[32519]

The fact that one bidding system is more popular than another doesn’t necessarily mean that it is better e.g. 2/1.
Neither the fact that the system may be easy to understand, again, e.g. 2/1.

A defining characteristic (for me anyway) making one system better than another would be the ability to describe a wider range of hand patterns no matter how infrequently those patterns may occur. Shape is often cited as more important than HCP, these hand types being more suitable for offense than defense. A wide range of conventions have been created attempting to describe shapely hands. The problem however is that you often have to make a choice between which conventions to include and which ones to exclude. (This theme is to form “Part 2” of this thread).

I don’t have any experience with relay systems, but am told that they are good to determine partner’s exact distribution. How relay systems are affected by opposition interference is an even greater mystery for me.

So then, what does make one bidding system better than another?

[the hog]

"So then, what does make one bidding system better than another?"

The fact that you and your patrner are comfortable in playing it. Conservative players should not use a highly aggressive system, such as "The science" for example, and aggressive players should not play "Roth Stone".

[Siegmund]

Quote

The fact that one bidding system is more popular than another doesn’t necessarily mean that it is better e.g. 2/1.

True.

Quote

A defining characteristic (for me anyway) making one system better than another would be the ability to describe a wider range of hand patterns no matter how infrequently those patterns may occur.

I don't think you will get much agreement for that.

From a matchpoint perspective, the best bidding system is the one that gets you to the best-scoring contract as often as possible. (For IMPs the concept is similar but harder to put into one readable sentence.) That may very well mean deliberately providing several sequences to distinguish small variations between common hand types, and providing no sequence at all except "close your eyes, pray, and blast" for really rare hand types.

It is very hard to quantify differences in quality between systems that solve the same problems in very different ways. Generally there are a few methods out there that are so bad it's easy to prove they are bad, but among the leading contenders the decision is usually made either based on taste, or based on how a few "pet" problems that particulary worry a pair are handled.

There are also often outside constraints - ease of memory, compliance with system regs, etc. - that cut down the universe of allowable systems in odd ways. In the ACBL, for instance, you are allowed to use transfer responses after 1NT and higher openings, and to use transfer rebids anytime, but not allowed to use transfer responses to 1-of-a-suit openings, in General Convention Chart events... so in a strong club system, you may not be able to play the same responses to 1S as after 1C(strong)-1D(weak)-1S, even if you want to. Play the suboptimal but always-legal method, or memorize two different structures?

---

The above mostly applies to players above a certain level of mediocrity.

With a random flight B players grabbed at the partnership desk, the big advantage of 2/1 is that fun scores like +230 that result from "is it forcing?" accidentally happen less often in 2/1-without-discussion than in SA-without-discussion. 2/1-with-careful-agreements, SA-with-careful-agreements, Polish-club-with-careful-agreements, etc, all do better than any sloppy system, and approximately equally well as each other.

[Flameous]

Bidding has many conflicting goals.

First we can split it to constructive bidding and destructive bidding.
Here I mean most one level openings to be constructive and all pre-empts destructive although that's a bit too black and white.

For some little extra goals we have lead directing bids and such, but I don't consider them now.


Then we have couple of goals in constructive bidding that conflict.

One is to conserve room to exchange more information with partner, other is to bid to the right spot asap to tell opps the least information. Now I'd say that best bidding system is the one that optimizes the harmony of these two goals, but I doubt there's any easy/good way to measure it.

For some systematic things I happen to like very much and seems to follow the harmonizing idea:

Relay (precision) with some transfer scheme, meaning that when opponents don't go around making our life ridiculously difficult, we can make most of the bidding room and give one sided information so that opponents know the dummy even before it hits the table but have little idea what declarer holds.

The principle of fast arrival.

Having many forcing sequences, often initiated by one bid (GF relay/puppet) because having more forcing sequences gives you more bidding space to exchange information, on the other hand having them go through minimum number of bids means that you can more often bid things to play. (For example XYZ/NT)


Then there is probably the most important factor that is even harder to measure. How well does your system handle opposing pre-emption, mainly meaning that how informative your opening bids are. Usually making more informative opening bids means you are not making optimal use of the bidding room, on the other hand if you were going to get pre-empted, it's better to give more accurate description of your hand with the one bid you get. Some smarter people have written about one bid and two bid hands which reflects this situation pretty well.

