[steve2005]

deleted duplicate post

[steve2005]

glen, on 2013-October-06, 14:55, said:

so enlighten me you have 15 hcp 5♠3♥2♦3♣
you open 1NT responder with balanced 5-6 hcp and 3♠ must surely pass and you miss 5-3 fit

but many world class partnerships open 1NT with a five card major, how can they not see your viewpoint?

There is a big difference between opening 1N with a 5-card major which is acceptable (though many don't) and opening 1♣ on a 5 card major.
[edit] with 1♣ your more likely to face interference than 1N, so opening 1M rather than 1♣ puts you in better position to handle interference [/edit]

with more points from a 1N opening will have better chance at making 1N

[glen]

steve2005, on 2013-October-06, 15:20, said:

There is a big difference between opening 1N with a 5-card major which is acceptable (though many don't) and opening 1♣ on a 5 card major

with more points from a 1N opening will have better chance at making 1N

so enlighten me Fantoni has 13 hcp 5♠3♥2♦3♣
he opens 1NT and Nunes, responder, with balanced 5-6 hcp and 3♠ must surely pass and they miss 5-3 fit

they don't have "more points from a 1N opening"

[awm]

glen, on 2013-October-06, 16:00, said:

so enlighten me Fantoni has 13 hcp 5♠3♥2♦3♣
he opens 1NT and Nunes, responder, with balanced 5-6 hcp and 3♠ must surely pass and they miss 5-3 fit

they don't have "more points from a 1N opening"

It does seem to be more effective to open strong notrump with a 5-card major than weak notrump. There are a couple basic reasons for this:

1. Opening 1M is more important when the strength is fairly even between our side and theirs, because we are more likely to struggle in 1NT, and because the auction is more likely to be competitive. The strength being evenly divided happens more often when the 1NT opening is weaker.
2. Opening 1NT allows opener to declare more of the game contracts, which is advantageous when opener tends to be the stronger hand but more randomizing when the hands are equal in value.

I've seen a lot of pairs who open weak notrump but never (or almost never) have a five-card major, whereas the strong notrump pairs very much tend the other way.

[rhm]

mgoetze, on 2013-October-05, 10:46, said:

I've always meant to finish this writeup of the system I stole from karlson and others before publishing it but I haven't made any serious progress in the last two years so I may as well show what I have so far...

http://bridge.mgoetze.net/21twalsh.pdf

It's horribly incomplete but oh well.

When designing a system you want to split the load of the hands over the available bids, but since you can have more sequences over lower bids, subject to no interference, you like to put more hands to lower bids.
This allows for maximum differentiation and information exchange in uncontested auctions.

If you look at the proposed system you find a very high load for 1♣ and a extremely low frequency for the unbalanced 1♦.
As far as 1♦ has less number of hands than 1♣ it is fine, but that 1♦ occurs less frequently than 1M or 1NT does not make sense from a general point of exchanging maximum information.
Mgoetze write up claims the relative frequency of hands opened would be:

1♣: 37.9%
1♦:9.8%
1♥:14,8%
1♠:19.9%
1NT:14.6%

The remainder being opened with higher bids.
I refuse to believe that this is an optimal system, no matter how much Bridge logic has gone into the split up of the bids, which of course is important.

FromageGB system looks even worse.

Rainer Herrmann

[fromageGB]

We seem to have degenerated into a completely different topic. Can anyone add to the discussion on useful meanings for the 1♠ 1NT 2♣ 2♦ replies to the 1♣ open? This seems to be the area of major variability in the use of transfer walsh.

What works best?
What should be adopted as "standard" for the purposes of teaching it to newcomers?
(They may of course be two different answers.)

[fromageGB]

rhm, on 2013-October-07, 03:54, said:

When designing a system you want to split the load of the hands over the available bids, but since you can have more sequences over lower bids, subject to no interference, you like to put more hands to lower bids.
This allows for maximum differentiation and information exchange in uncontested auctions.
...
I refuse to believe that this (the Goetze method) is an optimal system, no matter how much Bridge logic has gone into the split up of the bids, which of course is important. FromageGB system looks even worse.

