[MickyB]

It seems wrong to me to restrict the use of different methods opposite a passed hand. An "opening pass" communicates vital information; saying that a 1st/2nd seat 1D opening should mean the same thing as P:1D makes about as much sense as saying that P:1D should mean the same thing as 1C:1D.

[joshs]

awm, on Jan 14 2008, 11:29 AM, said:

Here's a recent observation from when I was simulating something else.

Suppose that the folks in 1st-3rd seat have a fairly standard opening style. Then in 4th chair, it is more frequent to have a strong (15-17) notrump opening than a weak (12-14) notrump opening.

This is not true in any other seat.

This actually gets more extreme if you play light openers.

Without simulation results to back me up, I make the following claims:

Suppose you play a 15+ strong club in 1/2, with other openings being 10-14, and you open most 10's.
Suppose the opps play a standard system (rule of 20ish openers).
Let P=Probability of holding a 15+ hand in first seat
Let Q=Probability of holding a 17+ hand in 3rd seat, after the first 2 players pass.

Claim A: Q>P, and its not even that close

Claim B: In this context, after only 2 passes, a strong NT (15-17) is more common than a weak NT (12-14) [unlike standard]

What hands you open in 1/2 has a big effect on the distribution of hands you have to deal with in 3/4, and thus its pretty clear that playing the exact same methods (or at least ranges) is not efficient.

For instance,
Claim C:
If you play standard methods, and a 15-17 NT in 1/2, its actually more efficient to play 14-16 in 3/4. I will let the simulators work this one out...

[Echognome]

joshs, on Jan 15 2008, 11:33 AM, said:

Claim C:
If you play standard methods, and a 15-17 NT in 1/2, its actually more efficient to play 14-16 in 3/4. I will let the simulators work this one out...

Define "efficient".

I am not disputing the claims or even trying to be pedantic. I'm just trying to clarify what you are getting at.

[joshs]

Echognome, on Jan 15 2008, 02:49 PM, said:

joshs, on Jan 15 2008, 11:33 AM, said:

Claim C:
If you play standard methods, and a 15-17 NT in 1/2, its actually more efficient to play 14-16 in 3/4. I will let the simulators work this one out...

Define "efficient".

I am not disputing the claims or even trying to be pedantic. I'm just trying to clarify what you are getting at.

Efficient-"performing or functioning in the best possible manner"

Now to evaluate efficiency you need a metric for "best", which in the case of bridge could be to
a. maximize your score per hand (in a given scoring method)
b. maximize you chance of winning a match or event (could be very different then a if you are a big favorite or underdog)
c. how much fun you have, in which case ignore my claims at efficiency, since clearly playing 0-12 NTs is more fun


I.E. An efficient method is one that is good at achiving your goals...

[kenrexford]

I agree whole-heartedly with the nonsense of requiring the same "system" for all seats.

Suppose that you play LIA, opening all 8-counts. Perfectly legit, right?

Well, why the heck would you use the same approach in fourth seat? That's just stupid.

I cannot imagine how the definition of a "system" would mean that you must ignore one action within the system, namely the pass, as suggested. A pass is not simply a waiting action, like a check in poker. A pass conveys meaning. Therefore, "responding" to the pass, even if it is a standard, non-forcing pass, cannot logically be restricted in any manner that makes logical sense, to me, so long as the structure of "openings" is one that would be allowed in first or second seat, perhaps.

[awm]

I have a friend who believes that playing roughly standard methods in 1st-3rd seat, it's a good idea to play a weak notrump in fourth chair. Basically his notrump range in fourth is "up to 14 hcp" with the bottom end being whatever his minimum opening would be (he states it as 10-14 but obviously many balanced 10s will not open in 4th chair).

His reasoning is that when he holds this hand, the points are usually pretty evenly divided around the table, and he wants to buy the hand in 1NT as often as possible (hoping to go plus). He figures that a lot of people would open (either one-level or a preempt) with hands that are suitable to bid over a weak notrump, so by opening this way he talks the opponents out of the auction. If he opened one of a suit, the opponents might overcall a five-card major (on a hand too weak to open) and thus find a fit and compete the hand past 1NT.

While this argument makes some sense, I would counter it by saying that whether opponents have a bid depends a lot on how aggressively they open shapely hands. When I've defended against this fellow, I find that I fairly often do have a bid over 1NT, because I often pass with 5-5 ten counts (for example) in first chair. When I do bid, it is quite often his side that's unable to find its best fit and compete the hand to the three-level, because of the 1NT opening call. A second argument is that if the points are roughly evenly divided, there is no particular reason to think that 1NT will make, and it might be better to try and find a suit fit (if one exists). Finally, a strong notrump is more common than a weak notrump in fourth so he's actually opening 1NT less often than the field, and with hands where partner will not be strong enough to look for a 4-4 major fit (because partner is unlikely to have an invite opposite 10-14 notrump).

