[Free]

kgr, on Jun 11 2004, 09:25 AM, said:

Suppose partner has following patern:
xx-xxxx-xxx-xxxx
with a limit hand (around 11/12 HCP), but upgraded to GF with H-fit.
After 1D-1H-2H he will bid 3NT or 4H and you will be too high.

I think you're wrong here. If you make frequent support on 3-card, partner won't upgrade his hand until he's sure he has a 4-4 fit. And he can examine that by bidding 2NT, invitational with exactly 4 ♥s. Opener will pass or bid 3NT with 3 card support, bid 3♥ or 4♥ with 4 card support.

[kgr]

Quote

Suppose partner has following patern:
xx-xxxx-xxx-xxxx
with a limit hand (around 11/12 HCP), but upgraded to GF with H-fit.
After 1D-1H-2H he will bid 3NT or 4H and you will be too high.

Quote

I think you're wrong here. If you make frequent support on 3-card, partner won't upgrade his hand until he's sure he has a 4-4 fit. And he can examine that by bidding 2NT, invitational with exactly 4 ♥s. Opener will pass or bid 3NT with 3 card support, bid 3♥ or 4♥ with 4 card support.

Suppose you agreed with your partner that you can sometimes (even regurarly) support with a 3 card.
Can you imagine that after bidding 1D-1H-2H your parnter has only a 3 card?
Why didn't he bid 1S, 2C, 2D, 1NT?
I think after 1D-1H-2H you should be sure that he has a 4 card.
Otherwise that will give you too much bidding problems:
Axxx-Kxxx-QJxx-x
a) after 1D-1H-2H: you want to bid 4H, 3NT is not a good alternative
after 1D-1H-1S: no problem, 4S

<void>-AJxxx-xxx-Jxxxx
a) after 1D-1H-2H: according to LOTT you want to preempt with 3H to avoid that opps find their S fit
after 1D-1H-1S: you bid 2C

Jxx-Axxx-xx-Jxxx:
a) after 1D-1H-2H: pass
after 1D-1H-1S: pass (still 4/3 fit, but played by strong hand).

With 10+points and 5 card H responder will be able to ask for 3 card H support on the next round.

...I don't like to support with 3 card, certainly not after:
1C-1D
1D-1H
1H-1S
..sometimes after 1C-1S or 1D-1S.

[hrothgar]

kgr, on Jun 11 2004, 02:51 PM, said:

...I don't like to support with 3 card, certainly not after:
1C-1D
1D-1H
1H-1S
..sometimes after 1C-1S or 1D-1S.

There are advantages and disadvantages to raising parter with 3 card support. However, there are advantages and disavantages to any method. The important question is whether the relative gains outweight the losses.

The most important gain for making an aggressive raise style based on 3 card support is related to the same LTT that you are trying to use to bolster your own arguments. It is vital to realize that the opponents are every bit as capable of counting trump as we are. If your raise promises that you are playing in an eight card suit at the two, the opponents have all the information that they need to scramble. Its FAR more useful to have the direct raise leave some degree of ambiguity regarding the trump fit.

As Free has pointed out, there are any number of good checkback schemes available after 1m - 1M - 2M that allow pairs to investigate strain and level.

[luke warm]

usually, after 1m : 1M : 2M responder passes with an average hand and only bids on with invitational+, so i'd expect a 2NT bid with a balanced hand holding four cards in his M, else 3m/M..

[PriorKnowledge]

Instead of writing flowing prose with ideas on general principles and philosophy, how about listing hands and bidding and see what happens? Philosophy is fine for some things, but bidding is more science and mathematical than art.

First: Give responder 8 possible hands:
4s - weak
4s - inv
4s - GF
5s - weak
5s - inv
5s - GF
4s - slam possibility (will we find best fit for slam in major or minor?)
5s - slam possibility

(I can say one thing, immediately, for the last two hand types: If we show 5-4, then raise responder for our 3rd bid, we will show our 5-4-3-1 perfectly and have the best chance for slam. If we don't do that in our first 2 bids, we can never show it.)