Your approach to design is different to mine. For your theoretically better approach to work, you would have a relay system where the actual bids bore little resemblance to the suits held. While you can do this with game hands, where there is room for artificial shenanigans, it cannot work at low partscores. You therefore need an artificial strong 1♣ or similar.

My approach is to have predominantly natural bids with a very limited artificiality, giving preference to finding the major fits. In terms of frequency, if you want a comparison, taking a 12-14 range so that you ignore higher bids, my hierarchy appears to work out at
1♣ : 51.8%
1♦ : 13.3%
1♥ : 16.0%
1♠ : 17.7%
with the inevitable approximation rounding errors. However, I don't think the percentage matters greatly.

[rhm]

fromageGB, on 2013-October-07, 05:00, said:

Your approach to design is different to mine. For your theoretically better approach to work, you would have a relay system where the actual bids bore little resemblance to the suits held. While you can do this with game hands, where there is room for artificial shenanigans, it cannot work at low partscores. You therefore need an artificial strong 1♣ or similar.

My approach is to have predominantly natural bids with a very limited artificiality, giving preference to finding the major fits. In terms of frequency, if you want a comparison, taking a 12-14 range so that you ignore higher bids, my hierarchy appears to work out at
1♣ : 51.8%
1♦ : 13.3%
1♥ : 16.0%
1♠ : 17.7%
with the inevitable approximation rounding errors. However, I don't think the percentage matters greatly.

You are wrong.
Relay systems make better use of lower bids not worse.
Therefor in a natural system with little artificiality putting almost 4 times more hands into 1♣ than 1♦ makes even less sense.
What is predominantly natural can be argued.
Transfer responses are hardly natural, in fact few of the response to 1♣ are. Accordingly when you open more than half your hands with 1♣ at least the frequency of artificial bids is high in this system.
In fact because non natural transfer responses make more sense over 1♣ than 1♦ and people seem to desire that. this seems to be a major reason for the skew.

Rainer Herrmann

[mgoetze]

fromageGB, on 2013-October-07, 04:02, said:

We seem to have degenerated into a completely different topic. Can anyone add to the discussion on useful meanings for the 1♠ 1NT 2♣ 2♦ replies to the 1♣ open? This seems to be the area of major variability in the use of transfer walsh.

What works best?
What should be adopted as "standard" for the purposes of teaching it to newcomers?
(They may of course be two different answers.)

Well, uh, I did contribute a whole PDF on the subject, so if you want me to discuss it further you'll have to comment on that first.

[mgoetze]

rhm, on 2013-October-07, 03:54, said:

If you look at the proposed system you find a very high load for 1♣ and a extremely low frequency for the unbalanced 1♦.
As far as 1♦ has less number of hands than 1♣ it is fine, but that 1♦ occurs less frequently than 1M or 1NT does not make sense from a general point of exchanging maximum information.
Mgoetze write up claims the relative frequency of hands opened would be:

1♣: 37.9%
1♦:9.8%
1♥:14,8%
1♠:19.9%
1NT:14.6%

The remainder being opened with higher bids.
I refuse to believe that this is an optimal system, no matter how much Bridge logic has gone into the split up of the bids, which of course is important.

Well, you see, at least I'm honest and up-front about it. Anyway this has bothered my slightly from a theoretical standpoint too, but from a practical standpoint I can't see many obvious improvements without changing too much. Consider just 1♠ and 1♥. The bids are defined symmetrically, and the 5% difference is due solely to opening hands with 5-5 or 6-6 in the majors 1♠ rather than 1♥. I'm sure I don't need to explain to you why one would open these 1♠ rather than 1♥ in a natural system. So if you want to play natural and with decreasing frequencies as the bids get higher, you'll have to start by doing something like opening 4-card heart suits 1♥...

Since I'm not playing a bunch of relays anyway I'd rather keep it simple.

