[awm]

Non-forcing 2/1 is another story. I'll compare against GF 2/1, since they are similar in that you cannot play 1NT.

GF 2/1 is generally better on part-score hands. The exact issue depends on how you define your NF 2/1s. Let's consider responder with a 2434 hand; what is his bid? Bidding 2♠ on this hand is obviously poor, missing the possibility of a heart fit and misleading opener by making the same call on hands with ♠xx and with a real fit. So let's assume that you are supposed to bid 2♣ (natural). It seems that this 2♣ response could be as few as four and could include a doubleton spade. What's opener's rebid on 5(332)? Passing is dubious, especially if clubs are the doubleton... but even with three clubs, you would usually rather play a 5-2 spade fit than a 4-3 club fit. Rebidding 2♠ will get us to 2♠ opposite a possible 1444 or 1(43)5 pattern, violating Burn's law. Perhaps the least of evils is to agree that opener's 2♦ rebid could be three, and to pass 2♣ with 5233 (accepting the possible 4-3 club fit situation)... but clearly this creates a number of issues.

There are further questions about when opener is supposed to bid on. For example, suppose the auction starts 1♠-2♦ and opener holds the black suits. Should he pass 2♦ with a singleton? Keep in mind that 2♦ might even be a four-card suit (i.e. 2443, unless this creates further issues by bidding 2♣) and will often be a five-card suit. If he doesn't pass, should he rebid a five-card spade suit or try 3♣? Does 3♣ show five, does it show extras? These sorts of problems don't exist playing 2/1 because you start with 1♠-1NT-2♣-2♦ and now opener has limited his hand (by failure to jump shift) and can show extras via 2♥ (the fourth suit) or 2NT, or rebid clubs to show an extreme two-suiter without much extra.

When responder has a game-force, you are basically playing relay by responding 1NT. There are definitely advantages and disadvantages to relay when compared to a natural GF approach. Some disadvantages include vulnerability to interference (and you don't necessarily even have a forcing pass available since 1NT is only inv+) and shapely responder hands that might do better showing rather than asking. Of course there are also advantages to relay (especially when opponents are passing and responder is balanced). In any case it's not clear-cut that this is a win or a loss, and it will also depend on precisely what your follow-up structure is.

Invitational hands will depend a lot on your continuations after 1M-1NT. Can you find your minor suit fits? The suggestion of 1♠-1NT-2♣ being "min without four hearts" doesn't seem to specify opener's minor suit length very well, which could be a loss. But obviously many structures are possible.

[Zelandakh]

An excellent write-up Adam, as always. A couple of things I will address. I play that Responder bids 4 card suits up the line with a weak hand, so 2434 is a 2C response for me. However I do not open 5(332) hands 1S - 1S is for me specifically unbalanced with a max of 17. This solves this particular issue by creating another one, but the other problem is (usually) easier to handle. So Opener always has either a second 4 card suit or 6 spades making the bidding after 1S - 2C very simple. You are quite right in highlighting that this approach does not guarantee being in the best part score though. In fact part of the mentality of the system is to stop as quickly as possible in (almost) any playable contract when it is clear no game is possible. That means playing in a 4-3 club fit rather than a 5-4 heart fit is also possible(!) at least if the opps judge to pass their 10(ish) card diamond fit.

Passing 1S - 2D with a singleton in a possible 4-1 fit seems distinctly odd to me. Yes 3C shows extras (it is invitational). It is true that 3C may be better than 2S when Opener has the very weak black 2-suiter. These hands are the kind of thing I was meaning when I said I can come up with hands where non-forcing is worse. I think it also highlights something I said in another thread quite a long while ago - that non-forcing is best when playing limited openings. That in turn goes back to something Adam already said, the weaker your opening bids the better the 2/1 approach is.

Of course you are also correct about the next step relay. There are pros and cons. I would not like to say whether my follow-ups are good or bad. I use relay breaks primarily for stopper asks and so sacrifice the ability for the distributional hands to show. Again, pros and cons to this approach. Finally, the second step (1S - 1NT; 2D or 1H - 1S; 2C) is used to show the other major. The first step is a minimum without the other major and the other bids are maximum without the other major. Thus 1S - 1NT; 2H shows a maximum with 4+ clubs. Obviously the "minimum without" first step does not show minor suit lengths which sometimes means having to play 2NT rather than 3m. Again a sacrifice of part-score precision.