/wall of text ends/

[Vampyr]

 32519, on 2012-January-29, 01:00, said:

The fact that one bidding system is more popular than another doesn’t necessarily mean that it is better e.g. 2/1.
Neither the fact that the system may be easy to understand, again, e.g. 2/1.

This depends on what you mean by "better". For a practised partnership willing to work hard popularity and ease of understanding are irrelevant; however these are important factors in allowing one to play with casual or new partners. So 2/1GF may be "better"; it depends on your goals.

EDIT: Posted before reading the previous two long posts; now I have read them and it turns out that they are excellent, and that Siegmund has expressed the above idea much better than I have.

He has also mentioned the very important concept that bidding agreements are very often decided based on frequency grounds, with the rare hand types left to fend for themselves. I, at least, fully agree, as predicted.

[pooltuna]

the obvious answer is if you use it you score more IMPS/MPs/total points. of course it is possible to find a system that scores better in IMPS than MP than total points

[hrothgar]

The Hog and Pooltuna have (essentially) nailed this...

Ultimately, the proof is in the pudding.

You want a system that serves you well, either measured in terms of personal utility or board results

[rhm]

The true answer is success at the Bridge table.
A lot of bidding systems, which were once popular, some due to their success in the past, are now more or less obsolete.
For example Roman or Blue club were much more popular 30 years ago.
Others have evolved, some in a way that their founders would not recognize them today.
A good example for the latter would be ACOL. When you find them at the tournament table today, Acol has little in common with what was played half a century ago.
I can tell because I grew up with an Acol book called "Bridge is an easy game".

Bridge is a very competitive game, and what is not optimal gets sorted out sooner or later.
Take for example a convention like ace asking.
I remember having played Culbertson 4/5 NT convention, various forms of Blackwood, Byzantine etc.

Today almost everybody plays a sort of Roman Keycard Blackwood simply because it seems to be more effective and simpler than some of its predecessors.
Simplicity is quite important in practice and one reason relay systems are not that popular.

Of course what is optimal for you may not be optimal for others.
What is important here is practical optimality, that is success at the Bridge table, not theoretical optimality, but error prone.
It may depend on your brain power, memory capability and your judgment and let's not forget your partner.
Some are notoriously bad in judging their own capabilities.
Ambitious Bridge player, just past the beginner stage, for example, are often mesmerized by conventions and tend to believe conventions are a solution to every bidding problem they encounter.

Rainer Herrmann

[mike777]

 32519, on 2012-January-29, 01:00, said:

The fact that one bidding system is more popular than another doesn’t necessarily mean that it is better e.g. 2/1.
Neither the fact that the system may be easy to understand, again, e.g. 2/1.

A defining characteristic (for me anyway) making one system better than another would be the ability to describe a wider range of hand patterns no matter how infrequently those patterns may occur. Shape is often cited as more important than HCP, these hand types being more suitable for offense than defense. A wide range of conventions have been created attempting to describe shapely hands. The problem however is that you often have to make a choice between which conventions to include and which ones to exclude. (This theme is to form “Part 2” of this thread).

I don’t have any experience with relay systems, but am told that they are good to determine partner’s exact distribution. How relay systems are affected by opposition interference is an even greater mystery for me.

So then, what does make one bidding system better than another?

Here is Eric Rodwell's take from BW

main system principle is logic, in addition:

1) be able to remember the system
2) have firm rules
3) greatest extent possible be impervious to preemption
4) long auctions are bad dont exchange any more info than necessary

He goes on to say many system-makers violate one or more of these principles.

[FrancesHinden]

Yes, the easy answer is that the best bidding system for my partnership is the one that gives us better scores than an other. This may not be the best system for your partnership.

There are both abstract, theoretical considerations:
- constructive vs destructive
- scientific but revealing vs guesswork but less revealing
- aimed at partials, games or slams

and practical things to worry about
- legality in the majority of events you play in
- memory strain
- time you have to put into developing parts in detail
- how well it meshes with your and your partner's style

so whether a relay system is 'good' or not depends at least as much on whether you want to put the work into writing and learning it, and whether you are of such a scientific bent, as whether it is theoretically superior to a natural system.