Second: our strength - give us 3 possible hands: min (13hcp), intermediate (16hcp), and strong (18 HCP)

Can we find the right game, chance for slam, or reasonable partial in the majority of cases?

Repeating my recommendations:

Case 1: 3-1-5-4 after 1D - 1S. Rebid 2C with any non-GF (up to about 17HCP). 3C is a GF.

Case 2: 3-4-5-1 after 1D - 1S. Without strength to reverse (less than about 15HCP), rebid 2S. With reverse strength (15+) rebid 2H. Responder will show 5, if responder has it.

Case 3: 4-3-5-1 after 1D - 1H, raising hearts on 3 when you have 4s is terrible. If responder is 44 in majors, you will lose the spade suit.

Sorry... too involved, have to go back to work now... I'll have to work out the details later. If someone else wants to try, go ahead.

I think you'll find that the methodology I recommend (which is used by most experts) will get you to the best contract most often.

[1eyedjack]

I may come back to PriorKnowledge's latest post in more detail later ... certainly I applaud his methodology, if you have the time, although it is quite complicated to be thorough, because (a) you must not treat each possible hand as of equal value, since (i) the likely swing at stake will not be equal and (ii) their frequencies are not equal, and (b) you have to take into account the possibility (and consequences) of a contested auction. This is probably why we try to distil theoretical principles, in addition to number crunching.

In this post I will simply point out one logical consequence that PK probably did not intend from his post, as it points to a contrary conclusion:

Specifically, compare the following comment:

Quote

4s - slam possibility (will we find best fit for slam in major or minor?)
5s - slam possibility

(I can say one thing, immediately, for the last two hand types: If we show 5-4, then raise responder for our 3rd bid, we will show our 5-4-3-1 perfectly and have the best chance for slam. If we don't do that in our first 2 bids, we can never show it.)

with the subsequent comment in the same post:

Quote

Case 3: 4-3-5-1 after 1D - 1H, raising hearts on 3 when you have 4s is terrible. If responder is 44 in majors, you will lose the spade suit.

The suggested treatment in Case 3 is to raise Hearts directly with a weak hand and to give delayed support (via 1S) with extra values. Now I ask you: on which of these two hand types (if either) are you going to be making slam? That is a rhetorical question, by the way. Anyway, under the suggested method your slam bidding will be more accurate because not only will you have shown accurate shape on the slam hands but you will also have more narrowly defined your values - the latter factor absent if you routinely rebid 1S.

Indeed, as a general principle (sorry) we can disregard most hands in which responder has at least a game try. There is plenty of room to clarify whether opener has 4 card support or 4 cards in an unbid major, and either method is playable. This is one of the advantages of talking philosophy before you waste a lot of time number-crunching irrelevant hand types.

[1eyedjack]

I think that we need to be wary of too much reliance on LOTT as a justification for any action here.

The best time for "mixing" it based on LOTT theory is when one of your side (partner "A") has shown precise length and the other (partner "B") has undisclosed support. Provided that "A" leaves competive decisions up to "B" you have the edge. "B" knows precisely the level of the fit, and opponents have no clue. So you might raise on 7, 8, or 9 card fit to any level an leave the opponents guessing. If they act then you know whether to leave them in it or continue.

But it is rather different if both partners on your side have variable suit length. Then each partner (and the opponents) will assume that the other partner has the most likely length for the bidding to date. So, how does this apply to case 3?

If you agree to raise with tripleton on certain hands the fact remains that you are odds on favourite to have 4 card support. Responder will therefore assume 4 card support in any subsequent defensive action (NB no such assumption necessary if making or accepting a game try - room to enquire). Opponents will also assume that opener has 4 card support for responder.

So, opponents balance. Who is the winner?