I find that your post really focuses on the (bad) parts of the system which do not get seen often in forum hands. I have mentioned this sort of thing before too. The question as always is whether the benefits occur often enough to offset the deficiencies. For my system with light and limited unbalanced openers I find that it is an improvement overall. For standard methods I do not believe it is better. In systems with super-solid opening bids like Fantunes it is a no-brainer for me to play 2/1 GF. In other words, if you like the idea of non-forcing 2/1s it is competitive enough to be worth trying out. But do not be surprised if you find it is better to go back to something more traditional unless you build the structure to be as harmonious as possible with them.

[nige1]

Pre-empts work.. oops... wrong thread Like most players, past the first blush of youth, I've played lots of different systems. Currently I play

• A strong ♣ system.
• Acol with 10-12 notrump in first two seats, non-vul.
• 2/1 (but with non-forcing 1N reply to 1-openers, 4-suit transfers, Rubensohl, Bergen, Gazzilli, Drury, 2-way check-back, and so on).

Like Frances, I'm not sure which I prefer. I find strong ♣ and Acol-like systems are more "natural" than 2/1. I still think that the main advantage of 2/1, especially for the casual player, is that so many expert system-builders have spent so much time and effort, grafting on, integrating, refining, and publishing compatible conventions. The result is a widely played, proven, internally consistent, powerful, and effective off-the-shelf system.

[phil_20686]

I actually find that 2/1 is much harder to play with pick up partners, there is a lot more that needs to be agreed.

For exapmle - is 2/1 always forcing?
Forcing or semi forcing NT?
Does patterning out show extra values?
Does reversing show extra values?
If you agree a suit at the 2 level after a 2/1, do you show shape or values?
What do jumps mean in 2/1 GF?
Is a 2/1 4+ cards, or are they all 5+ except for 2c= semi balanced?
Do you play different cue bidding styles depending how many rounds of cue bids are available below game?
what are 1M-3m?
What is 1M-3N?

Etc Etc, all but the last one seem to have clear answers in acol/sayc, but are matters of agreement in 2/1.

[nige1]

 Vampyr, on 2011-November-17, 19:29, said:

On the other hand, perhaps 2/1 GF players have it all wrong. If you are playing a 1NT response as forcing, why not play NF 2/1s?

Good point. I used to play that and it worked fine in the context of a strong ♣ with other 1-openers limited.

[xxhong]

2/1 makes slams and game bidding rather simple. Typical sequence in 2/1 gf is to find a trump fit at 2 or 3 level, then cuebid for slams. non 2/1 systems frequently need good guesses in sequences like: 1S 2H 4H ?

 Jinksy, on 2011-November-13, 16:36, said:

I've seen a lot of assertions that it's better, but very little reasoning to show it.

I recently read Bridge in the Enigma Club by Peter Winkler, which recommends non-forcing 2/1s; he's a cryptographer rather than a world class player, but I didn't find his arguments any less persuasive than those I've heard for the reverse.

I assume it's agreed that when playing 2/1 you gain vs the alt whenever you make a 2/1 bid (except on marginal hands where you have to force to game without knowing enough about P's hand to know if your values are working), and lose whenever you bid 1N (or canape as a result of insufficient strength). Most hands presumably fit into the latter category, so presumably the argument is that the gain is sufficient to outweigh this.

But it only seems to help significantly on slam-seeking hands, which are a pretty small minority. Sure, you might bid games more accurately on occasion, but that seems rare, and is offset partially by the fact that since the defence plays a greater role in games than slam, you're also helping them when your bidding is more detailed.

I'm not against 2/1 as a GF per se - I don't really have a view. But I'd really like to hear some decent arguments either way.

[awm]

 phil_20686, on 2011-November-18, 07:01, said:

I actually find that 2/1 is much harder to play with pick up partners, there is a lot more that needs to be agreed.

