[jogs]

aguahombre, on 2011-December-08, 23:17, said:

You are getting beyond the difference in concepts, and into your preference for partnership agreements in the two styles.

Don't think so.

♠ Q65432 ♥ AKJ ♦ AKJ ♣ 3

1♠ - 2♣, 2♠

Playing SAYC with a regular partner, we may agree that this auction is forcing.
Playing SAYC with a pickup, would you dare rebid only 2♠. This partner
may pass. Forced to rebid 3♠.
Playing 2/1 with a pickup, you can safely rebid 2♠.

[Zelandakh]

You can also rebid in diamonds (style), or use transfer rebids (system), or use an artificial 2D (gadget), or ... (etc). Thus the follow-up methods and general style matter here just as much as whether 2C was GF or not.

[Cascade]

jogs, on 2011-December-09, 08:20, said:

Don't think so.

♠ Q65432 ♥ AKJ ♦ AKJ ♣ 3

1♠ - 2♣, 2♠

Playing SAYC with a regular partner, we may agree that this auction is forcing.
Playing SAYC with a pickup, would you dare rebid only 2♠. This partner
may pass. Forced to rebid 3♠.
Playing 2/1 with a pickup, you can safely rebid 2♠.

The problem with this sort of example is that it is extremely low frequency.

I dealt 10000 hands with 18 hcp and 6331 distribution.

Just over 6% of the time partner had a 2♣ GF response.

Less than 1% partner had a Qxxxxx suit - even less if you excluded suits like Q109xxx or Q108xxx.

Around 2% partner had one particular red suit AKJ (4% for either red suit).

Not once in 10000 hands did partner have both red suits AKJ.

I don't think these low frequency hands are the reason many think that 2/1 has an advantage over non GF 2/1 systems.

[EricK]

jogs, on 2011-December-09, 08:20, said:

Don't think so.

♠ Q65432 ♥ AKJ ♦ AKJ ♣ 3

1♠ - 2♣, 2♠

Playing SAYC with a regular partner, we may agree that this auction is forcing.
Playing SAYC with a pickup, would you dare rebid only 2♠. This partner
may pass. Forced to rebid 3♠.
Playing 2/1 with a pickup, you can safely rebid 2♠.

But the advantage you are talking about here comes about from simply having an agreement over how far 2/1 is forcing rather than on the specific merits of one agreement over another. Actually, in "true" SAYC, the sequence is forcing (a 2/1 bid promises a rebid over opener's minimum replies). But anyway, you still need some agreements even playing 2/1GF with a pick-up partner, which you don't need if playing 2/1NGF. Are you forced to game or just 4m? Is responder's rebid of a suit forcing or not? What is 1M 3m?

[aguahombre]

Focus

Jinsky started this thread to discern the merits of a "sensible" 2/1 NGF system versus a "sensible" 2/1 G.F. system.

SAYC is designed as a practical thing for people to play with pickup partners and has nothing to do with the OP. Regular partnerships will sensibly adopt gadgets/tools to create a workable style, no matter what basic premise they have. SAYC is a rigid set of agreements with very few of those tools.

I have the opinion that NGF requires more artificial gadgets and more hard work by a partnership to cover the holes than does GF. But, I am prejudiced in this regard because I haven't tried to make a NGF style workeable and have done so with GF.

Specific hands, as requested by the OP, won't really illustrate the advantage of one choice over the other. They will merely identify an area which needs refinement in one style or the other.

[awm]

It's hard to give examples without specific systems in place. Most particular problem hands (in either style) can be "fixed" by adopting some conventional method, which in turn hurts you on other hand types.

However, the following general statement will always apply. Playing 2/1 GF gives you many more forcing sequences! If 2/1 is not game-forcing, there will be at least some sequences after 1M-2x that are not forcing. Having more forcing sequences will help you when you have a hand that wants to force. This is a big advantage when partner opens 1M and you hold a game-forcing hand, and typically the wins will show up in slam bidding.

The flip side is that playing 2/1 GF means you have fewer non-forcing sequences. This inevitably makes it harder to bid (some) hands where responder does not have enough for an immediate game force.

So why is 2/1 GF "better" or is it? One line of reasoning is that the wins (responder has immediate GF) are usually high-scoring hands which produce slams (or at least games). The losses are quite often part-score hands. At IMP scoring, the magnitude of wins/losses is important, so playing in the wrong partial occasionally is nowhere near as bad as occasionally missing a good slam or playing in the wrong game. The second argument is that the 13-20 point range for responder (where 2/1 GF has an advantage) is more frequent than the 11-12 point range (where 2/1 non-GF has an advantage). Of course you must adjust these point ranges based on your opening style!

