[nullve]

Many of the opening structures posted on BBF make heavy use of two-bids for hands with opening strength, as if the cost of not having those two-bids available as weak preempts is either negligible or more than made up for by the structure as a whole, whatever that means. Of course, that raises the question how to measure things like that, but here I want to ask a different question:

Suppose A and B are systems that perform equally well after opening bids promising "opening strength", where 'opening strength' is defined the same way in A and B. Also suppose that

* A and B use the m and n first opening bids, respectively, for hands with "opening strength"
* in B, the m-n next opening bids are available as weak preempts (yes, assume m>n)
* the remaining opening bids are the same in A and B

My (not necessarily rhetorical) question, then, is: Wouldn't you rather play system B than system A?

[hrothgar]

nullve, on 2015-December-19, 12:24, said:

Many of the opening structures posted on BBF make heavy use of two-bids for hands with opening strength, as if the cost of not having those two-bids available as weak preempts is either negligible or more than made up for by the structure as a whole, whatever that means. Of course, that raises the question how to measure things like that, but here I want to ask a different question:

Suppose A and B are systems that perform equally well after opening bids promising "opening strength", where 'opening strength' is defined the same way in A and B. Also suppose that

* A and B use the m and n first opening bids, respectively, for hands with "opening strength"
* in B, the m-n next opening bids are available as weak preempts (yes, assume m>n)
* the remaining opening bids are the same in A and B

My (not necessarily rhetorical) question, then, is: Wouldn't you rather play system B than system A?

Impossible to say unless you know the expected value and standard deviation for the preempts

For example, assume that A and B are identical, aside from the definition of the 7N bid.

In system A, 7N shows a balanced 36 count
In system B, a 7N opening shows a balanced 9-11 count.

I would prefer to be playing system A

[gwnn]

oops, misread.

[nullve]

hrothgar, on 2015-December-19, 14:06, said:

For example, assume that A and B are identical, aside from the definition of the 7N bid.

Maybe I wasn't clear. By 'first' I meant 'lowest', assuming the usual ordering of opening bids, so the meaning of 7N will be the same in A and B unless m=35.

[gwnn]

If system A can do only as well as system B on opening hands despite having more opening bids, that just makes system A poorly designed (by definition) so I wouldn't want to play it.

[mgoetze]

gwnn, on 2015-December-19, 15:19, said:

If system A can do only as well as system B on opening hands despite having more opening bids, that just makes system A poorly designed (by definition) so I wouldn't want to play it.

Yes, this is the obvious answer. It's so obvious that I almost feel like I've not understood the question.

[nullve]

gwnn, on 2015-December-19, 15:19, said:

If system A can do only as well as system B on opening hands despite having more opening bids, that just makes system A poorly designed (by definition)

Agree. (We're not comparing A and B's user-friendliness here.)

mgoetze, on 2015-December-19, 15:40, said:

Yes, this is the obvious answer. It's so obvious that I almost feel like I've not understood the question.

It's obvious to me as well, so it always baffles me a bit when someone posts a new opening structure without trying to justify the scarcity of weak preempts in it. After all, it's hardly evident from the opening structure alone that there will be (more than) enough compensation.

[glen]

nullve, on 2015-December-19, 17:11, said:

It's obvious to me as well, so it always baffles me a bit when someone posts a new opening structure without trying to justify the scarcity of weak preempts in it. After all, it's hardly evident from the opening structure alone that there will be (more than) enough compensation.

My wife and I play a version of standard having weak twos with five or longer suits. Now most people require six card suits for a weak two. Who needs to justify their approach, does the more freewheeling style need to discuss compensation or does the more disciplined side need to justify their approach? Likewise if you play modified standard where a 2 club opening was a natural weak two, do standard folks need to justify their relative scarcity of weak preempts?

[gwnn]

I mean not sure who has the burden of proof here. More preempts (other things being equal) is more effective, I agree. But also more opening bids (for opening hands) is more effective for the oprning hands, other things being equal. So it's just a tradeoff and both sides should justify themselves, not one side in particular. It's a different question that usually neither side does.

[nullve]

glen, on 2015-December-19, 18:44, said:

My wife and I play a version of standard having weak twos with five or longer suits. Now most people require six card suits for a weak two. Who needs to justify their approach, does the more freewheeling style need to discuss compensation or does the more disciplined side need to justify their approach?

I don't know, although from the quote you might get the impression that I think it's the sheer number, not the quality, of weak preempts that matters. But the conditions in the OP are such that whenever A and B use the same opening bid as a weak preempt, the meaning of that opening bid will also be the same. In fact, A's weak preempts are just a subset of B's.

Quote

Likewise if you play modified standard where a 2 club opening was a natural weak two, do standard folks need to justify their relative scarcity of weak preempts?

Yes, provided the conditions in the OP are met. There's nothing sacred about standard.

[fromageGB]

B, but I look forward to your system where n=1 that is just as effective as A.
In practice, I don't think this is a valid question. A will always be better than B for opening hands, and B will always be better than A for preempts, if the systems are properly designed.

