over 1♥, the 2♣ bid is a LOT more well defined than over 1♠. However, we also can show 5♥ and a weak hand directly over 1♠ which is nice. Majors rule, minors drool... you know how it goes.

Well I don't agree with your interpretation of the rules. There are many non-GF relays in bridge (stayman for instance). What was illegal in the midchart was a sequence of relay bids that do not promise game forcing values. Its sort of unclear what a sequence is in the definition, but its certainly not 1 bid.

So I see nothing wrong with
1S-1N(10+ HCP, any shape)-2C(any min)-2H(Non-forcing, about 10-12) in the midchart.

(Note: I have always interpreted GFing to mean forcing to 3N or higher, so stopping in 4m is ok, but that more based on general bridge principles then anything that is explictly written anywhere).

Aloso, on occasion, a sequence can technically be GFing, but a player gambles and passes. I don't think there is a problem with that.

Anyway, here is my system suggestion:
1H:
1S Forcing 1 round
1N:INV+ Relay
2C Any Min (2D=GF Relay, others NF)
2D+ GFing
2C: Constructive, both minors (4+4+)
2D: 5+D, NF
2H: Normal raise
3C: 6+C, NF

and I will leave it to the reader to figure out the other bids.

Over 1S things are more interesting:
1S-1N(INV+ Relay)
2C Any Min (2D=GFing relay)
2D+GFing
1S-2C: Constructive, 0-2 spades, usually 3+ cards in the other suits, but could be 1525 (will then correct 2D to 2H)
1S-2red: NF
1S-3C: 6+C, non-forcing

Now in this scheme the main problem is if opener is a min, and responder has an Invite with exactly 4 hearts you might lose a 4-4 heart fit over 1S-1N-2C.

Consequently, I think the rebid that shows side hearts over 1S-1N should not promise any extras, but thus needs to be either 2D or 2H so responder can bid 2N on up as INV.

[glen]

joshs, on Jun 17 2008, 05:16 PM, said:

Well I don't agree with your interpretation of the rules. There are many non-GF relays in bridge (stayman for instance). What was illegal in the midchart was a sequence of relay bids that do not promise game forcing values. Its sort of unclear what a sequence is in the definition, but its certainly not 1 bid.

...
1N:INV+ Relay
...

If you define 1NT as "INV+ Relay", which to TDs will mean "Relay" and "Invite or bettter", I promise you ACBL TDs will decide you are playing a "Relay" without GF values, and will not allow it in Mid-chart events. In other words, they will not agree with your "interpretation of the rules".

Edit: Here's a term that might produce another reaction: 1N: Waiting, Inv+. No TD, not a relay, just one of those waiting bids everybody uses. Oh, you read the non-natural & out-2-lunch BBO forum - I didn't know TDs did that

[joshs]

glen, on Jun 17 2008, 05:55 PM, said:

joshs, on Jun 17 2008, 05:16 PM, said:

Well I don't agree with your interpretation of the rules. There are many non-GF relays in bridge (stayman for instance). What was illegal in the midchart was a sequence of relay bids that do not promise game forcing values. Its sort of unclear what a sequence is in the definition, but its certainly not 1 bid.

...
1N:INV+ Relay
...

If you define 1NT as "INV+ Relay", which to TDs will mean "Relay" and "Invite or bettter", I promise you ACBL TDs will decide you are playing a "Relay" without GF values, and will not allow it in Mid-chart events. In other words, they will not agree with your "interpretation of the rules".

Edit: Here's a term that might produce another reaction: 1N: Waiting, Inv+. No TD, not a relay, just one of those waiting bids everybody uses. Oh, you read the non-natural & out-2-lunch BBO forum - I didn't know TDs did that

I repeat, relays are allowed, just not relay systems (defined as a sequence of relays) that do not promise a game force.

Can you tell me the difference between:
a. 1S-1N(Forcing, 5-11, any shape)
b. 1S-1N(Forcing, 12+ Any Shape)
c. 1S-1N(Forcing, 5+ Any Shape)

These bids are either all relays, or all not relays depending on whether you think the point count shown means its descriptive in some way. Or the negative inference of the hands that would bid something else provides significantly more definition (it provides some extra definition in each of these but not a lot)

Some people even play
d. 1S-1N(Forcing, 0+ Any Shape).... [Who amoung you has not bid 1N forcing on xxxx xxxx xxx xx over 1S?]


