[barmar]

What that example shows is that there's a middle ground. With too few conventions, you run into problems on many common hands, and this is a problem. It's like trying to have a conversation with only the vocabulary of a 3-year-old.

So there's a good reason why the conventions that have become extremely popular have done so. Transfers solve a real problem due to the preemptive nature of opening 1NT -- there isn't enough room to show all the combinations of weak/invitational/forcing hands and 5/6-card suits with natural responses. New Minor Forcing is also a simple solution to a common problem.

But as you continue to add conventions, there may be diminishing returns. I play Mexican 2♦ with my regular partner, but it sometimes seems like we go weeks without it ever coming up. And when it does come up, we don't always remember some of the uncommon followup bids correctly.

[awm]

It can often be hard to determine whether system caused a success. To give a few examples:

1. I opened 2c showing 6+c, 10-15 hcp. LHO overcalled 2d. With four clubs and a diamond void, partner immediately went slamming and we found an excellent 6c. If I had opened a standard 1c, partner would surely bid 1s rather than raise clubs opposite what is often a balanced hand. Opponents would preemptively raise to three or four diamonds. Do we get to 6c? Maybe... but its tougher. Most of the field didn't, but we are a better pair than most of the field so maybe not a good comparison.

2. I opened 1s on 9 hcp (systemic). LHO bid 1nt (a weak NT in their methods). Partner bid 4s on a flat 13 with four spades, which was cold. If I had passed on first seat (as I would in standard) LHO opens weak notrump. Do we get to game? Well it is possible I balance despite having a crummy spade suit (and it may depend on methods there too) and its possible we reach game... but there are some judgment calls so who knows.

3. RHO opens 1c strong and I bid 1d showing diamonds or majors. LHO doubles showing any 5-7 hcp. Partner bids 2d pass or correct. RHO has a 4414 but their agreement is that double here is penalty. So he bids 2s and they play there, partner having a 2533 six count with 6h making. Bad bidding by opponents obviously. Do they reach game or slam if I bid 1d (natural)? Maybe... but I will never know.

[CSGibson]

This is something that's individual, and not a community decision. I know a player who plays a complex relay big club canape system, who never messes up his system despite 700 pages of system notes. He obviously is not suffering from memory burden, and if he were to find a partner with a similar ability to remember, than memory burden would not be a factor (or would be a very low factor, possibly with fitness in terms of potential mental fatigue).

I play a complex and highly artificial 2/1 system with a partner which has about 130 pages of notes - and we could probably add another 30 pages if we were so inclined. I don't have a problem remembering the system with minimal effort. My partner sometimes does (he's 40 years older), but not frequently enough to outweigh the system gains. We also play at least twice a week, have been partners for more than 2 years, and regularly discuss system, follow-ups, and the like outside of our playing session. I suspect that he also studies the system notes daily as part of his gym regiment (he's said something similar to that).

Your memory-burden vs artificiality "argument" is really a question of how much effort you need to remember stuff and how much effort you are willing to put in. If the answer is a lot, and not a lot, then go with non-memory intensive stuff. If you are willing to put in the time necessary to learn it, then memory intensive is not an issue, and you can look strictly at the bridge losses & gains of a convention in order to decide whether to play it.

[glen]

I think the "natural" is misleading here. For example, Stayman is not natural, but playing a natural approach does not work as well.

[chasetb]

I agree with all 4 posts above. The complexity of the system does depend on the person - I know with the right partner and a few hours a week to practice the system, I can handle the system Chris mentioned above (can you give me his contact info?). I also know that except my Precision partner who I won Flight C NAP with, no one else I play with can handle more than 20 pages; age and desire being the key factors. No one else even has a write-up.

There are a few bids/conventions, that while aren't 'Natural', are so commonplace and well-known they have to be used. You're nuts if you don't play Stayman, Jacoby transfers over a Strong NT (Keri if you have the fortitude), Negative Doubles, and Blackwood. A few others may be artificial and a tad complex, but the gains from them are too great to ignore. A Larry Cohen link a friend gave me covers everything in a nutshell. http://www.larryco.c...px?articleID=54

I think one of awm's examples isn't quite ideal, though his post is truthful. #1 is a clear system success. I love the 2♣ bid, and while I'm a dinosaur because I like to open it on a GREAT 5♣ - 4M rather than requiring 6+, my experiences are that it is a huge winner, and it's because of knowing distribution and how to bid after it. Partner and I found 4♥, 4♠ in competition, 5♣ when 3NT doesn't make, and +800 because of the bid (at Memphis).