[Free]

nullve, on 2015-December-19, 12:24, said:

Suppose A and B are systems that perform equally well after opening bids promising "opening strength", where 'opening strength' is defined the same way in A and B. Also suppose that

This is too hypothetical to have any meaningful discussion. 2 good systems simply can't perform equally well if one of them needs more constructive opening bids. With a false premise, you can prove anything...

[nullve]

fromageGB, on 2015-December-20, 06:16, said:

A will always be better than B for opening hands, and B will always be better than A for preempts, if the systems are properly designed.

Free, on 2015-December-20, 10:56, said:

2 good systems simply can't perform equally well if one of them needs more constructive opening bids.

Agree, at least if 'properly designed'/'good' means 'optimally designed' in this context.

fromageGB, on 2015-December-20, 06:16, said:

In practice, I don't think this is a valid question.

Free, on 2015-December-20, 10:56, said:

This is too hypothetical to have any meaningful discussion. [...] With a false premise, you can prove anything...

I never said anything about A and B being optimally designed. Clearly, A cannot be.

[PhilKing]

nullve, on 2015-December-19, 17:11, said:

Agree. (We're not comparing A and B's user-friendliness here.)


It's obvious to me as well, so it always baffles me a bit when someone posts a new opening structure without trying to justify the scarcity of weak preempts in it. After all, it's hardly evident from the opening structure alone that there will be (more than) enough compensation.

I think you are missing a key point.

It's not so much that opening at (say) the two level may may the system better on constructive hands. The gains come potentially come in competitive auctions. Fantunes two openings are probably a theoretical and practical loser, but the stats suggest that Flannery, for example is a big winner on it's own, and helps improve other sequences. Essentially it gains through preemption. Strong preempts can be just as effective as weak ones.

Comparing the trade-off from losing a multi 2♦ or a weak 2♦ in order to play a constructive 2♦ of some sort is not easy, but I would not start from here.

[awm]

It's important to consider competitive auctions in this setting as well. To give a simple example, we might have:

A = 2/1 with 3NT = 12-15 hcp and a 7+ card major
B = 2/1 with 3NT = 5-9 with 6/5 or more in the minors

If opponents are passing, it's likely that B does just as well as A on the "12-15 hcp and 7+ major hands"; in fact you have more space for slam bidding and may do better in those auctions. However, in real bridge the 3NT opening makes it quite difficult for opponents to find their sacrifices or even making games, by taking away all their space. Further, the "6/5 minors weak" meaning is sufficiently rare that the benefits from having that call available to system B will be pretty small. So it's easy to believe that A is "better" even though it has one less weak preempt and the 3NT hand type is better handled (assuming opponents silent) by opening 1M.

[ulven]

I'm pretty done with the weak opening preempts. Constructive intermediate 2-openings is my choice now.
And a quick glance at the Denver Sun-Mon BAM showed that six 2-level openings resulted in six wins (against competent opps, such as Gawrys, chinese etc).
I remember two occasions where we didn't have a weak 2M available and had to come in later into the auction. That was another two wins.
So, no fancy weak for me. How do I measure? With at-the-table results against good opposition.

[straube]

fred, on 2008-September-26, 10:20, said:

I recently spent a few happy years playing on a team that included Russ Ekeblad and Ronnie Rubin. They use a strong club canape system that Russ devised which utilized 2-level openings to describe various 11-15 HCP hands. There were no weak 2-bids in their system. When they were dealt an appropriate hand for a weak 2, they would either Pass, open 3, or open 1.

I can't make any claims about the overall effectivenes of their approach. For sure their system gained on a lot of hands and for sure their results as a partnership were consistently strong (though probably this had more to do with the fact that they are both excellent players who worked hard on their partnership than anything else - I am sure they would have done just fine if they played "standard").

But I can tell you that whenever I was playing at the other table and one of my opponents opened a weak 2-bid, I was terrified. We regularly lost IMPs on these hands. We did gain plenty of IMPs as a result of other aspects of the Ekeblad-Rubin system, but I doubt their 2-bids directly contributed much in this regard. My sense is that at best we broke even when their system forced them to open at the 2-level.

That is because the particular 2-bids they played were designed to "fill holes" (ie providing them with *some* way to bid hands that would be easier to bid if you could start at the 1-level as you could in a natural system) rather than to give them any inherant advantage over people using "normal" systems when hands appropriate for their 2-bids were dealt.

Of course it could well be the case that this allowed them to gain more from their 1-level openings that the IMPs they lost from their 2-level openings (and lack of weak 2-bids).

My teammates inability to open weak 2s was especially worrisome when we were playing against teams that we expected to beat (ie most of the time). Under such circumstances it is in the interest of the favored team to decrease volatility. Creating swings on normal and common hands is not a good thing in this regard. I know my friend Russ would respond to this point with something like "I don't care if we lose the occasional match to a bad team if it increases our chances of beating Nickell when we face them". Maybe he is right...

In case this post is read as being critical of Russ and Ronnie, I should emphasize that this was not my intention. They were great teammates who produced excellent results over the course of several years. Their system is very clever and resulted in a LOT of gains for our team.