If you only had a 4-3 fit then you win the exchange. The opponents have balanced at an inappropriate time. Responder, expecting an 8 card fit, will not generally take the push to the 3 level. He will expect a par result by defending, and below par by bidding. You (opener) will expect a better than par result by defending. You will therefore end up defending and responder will be pleasantly surprised with the result.

If however you had a 5-3 fit then you lose the exchange. The opponents accurately assess the size of the fit (although they may expect it to be 4-4 rather than 5-3). Responder however will assume (as is most likely) that you (opener) have 4 card support and he will accept the push to the 3 level, which in the long term will be a losing action (kgr mentioned this).

If you have a 6-3 fit then the position is probably fairly neutral, although I have not really thought that one through.

So, the question remains, if it goes 1D-1H-2H (uncontested) where you conventionally raise on a 3 card suit in the situations under discussion, which is more likely (and by how much), between a 4-3 fit and a 5-3 fit? I have repeatedly asked this question in this thread in a roundabout fashion but so far no takers.

[hrothgar]

Hi All

I wrote a quick and dirty dealer script to model the following:

North has a 1D opening and
South has a 1H response and
North has a hand where I would raise to 2H

(North has 3 card heart support and
North does not have 4333 shape)

Under these circumstances, we will miss a 4-4 spade fit ~ 1.8% of the time.
Low enough percentage that I really am not going to get worked up about it.

The 2H raise will place us in a 7 card fit ~ 55.7% of the time (4-3 fit)
The 2H raise will place us in an 8+ card fit ~ 44.3% of the time (5+ - 3 fit)

If I relax the conditions and cosnider cases where North is rasing on 3+ Hearts

The 2H raise will place us in a 7 card fit ~ 42.7% of the time
The 2H will place us in an 8+ card fit ~ 57.3% of the time

I'm attaching the script in case anyone cares

__________________________
no_trump =

hcp(north) >= 15 and
hcp(north) <= 17 and
shape(north, any 4432, any 5332, any 4333 - 5xxx - x5xx)

diamond_opening =

hcp(north) >= 11 and
not no_trump and

(

(
diamonds(north) >= clubs(north) and
diamonds(north) > hearts(north) and
diamonds(north) > spades(north) and
diamonds(north) > 3
)

or

(
shape(north, any 4432 - any xx2x)
)

)

one_heart =

hcp(south) >= 6

and

(

(
hearts(south) >= 4 and
hearts(south) > spades(south)
)

or

(
hearts(south) == 4 and
spades(south) == 4
)

)

two_hearts =

hearts(north) == 3 and
not shape(north, any 4333) and
hcp(north) <= 14 and
hcp(south) <= 10

trouble = spades(north) == 4 and spades(south) == 4

condition

diamond_opening and
one_heart and
two_hearts

action

average trouble

[mikestar]

hrothgar, on Jun 11 2004, 08:16 PM, said:

I wrote a quick and dirty dealer script to model the following:

North has a 1D opening and
South has a 1H response and
North has a hand where I would raise to 2H

(North has 3 card heart support and
North does not have 4333 shape)

Under these circumstances, we will mis a 4-4 spade fit ~ 1.8% of the time.

The 2H raise will place us in a 7 card fit ~ 55.7% of the time (4-3 fit)
The 2H raise will place us in an 8+ card fit ~ 44.3% of the time (5+ - 3 fit)

...

This convinces me. The chance of missing 4-4 spades is too low to worry about vs. the chance of getting to the 5-3 heart fit we might well miss if I don't raise. Even if we'd only miss one 5-3 fit in ten (and I'm sure we'd miss more) this is a substantially greater risk than missing the spade fit.

As for LOTT decisions, responder assumes 3 1/2 card support--that is he assumes three if there are negative factors (minor honors in enemy suit, flat hand, being vulnerable) and assumoing four if there are positive factors (minor honors in our suits, good shape, being NV).