[Zelandakh]

RunemPard, on 2013-October-03, 09:50, said:

[1]Does putting all balanced 5 card ♦♥♠ hands into 1C hurt at any scoring?
[2]How much does the unbalanced only openings help?
[3]If I do put all 5 card balanced hands into 1♣, how do we find them?
[4]Should we reply 1M after partner bids 1♦/1♥ with 2-3 and a 12-14, or only 3 card support?
[5]If we do remove all balanced hands, what are good uses of the 2NT rebid by opener in common situations?
[6]Good uses of 1♣-1♠, 1♣-1NT, and 1♣-2x

1. Yes, it hurts sometimes. You can bury a fit or let the opponents in (especially when our major is spades) or provide a roadmap for the defence. You also need to modify other parts of the 1♣ structure to get the most from this, which is not to everyone's taste.
2. I like them a lot, although I have only played unbalanced 1M openings in the context of limited openings. I think the benefits outweigh the losses from the 1♣ opening. Within a natural structure it is possibly better to opem 5♥332 with 1♣ and 5♠332 with 1♠, although that makes the weightings Rainer has brought up look even worse.
3. There are lots of options here. The simplest is just to play a Puppet Stayman structure when Opener shows a balanced hand with the rebid, although there are better possibilities with a TWalsh structure.
4. I think the 1M rebid with a weak balanced hand is better than any 3 cards. I put the latter down to early experimentation. I have another option that I am (loosely) working on (completing with <3) but have not had time to fully work out any details. I will post it here in a few years' time...
5. The obvious use is as a strong 3 card raise of partner's suit. Some use it as an artificial GF.
6. I think Andy's use of 1♠ - a weak hand or a GF with primary diamonds is about as standard as it comes and his structure looks generally sound (as you would expect!). There are other possibilties though, for example 1NT can be a natural invite with 2♣ inverted and 2♦ a natural invite. A similar idea to 4 suit transfers over 1NT is to switch the 1♠ response to showing clubs (weak or GF) with a 1NT response showing diamonds (weak or GF). Now the 2m responses are available as natural invites, although 2♣ can be 4 card support (unless you use Andy's 2♥ response too) which makes it harder to stop here than in a structure that contains a 'proper' invitational 1NT. I seem to recall that PK had a pretty nice TWalsh structure too - perhaps you can find a link if he does not post a reply.

[PhilKing]

Zelandakh, on 2013-October-07, 16:45, said:

I seem to recall that PK had a pretty nice TWalsh structure too - perhaps you can find a link if he does not post a reply.

In a nutshell:

1. Only one balanced range goes through 1♣ (with 18-20 open 1♦). It is important not to overload 1♣ - this enables responder to pass with balanced garbage with no long major rather than put his head on the block for fear of a 24/23 fit opposite the stronger balanced hand.

2. Play completing the transfer as forcing. Allowing the auction to stop in 1M is useless and takes away the huge potential of having an extra layer of bidding.

3. Use the extra space to handle various awkward "unbiddable" hands, and also to tighten up our raising ranges.

For instance, after 1♣-1♥:

1♠ = F1 various:

a) 5♣4♦ any range
b) 3 spades in an unbalanced good 3-card raise (circa 14-16 points, usually 5431)
c) a good club rebid, not quite suitable for 3♣, but less than a game force (a 2NT rebid with 6 clubs, a 3♣ rebid with ample strength but insufficient points, and good 3M6♣ hands)

With limited values responder continues as if facing clubs and diamonds, which makes those hands a doddle. The other hands reveal as appropriate. With game values, relay with 2♥.

Higher bids are thus revolutionised:

1NT/2♣ = natural (2♣ can be 5♣ 4♥, else 6+♣)
2♦ = a reverse in hearts. This enables on to stop in 2♥ facing dreck with three hearts and short clubs and occasionally responder floats with 4♠6♦
2♥ = a good raise to exactly 2♠ with exactly four trumps (don't overload this bid, since you will end up leaking information on standard hands)
2♠ = a very bad raise - min weak NT with four, but also 5431 mins with 3 trumps
2NT = FG with 6+ clubs
higher = relatively standard

This is not a guide - it is just an indication of what you can do by playing completing as F1.

