[Hawkster1]

SAYC, 5-Card Major Opening. Partner opens 1 ♣ or ♦. I respond 1 ♥. Partner supports by bidding 2 ♥. (No interference.)

In my opinion, the 2♥ by partner should promise 4-card support. However, I have observed that many, many players on this board will make that bid with only 3-card support. Sometimes, they even fail to bid a 4-card ♠ suite and support hearts instead with only 3 hearts.

When I have asked them not to support me with only 3 cards in that bidding sequence, they reply that their bidding is standard.

Certainly it is not standard in any system that I have knowledge of.

Is my bidding hopelessly out of date? If so, please explain the logic of supporting with 3 cards when my major suite response only promises 4 cards.

[Hanoi5]

I think supporting with 3 cards is an exception rather than a rule. I use it when I can't bid the other Major, or it looks better than rebidding NT or repeating your original suit. I do have a gadget for responder to ask how the support was.

[Fluffy]

Hawkster1, on 2014-March-17, 15:00, said:

SAYC, 5-Card Major Opening. Partner opens 1 ♣ or ♦. I respond 1 ♥. Partner supports by bidding 2 ♥. (No interference.)

In my opinion, the 2♥ by partner should promise 4-card support. However, I have observed that many, many players on this board will make that bid with only 3-card support. Sometimes, they even fail to bid a 4-card ♠ suite and support hearts instead with only 3 hearts.

When I have asked them not to support me with only 3 cards in that bidding sequence, they reply that their bidding is standard.

Certainly it is not standard in any system that I have knowledge of.

Is my bidding hopelessly out of date? If so, please explain the logic of supporting with 3 cards when my major suite response only promises 4 cards.

As far as I know supporting with 3 cards is standard in america, and it is also extensively done in England.

In Europe 2M raise shows 4 cards.

[steve2005]

if your playing SAYC there is no gadget for checking back for 5-3 fit on balanced hands. this is something most people have nowadays. Most people I know around here Canada and on BBO raise on 4 but missing a 5-3 fit is why some raise on 3.

In my opinion you should raise on 3 if 5-4 in minors 3♥ 1♠ a 2♥ raise is very practical and should play well in a 4-3 fit. Even with 2=3=4=4 same may hold true

raising with 4♠ 3♥ and therefore a balanced doesn't seem right, you can always bid 1♠ and show delayed support on next round

Opener with a balanced hand and only 4♥ that wants to go forward can bid NT which should keep you out of bad 4-3 fits

[1eyedjack]

I think that the problem here is that 3 card support is something that you should generally be reluctant to conceal, especially if unbalanced. By the same token, if you have (only) 5 cards in a suit in which you have already promised 4+ cards then you would rebid that suit to show the extra length only under duress.

Given that you desire to show support, the next question is when to do so. Specifically, whether immediately or delayed. Delayed support may suggest shorter trumps, consistent with 3 cards. But it might also suggest extra values, for example if you show delayed support in a nonforcing situation where you might have passed in a playable partial fit. If you reserve delayed support to show the weaker hand type you store up problems for yourself on those hands where you have 3 card support + extra values.

Some of us prefer a "get out quick" strategy with weak hands. Find a playable spot as fast as possible and give opponents the last guess on minimal data. When deciding whether to balance they will guess wrongly more frequently if their assessment of your trump fit has a higher variance.

I would not go so far as to say that either method is so standard as worthy of undiscussed assumption in a pickup. But both styles have substantial followings.

[MrAce]

Hawkster1, on 2014-March-17, 15:00, said:

Is my bidding hopelessly out of date? If so, please explain the logic of supporting with 3 cards when my major suite response only promises 4 cards.

No, your bidding is not hopelessly out of date. Imo your partners, since they accepted to play with you, should respect your demand. After all what you are asking is not absurd at all to begin with and any player who wants to win also knows their pd performs better when he/she plays something they are comfortable with.

Having said that, raising with 3 cards is something i also do, but i do not abuse it. I do not like doing it with 4432 or 5332 type hands. I do it with usually a stiff and 3 card support AND when i believe raising now makes the future of auction easier for me.