See Larry Cohen's LOTT books for more detail about the 1/2 trump concept.

[luke warm]

PriorKnowledge, on Jun 11 2004, 05:27 PM, said:

Repeating my recommendations:

Case 1: 3-1-5-4 after 1D - 1S. Rebid 2C with any non-GF (up to about 17HCP). 3C is a GF.

Case 2: 3-4-5-1 after 1D - 1S. Without strength to reverse (less than about 15HCP), rebid 2S. With reverse strength (15+) rebid 2H. Responder will show 5, if responder has it.

Case 3: 4-3-5-1 after 1D - 1H, raising hearts on 3 when you have 4s is terrible. If responder is 44 in majors, you will lose the spade suit.

I think you'll find that the methodology I recommend (which is used by most experts) will get you to the best contract most often.

case 1, opener 3154:
1D : 1S - my point here is, either responder has a hand with which he'll bid again or he doesn't... if i bid 2S, most good players will find another bid with invitational+ hands and 4 spades, such as 2NT... he'll know i *might* not have 4 card support and will find out shortly.. so with less than 15, 16 points i'm raising his spades with 3... i honestly don't think we're missing anything, and richard's script seems to bear this out... with a stronger hand i'd bid my clubs

case 2, opener 3451:
1D : 1S - agreed

case 3, opener 4351:
1D : 1H - i can't bring myself to agree with you, even if free is correct that 1S is forcing (and i'm not sure it is, or that it should be)... the same reasoning as in case 1 applies, responder either does or does not have a hand with which to continue bidding... again, richard's script seems to back this up... *any* bid by responder shows at least invitational strength...

as for your last statement, there are quite a few experts who read and post in this forum... some even read this forum who have written books on this and other subjects... in the books i've read, i've never seen that raising with 3 card support, even with 4 spades, is wrong.. it depends on the hand, and the partnership

[Free]

luke warm, on Jun 11 2004, 11:30 PM, said:

~snip~
case 3, opener 4351:
1D : 1H - i can't bring myself to agree with you, even if free is correct that 1S is forcing (and i'm not sure it is, or that it should be)... the same reasoning as in case 1 applies, responder either does or does not have a hand with which to continue bidding... again, richard's script seems to back this up... *any* bid by responder shows at least invitational strength...
~snip~

I never said 1♠ is forcing, but if you always bid 1♠ with that, it should be forcing. We had already another thread about 1♠ being forcing or not, and it's also a matter of preference

[luke warm]

oops, sorry.. yeah i remember the other thread... seems like the opinions were divided on it

[PriorKnowledge]

good.... just the 1D 1H when opener has 4351
How easily you deceive your partner about your distribution and miss (the 1.8%) chance for a 44 spade fit. For what?

1D 1H 1S
If responder is weak with 5+ hearts, will bid 2H, or 2D or 1N. no game, no bid deal
If responder is weak with 4h and club stopper will bid 2D, 1N, or even pass (playing 2H on a weak 43 fit sucks, esp when staring at the missed 44 spade fit)
If responder is inv with 5+h can bid 2N, 3D, 3H. If we are max, we can make a delayed raise of 3H.
If responder is inv with 4h can bid 2N, 3D.
If responder is game forcing will bid 2C, 4SF, and our delayed raise of 2H shows our distribution and gives us best chance to find right game or slam. (Initial 2H distorts our hand and prevents us ever finding the correct contract. Plus we will miss our 1.8% chance of 44 spade fit)

Your misguided "rule" distorts your hand.

My last post on this subject.

[mikestar]

PriorKnowledge, on Jun 12 2004, 04:33 AM, said:

1D 1H 1S
If responder is weak with 5+ hearts, will bid 2H, or 2D or 1N. no game, no bid deal
If responder is weak with 4h and club stopper will bid 2D, 1N, or even pass (playing 2H on a weak 43 fit sucks, esp when staring at the missed 44 spade fit)

Admittedly 4-3 fits can suck. 5-1 fits suck worse and 5-0 fits suck worse yet. All leading authourities avoid rebidding five card suits like the plague in sequences where a pass is at all likely. Alan Truscott goes as far as calling it the Golden Rule of Bidding.