Anyways, my point here is that it's far from obvious what is "most efficient" even assuming that we can compute frequencies, and that it may well depend on the opponents' opening style as well as your own side's.

[joshs]

Scetch of Claim A:

First we ignore pre-emptive openings:

Seat 1 :Let me assume that the Strong Club Pair opens all 10 counts, and 1/2 of the 9 count, but nothing lighter.

Seat 2: Let me assume that the Standard Pair opens all 12 counts, half the 11 counts, and nothing lighter

(Feel free to redo these caculations with your own distribution here, I am just trying to approximate things).

The player in Seat 1 having passed, has on average 6.19 HCP
The player in Seat 2 having passed has on averqage 7.35 HCP
(take a standard table of probabilities of HCP holdings, and take conditional expectations given the above assumptions)

Thus player 3 has on average (40-6.19-7.35)/2=26.46/3=13.23 HCP

Now you need to compute the probability of 17+ given a mean of 13.23 and compare it with 15+ given a mean of 10. to do this for real, you need to calculate the standard deviation in the 3'rd seat case (which is less than in the 1st seat case) but this has to change by an aweful lot to make up for the fact that you are 1.23 HCP closer to your objective value.

Now assume pre-empts:
Claim A prime:
At least with relatively normal pre-emption styles, the premeptive hands that are not good enough to open 1, are better than average for the range (so that we don't pre-empt on 0's and pass with 9 or 10's, but rather we pass 0's and preempt on 9 or 10). Thus removing the premepts from the passed hand actually decreases the average number of HCP given a pass (since the preempts were better than the average pass). Hence this further increases the expected HCP in hand 3 after 2 passes. Probably the correct value, given my assumptions, is almost 13.5.

Claim B: Note that 13.5 is only 1.5 short of 15, while 10 is 2 short of 12. I leave the calculations of the standard deviations as an excercise to the reader....

[Cascade]

kenrexford, on Jan 16 2008, 10:47 AM, said:

Suppose that you play LIA, opening all 8-counts. Perfectly legit, right?

Well, why the heck would you use the same approach in fourth seat? That's just stupid.

I actually came up against a pair many years ago playing 8-10 1NT in all seats.

When it came up in third seat opposite a known 0-8 hcp (but not 8 balanced) I devised an on the spot defense of "DOUBLE".

For some reason that I cannot quite put my finger on this pair now play a different system in 3rd/4th seats.

[Cascade]

joshs, on Jan 16 2008, 08:33 AM, said:

If you play standard methods, and a 15-17 NT in 1/2, its actually more efficient to play 14-16 in 3/4. I will let the simulators work this one out...

What do you perceive is the advantage of this?

What I can think of is that partner with a balanced 11 hcp that was passed will know immediately what to do opposite a 14-16. But is there something else.

I don't play 15-17 in 1st/2nd so haven't really thought about this much.

We play weak (something like) 10-13 favourable, (10)11-13 None, (11)12-14 All, 12-14 Unfavourable in first seat. In 3rd/4th seat we play 15-17. I am wondering if there is something that should convince me to play a slightly different range in 3rd/4th seats.

[Cascade]

kenrexford, on Jan 15 2008, 05:22 AM, said:

I never quite understood that "two systems" concern.

Suppose that you played "2/1 GF" in 1st-2nd seat and "canape" in 3rd-4th seat. What if you instead called the system "Noodle Scientific" and defined the system such that it paralleled 2/1 GF in 1st/2nd seat and canape in 3rd/4th seat?

Its just another example of a badly thought out regulation.

[Cascade]

awm, on Jan 16 2008, 10:51 AM, said:

I have a friend who believes that playing roughly standard methods in 1st-3rd seat, it's a good idea to play a weak notrump in fourth chair. Basically his notrump range in fourth is "up to 14 hcp" with the bottom end being whatever his minimum opening would be (he states it as 10-14 but obviously many balanced 10s will not open in 4th chair).

His reasoning is that when he holds this hand, the points are usually pretty evenly divided around the table, and he wants to buy the hand in 1NT as often as possible (hoping to go plus). He figures that a lot of people would open (either one-level or a preempt) with hands that are suitable to bid over a weak notrump, so by opening this way he talks the opponents out of the auction. If he opened one of a suit, the opponents might overcall a five-card major (on a hand too weak to open) and thus find a fit and compete the hand past 1NT.

I think this is wrong reasoning too.