For example bidding 2 suits as opener, w/o reverse, and then supporting pd's major over a simple preference shows extras. If i do not have those extras and not raising pd's major immediately could cause problem in our auction, then i raise. This does not mean i would do it auto with all hands that holds 3 card support and a stiff or void. I try to predict the most likely rebids and which bid i make now will place me better later.

[gnasher]

Fluffy, on 2014-March-17, 15:14, said:

As far as I know supporting with 3 cards is standard in america, and it is also extensively done in England.

In Europe 2M raise shows 4 cards.

Unless there's been some very unusual tectonic movement recently, England is still in Europe.

[mike777]

I think meckwell tends to raise less often on three cards and tends to rebid 1nt with more minimum type hands than others. So the methods here are playable.

[the hog]

I follow the same line of thinking as Mr Ace. Raise with 3 card support and a stiff. Very rarely do so with a 5332 shape. I think you need to ask yourself why you don't want to raise on 3 cards. Are you scared of playing Moysian fits?

[Stephen Tu]

Hawkster1, on 2014-March-17, 15:00, said:

If so, please explain the logic of supporting with 3 cards when my major suite response only promises 4 cards.

As mentioned it's a style thing with different regional preferences. As for the logic behind raising with 3:

- Sometimes it's just really the most palatable option. Like with a 3415 minimum after 1c-1s. Two hearts is a reverse and shows much more strength. 2 clubs is supposed to imply 6, and you might play some silly 5-1 or 5-2 club partial instead of a nice 5-3 spade fit.

- Responder can have shapely hands with 5/6 cd M that are marginal invites, but only if there is a fit (light on HCP). Knowing at least a 3 cd fit he can try for game and you can bid it. Without knowing the fit, sometimes it's too sketchy to take aggressive action & you stop in a partial, especially lacking certain tools being available like xyz/2 way puppet checkback etc.

- Raising with 4432 and 5332 type hands (usually only when holding small doubleton) is more controversial, less popular than raising with stiff. This is because 1nt rebid is possible and 1nt might be better contract than 4-3 M fit. But even this has advantages. Holding small doubleton, partner's holding in that suit is vulnerable to attack on opening lead. If you play 3nt, it might be better to play in it from his side, after 1m-1M-2M-3nt-p (he offers 3nt holding only 4 cd M, with 3 only you pass), than you playing it with the lead coming into your small doubleton. Plus, you miss fewer 5-3 M fits when partner doesn't move over 1nt because of lack of strength/suit quality. Now one *could* have 5 cd hands rebidding automatically over 1nt rebids, but that is sometimes wrong, and is rather incompatible with a style that allows 1nt rebids holding a singleton in partner's suit (1444 or 1435 after 1m-1S). This rebid 1nt with a singleton is also a style thing with pros & cons.

- sometimes 4-3 fit plays just better than 1nt with ruffs taken in the short hand

- with 5 cd suit and game values, partner can just jump to game, and you have achieved a less revealing auction than bidding something else and going through checkback sequences, which might help the opps on opening lead.

- sometimes even though your 4-3 fit is doomed to fail, one of your opps doesn't know and balances you out of it. Or the flip side, you really do have a 4-4 fit but the opp doesn't take a marginal balance that would have been successful because you might be in a 4-3.

So basically the logic is that the sum total of all these possible positive outcomes creates enough plus situations to counter the times you play a stupid 2M when something else, usually 1nt would have been better.

[jogs]

Similar to Mr Ace's thinking.


1♣(♦) - 1♥ opponents silent.
Raise with 3 card support, stiff AND an absolute minimum opening. With a stronger hand rebid 1♠.

1♣(♦) - 1♠ opponents silent.
Raise with 3 card support and a stiff. With 3=4=1=5, only rebid 2♥ with a hand worth a reverse.

[bluecalm]

Quote

In my opinion, the 2♥ by partner should promise 4-card support. However, I have observed that many, many players on this board will make that bid with only 3-card support. Sometimes, they even fail to bid a 4-card ♠ suite and support hearts instead with only 3 hearts.