And just because a director might or might not know the rules (I sadly would bet against most of them) the main question is what the rules say. Our announcement of "Forcing" or "INV+Relay" shouldn't matter. If we announced forcing, but the bid really promised 10 HCP, its still not GCC legal (1N response can not guarantee Inv values), although don't get me started on that rule.

[rbforster]

True a "relay system" is a pretty poorly defined, perhaps impossibly defined creature, at least if you define relay breaks for both bidders. Surely a 2/1 GF auction isn't a "relay system" so why would it this be if both sides could make any bid except that we've assigned artificial (rather than natural) meanings to those? I think most people think of a relay system as one where (say) responder keeps making the cheapest bid and opener keeps making artificial descriptive bids, but good systems will also have at least somewhat useful and somewhat frequent meanings for responder's non-step rebids so then the cheapest step is really a specific sort of ask rather than a relay.

joshs, on Jun 18 2008, 10:31 AM, said:

If we announced forcing, but the bid really promised 10 HCP, its still not GCC legal (1N response can not guarantee Inv values), although don't get me started on that rule.

As for the 1NT forcing not guaranteeing INV+ values, this doesn't take much creativity to avoid. Perhaps there is some hand that is willing to pass all of opener's minimum responses? Or maybe a weakish one suiter that is going to bid 4OM next time to play?

[benlessard]

in practice what is allowed (mid-chart) is that the first response can be artificial 6+ forcing "say like 1Nt forcing" and that your next relay is GF. What director wont allow you to play is 2 or 3 relays or waiting responses that dont show GF values or 0+ responses. After a strong C opening you can play whatever you want BTW.

What im 100% sure is that

1- after 1H opening play 1S as the relay and not 1Nt.

2- dont play light opening if you want to play 1NT as absolute GF, so when you say 10-14 its should be great 10 count with 2 aces & 6+ cards suit not overcall type of hands.

3- if you play superchart-- put non inv & non GF hands in your relay. Setting FP is nice but getting better partscore will be more important. Plus its tougher for the opponents to overcall.

1H--(pass)---1Nt-gf----(???) (here he can bid with crap just so that you dont relay)

1H-----1Nt (6+)----here he should overcall with sound values because it might be their hands.

4- its nice if your 1M openings are unbalanced. So if you are playing weak NT then 1Nt 12-14 with 5M possible & 1M opening as unbalanced is a better setup.

Our setup is

1H (12-14 or 18-22) 1S--- relay 6 +
???

1NT clubs 12-14 or jump shift strenght
2C diamonds 12-14 or jump shift strenght
2D 6H 12-14 or jump shift strenght
2H 5H + 4S 12-14
2S 5H +5S jump shift strenght (or 6H +4S 18-22 if you open 1S with 5-5)
2Nt 4522 jump shift strenght
3C 4513 jump shift strenght
3D 4531 jump shift strenght
rest is voids

[glen]

Certainly I like 1NT 6+, may be GF

1M-?
--1NT: 6+, forcing, less than GF will be flat or short in M with 3+ length in other 3 suits
---- after opener's rebid, cheapest bid by responder not M is artificial GF
---- pass, cheapest bid in M show less than an invite
---- raise of opener's last suit, jump to 3M, 2NT (if not artificial GF) are all invites
---- other bids, not including the artificial GF, are natural GF

--2X: 5 or longer, up to an invite, non-forcing

--2NT: natural invite, fewer than 4 in the other major

[awm]

I know a lot of people here love their 1NT response that shows any of three or more totally different hand types some of which are weak and some of which are strong. These kinds of bids seem nice when you're designing a bidding system in isolation, or when you play against opponents who always pass. But in real bridge, opponents tend to bid, especially when you are making nebulous calls that give virtually no information about shape or values. Even simple interference seems to be tough to deal with using these kinds of methods: take 1♥-P-1NT-2♠. If the 1NT bid could potentially be some 5-count without a real fit, you don't really want opener to be doing anything but passing here with the vast majority of 10-14 hcp hands. But if opener has to almost always pass in competition, then responder has no real information at second turn to bid, and pretty much has to double with a wide variety of hand types that include real values. In some cases it becomes hard to distinguish hands that were "invitational" from hands that are "game force" because your cheap relays are no long available. And it becomes very difficult to defend on any hand, because you have neither a penalty double nor a takeout double available.