To me opposite a passed hand with a weak NT hand the hand is all about part-scores. Yes 1NT will sometimes be the right spot but it is something of a crap shoot to open 1NT.

We comprimise a little and almost always open 1Maj with a four-card suit and respond 1NT with up to 11 hcp (actually at some vulnerabilities we may have opened 10-11 hcp hands). We also make a 2/1 by a passed hand with something that looks like a weak two - these are very non-forcing. This means we can very effectively get to 1NT and two of any suit very quickly.

When developing this my alternative was to develop some different system of responses over a weak NT in 3rd/4th seats. It just seemed inefficient (there is that word again) to have a structure opposite a passed hand that was designed primarily to find the best game.

[MickyB]

joshs, on Jan 15 2008, 07:33 PM, said:

Claim C:
If you play standard methods, and a 15-17 NT in 1/2, its actually more efficient to play 14-16 in 3/4. I will let the simulators work this one out...

I hinted at this in my original post - I'm now keen on using the same structure in 4th, as well as 3rd.

Opposite a passed hand, your no-trump ladder should be based on how high you want to be opposite a balanced maximum pass with no fit.

• Expects 3NT to be marginal at best
  
• Clearly wants to be in 3NT opposite a maximum pass
  
• Clearly wants to be in game opposite a hand an ace shy of opening

Opposite a partner who will open almost all flat 12s but pass most flat 11s, this means playing a 14-16 NT, or something thereabouts.

If you open four-card majors on weakish balanced hands, this is even more important - you can't open 1M on a balanced 14, because partner has to be able to respond 1NT on his balanced 11, over which you couldn't possibly entertain taking another call.

Of course, the exact range is a function of your aggressiveness and the conditions. Vulnerable at teams, some feel it's best to pass a fair proportion of flat 12-counts in 1st/2nd, but still want to be in a thin game with 12 opposite 12. This is somewhat problematic!

[joshs]

MickyB, on Jan 15 2008, 05:47 PM, said:

joshs, on Jan 15 2008, 07:33 PM, said:

Claim C:
If you play standard methods, and a 15-17 NT in 1/2, its actually more efficient to play 14-16 in 3/4. I will let the simulators work this one out...

I hinted at this in my original post - I'm now keen on using the same structure in 4th, as well as 3rd.

Opposite a passed hand, your no-trump ladder should be based on how high you want to be opposite a balanced maximum pass with no fit.

• Expects 3NT to be marginal at best
  
  
• Clearly wants to be in 3NT opposite a maximum pass
  
  
• Clearly wants to be in game opposite a hand an ace shy of opening

Opposite a partner who will open almost all flat 12s but pass most flat 11s, this means playing a 14-16 NT, or something thereabouts.

If you open four-card majors on weakish balanced hands, this is even more important - you can't open 1M on a balanced 14, because partner has to be able to respond 1NT on his balanced 11, over which you couldn't possibly entertain taking another call.

Of course, the exact range is a function of your aggressiveness and the conditions. Vulnerable at teams, some feel it's best to pass a fair proportion of flat 12-counts, but still want to be in a thin game with 12 opposite 12. This is somewhat problematic!

Yes that is the crux of my argument.

I don't think the difference between 14-16 and 15-17 matters that much for competative auctions (and the exact effect is pretty complicated since it depends on how likely the opps will enter over 1N as a passed hand, and what kind of hands they pass, since opening NT makes it harder for your side to find its own fits) but in constructive bidding it isn't even close. In standard ranges, you needlessly end up in 2N down much more often then you need to here...

Having a good range for 1N means that a passed hand 1N response can truely be semi-forcing e.g. No game opposite a balanced hand less than a 1N opening.
This means you do not have rediculous auctions like:
Qxxxx Axx Kx AJx
xx Kxx xxx KQxxx
1S-1N-2C-2S (or 3C)

Instead it goes, 1N-All Pass

If you don't have to worry about missing game with you 5332's you can then decide to bid or not to bid over 1N based on the nature of your values (how strong is the 5 card suit/ how likely is it that 1N is the best spot) instead of based on just your point count.

[awm]

Without making a general value judgement about the merits of system, what are the advantages and disadvantages of playing four-card majors? I'm not suggesting canape here, just that balanced hands with a 4cM and not in the notrump range would open 1M. It seems that:

(1) Four card majors can be problematic when the opponents intervene, especially at the three-level. For example 1♠-(3♦) and partner is a bit stuck with a decent hand including exactly three spades, where bidding 3♠ opposite a 5cM system is easy but opposite a 4cM system (especially 4cM with strong notrumps) is potentially bad.