[1eyedjack]

PriorKnowledge, on Jun 11 2004, 11:33 PM, said:

good.... just the 1D 1H when opener has 4351
How easily you deceive your partner about your distribution and miss (the 1.8%) chance for a 44 spade fit. For what?

1D 1H 1S
If responder is weak with 5+ hearts, will bid 2H, or 2D or 1N. no game, no bid deal
If responder is weak with 4h and club stopper will bid 2D, 1N, or even pass (playing 2H on a weak 43 fit sucks, esp when staring at the missed 44 spade fit)
If responder is inv with 5+h can bid 2N, 3D, 3H. If we are max, we can make a delayed raise of 3H.
If responder is inv with 4h can bid 2N, 3D.
If responder is game forcing will bid 2C, 4SF, and our delayed raise of 2H shows our distribution and gives us best chance to find right game or slam. (Initial 2H distorts our hand and prevents us ever finding the correct contract. Plus we will miss our 1.8% chance of 44 spade fit)

Your misguided "rule" distorts your hand.

My last post on this subject.

This adds nothing but repetition. Unfortunately this response in the main does likewise. All - ignore this:

"good.... just the 1D 1H when opener has 4351
How easily you deceive your partner about your distribution"

No deception. You and your partner are expected to be playing the same system.

"and miss (the 1.8%) chance for a 44 spade fit."

Only if responder is too weak to bid again.

"For what?"

I thought this was clear ... in order to secure playing in Hearts on the rather higher frequency when this would otherwise be missed.

"1D 1H 1S
If responder is weak with 5+ hearts, will bid 2H, "

Make that 6+ Hearts. With 5 I regret that you miss the Heart contract, unless you bid them yourself later on weak and strong openers alike. You have repeatedly ignored the problems that this in turn can create, perhaps because they cannot be refuted.

Remainder of quoted post deals with invitational or stronger responders, wherein there is no problem whichever method you play.

I should point out that I think it is a fairly finely balanced problem, just based on experience, despite the rather compelling stats posted by Hrothgar. My posts have been heavily supportive of the direct raise mainly because I find the counter arguments posted so far less convincing, and I am only responding to those.

One aspect that others have not really addressed is the MP v IMP issue. At MP I think bidding 1S makes a lot more sense in case 3. At IMP, if I am allowed to play in 2H opposite a weak responder in a 4-3 fit having missed a 4-4 S fit (and opps having missed a 4-4 minor fit) then I am content in the long term. Not so content at MP.

[1eyedjack]

mikestar, on Jun 11 2004, 04:53 PM, said:

As for LOTT decisions, responder assumes 3 1/2 card support--that is he assumes three if there are negative factors (minor honors in enemy suit, flat hand, being vulnerable) and assumoing four if there are positive factors  (minor honors in our suits, good shape, being NV).

See Larry Cohen's LOTT books for more detail about the 1/2 trump concept.

If you were habitually to raise on all hands with 3+ card support then I would go along with that. As it is, partner would be wrong to assume 3.5 card support. If he is going to attribute you with fractions of a trump card then the fraction should reflect the frequency. In fact you would quite rarely raise on tripleton, even if you are in the "raising" camp, so he should assume more likely 3.9 card support (rough guess). Or 4, to all intents and purposes.

Additional points (unrelated to above quote):

It is not inconceivable that opps will balance into 2S if you raise H immediately. In the long term I welcome that action by the opps. But they would be rather less likely to take that action if I start by rebidding 1S.

I confess to considerable surprise at the 1.8% frequency. I am not a statistician but instinct based on experience tells me it should be higher. I wonder whether opponents' failure to bid Spades to date might be a reason.

