Consistency in System Design

[the hog]

A number of posts over the last week made me somewhat curious. I am particularly referring to the following two:

1S 4H - splinter or not

(1m) P (1H) 2H - natural

 

When I was playing seriously a few years ago, partner and I had at least a 2 hour system/bidding session every week. One of the things we tried to do was to make the system as consistent as possible. As we sometimes played up to 60 boards a day in serious competition, we wanted to elimilate as many memory lapses as possible. We felt consitency was the answer -

 

so 1M -4level bid, (except 1H -4S of course), was ALWAYS a splinter.

(1m) P (1H) 2H or similar sequences - was ALWAYS a shapely 2 suiter.

There were many other similar situations. Perhaps our system was not absolutely optimum, but at least we went a long way to eliminating memory errors when we were very tired.

(Just as an aside here - we played a relay system and in contrast to the vies of many here, I really feel that a relay system is easier than say something like souped up 2/1. However this discussion is not the purpose of this post.)

 

What I am curious about is how many here regard this form of consistency in system design as important? Judging by the posts some of you make, I would say this is not an important consideration for many of you

[jdonn]

I don't think your examples point to inconsistency.

 

The problem with the 1♠ - 4♥ auction is that there are two logical, simple, and consistent rules that are common among many partnerships. Double jumps to the 4 level in a lower suit are splinters, and game bids at a player's first turn to call are natural. Obviously the problem is a partnership may have adopted both without realizing they conflict. I don't think there is any way around it but for it to occur to people to discuss this exact auction.

 

I also don't see inconsistency with playing 2♥ natural on the second auction. Isn't it just as consistent to say such a bid is always natural as it is to say the bid is always a 2 suited hand?

 

I think in most cases you have to start from a very consistent base to be successful. From there you can add exceptions as you desire, though I tend to think most partnerships are too quick to add them.

[Fluffy]

As general principle I trully believe in keeping thigns simple, but on those 2 concrete sequences, it just happens that whatever my partner is around my area, the standard for both os us will be natural and natural.

[andy_h]

I agree that consistency is a good starting point in designing a system. That way it's easier to remember while you can get used to it/learn the implications of (lack of) bids. But once you get used to it, you start to realise that many situations will come up in which you would need exceptions and bridge is always about exceptions. And soon enough maybe those exceptions will turn to "another consistency".

 

For example, the (1m) - p - (1H) - 2H is natural, so thereby you may have the tendency that the 4th seat's bid of opponents suit is natural. It also applies if the auction was like (1m) - X - (1M) - X/2M, meaning penalty or natural.

 

So this pretty much agrees with what jdonn has said.

[the hog]

I don't think your examples point to inconsistency.

 

The problem with the 1♠ - 4♥ auction is that there are two logical, simple, and consistent rules that are common among many partnerships. Double jumps to the 4 level in a lower suit are splinters, and game bids at a player's first turn to call are natural. Obviously the problem is a partnership may have adopted both without realizing they conflict. I don't think there is any way around it but for it to occur to people to discuss this exact auction.

 

I also don't see inconsistency with playing 2♥ natural on the second auction. Isn't it just as consistent to say such a bid is always natural as it is to say the bid is always a 2 suited hand?

 

I think in most cases you have to start from a very consistent base to be successful. From there you can add exceptions as you desire, though I tend to think most partnerships are too quick to add them.

Perhaps I didn't make myself as clear as I should have.

By consistency I would mean a rule similar to say "A bid of the opponent's suit is ALWAYS natural (conventional?), at the one or 2 level regardless.

 

Or say, a double jump is ALWAYS a splinter/fit showing/natural.

 

I also don't mean to harp just on these two examples but on the principal of consistency as a whole.

[awm]

I'd look at this slightly differently, and say that it is valuable to have simple rules which are as general as possible. Such a structure means that in an unfamiliar auction, you can apply the rule and be likely to be on the same page with partner, even if the meaning that the rule gives you is not really "optimum." As many people have commented before, there might be slight advantages to one set of agreements over another, but there is a huge advantage to having agreements over not, and being on the same page with partner.

 

Anyway, in the case of the cuebidding auction, I prefer a set of rules which is a bit more complicated than "cuebid always shows a two-suiter with unbid suits" just as for doubles my rule needs to be a bit more complicated than "double is takeout." My rules are:

 

(1) If partner has bid a suit naturally, then a cuebid of a suit bid naturally by opponents cannot be natural. Usually such a cuebid shows either some good raise of partner's suit, or a strong hand with no clear direction.