In a typical five-card major system, the 1♠ opening is more frequent than the 1♥ opening. This is because 5-5 major hands open 1♠, and otherwise the hand types are symmetric. So I don't see a particularly strong reason that the relay over 1♥ needs to be 1♠.

Josh's structure using 2♣ as "all minimums" doesn't seem that great to me. You've basically lost a full step in your relay resolution because of this (okay 2 steps if opener has the minimum). So you might get cheaper resolution by using 2♣ as a fully game forcing relay (which gets you back a "normal" forcing notrump response). The fact that 1NT isn't game forcing also loses some forcing pass situations which can help you in competition after a game-forcing relay. And my feeling is that shape is very important on invitational hands (often a good minor suit fit upgrades your hand to bid a light 3NT or try a 5m game, or a 4-4 heart fit upgrades you to game on a pair of shapely hands without a lot of points). If you're going to bid 1♠-1NT-2♣(any min) and now you have to bid 2♥+ to break the relay, it seems like it will be very hard to find these fits. For example say opener has 5-1-4-3 and responder 2-3-5-3. Are you even going to find the diamond fit, or are you going to bid 1♠-1nt-2♣-2♠-pass? It's not even totally clear that you find 5-5 side-suit fits on this bidding (give opener 5-1-5-2 minimum, he might bid 3♦ over 2♠ but then again, can't responder be 2-4-2-5?).

My feeling is that relays really shine on slam decisions. Usually when opener is pretty tightly limited and responder can't game force opposite opener's minimum, you're not really making a slam decision. It'd be nice to have a system where relay is available for possible slam decisions, but where you're not trying to multiplex virtually every constructive response to the opening bid into some kind of relay (which makes you very vulnerable in competition, forces you to relay on game-only hands where natural bidding is often better, and loses you steps because you have to be able to bail out of the relay early when you don't even have game).

[hrothgar]

Last couple weeks at work have been quite hectic, so I haven’t had much time to post on this topic. However, I do have some (fairly) strong opinions.

First: I pretty much agree with a number of AWM comments. Relays are great for some things. For example, they are fantastic for intelligently investigating slams. However, relays also suffer from some very significant flaws. Most notably, relays are great for discovering shape; however, many pairs will have trouble investigating stoppers below 3N. There are ways to compensate for this (many pairs will build in a relay break to switch from asking about shape to asking about stoppers, but this increases memory load)

Second: I think that a lot of the discussion here puts the cart ahead of the horse. I view relays in much the same way that I view strong club openings. I hate opening 1C playing MOSCITO. I expect to score badly. (I try to make sure that my 1C opening is as effective as possible, but I still consider it a weak point of the system) Even so, I am very happy to play a strong club system because I think that the benefits from the light / limited openings more than compensate for the costs.

In much the same way, I don’t play relays over limited openings because I believe that relays are the best way to bid. Rather, I play use relays because I don’t want to waste 5+ responses to my 1H opening showing various game forcing hands. If I open 1H (showing 4+ Spades and ~ 9 – 14 HCP) my primary goals are:

Preemption with a fit and no game interest
Intelligent game investigate with a fit
Identifying the best part score with weak hands

Playing relay methods allows me to free up a whole lot of bidding space all sorts of useful stuff. I’m willing to sacrifice a bit with my game forcing hands (and game invites with no fit) to facilitate more useful parts of the system.

In short: I don’t really understand a conversation that starts with the assumption that you want to play relays and then asks what to do with the remaining bids. If you’re asking this question, you probably don’t want to play relay methods.

[awm]

There are obviously some other decisions/assumptions being made here. Let me try to outline:

I like five-card major methods. In competitive situations I think there is a big advantage to opening hands where you have a long suit (say a 5cM or 6cm) with a bid that shows substantial length in that suit! This helps partner figure out whether to raise and/or double the opponents and also to find a lead. There are many competitive sequences where a system that frequently opens weak four-card majors is at a substantial disadvantage. In general I like my opening structure of 1M 5+ intermediate, 2m 6+ intermediate, 1♦ as a "catch-all" without a 5M or 6m. Distinguishing "square" hands from hands with one or more long suits seems more effective to me than showing the longest major even if it is four cards with a much longer/stronger minor suit. I am also somewhat restricted by the ACBL mid-chart, which does not permit transfer-style opening bids.