I did enough thinking about it to be confident in my conclusions:

-bidding 2H with 4S is terrible bridge
-3card raise with with 533-xx is imo bad but playable
-supporting with 5431's (like 1345 after 1H) is probably a bit better than bidding 2m
-3card raise with any other shape is terrible

It's more common, maybe standard to support with 3 cards in America/Sayc world but there are a lot of very bad plays which are standard so don't worry about it.
The best bidders in the world rarely support with 3 cards. Meckwell do it only with 5431's, Italians sometimes with perfect 533-xx's but my records show that this is at best break even for them and hurts slam investigation after a raise. Polish players do it very rarely, even 5431's are considered close here.

[SteveMoe]

Something from Steve Weinstein that you might find interesting:
http://bridgewinners...1m-1m-now-what/

[Vampyr]

gnasher, on 2014-March-17, 17:15, said:

Unless there's been some very unusual tectonic movement recently, England is still in Europe.

Well, technically.

[jmcilkley]

Some people play that 5-card majors applies equally to the responder - possibly resulting from a misunderstanding. I can't see any good reason to support with 3 if the 1h bid can be only 4. With 4 spades there is no problem in bidding 1s and if not then the hand may be 3-3-4-3 and a 1nt reply seems fine to me

Hawkster1, on 2014-March-17, 15:00, said:

SAYC, 5-Card Major Opening. Partner opens 1 ♣ or ♦. I respond 1 ♥. Partner supports by bidding 2 ♥. (No interference.)

In my opinion, the 2♥ by partner should promise 4-card support. However, I have observed that many, many players on this board will make that bid with only 3-card support. Sometimes, they even fail to bid a 4-card ♠ suite and support hearts instead with only 3 hearts.

When I have asked them not to support me with only 3 cards in that bidding sequence, they reply that their bidding is standard.

Certainly it is not standard in any system that I have knowledge of.

Is my bidding hopelessly out of date? If so, please explain the logic of supporting with 3 cards when my major suite response only promises 4 cards.

[gnasher]

bluecalm, on 2014-March-17, 18:32, said:

I did enough thinking about it to be confident in my conclusions:

-bidding 2H with 4S is terrible bridge
-3card raise with with 533-xx is imo bad but playable
-supporting with 5431's (like 1345 after 1H) is probably a bit better than bidding 2m
-3card raise with any other shape is terrible

Are you going to support these assertions with any actual arguments, or are you just asking us to trust you to be right?

[bluecalm]

Quote

Are you going to support these assertions with any actual arguments, or are you just asking us to trust you to be right?

When it comes to hand records/going through hands thing you will have to trust me as putting that argument in text form would require many hours of work.
The thing about hurting slam bidding is obvious (you have more hand types to handle after 1m - 1M - 2M)

When it comes to simulations, here are some to start:

1)
S have KQ6 T943 AK65 43
N have 4spades, no 4hearts, balanced hand, 6-10hcp (below invite)

Strategy of bidding 1N and then playing either there or 2D if N have 4+diamonds compared to playing 2S:

vulnerable: o
IMPs: 0.87
MPs: 0.587
vulnerable: NS
IMPs: 1.202
MPs: 0.587
vulnerable: WE
IMPs: 0.87
MPs: 0.587
vulnerable: WNES
IMPs: 1.202
MPs: 0.587

bidding 1NT is on average 1imp/hand better

2)
Same hand for S.
N has any hand in 6-10hcp range with 4 spades without 4 hearts (might be unbalanced).
The strategy of bidding 1NT and then signing off in 2D with 4+diamonds or in 3C with 6+clubs (4-x-y-6, assuming somewhat standard methods we can't sign off in 2C with that but we can in 3C) compared to just playing 2S:

vulnerable: o
IMPs: 0.88
MPs: 0.5775
vulnerable: NS
IMPs: 1.211
MPs: 0.5775
vulnerable: WE
IMPs: 0.88
MPs: 0.5775
vulnerable: WNES
IMPs: 1.211
MPs: 0.5775

In general double dummy simulations are very close to practical play (http://www.rpbridge.net/9x29.htm) so I see it as strong argument.
Going into invitational hands complicates matters but I think it's even worse for 3card raise there. Especially disastrous is playing in 3M compared to 2N/3m if responder has invitational hand.
That shapely weak hands are worth a lot opposite 4card fit but much less opposite 3card one is another problem.