 

(2) Otherwise if we have bid a suit naturally, then bidding a suit shown by the opponents shows a strong hand.

 

(3) If our side has never bid a suit, and opponents have bid only one suit naturally, or have bid and raised a suit, then bidding their suit cannot be natural. In this case a cuebid at the two-level is michaels, a cuebid at the three-level is a stopper-ask, and a cuebid at the four-level shows a (possibly unspecified) two suited hand.

 

(4) If opponents have bid two or more suits naturally, then cuebidding a suit which they have bid but not raised is normally natural (or if this does not strictly make sense, it shows values in that suit for a possible notrump contract).

[PrecisionL]

Playing a strong club system, we (Keylime & I) keep reviewing it for what we call "Parallelism."

 

Many sequences are similar to ease the memory load:

 

1a) 1C - 1M - Relay - all 2-level rebids by responder are similar distributions no matter which Major his first response was in.

 

1b) A 2NT rebid by responder designates a freak hand: 6-5 / 6-4

 

2) The same design was used for our intermediate two bids and the GI or better response of 2NT. Rebids by opener show if he is minimum or maximum.

 

3) 1M - 2C = A & GF is similar to 1C - 1M - Relay, but 1-step higher.

 

Larry

[barmar]

Using the 2nd sequence to show a 2-suiter is likely to be redundant in many systems. Just about everyone plays Unusual NT to show the unbid suits. And if it bothers you that this forces you to the 3 level if partner prefers spades, you can agree to play Sandwich 1NT, which allows your side to stop in 2om or 2♠.

[gnasher]

Personally I'm a strong believer in consistency of methods, and am happy to play something slightly inferior if it makes the entire system easier to remember. I'd rather not have to spend time and energy trying to remember how we play a particular sequence.

 

For example, I've just agreed with one of my partners that when we use two artificial bids to distinguish between a minimum and a maximum, the lower one always shows the weaker hand. In some sequences that's necessary, because we need to be able to pass opposite the weaker hand, or because having moderate values means that we need more room to investigate the best game. In other sequences it's probably inferior, because the extra space would be more useful opposite the stronger hand. However, I think that the benefits of making the two consistent easily outweigh any disadvantages.

 

Regarding what people post in these forums, I can't speak for others, but I sometimes suggest methods that I wouldn't play myself, because I'd find the complexity hard to cope with.

[whereagles]

I am willing to waive some precision to make it easier on my memory. My pard sometimes thinks otherwise, though... lol.

[Codo]

I belive that even complex systems should be very well designed regarding to consistence. I am sure that my play is significant worse, when I am playing a system, where I have to remember many different pathes.

 

During the years, my system notes got fewer and fewer pages and more and more as Gnasher wrote: Easier to remember but maybe slightly inferior.

[kenrexford]

Consistency should always be a factor in deciding how to structure a particular set of bids that could be structured in more than one way. If it truly does not matter how you structure things, then consistency/parallelism should be the default.

 

Consistency occasionally yields non-optimal conclusions. When that occurs, the solution is (1) consider changing the existing structure to account for the new instance (if not overwhelming), (2) consider a discussed exception, or (3) forfeit the possible advantage from a re-structure because the gain of consistency and memory outweighs the risk of confusion in this and the other sequences.

 

That said, consistency can be overdone. Obviously, treating any 2♣ bid, whether opening, responding, or overcalling, as 21+ HCP would be consistent, but stupid. One could say that a system improves in utility as the rules become less and less consistent, or diverse and situation-specific, with a constant counterweight of memory. The key is balance. Where to strike that balance is always a partnership choice (unless the partnership has random inconsistencies, which is without debate bad). Arguing between different lines is like Dems and Reps claiming that the other side is completely liberal or completely conservative.

[jdonn]

Perhaps I didn't make myself as clear as I should have.

By consistency I would mean a rule similar to say "A bid of the opponent's suit is ALWAYS natural (conventional?), at the one or 2 level regardless.

 

Or say, a double jump is ALWAYS a splinter/fit showing/natural.

 

I also don't mean to harp just on these two examples but on the principal of consistency as a whole.

I don't think it's that simple. What is the opponents' suit? If they make a negative double is the unbid major their suit? What about an unbid minor? What if they open a 2+ 1♣? What if they open 1♣ and later guarantee exactly a three card suit? What if your partner shows a suit with a lower bid and the opponents overcall it naturally on the one level? And even if it's always natural/convention, what sort of hand does it show? (Alert, partner's bid is conventional but I don't know what it means!) I'm not saying you couldn't answer these questions (even easily answer them), just that making a giant blanket rule is rarely the answer.