Having decided on this opening structure, I want to know what the best follow-up structure is for the 1M openings (something like 10-14 hcp with 5+M). Even if we agree that 2/1 GF and a forcing 1NT response is a good approach in a "standard" system, these openings have a weaker lower-limit and a much more narrow range. It seems that devoting all three two-new-suit sequences to game forcing hands and relegating all non-fitting non-GF hands (as well as some fitting hands) to the 1NT response is not as good an idea here. Reasons for this: (1) GF hands are less frequent than opposite sounder openings (2) the 1NT response becomes much more frequent and wider ranging, especially if we don't routinely pass the opening on hands in the 6-8 hcp range (3) opener is way ahead of responder in describing the hand because of the narrow range, which makes assigning captaincy to responder and using a relay style of slam bidding more appealing.

I personally think that the "invitational" hand types are very important. These are hands in a point range something like 11-13, which while somewhat narrow is among the most common strengths to hold. A lot of IMPs are won and lost on game vs. partscore decisions, and it would be nice to get these right consistently. I don't think relays are the best way to make these decisions, especially relays that emphasize distinguishing max/min rather than shape at opener's second call.

Basically I can see four options here given that I don't like "INV+ relay": (1) continue to use 2/1 GF (2) switch to a method where 2/1 bids are INV+ which ups their frequency and helps with the INV hands, but makes bidding some slam hands more awkward (3) use 1NT as GF relay and the 2/1 bids with INV and weak hands (4) use 2♣ as GF relay and other calls to handle INV and weak hands.

It seems from this discussion that (3) is probably not a good option for me. All the methods proposed seem to do substantially worse on the INV and weak hand types than any of (1), (2), (4).

[hrothgar]

For what its worth, here's the response structure that I use over MOSCITO's 1♥ opening.

2/1s are Natural and non-forcing. They promise 5+ Cards in the bid suit and deny a fit.

A 2/1 promises real values. Opener is encouraged to double for penalties in non-fit auctions.

1♥ = 4+ Spades, might have a long suit

4♠ = to play
4♥ = Splinter
4♦ = Splinter
4♣ = Splinter
3N = To play
3♠ = Value raise
3♥ = fit jump (game invite with 6 hearts and 3 spades)
3♦ = fit jump ((game invite with 6 diamonds and 3 spades)
3♣ = fit jump (game invite with 6 clubs and 3 spades)
2N = Limit raise+ with 4 trump
2♠ = value raise, 3 card Spade support
2♥ = NNF (shows 5+ Hearts and ~ 7 - 11 HCP)
2♦ = NNF (shows 5+ Dimaonds and ~ 7 -11 HCP)
2♣ = NNF (shows 5+ Clubs and ~ 7 - 11 HCP)
1N = NNF
1S = Relay

I've given serious consideration to the following modification:

1H - 3C = natural and non-forcing with 6+ Clubs

1H - 2C = 4+ Clubs. Promises either a two suited hand or a three suted hand with short spades

[rbforster]

awm, on Jun 18 2008, 02:59 PM, said:

Basically I can see four options here given that I don't like "INV+ relay": (1) continue to use 2/1 GF (2) switch to a method where 2/1 bids are INV+ which ups their frequency and helps with the INV hands, but makes bidding some slam hands more awkward (3) use 1NT as GF relay and the 2/1 bids with INV and weak hands (4) use 2♣ as GF relay and other calls to handle INV and weak hands.

There's also the version where you use 1NT as weak or GF, and 2/1 for invitational hands (F1 or NF to taste, probably 2♣ is F1 since it might be balanced). This has several advantages -

1. you have a bid for constructive flattish hands without a fit - bid 1NT and either pass, or possibly correct to 2M. This will get you to better partials than just always passing these.

2. handling interference over 1M-1N is better but not great - here there's a big values gap between the constructive 6-9 hands and the 14+ GF hands. There's enough weak hands that crazy interference isn't indicated, and while you lose a forcing pass by opener you can still play either penalty or takeout doubles by responder (to taste) with X and all other bids being unambiguously GF. This is better than dealing with an invitational or maybe GF hand in the same situation where you'll basically be forced to underbid or overbid for lack of space.

3. you can handle invitational hands fairly well in a "Standard style" context, while not being absolutely forcing if opener wants to pass a dog minimum with 3+ cards in your 2/1 new suit.