With one partner I play quite a complicated system. It is based on strong NT and 2/1 game forcing. I don't necessarily think that it is the 'best' system in any sense*. We started with that because it was easy to agree on as a somewhat occasional partnership. When it become a more serious partnership we worked on improving the already-agreed basic system because we both had come to know how we approached auctions in this system and how our style fitted and we understood inferences. Changing basic system requires a huge amount of work if you are taking it seriously, because there are so many inferences from partners choice of calls and they change completely if you switch to a different basic approach to bidding.

*Actually I'm pretty certain it isn't. But given where we are now it's the best system for us.

[Free]

Whatever works for your partnership against your opponents.

[Statto]

 32519, on 2012-January-29, 01:00, said:

So then, what does make one bidding system better than another?

The partnership playing it. And I'm not even being flippant here. I've seen arguments that Acol is no good and Std Am or 2/1 or Blue Club or whatever must be better because that's what the world champions have been playing. I don't buy that. That they're very good players who know their system inside out is what makes them world champions.

[Zelandakh]

I think the basic premise of the OP is wrong. Being able to show lots of different hand patterns is of little benefit in practrise compared to being able to bid the more common hand patterns effectively. My bidding system is one that can show alot of different hand patterns, all the way to 6511 or 7411. But the only reason I do that is because there are spare bids for them - when I designed the system I simply called them "freaks" and did not care about them at all. Having made the change the difference is minimal. I would far rather have a tighter hcp range for 4432 hands than be able to show 6511 vs 6520 vs 6502. Shape is important of course, but getting across exact shape on freak hands is not.

As for what makes a system good, well others have pretty much covered this. Obviously the system which gives the best results but this is a bit abstract. Some useful things to consider when building a system:-

Homogeneity: The more homogeneous your bids are the more efficient it will tend to be and, most importantly, the better it will work in competition. When considering how homogeneous a complex opening bid type is you should compare with the most common and weakest possibilties. A homogeneous opening bid does not necessarily mean uniform; a Swedish Club 1♣ opening has good homogeneity because the useful information opposite a weak NT (common type) is similar to what is useful opposite the strong types. Homogeneity is also the reason why opening with your longest suit is typically a good option.

Efficiency: A theoretically optimal bidding system would use every bidding sequence and do so in proportions close to the Fibonacci sequence. Or at least it would in the absence of opposition bidding! There is practically no bidding system in the world that comes close to this. Similarly, certain information is nearly always useful to have - does partner have a 6 card major for example. A good system should be able to get key information quickly for all hands where this might be useful.

Resilience: How well does your system stack up against a barrage from the opponents? How resilient a system is tends to be highhly dependant on how homogenous it is. It is often better for a system to be less efficient and more resilient.

Pressurising: In some ways the opposite of efficiency, a system should be capable of applying pressure to the opponents whenever they might want to be competing. Systems which tend to open high often and pass or 1♣ very rarely are examples of pressurising systems. Acol with weak 2s applies more pressure (on opening bid) than SAYC for example.

Safe: Added here to be the opposite of pressurising. A pressurising system might achieve this effect by taking risks that are too high. A classic example here is the weak NT - high pressure, more risk - although clearly there are more factors to consider in this decision.

Accuracy: How often does your system reach the theoretically otimal contract? Clearly we would like this to be true more often than not. An efficient system will tend to be accurate when the opposition are silent but may be wildly inaccurate in competition. A resilient system will be accurate in competition. In my opinion it is most important to be accurate in the game zone but this is less clear and (especially) MP systems sometimes sacrifice a little game accuracy for part-score effectiveness.

Information Leakage: Ideally we would like to reach our final contract while giving the opposition as little information as possible, especially about declarer's hand. Similarly, when the auction is competitive we would like to give the opposition as little to go on as possible once we know our final spot. It is often better to be in a playable spot with low information leakage than the optimal spot with the defenders playing double dummy.

Memory: An issue that might be the top priority or might be of quite low importance depending on the players involved. Even if this is a very low factor for you it does have an effect over a long session and can affect other aspects of your game. A single "forget" in a complex system may very well cause a result which more than offsets all of the gains the system has given on the preceding hands.

I am sure I have forgotten a couple of things but I think this list is a reasonable starting point. I think the most important factors are memory, homogeneity, resilience and accuracy. Obviously many of these design crieteria are contradictory and mutually exclusive so you have to make some design decisions and accept that your system will be good on some hands and bad on others.