(2) Four card majors can make slam decisions difficult, since your trump suit provides less safety. If you have the trump AKQ and a nine-card fit, then the chances of a trump loser are close to nil. In an eight-card fit you will have a trump loser fairly frequently (about 1/3 of the time if you don't have the jack or ten).

(3) Four card majors don't interact well with forcing notrumps; playing 5cM it is rare that the major suit isn't at least an okay fit so partner can correct to the 5-2 on many hands, whereas the 4-2 major fit is likely to be a silly spot.

(4) Playing four card majors can help substantially when you do open 1m since partner knows you have a real suit there and can freely raise on four cards, etc. Also 1m-all pass is less likely to be a ridiculous spot.

(5) Playing four card majors means you open 1M more often, which is harder for opponents to bid over (especially 1♠), causing them to miss out on some fits they'd have found over the 1m opening.

(6) Playing four card majors can give less information to the opponents in the (not unlikely) situation that you wind up playing in the major or in notrump.

It seems like in fourth chair, the big disadvantages (1-3) are minimized. The opponents are unlikely to intervene at the three-level having not opened a preempt. You are unlikely to have a slam when you hold a balanced hand of less than 20 hcp opposite partner's initial pass. Most people don't play forcing notrumps by passed hand, and a passed hand is much more likely to be balanced (forcing notrump generally helps you when responder is shapely, not so much when responder is balanced).

So maybe 4cM are best in fourth position? I know a fair number of people will open a 4cM with a "sub-minimum" but it could easily pay to open a 4cM in any case, planning to try 2NT next with the "big balanced hand" or pass partner's 1NT or 2M response with the "weak notrump" hand. This is especially good with Josh's 14-16 notrump suggestion, since you can happily pass/bid 2NT over partner's 1NT response.

[MickyB]

joshs, on Jan 15 2008, 10:54 PM, said:

Yes that is the crux of my argument.

I don't think the difference between 14-16 and 15-17 matters that much for competative auctions (and the exact effect is pretty complicated since it depends on how likely the opps will enter over 1N as a passed hand, and what kind of hands they pass, since opening NT makes it harder for your side to find its own fits) but in constructive bidding it isn't even close. In standard ranges, you needlessly end up in 2N down much more often then you need to here...

Yup. It reminds me somewhat of the bizarre WJ05 sequence -

1♣:1NT, 2NT:P

1♣ = a variety of hands, most probably a weak NT
1NT response = 9-11 balanced
2NT = Balanced 14
Pass = Balanced 9/bad 10

Why make partner think there might be a game on when you know that there isn't?

[MickyB]

Cascade, on Jan 15 2008, 10:30 PM, said:

joshs, on Jan 16 2008, 08:33 AM, said:

If you play standard methods, and a 15-17 NT in 1/2, its actually more efficient to play 14-16 in 3/4. I will let the simulators work this one out...

What do you perceive is the advantage of this?

What I can think of is that partner with a balanced 11 hcp that was passed will know immediately what to do opposite a 14-16. But is there something else.

I don't play 15-17 in 1st/2nd so haven't really thought about this much.

We play weak (something like) 10-13 favourable, (10)11-13 None, (11)12-14 All, 12-14 Unfavourable in first seat. In 3rd/4th seat we play 15-17. I am wondering if there is something that should convince me to play a slightly different range in 3rd/4th seats.

Playing a 15-17 NT in 1st/2nd is only relevant in that it implies passing flat 11s and opening flat 12s. The inferences are the same with a 12-14 NT opening.

If you routinely open flat 11s in 1st+2nd, a 15-17 NT is probably best in 3rd+4th.

Am I right in thinking that you play a weak NT 3rd NV, but sometimes open 1M on a weak NT? It sounds to me like you'll have a lot of difficulty responding to such a 1M opening.

[MickyB]

awm, on Jan 16 2008, 12:03 AM, said:

Without making a general value judgement about the merits of system, what are the advantages and disadvantages of playing four-card majors? I'm not suggesting canape here, just that balanced hands with a 4cM and not in the notrump range would open 1M.

I do like to open the major on minimum hands (max 13 points) with 4M5m when using this structure in 3rd seat, but in 4th seat you are perhaps more likely to be preempting your own auction.

[snip advantages+disadvantages of playing 4cM]

IMO the most significant disadvantage of playing strong NT+4cM in 1st+2nd seat is that both

xxx Axxx xx Jxxx
and
Qxx KQxx x Kxxxx

look like single raises of a 1♠ opening - if partner has 12-14 balanced with only four spades, you don't want to be anywhere else. You can try putting one or the other through a non-forcing 1NT response, but you don't really want to play 1NT in either case.