The main disadvantages of this of course is that your GF relay suffers some and interference is a little more troublesome. Specifically -

1. you lose a forcing pass (FP) over interference. you seem to think this is important, but in my mind it only matters when a) they bid (not that common against me anyway), we had a hand for whichever of takeout or penalty responder can't show, c) we would have sat for the double, and d) it would have been right. I guess you can start using the FP to describe different hand types that aren't going to penalize too, but I don't think this is the huge loss you think it is (and the field will have your problem too most likely).

2. your cheapest relay bid might be 2M in a purely 1NT GF relay system, but now you might want to allow a preference with a flattish constructive hand. So your relays are up a step sometimes. You can fix this by only bidding 1N on flattish hands that intend to pass rather than pass-or-preference, although I'm not sure this is better for finding good partials.

3. since responder might have a constructive flattish hand, you'll be constrained to have opener's rebids in response to 1NT to be more-or-less natural. High jump bids may be hard to use if they can't tolerate pushing that high when responder is on the weaker hand. You won't be able to use transfers which might have helped on slam sequences.

4. you can't sign off in with a weakish long suit (although I'm not sure you could have done this in a 1NT GF system either, but you can in 2/1). This is much less of a loss playing a limited bid system than it is in 2/1 since in 2/1 opener can have a really big hand and keeping the bidding alive with a random 5-6 count and a long suit might let opener get to game. This pretty much never happens opposite a 10-14 opener, so the only loss in passing is that 1M might be a worse contract than 2X or 3X in your long suit. This might happen sometimes, but the opponents will often balance (since they've got at least half the deck) and then you're off the hook.

Still, I think this is a decent alternative and pretty straightforward. If your priorities are mostly for relay slam bidding and good partials, I think this is a good compromise. You probably get to the field partials, and while your relay isn't always as efficient as the pure GF version, you'll still be starting your relay so low that I expect you can reasonably investigate for slam effectively.

[benlessard]

Quote

I know a lot of people here love their 1NT response that shows any of three or more totally different hand types some of which are weak and some of which are strong. These kinds of bids seem nice when you're designing a bidding system in isolation, or when you play against opponents who always pass. But in real bridge, opponents tend to bid, especially when you are making nebulous calls that give virtually no information about shape or values. Even simple interference seems to be tough to deal with using these kinds of methods: take 1♥-P-1NT-2♠

. The more type of hands you have in the relay the more precise the other bids will be so in the end the other bids will be less vulnerable to preemption. The important thing is that the nature of the bids that are in the relay are compatible. By compatible i dont mean similar. A michael cue bids with 2 range that dont touch each other (4-8, 16+) will play better then a wide range (6-14). Polish club 12-14 balanced (or 18+) will play better then a 14+ big club (both openings will have similar frequency) Another example is Multi-2D. If you put lightish (5-7) balanced hand in the relay you will have no problem at all with most interference. Opener with a monster will GF anyway. With a weak or medium hand he doesnt need to bid unless extreme shapes and want to sacrifice. This is because a weak balanced hand is compatible with a GF hands.

However if you put invitationnal hands in 1Nt youre going to have some problems. Invitationnal hands are not compatible with GF hands. Take for example this setup.

1S----2S (6-9)
1S----2H (limit or Gf with a S fit)

(here if opener make a splinter whatever the follow up it will be unclear if the responder is limit or GF.)

If you reverse the 2S regular raise and 2H

1S----2H (GF raise or 6-9)
1S----2S (limit)

Then you wont have problems. If partner is making a slam or bidding game facing a possible 6-9 and i was GF there will be no problems.

There is no doubt in my mind that 1Nt as GF only is inferior and the only reason for playing it is legal matters.

Quote

In much the same way, I don’t play relays over limited openings because I believe that relays are the best way to bid. Rather, I play use relays because I don’t want to waste 5+ responses to my 1H opening showing various game forcing hands.

Then my guess is that your relays are not optimal. If i have a GF hand and a partner showed a fairly precise opening I dont want to show anything, i want to listen to partner. The problem i see with many system here is that they come from a system with a "natural core" in wich we add some relays to improve for slam bidding. Instead of starting from the natural system and adding some relays, start from a already made relays system and try to naturalized it to your taste.

I agree with 95% of -- rob f -- last post

[hrothgar]

benlessard, on Jun 19 2008, 04:50 PM, said:

Quote

In much the same way, I don’t play relays over limited openings because I believe that relays are the best way to bid. Rather, I play use relays because I don’t want to waste 5+ responses to my 1H opening showing various game forcing hands.

Then my guess is that your relays are not optimal. If i have a GF hand and a partner showed a fairly precise opening I dont want to show anything, i want to listen to partner.