But their lack of weak 2-bids, more than ANY other aspect of their system, did cause me to lose some sleep at night <img src='http://www.bridgebase.com/forums/public/style_emoticons/<#EMO_DIR#>/smile.gif' class='bbc_emoticon' alt='B)' />

Fred Gitelman
Bridge Base Inc.
www.bridgebase.com

[nullve]

PhilKing, on 2015-December-20, 12:23, said:

I think you are missing a key point.

It's not so much that opening at (say) the two level may may the system better on constructive hands. The gains come potentially come in competitive auctions. Fantunes two openings are probably a theoretical and practical loser, but the stats suggest that Flannery, for example is a big winner on it's own, and helps improve other sequences. Essentially it gains through preemption. Strong preempts can be just as effective as weak ones.

Comparing the trade-off from losing a multi 2♦ or a weak 2♦ in order to play a constructive 2♦ of some sort is not easy, but I would not start from here.

I'm certainly open to the possibility that if e.g.

S = 2/1 with Flannery (e.g. Levin/Weinstein)
S' = 2/1 with Weak 2♦ (e.g. Chagas/Villas-Bôas),

then S might well be the superior system partly because of the decision to play Flannery (strong preempt) instead of Weak 2♦. In my defence, the (somewhat artificial) assumption was that A and B perform equally well after non-weak openings, which with A = S and B = S' could then only mean that A suffered from poorer design after other non-weak openings. (Hence the artificiality of the assumption.)

awm, on 2015-December-20, 12:24, said:

It's important to consider competitive auctions in this setting as well. To give a simple example, we might have:

A = 2/1 with 3NT = 12-15 hcp and a 7+ card major
B = 2/1 with 3NT = 5-9 with 6/5 or more in the minors

If opponents are passing, it's likely that B does just as well as A on the "12-15 hcp and 7+ major hands"; in fact you have more space for slam bidding and may do better in those auctions. However, in real bridge the 3NT opening makes it quite difficult for opponents to find their sacrifices or even making games, by taking away all their space. Further, the "6/5 minors weak" meaning is sufficiently rare that the benefits from having that call available to system B will be pretty small. So it's easy to believe that A is "better" even though it has one less weak preempt and the 3NT hand type is better handled (assuming opponents silent) by opening 1M.

Agree with everything, but see my reply to PhilKing, who I basically also agree with.

[kenrexford]

There are several things to keep in mind, IMO.

First, systems that have intermediate 2-level openings tend to be strong club systems with limited openings. As such, the tighter range for the opening bid means a tad more flexibility in opening lighter hands generally. Thus, some of the high-end weak two's (10-11 for sure) can be and usually are opened 1♥ or 1♠ anyway.

Second, the mere fact that you can open more hands is not necessarily a plus, or a clear plus. The assumption is made that an ability to open a weak 2♥, for instance, is a plus. While this seems obvious, because most people play a weak 2♥, this is not as obvious as it seems. Results determine whether having the ability to open a weak 2♥ actually benefits anything. My experience is that the weak 2-bid, while occasionally gaining, is actually quite rare as a noted benefit. When I do not open weak twos, the end contract is often the same, because all roads often lead to Rome. Some passes with these hands work magic, where we get to a better partscore or stay out of the way. I only recall a small handful of deals where we lost something, and usually not much loss, and often because one of us made a different miscalculation that contributed to the problem. In other words, the weak two often makes easy that which can still be done anyway.

Third, for what it is worth, I included a lengthy analysis of this issue in my MICS book. Not all of us who use two-level intermediates ignore this potential issue.

Finally, while I agree with Fred that many systems have two-bids as system purifiers, and thus the two-bids suck, other systems are not so structured. I have added most of the MICS two-bids into natural systems like 2/1 in various partnerships with great success because I like them so much and because they are effective bids in their own respect. Those systems that have hole-filler two-bids are, IMO, a result of people not liking to open an intermediate 2-bid, or bending backwards to try to keep weak two's, and thus the structure sucks. If you embrace the intermediate and actually try to open these as often as possible, in a manner that is sound judgment, you end up with a better structure and better results.

[nige1]

nullve, on 2015-December-19, 12:24, said:

Many of the opening structures posted on BBF make heavy use of two-bids for hands with opening strength, as if the cost of not having those two-bids available as weak preempts is either negligible or more than made up for by the structure as a whole, whatever that means. Of course, that raises the question how to measure things like that, but here I want to ask a different question: Suppose A and B are systems that perform equally well after opening bids promising "opening strength", where 'opening strength' is defined the same way in A and B. Also suppose that

* A and B use the m and n first opening bids, respectively, for hands with "opening strength"
* in B, the m-n next opening bids are available as weak preempts (yes, assume m>n)
* the remaining opening bids are the same in A and B

My (not necessarily rhetorical) question, then, is: Wouldn't you rather play system B than system A?

System B seems preferable (given that opening strength bids are equally effective in systems A and B; also given that system B's weak pre-empts are sensibly defined -- for example, you don't open 7N on a Yarborough). This conclusion seems inevitable unless you believe some pre-empts to be a liability.