[nikos59]

IMHO the 1,8% figure is totally misleading.

What I want to know is:
after 1D - 1H
if I hold four spades and three hearts and raise partner at 2H,
what chances do I have to miss a 4-4 spade fit?
Or, to put it plainly, what are the chances for pard to have
four spades?
The answer to this, unless I misread the script,
is *not* 1.8%

If I read well the dealer script, 1,8% is the percentage of the
"trouble with 4-4 spades" hands averaged on ALL North hands that open 1D,
but this is irrelevant because it includes all the North hands that
open 1D and raise to 2H but do not have four spades.

Or am I wrong?

n.

[hrothgar]

nikos59, on Jun 12 2004, 01:41 PM, said:

IMHO the 1,8% figure is totally misleading.

What I want to know is:
after 1D - 1H
if I hold four spades and three hearts and raise partner at 2H,
what chances do I have to miss a 4-4 spade fit?
Or, to put it plainly, what are the chances for pard to have
four spades?
The answer to this, unless I misread the script,
is *not* 1.8%

If I read well the dealer script, 1,8% is the percentage of the
"trouble with 4-4 spades" hands averaged on ALL North hands that open 1D,
but this is irrelevant because it includes all the North hands that
open 1D and raise to 2H but do not have four spades.

Or am I wrong?

n.

Nikos raises an important point that I should have made more explict.

As Nikos notes: The 1.8% figure assumes the following

(1) A 1D opening
(2) A 1H response
(3) A 2H raise on 3 cards by opener
(4) A pass by opener

If we are doig a 1:1 comparison, its probably best to also include the assumption that opener holds 4 Spades as well. This increase the percentage chance that we hold a "trouble" type by roughly an order of magnitude.

The percentage chance that we hold a 4-4 Spade fit is now ~17.4%
About 67% of of the time, we're in a state that I call "real trouble" in which
we are playing in a 4-3 heart fit rather than a 4-4 Spade fit.

17.4% still isn't enough for me to worry about.
IMHO, the gains from

(a) Immediately limiting my strength outweight the losses
(b) Promising greater strength with a delayed Heart raise
© Forcing the bidding to an acceptable contract ASAP

Outweigh occasionally missing a better contract.

[hrothgar]

PriorKnowledge, on Jun 12 2004, 07:33 AM, said:

good.... just the 1D 1H when opener has 4351
How easily you deceive your partner about your distribution and miss (the 1.8%) chance for a 44 spade fit. For what?

1. The direct raise to 2H immediately limits the strength of my hand.
Partner will be well positioned to determine whether to bid game, invite game,
or pass. In contrast, the style that you advocate in which responder
initially shows Spades and then shows Heart tolerance is necesarily much
more ambiguous with respect to strength.

2. The style that I advocate also has significant advantages if the auction starts
with a delayed heart raise. Here once again, accurately conveying range
information can be critical.

3. As The_Hog noted, bridge is a 4 hand game. Maximizing pressure on the
opponents often needs to take priority over perfecting constructive bidding.
From my perspective, expert bidding is moving to a style in which pairs
"blast" to acceptable contracts as quickly as possible, even at the expense of
occasionally missing an "optimal" contract.

[Free]

I'm used to play 1m-1♥-1NT with possible 4 card ♠, since the chance is indeed not thàt big to miss a ♠ fit, and NT range is more important to us! This also means 1m-1♥-1♠ is unbalanced! This way I can show my hand just perfect: when p rebids 1NT or so, I support his ♥s and it should be clear I have 5+m, 3♥s and 4♠s. Isn't that nice? Showed my hand at 2♥! If p bids 4th suit, I can also support his ♥ and my hand is again shown. And we use 1♠ as forcing, which balances the system.

Never gave me problems, we still play in ♥s when there's no ♠ fit, and we have time so why waste it?