EDIT - Larry also has a post where he supports Precision. I just found it while looking through his stuff. WOOHOO! http://www.larryco.c...px?articleID=82

[JLOGIC]

Having numerous agreements and a reasonable system is hugely important if you want a lot of success. Which of the possible reasonable agreement/system you play probably doesn't matter much.

[FrancesHinden]

32519, on 2012-April-24, 00:09, said:

At the end of the day, when you sit examining the traveller (at match points), how many times did you end up in a superior contract versus the rest of the field using scientific methods?
What about you?

My system is designed for IMPs, not for matchpoints.
This is a serious comment: complex agreements are far less important at matchpoints than they are at imps. Having a fast auction to the normal contract and gaining an overtrick by giving less away is worth just as much at matchpoints as having a delicate auction to a minor suit slam. At imps one is 13 imps the other is 1 imp. At matchpoints, if you aren't sure whether 6M or 6NT is going to be a better spot, you bid 6NT. At imps it can be worth another 13 imps getting that one right. Big pairs events are often won by scratch or semi-scratch partnerships (between two very strong players), big long teams events rarely so.

[FrancesHinden]

JLOGIC, on 2012-April-24, 13:14, said:

Having numerous agreements and a reasonable system is hugely important if you want a lot of success. Which of the possible reasonable agreement/system you play probably doesn't matter much.

I was going to say something like this (for some definition of 'reasonable'), although I think it does matter.
We have close on 100 pages of detailed agreements, but at least a third of these are probably covered by what you might consider 'natural' (ish) bidding, it's just we've discussed things and written them down. (When they open a short club or precision diamond, when are our club/diamond bids natural? What are our responses to a sandwich NT, and why they are different against Polish club pairs? What does a 4-level fit bid look like? What's the difference between a 4-level fit bid followed by a double and a 3-level fit bid followed by a double if both were available? )

Also, on a different tack, the following two statements are both true:

- Once your agreements are better than a certain minimum level, expert cardplay and defence are good judgement are worth more imps than a more complex system
- Detailed scientific agreements gain imps

There's a limit to how much and how quickly I can improve my cardplay. However, I have no trouble remembering system, so one way my partnership generates swings in is by having very good and thorough agreements. The two are not exclusive.

[mgoetze]

FrancesHinden, on 2012-April-24, 13:41, said:

What are our responses to a sandwich NT, and why they are different against Polish club pairs?

Because they've only bid one suit, so system on makes more sense?

[phil_20686]

Obviously Justin basically nailed it. If you think natural methods means you dont need about a million agreements then you probably just don't bid very well.

Take an easy auction like 1H-2d-2H in 2/1.

Here are some questions:
(1) Have you denied a 4522 12 count?
(2) Have you promised a sixth heart?
(3) Can you be 46 minimum?
(4) Would you bid a 46 with extras differently from a 46 minimum?
(5) Can partner have 4D, or has he promised 5.
(6) Can partner have a 4342 shape?
(7) If I bid 3D instead of 2H does that show extras?
(8) if I bid 3D instead of 2H does that promise 4D?
(9) If I can raise 3D on three is a fourth diamond always extras, or is it HCP orientated?
(10) If responder rebids 2N does that deny 4 spades?
(11) If partner bids 4H is that fast arrival, or does it have a specific meaning.
(12) Can opener have a hand with AKJTxx heart and 15-18 HCP?
(13) Can opener have a hand with Kxxxxx hearts and a 19 count?
(14) Can opener have a weak NT?
(15) If opener bids 2N over 2D is that natural, and what is its strength.
(16) What to all the jump bids mean now? (3S/4C/4d)
(17) If partner bids 3C over 2H, does that show extras? Does it carry the implication of extra shape?
(18) If I rebid 3N over 2d, what does that show?

Here are a bunch of questions, about one pretty simple auction, and even if you play natural I need to have agreements about all these sequences. If I play artificial methods, I have to answer the same 18 questions about how to bid certain kinds of hands. Maybe I think oddly by thinking of my agreements in terms of how you bid classes of hands, rather than in terms of "what does sequence x show", but doing so makes it clear that the key is to differentiate as many hands as possible, and if a system has less memory load, it can only be because it combines the same set of hands into fewer sequences, reducing the scope of judgement. It seems to me then, that natural and artificial methods with the same level of definition, necessarily have the same memory load.

Of course, humans are not so simple, and repetition in agreements is important in reducing memory load, but natural and artificial systems seem to have identical scope. If I give my relays all the same answers, I (normally) lose some efficiency but make it easier to learn. The same thing happens in natural systems. I choose to have similar agreements after 1h-1N-3d and 1S-1N-3C even though in the second auction my continuations could make better use of the space.