I would be happy to have some discussion about it if you have some ideas for simulations serving as convincing argument for you.
I did more of those in the past and I don't think it's close tbh.

[hrothgar]

bluecalm, on 2014-March-18, 07:13, said:

In general double dummy simulations are very close to practical play (http://www.rpbridge.net/9x29.htm) so I see it as strong argument.

I don't object to the use of double dummy simulations.
I do object to strong claims being made based on simulation where no code is provided.

I have seen far to many examples where extraneous constraints or poorly written code have biased the results.

As a practical example:

Do your simulations include issues like suit quality?

I am happy to raise on three pieces. However, ideally I would want Axx in the three card suit

I don't want HHH because I'll be ruffing with honors that I need to pull trump
I don't want xxx because it risks the opponent's being able to kill my trump before I can take necessary ruffs

These types of issues are crucial to make an informed decision about raising on three pieces and if your simulations don't take this into account, then they suffer from GIGO...

[jvage]

The problem with your simulations as I understand them is that they are only relevant to the case where responder got a balanced hand with exactly 4 spades. If we knew that was the case I am sure far fewer players would suggest supporting with a 3 card suit. The main advantage of supporting is probably to releave partner of the problem of wether to rebid a fivecard-suit after openers 1NT (or 2m) rebid. I suspect this is more difficult to simulate, partly because it probably depends on how often opener actually got 3 card support (where some would have supported directly) and also how likely a singleton or a semibalanced doubleton are (are say A, AKxx, Qxx, xxxxx or xx, Kxxx, AQ, Kxxxx possible after 1♣-1♠-1NT?).

While we don't have to repeat the debate about double dummy simulations versus real play, I suspect this affects the results more on this simulation than on average. Playing and defending 4-3 fits are often tricky, starting with the opening lead. If I knew in advance that declarer was on a 4-3 fit with a semibalanced dummy I would for example be more likely to lead a trump, but in practice of course you don't know this.

[bluecalm]

Quote

Do your simulations include issues like suit quality?

Of course not, as I provided example hand (perfect one for 3card raise) and 1S is bid regardless of suit quality.

Quote

I am happy to raise on three pieces. However, ideally I would want Axx in the three card suit

Again, I gave example hand: KQx Txxx AKxx xx. Raising with this hand is very bad if partner is in 6-10 range.

Quote

These types of issues are crucial to make an informed decision about raising on three pieces and if your simulations don't take this into account, then they suffer from GIGO...

Yeah, I just made two. I made many more in the past and I am confident about my conclusion. I am happy to do more if you have specific examples but in the end you will probably have to believe me or do research yourself.

Quote

The main advantage of supporting is probably to releave partner of the problem of wether to rebid a fivecard-suit after openers 1NT (or 2m) rebid. I suspect this is more difficult to simulate, partly because it probably depends on how often opener actually got 3 card support (where some would have supported directly) and also how likely a singleton or a semibalanced doubleton are (are say A, AKxx, Qxx, xxxxx or xx, Kxxx, AQ, Kxxxx possible after 1♣-1♠-1NT?).

Yeah, you can do any strategy you like but there will always be concerns that it's not exactly what you have in mind.
In the end you can believe me or any other person. I am sure I would win the argument (because I doubt there was anybody else in the world having that much time and as good tools for that as I have) if it came to it but that requires a lot of effort to cover all cases/make all points.
The thing is I am not very motivated to debate it for two reasons:
-I don't think it's close
-I don't see any effort from 3card raisers to argue for their position
-I see them losing in practice and top players moving away from 3 card raises

That's why I offer my view with confidence level have in it. I think both are useful information.

Quote

While we don't have to repeat the debate about double dummy simulations versus real play, I suspect this affects the results more on this simulation than on average. Playing and defending 4-3 fits are often tricky, starting with the opening lead. If I knew in advance that declarer was on a 4-3 fit with a semibalanced dummy I would for example be more likely to lead a trump, but in practice of course you don't know this.

It's known that 1NT is one contract declarer does significantly better at in practice than in double dummy world so if anything there is bias in other direction.