For exapmle - is 2/1 always forcing?
Forcing or semi forcing NT?
Does patterning out show extra values?
Does reversing show extra values?
If you agree a suit at the 2 level after a 2/1, do you show shape or values?
What do jumps mean in 2/1 GF?
Is a 2/1 4+ cards, or are they all 5+ except for 2c= semi balanced?
Do you play different cue bidding styles depending how many rounds of cue bids are available below game?
what are 1M-3m?
What is 1M-3N?

Etc Etc, all but the last one seem to have clear answers in acol/sayc, but are matters of agreement in 2/1.

I'm pretty sure the "undiscussed 2/1" answers to some of these are; (1) 2/1 is always forcing to game (2) 1NT is forcing by UPH and semi-forcing by PH (3) Patterning out does not show extra values (4) A 2/1 is 4+ cards (5) Cuebidding styles don't change based on the auction. This is not to say that alternative answers might not be better, but this is all what I'd assume (and have never really been disappointed on) with a pickup playing 2/1.

I agree that it's not clear whether reversing shows extras in 2/1, what jumps mean in 2/1, and what the meaning is of 1M-3m and 1M-3NT. However, agreeing to play acol/sayc you also have to discuss what jumps mean in 2/1 sequences (are they just extras, is there a suit quality requirement, if in a new suit are they splinters or natural), what jump shifts mean (I know plenty who play "sayc with weak jumps"), and what 1M-3N means. You don't have to discuss whether reverses show extras in 2/1 sequences, but there are plenty of 2/1 sequences where not only is it unclear whether they show extras, but it's unclear whether it's even forcing. Common examples are 1M-2m-3m and 1M-2m-2NT, and possibly even 1M-2m-2M. It is true that much of this is formally defined in the sayc document (which most sayc players have never read) but you could say the same about them being defined in the lawrence 2/1 book (which many 2/1 players have never read)... it doesn't mean a pickup partner will play them the same way you do.

[FrancesHinden]

 xxhong, on 2011-November-18, 16:03, said:

2/1 makes slams and game bidding rather simple. Typical sequence in 2/1 gf is to find a trump fit at 2 or 3 level, then cuebid for slams. non 2/1 systems frequently need good guesses in sequences like: 1S 2H 4H ?

This is a typical comment about 2/1, that it makes slams and games 'simple'. It's true that when responder has a game force, then showing it immediately at the 2-level makes game and slam bidding easier (assuming that you are actually playing a fully-worked out system and know what shows extra values etc, but that applies to any system not just 2/1).

One can say equally well that playing lighter 2/1s makes slams and game bidder simple. They work much better when responder doesn't have a game force and opener has a strong hand. You can play all sorts of clever methods after 1M - 1NT, but these are all to recover from the fact that responder has an extremely wide range of possible hands (even more so if you play a style where responder basically never passes 1M). Playing non-FG 2/1 you can have auctions like 1S - 2H - 3H - 4H. Playing 2/1 you might bid 1S - 1NT - Pass on some hands. And when you have a minor suit fit you really want to get responder's minor bid nice and early in the auction. If opener has a strong 5413 and responder is 1336 then you might start 1S - 1NT - 2H - and either 3C is undefined in strength or you can't bid it on a weakish hand... so it starts 1S - 1NT - 2H - 2S - 3C and you only know there is a possible fit at the 4-level, and responder can't bid 4C forcing so has to guess the final contract.

[mycroft]

AWM: I would suggest that you are a westerner. Here, I agree with you. Out east, though, the standard was:

1M-2m only is GF (sometimes 1S-2H as well); rebidding the Major is a "catchall"; 2NT may or may not show extras, but it tends to show 5332; any other bid past 2M is extras; 1S-2H is 5, anything else can be 4 (and the rest like you).

Also, 1m-2m (and j/s in om) are going to mean different things by default; the question of Constructive single raises or no comes up; and I'm sure there are a couple of things I've forgotten.

I will admit I felt much more comfortable making "investigate slam or not" decisions playing Ontario-style 2/1 than I do here, where we can have 4 rounds of bidding and be at 4 of a minor, and still have no idea either if I've shown extras or partner has extras. In exchange, of course, we find our fits better. But I've had a number of 5M-1 that would be 4M= (or 4M+1 because we haven't pinpointed the defence) in Ontario.