[peachy]

I don't think your examples point to inconsistency.

 

The problem with the 1♠ - 4♥ auction is that there are two logical, simple, and consistent rules that are common among many partnerships. Double jumps to the 4 level in a lower suit are splinters, and game bids at a player's first turn to call are natural. Obviously the problem is a partnership may have adopted both without realizing they conflict. I don't think there is any way around it but for it to occur to people to discuss this exact auction.

 

I also don't see inconsistency with playing 2♥ natural on the second auction. Isn't it just as consistent to say such a bid is always natural as it is to say the bid is always a 2 suited hand?

 

I think in most cases you have to start from a very consistent base to be successful. From there you can add exceptions as you desire, though I tend to think most partnerships are too quick to add them.

Perhaps I didn't make myself as clear as I should have.

By consistency I would mean a rule similar to say "A bid of the opponent's suit is ALWAYS natural (conventional?), at the one or 2 level regardless.

 

Or say, a double jump is ALWAYS a splinter/fit showing/natural.

 

I also don't mean to harp just on these two examples but on the principal of consistency as a whole.

Your suggested "rule" is not consistent anyway, in fact I would consider it unplayable because it changes the bidding structure so much. If [by your suggested rule] bidding the opponent's suit on the one or two-level were always natural, then you cannot play Michaels and you cannot make a forcing bid by cuebidding the opponent's suit lower than the 3-level, for couple of quick examples.

[DrTodd13]

Playing a strong club system, we (Keylime & I) keep reviewing it for what we call "Parallelism."

"Symmetry" is the word that is most often used for this.

[Free]

I like consistency. For example, showing shortness is done the same way throughout the whole system. I think it's important to lower the memory load.

[firmit]

I like consistency. For example, showing shortness is done the same way throughout the whole system. I think it's important to lower the memory load.

I agree. I am no expert - faaaar from it. I have only been playing for a couple of years. However - I do have some feelings with regards to consistency with the system I'm using.

 

(1) Every jump in partners suit is pre-emptive ( obviously ) - regardless of opps interference

(2) Bid vs dbl vs pass - is respectively (normally) patterning/weak/strong if initially partner made ANY forcing bid, after opps make any action besides pass.

(3) When competing - if new suit is bid by advancer, one always have secondary support in overcallers suit.

 

Everything else is common sense! This works for me. I may be way off, though! ;)

[hrothgar]

Playing a strong club system, we (Keylime & I) keep reviewing it for what we call "Parallelism."

"Symmetry" is the word that is most often used for this.

Not sure whether or not this is true:

 

I might be projecting too much of my own experiences with relays onto this, however, from my perspective, "symmetry" has to do with

 

1. Resolving the same shape at the same level

2 Using a "reverser" bid

 

I think that you can achieve parallelism without symmetry... Moreover, in many cases, achieving perfect symmetry requires that you abandon parallelism.

 

Case in point... Consider the following bidding structure over a strong club opening:

 

1C - 2H = GF with 4 Diamonds and 5+ Clubs (Reversers)

1C - 2S = GF with at least 5-5 in the minors

1C - 2N = GF with 4 Clubs, 5+ Diamonds and 0-1 Spades

1C - 3C = GF with 2=2=5=2 shape

1C - 3D = GF with 3=1=5=4 shape

1C - 3H = GF with 2=1=6=4 shape

1C - 3S = CF with 2=0=7=4 shape

 

This structure is parallel and symmetric

 

Consider the following

 

1C - 1N = GF with a balanced hand

 

1C - 1N - 2C

 

2D = 4+ Hearts

2H = 4+ Spades (0-3 Hearts)

2S = Any 4-3-3-3

2N = 5332 with long Diamonds

3C = 2=3=4=4 OR 4=2=4=4

3D = 2=3=3=5 shape

3H = 3=2=3=5 shape

3S = 3=3=2=5 shape

 

This structure is parallel but NOT symmteric

[DrTodd13]

To me, symmetric is the general sense just means that several different starting points converge on identical or similar sequences with perhaps suits shifted around. In so doing, you minimize the memory load. Again, to me, symmetric doesn't have anything to do with the way you choose to divide and express hands but only speaks to converging on that methodology as often as possible. So, to me, citing one way of expressing balanced hands doesn't prove whether your system is symmetric or not. I you mean something else by parallel then I don't think I've seen a definition of it.

[mikestar]

As long as it is tempered by common sense, striving for consistency in system design is a great virtue IMHO.