[fromageGB]

mgoetze, on 2013-October-07, 12:18, said:

Well, uh, I did contribute a whole PDF on the subject, so if you want me to discuss it further you'll have to comment on that first.

I read it of course, and am happy to give you my comments of how it seems to me. First, a quick summary for convenience: responses to 1♣
1♦ = hearts
1♥ = spades
1♠ = clubs
1NT = diamonds, weak. All the above are 4+ so all hands basically covered.
2♣ = long diamonds weak, or GF
2♦ = long diamonds invitational
2♥/♠ = both minors, bidding a shortage
2NT = natural

It covers the minors well. It is very poor with less than game major hands, as is inevitable when there is only one bid to show a major and no bids for both majors. If the opponents enter the bidding, the other major is lost. Even with no opponents, an invitational responder with just one major cannot distinguish whether he has 5 or 6. If he is invitational with both he cannot show the length of the second, and will possibly end up in the wrong one, or at the 3 level. If he is weak with both majors and no opponents he appears not to be able to distinguish between 44 and 45, or 54 and 55.

The other methods proposed so far are from paulg :
1♠ = 5-12 hcp with no major
1NT = relay GF
2♣ = diamonds GF
2♦ = 6 card major, 3-8
2♥ = 5♠ 4+ ♥, up to 8
2♠ = 5♠ 4+ ♥,9-11
2NT = 6 clubs, up to 7
3♣ = 6 clubs, 8-10

from me :
1♠ = relay, any strength balanced or minor
1NT = 54 or 45 majors, weak
2♣ = 44 majors invitational +
2♦ = 55 majors any strength
2♥/♠ = 6 card up to 8 hcp
2NT = 55 minors any strength
3m = preempt 6+

gnasher:
1♠ = no major weak, or diamonds GF
1NT = clubs invitational +
2♣ = diamonds inv
2♦ = long major weak, or no major slam try
2♥ = no major invitational +
2♠ = weak both minors
2NT = clubs weak.

karlson: as mgoetze

steve2005:
1♠ = relay, including balanced, or weak clubs, or weak diamonds, or inv diamonds
1NT = 54 or 45 majors, weak

PhilKing:
1♠ = usually 5-12 no major but can be 13+ if (2)(3)44 or 33(4)(3)
1NT = FG clubs
2♣ = FG diamonds
2♦ = multi
2♥ = 7-10 54MM or 6-9 55M
2♠ = 7-9 6 clubs
2NT = 55mm <11
3♣ = weak

jallerton:
1♦ = 4+ hearts
1♥ = 4+ spades
1♠ = 4+ diamonds
higher bids as you wish

Any others - particularly from people who regularly play twalsh?

This post has been edited by fromageGB: 2013-October-10, 06:47

[paulg]

Technically my 1♣-1♠ is not a relay, it is 5-12 with no major.

We also play:

2♦ - 6-card major, 3-8 HCP
2♥ - 5♠, 4+♥, weak
2♠ - 5♠, 4+♥, 9-11 HCP
2NT - weak with six clubs
3♣ - six clubs, 8-10 HCP

[PhilKing]

paulg, on 2013-October-09, 05:45, said:

Technically my 1♣-1♠ is not a relay, it is 5-12 with no major.

We also play:

2♦ - 6-card major, 3-8 HCP
2♥ - 5♠, 4+♥, weak
2♠ - 5♠, 4+♥, 9-11 HCP
2NT - weak with six clubs
3♣ - six clubs, 8-10 HCP

I play more or less the same structure of bids as Paul from 1♠ up:

1♠ = usually 5-12 no major but can be 13+ if (2)(3)44 or 33(4)(3)
1NT = FG clubs
2♣ = FG diamonds
2♦ = multi
2♥ = 7-10 54MM or 6-9 55M
2♠ = 7-9 6 clubs
2NT = 55mm <11
3♣ = weak

I like the idea of being able to respond 1NT with the 54M hand, but in practice, 2♥ has worked fine.

I would add that those who play that the FG minor-suit response denies a major are missing out - there is no problem bidding the longer suit first.