Please...

Do you genuine believe that relay methods are the best way to make an intelligent decision between 3N and 4M?

Relay's are great for exploring shape.
I love knowing partne's precise shape by the time he has bid 3♦ or 3♥.

However, precise information about shape is useless if I need to know about stoppers. The difference between xxx and Kxx can make an enomous difference in
making an intelligent decision whether or not I want to play 3NT or not. This is where relay methods typically fail.

I readily admit... It's possible to build in relay breaks that allow the partnership to switch from showing shape to showing stoppers. (If you look at my first post, I already commented on this). However, I don't go arround deluding myself that this is going to work as effectively as natural bidding.

For what its worth, I've had some extensive discussion on this topic with some good relay pairs. They pretty much agree that they prefer natural bidding to explore the best game contracts.

Paul Martson is quite specific on this topic. His philosophy is as follows

Relays for slam exploration
Natural bidding to explore the best game
Bashing for part score contracts

[rbforster]

I'm not sure I entirely agree with Richard's post that relay can't find you stoppers if you do a careful job of designing good relay breaks, but admittedly most people don't get that far into designing their system, just get some shape relays and always relay as responder. I haven't fully explored the best way to do all these stopper asks, but I agree their important/useful. I'm not sure exactly how useful since unlucky leads vs an unknown relayer's shape and non-perfect defense often sees through many a dubious 3N contract and of course most of the time it's a good contract when you don't have a fit elsewhere so I'm not sure if this really happens enough to be a serious consideration in your overall system design.

If it turns out that stopper exploration and/or natural bidding is important to you and is incompatible with your 1NT GF relay methods, you could always use

1M-?

1N GF with extras and slam interest (17+), or flattish constructive (6-10)
2X F1, invitational or min GF values "Standard style" (11-16)

This way on invites and minimum GF hands you get your natural bidding and can explore for 3N more easily, and use relays only when you've got extra values (say 17+ or so). Now at least if you find out you want to play 3N your extra values will probably make it a favorite to make and if opener's shape makes 1N relayer's hand look unsuitable for NT maybe you'll have the strength to make 4M on a 4-3 or 5m.

[benlessard]

Quote

Do you genuine believe that relay methods are the best way to make an intelligent decision between 3N and 4M

IMO good relay is better or equal to natural bidding for COG in the majority of cases.

Noble showed a hand that is usually tough for our system.

AKQTx
QJT
J
xxxx

xx
Axxx
xx
AKxxx

1S-----1Nt (relay)
2S -----???

2S showing 5S + 4C (12-14) wich is our worst hand.

now
pass= i think 2S is better then 3C
2NT GF further relay / all slammish hands
3C to play
3D relay break asking for a H stop
3H relay break asking for a D stop.

If partner is 6??4 then 4S will probably be the best game.
IF partner is 5224 then i need 5 clubs trick a D stopper and 2 trick elsewhere for 3Nt and if partner deosnt have a stopper 5C is probably hopeless.
if partner is 5??5 then 5C is probably the best spot.

But if you bid 2Nt to know more about shape you might be in the dark.
IMO the best way is to break the relay

----------3H (relay break asking for a D stopper)
3S (5 good spades or 6 weakish spades doesnt deny a D stopp)----------3Nt not enough to raise S
4H (good H fragment D stiff denies a 5 club suit/beefy 4 card suit)-----------???

Now here you are pretty well placed

If the H would have been better (AT9x) you would have played 4H
You can bid 4S showing a weak doubleton and suggesting 5C or you can just bid a plain 5C.

[awm]

Let me compare the following two similar ideas in response to 1♠:

(1) 1NT is 17+ or 6-9; 2♣/2♦/2♥ are 11-16 natural and forcing

(2) 1NT is 6-9 or 11-13 with ♣; 2♣ is 14+ clubs or 17+ any; 2♦/2♥ are 11-16 natural.

There are four hand types where we will see a major difference in these structures:

A: Hands in the 6-9 range. Here structure (2) will be superior, because you have the option to bid your own diamond suit over 1♠-1NT-2♣ or heart suit over 1♠-1NT-2♦ (these would be the relay in the first structure). It is also likely that opener's 2♣ rebid is "more natural" in structure (2) because you don't need to pack a relay structure in. And you can create useful jump shifts for opener on shapely hands in structure (2) to find the occasional game opposite 6-9, which is harder to do in structure (1) because you have to make room for relay. Also, you may be able to introduce a long suit when opponents intervene without gadgets or ambiguity (i.e. if opponents overcall 2♥ over 1NT and you back in with 3♦ it is definitely weak).