Opposite a 3rd or 4th seat opening, Drury solves this problem comfortably. Even after an overcall, you should still have two ways to raise to 2♠ available.

Quote

So maybe 4cM are best in fourth position? I know a fair number of people will open a 4cM with a "sub-minimum" but it could easily pay to open a 4cM in any case, planning to try 2NT next with the "big balanced hand" or pass partner's 1NT or 2M response with the "weak notrump" hand. This is especially good with Josh's 14-16 notrump suggestion, since you can happily pass/bid 2NT over partner's 1NT response.

That's certainly playable, but I prefer my suggestion in the original post - only open 1M on a four-card suit on hands with 13 points or fewer, open 1♣ on all 17-19 balanced and rebid 1NT to show this hand. It gets the stronger hand declaring more frequently, reduces disclosure of unnecessary information about suit lengths on the way to 3NT and keeps the auction low in case partner is weak or you need the space to explore the best game. It also frees up the 2NT rebid to show a different handtype.

[awm]

Some issues with opening 1♣ on 17-19 balanced:

(1) You invite opponents into the bidding. If they couldn't open, they probably can't make a safe two-level overcall either. But there are plenty of hands that can overcall 1♥ but can't open.

(2) You don't necessarily right-side the contract. Even if you play transfer responses, these are not necessarily on if opponents overcall (i.e. P-P-P-1♣-1♥ and you may well play spades or notrump from the "wrong" side).

(3) If 1♣ passes out, you could be in an awful spot. It's certainly possible for partner to have 0-3 hcp if you have 17-19. Does partner have to bid on this hand, and if so, won't it give you trouble on other sequences? If partner can pass, you could play in a 2-2 club fit with 9 diamonds between you.

(4) If opponents do compete, you're at a disadvantage by not having named a real suit. Say the auction goes P-P-P-1♣-1♠ to partner. If 1♣ guaranteed four (or even three) partner could raise. When 1♣ is frequently two, partner will much more often get shut out. And if the other opponent raises 1♠ to 2♠, you're truly in the soup (could even have a game, but rebidding 2NT opposite a partner who could still be broke seems like suicide).

(5) What are you doing on the weak notrump (say 13 hcp) hands? Do you open and then pass partner's response in your doubleton (4-2 fit here we come)? Again transfer responses might help you if opponents pass but who says opponents have to pass? And again, if you open a balanced 13, the points will often be fairly evenly divided. Your chances in a competitive auction are a lot better if you start the bidding by showing a real suit rather than making a noise.

[MickyB]

awm, on Jan 16 2008, 02:38 AM, said:

Some issues with opening 1♣ on 17-19 balanced:

(1) You invite opponents into the bidding. If they couldn't open, they probably can't make a safe two-level overcall either. But there are plenty of hands that can overcall 1♥ but can't open.

This works both ways. I don't mind finding out that the hearts split 5-x - it will help me decide how to bid the hand and how to play it. I certainly think that the room is more valuable to us than it is to the opposition.

Quote

(4) If opponents do compete, you're at a disadvantage by not having named a real suit. Say the auction goes P-P-P-1♣-1♠ to partner. If 1♣ guaranteed four (or even three) partner could raise. When 1♣ is frequently two, partner will much more often get shut out. And if the other opponent raises 1♠ to 2♠, you're truly in the soup (could even have a game, but rebidding 2NT opposite a partner who could still be broke seems like suicide).

1♣ can be two or three cards only if strong. Otherwise, it is almost certainly 5+clubs. Any seven count with three+ clubs can raise happily, knowing that we are on our way to 3NT if pard doesn't have real clubs.

Quote

(5) What are you doing on the weak notrump (say 13 hcp) hands? Do you open and then pass partner's response in your doubleton (4-2 fit here we come)? Again transfer responses might help you if opponents pass but who says opponents have to pass? And again, if you open a balanced 13, the points will often be fairly evenly divided. Your chances in a competitive auction are a lot better if you start the bidding by showing a real suit rather than making a noise.

With four cards in another suit, you'd normally open that suit, so the only weak NTs that would open 1♣ are 3334 and (233)5 - so I'd expect the frequency of -

• Holding a (233)5
  
• Partner not having a five-card suit
  
• Partner responding in your doubleton
  
• Neither opponent taking a call, and
  
• This leading to a bad board

to be very low.

[dake50]

Why think 1NT in 3-seat is constructive? Then what range?
Obstruct is best on 3-2Majors with 4-4minors and okay on 3-2Majors and 5-3minors as a 1-level preempt of oppts likely M-partial. Partner invited to 3m on 5+suit to barrage. So, minor suits suggested 10-12.