As far as I can see then, whether I play natural or artificial methods I will always have agreements up to the level of detail I am comfortable with. I think I am about average in respect of the level of detail i can handle.

=====================

A slightly unrelated point. I have found that memory load is almost never the issue. Rather the issue is that in unfamiliar systems I do not always draw all of the available inferences. I have never had a problem with "what do my continuations mean now" in my regular partnerships. We have however fairly often had disagreements about whether you would/should/could have bid differently with a particular hand. This is partly a system disagreement, and partly a judgement issue, and occasionally a case of "i thought of sequences a and b, but not c" and so we bid a bit wrong. However, I think this issue of inferences is what most people mean when they think of memory load. If you have to sit at the table and work out these inferences, its very tiring. It distracts from your table focus, and cardplay suffers. However, I think this is a slightly different problem from what average players think of as memory load, and I think its much more important.

[mikeh]

My own experience is that having extremely detailed agreements, whether they include exotic methods like relay or simply 'naturalish' discussion about what to do in low frequency occurrences or against methods one rarely encounters, is hugely important.

However, I also think that a lot depends on whether you are playing primarily in one partnership. If you are playing, on a regular basis, with multiple partners, then try to have everyone playing as much the same as possible, since otherwise you may get confused as to which agreement you have with this particular partner even, and probably especially, in the agreements in effect are 'naturalish'.

You aren't apt to confuse the agreements you have in a strong club method with the agreements you have in an Acol method, at least in terms of your opening bid and response auctions.

If you are playing a complex artificial system, then you should (in my experience) expect your cardplay and judgment to be adversely affected until the system is internalized....memorized to the point that it seems easy. I played what I am sure was the most complex method ever played by a Canadian pair (until recently), and my game suffered terribly for the first two years, and then it seemed to click, and became so easy that I still look back to those days and regret that I can't get anyone to put in that kind of effort...tho I doubt that I could either, now.

So: if you have a keen mind, a good memory, and a partner willing and able to do the work, and you are really serious about the partnership and play or practice frequently, then the more complex (but logical) the method, the better.

Having said that, the reality is that a lot of mediocre players seem to think that building a complex system will turn them into experts, and that is not true in the least. Experts who play complex methods win primarily because they are experts.

[nige1]

It seems bizarre to apply the term Natural to a game. In Bridge "Natural" seems to be a euphemism for "what I'm accustomed to play" or "what I like". For example, the ACBL -- and now the EBU -- define bidding a three-card suit as "Natural", however unlikely it is to become the final contract. In their seminal "Partnership Bidding at Bridge", Segal and Robson describe "Fit jumps" as "Natural". Some even claim cue-bids and splinters to be "Natural".

For decades, the Sharples brothers were the best bidders in Britain. They played a "Natural" system based on Acol. But they continuously honed their understandings and their methods became as sophisticated as any "Scientific" relay system.

It does take time and effort to learn a system. A catastrophic mistake with a conventional call can cancel any small gains that it achieved. The Sharples practised a new version of their 2♦ bid for a year before they risked it at the table.

Nevertheless, many "Artificial" calls have obvious advantages if you can remember them. (e.g. Take-out doubles, Stayman, Blackwood, Transfers, Cues).

All Bridge innovations are resisted as tantamount to cheating. Resistance has progressively hardened. The game is now static and stultified. The enthusiastic young system guru will find few opportunities to practice his forcing pass, multi-2♠ or encrypted continuations

The Italians were World Champions for thirteen years, using "Artificial" systems. They benefited from opponents' unfamiliarity and lack of preparation.

IMO, however, the main edge of successful bidders is that their comprehensive understandings eliminate a lot of "judgement" (aka "guesswork"). This saves them time and conserves their adrenalin for difficult defences, hard declarer plays, and complex competitive auctions. World-championship records provide ample evidence of this.

[mgoetze]

nige1, on 2012-April-24, 18:40, said:

The Italians were World Champions for thirteen years, using "Artificial" systems. They benefited from opponents' unfamiliarity and lack of preparation.

But they also benefitted from knowing whether or not partner has the king of trumps...

[barmar]

glen, on 2012-April-24, 12:00, said:

I think the "natural" is misleading here. For example, Stayman is not natural, but playing a natural approach does not work as well.

Right. Natural vs. scientific is not a dichotomy, it's a continuum. Some systems and styles are more natural than others.

But I think players do have a general feeling that some systems emphasize natural bidding as their baseline, while others make heavy use of artificiality. E.g. systems where relays are used extensively after establishing a game force.