B: Hands in the 11-13 range with 4+♣. Here structure (2) should be superior as well, because opener's rebid is more descriptive. In particular 1♠-1NT-2♠ is definitely six (or more) in structure (2) whereas 1♠-2♣-2♠ might have to include 5(332) hands and/or hands with 4♣ that are too weak to force game in structure (1). Obviously you can jigger your rebids in structure (1) however you like, but the point is that you have lost a step for opener to describe so you will probably end up doing slightly worse than a similarly optimized structure for (2).

C: Hands in the 14-16 range with 4+♣. Again structure (2) should be superior. Here you have the option to relay when this seems best, or to break relays in a natural game-forcing way (i.e. by bidding 3♣ if it's not the step) when that seems more effective. In structure (1) you are locked in to "natural bidding" and may have to take some awkward actions to force if a lot of sequences show the 11-13 range (i.e. 1♠-2♣-2♠-3♣ may be NF).

D: Hands in the 17+ range. When opponents interfere with the auction structure (2) should still win (assuming the same interference in either case). In fact you can simply continue to relay if opponents double or overcall 2♦/2♥ having lost nothing. And you have forcing pass available in other auctions to help clarify the hand (for example you could play that a direct bid by opener after interference shows 0-1 cards in the opposing suit, double shows two cards, pass shows 3+). When opponents don't intervene is the only case where structure (1) is a win. But it doesn't win as much as you might think. You gain a bunch of sequences by starting relays with 1NT (if you are measuring sequences ending in 3♥/♠/NT then you have 89 instead of 55). But then you lose some of these because 2♠ is not the relay over 1♠-1NT-2♥, and you lose some because you can't make efficient use of opener's jump shifts (because of the 6-9 possibility), and then you lose some more because you can't pack enough hands into opener's 2♣ rebid and still have it be fairly "natural." I suspect you'll end up with more like 65-70 sequences rather than 89. And even to do this may require rewriting your relay structure (i.e. it won't strongly resemble whatever structure you use over 1♣ strong) thereby greatly increasing memory load. Another point is that you can "relay break" after 1♠-2♣-2X in method (2) whereas most such breaks in method (1) presumably show the weak option.

So I prefer structure (2) from these possibilities. How many 17+ hands do you get anyway?

[benlessard]

Quote

(1) 1NT is 17+ or 6-9; 2♣/2♦/2♥ are 11-16 natural and forcing

to wait for 17+ to relay say that you should play a simple transfer responses instead.

Ive never saw a relay system that allow for LOB that satisfy me because the requirement for GF are just too harsh.

(2) 1NT is 6-9 or 11-13 with ♣; 2♣ is 14+ clubs or 17+ any; 2♦/2♥ are 11-16 natural.
A simple 2/1 structure with 2C as artificial seems better then this. Again the requirement to relay is too high to my taste.

My philosophy is this loosen the requirement for a weak 2 & for 1Nt openings thighten the requirement for a 1M opening so that the response after 1M work like magic. Im not a great believer in LOB so my judgement is tainted on that subject.

[awm]

benlessard, on Jun 19 2008, 08:39 PM, said:

My philosophy is this loosen the requirement for a weak 2 & for 1Nt openings thighten the requirement for a 1M opening so that the response after 1M work like magic. Im not a great believer in LOB so my judgement is tainted on that subject.

My post was more a response to Rob Forster's suggestion than yours. I will note that 14 hcp is probably enough for a game force opposite 10-14.

Structures like the one Ben Lessard suggests are very nice when opponents are passing. But I doubt in competitive auctions that they work so well. You will often have auctions such as:

1♠ (some split range, natural)
Pass
1NT (some split range)
3♣ overcall
....

Now presumably if opener bids again it shows the higher of the two ranges (otherwise it is not clear how you will ever distinguish now that opponents have crashed the auction). But this means you have to pass with some shapely hands where you might like to bid in order to avoid showing the upper range.

So the auction passes back to responder. He is in a similar situation; presumably bidding here shows the game forcing 1NT range. So you could have an auction where opener is 5-5 in the majors with a minimum and responder is 2-5-3-3 with the weaker option and you pass out 3♣? Or am I missing something here?