[jallerton]

PhilKing, on 2013-October-08, 05:38, said:

In a nutshell:

1. Only one balanced range goes through 1♣ (with 18-20 open 1♦). It is important not to overload 1♣ - this enables responder to pass with balanced garbage with no long major rather than put his head on the block for fear of a 24/23 fit opposite the stronger balanced hand.

2. Play completing the transfer as forcing. Allowing the auction to stop in 1M is useless and takes away the huge potential of having an extra layer of bidding.

3. Use the extra space to handle various awkward "unbiddable" hands, and also to tighten up our raising ranges.

Re point 1, don't you just move the same issue into your 1♦ opening?
Re point 2, I can recall a number of good results from being able to stop in 1M opposite a weak NT hand, especially when our side is vulnerable. I think your suggestion works better with weak NT openings, as then a 1NT rebid can show 15-17 balanced.

[jallerton]

fromageGB, on 2013-October-09, 03:47, said:

Any others - particularly from people who regularly play twalsh?

1♦=4+♥
1♥=4+♠
1♠=4+♦

The practical advantage of this relatively simple structure is that you can use responses of 1NT+ to mean whatever you used to play when 1♦/♥/♠ responses were natural.

[PhilKing]

jallerton, on 2013-October-09, 13:27, said:

Re point 1, don't you just move the same issue into your 1♦ opening?
Re point 2, I can recall a number of good results from being able to stop in 1M opposite a weak NT hand, especially when our side is vulnerable. I think your suggestion works better with weak NT openings, as then a 1NT rebid can show 15-17 balanced.

Point 1. No.

a) Passing 1♣ allows you to play any 1-level contract. When you pass with weakness opposite a likely weak no trump (the alternative is that partner has, er, clubs), you are freerolling their plus score. That is not true opposite the stronger balanced hand, where you need to escape from the non fit.

b) Bidding on rubbish is fine opposite 1♦ when pard has the balanced strong hand, whereas it is a suicide job opposite the weak balanced hand. When pard is unbalanced, you just don't get hammered - all the time I have been clobbered are when pard responds to 1♣ for no reason whatsoever.

Point 2, it is not my experience, but that may be because I can't stop. Are you sure this does not just apply versus plankton? Anyway, I like the idea of 1NT rebid as 15-17 nv, since I can easily incorporate that into my raise to two structure when we find a fit.

[fromageGB]

PhilKing, on 2013-October-09, 17:09, said:

Are you sure this (it's good to be able to pass a completed major transfer) does not just apply versus plankton? Anyway, I like the idea of 1NT rebid as 15-17 nv, since I can easily incorporate that into my raise to two structure when we find a fit.

If you have a weak NT and sometimes complete the transfer with a 15-17 balanced hand, then I guess you do need to have it forcing. For me completing the transfer is 14 max, so passing with any 5 card major less than invitational works fine. Sometimes it may resemble plankton, sometimes it may be a 10 count, but either way it denies game. In this style, you assume that a 5-2 or 5-3 major fit is worth more MP than playing 1NT, which it commonly is when the long suit is in the weaker hand. If 4th seat protects, you know where you are : 2M from opener is 3, and if he passes, responder is in position to double for penalty or pass accordingly. Or bid the major or a 5 card minor etc.

[PhilKing]

fromageGB, on 2013-October-10, 06:27, said:

If you have a weak NT and sometimes complete the transfer with a 15-17 balanced hand, then I guess you do need to have it forcing. For me completing the transfer is 14 max, so passing with any 5 card major less than invitational works fine. Sometimes it may resemble plankton, sometimes it may be a 10 count, but either way it denies game. In this style, you assume that a 5-2 or 5-3 major fit is worth more MP than playing 1NT, which it commonly is when the long suit is in the weaker hand. If 4th seat protects, you know where you are : 2M from opener is 3, and if he passes, responder is in position to double for penalty or pass accordingly. Or bid the major or a 5 card minor etc.

Yes, I understand that. You have given up the benefit of having two layers of rebid (via a multi-way forcing completion) for the ability to stop in a triumphant contract of 1M.

Anyway, 1M is forcing for me because it is consistent with a 21 count with clubs and diamonds.