Given the above, why do some good players choose to play 2/1 non-GF? First, if you play a sizable amount of MP or BAM scoring, the magnitude of wins doesn't matter so much. Second, the lighter you open 1M (and light openings are all the rage these days) the less often you can bid 2/1 GF (so frequencies start to swing the other way). Third, there are arguments that most of the strong hands are biddable in 2/1 non-GF, albeit with a bit more guessing involved.... whereas certain invitational responder hands are almost impossible to communicate in 2/1, at least without special gadgets (an example of this is a hand with long clubs after 1♠-1NT!-2♥).

--------------

The problem of what to play in a "pickup" partnership without a lot of discussion is almost totally unrelated to the question of what to play in a serious pair. This often revolves around the two players' level of experience/comfort with methods, and in the US most advanced-expert players have more experience with 2/1 GF than otherwise. This could easily be the opposite in other countries. A secondary concern is that you are unlikely to be able to conduct delicate slam auctions in a pickup partnership regardless of your general approach (because you have not discussed cuebidding style in detail or which actions show extras) but playing 2/1 GF at least guarantees that you will never be dropped in an auction that you thought was forcing.

[mike777]

Perhaps the biggest plus of 2/1 gf vs nongf is there are alot more software updates to improve the system.

Same thing playing strong nt vs weak nt...just more software updates...

[Vampyr]

EricK, on 2011-December-10, 04:24, said:

But the advantage you are talking about here comes about from simply having an agreement over how far 2/1 is forcing rather than on the specific merits of one agreement over another. Actually, in "true" SAYC, the sequence is forcing (a 2/1 bid promises a rebid over opener's minimum replies).

I think that it is very true that having an agreement is the main thing.

I honestly do not know how SAYC can be played in anything but a serious partnership who have discussed a lot of auctions. 2/1 non-GF is fairly simple to understand and play, 2/1 GF even more so. But a halfway house where 2/1 promises another bid, or is forcing to 2NT etc seems much more difficult, since forcing and non-forcing auctions are not clearly defined.

The above probably explains, to some extent, the popularity of 2/1 GF in the US -- there is a tradition there of strong 2/1s, so strengthening them that little bit extra and making them GF resolved a lot of ambiguity.

[myprac]

The Grant/Rodwell book 2 Over 1 Game Force discusses at some length the potential advantages of GF with sample hands in various categories (finding the best game, exploring for slam) and compares alternatives (why GF works better than FSF in some situations, for example). I imagine some here would consider it too basic, but it's an excellent book and directly responds to the OP.

[han]

Cascade, on 2011-December-09, 14:10, said:

The problem with this sort of example is that it is extremely low frequency.

I dealt 10000 hands with 18 hcp and 6331 distribution.

Just over 6% of the time partner had a 2♣ GF response.

Less than 1% partner had a Qxxxxx suit - even less if you excluded suits like Q109xxx or Q108xxx.

Around 2% partner had one particular red suit AKJ (4% for either red suit).

Not once in 10000 hands did partner have both red suits AKJ.

if this a serious post could you give us a specific hand that is high-frequency?

[mycroft]

I play 2/1 and like it.
I play K/S (with 2/1 GF unless suit rebid) and like it.
I play Precision (with "open all 10s that look like 11) without 2/1 and like it.
I have played K/S with 2/1 "promises a rebid" and liked it.
I have played Standard with 4 or 5 card majors, and with the appropriate gadgets assumed in 2/1 to fill in the holes, and like it (but not as much, because usually I'm playing with people who don't know how to re-evaluate their hand. When I do play Standard with decent players, it's much more fun).

They each have different strengths and weaknesses, and when I get a hand that plays to the system's strengths, I'm very happy, especially if it isn't the field system; when I get a hand that preys on the system's weaknesses, I do my best to deal with it, and hope that I can claw my way to average.

Whether one is "better" or not, I don't know. 2/1 is great in the ACBL because you don't want to play with the players that can't play 2/1 (and if you're the weak(er) player, you won't get good players to play with you if you don't play 2/1); conversely, given the homogenous(*) nature of ACBL bridge, it's fun playing *anything else*, just to watch the run of the mill's brains explode when you bid "wrong" (or watch them panic over our strong club when we play Precision, and not even think about our 1M and 1♦ bids, frankly more important to be able to deal with).

I will say that the hands where the system is your judgement playing GF 2/1 that you do playing NGF 2/1 are more spectacular and more obvious than the hands where playing NGF 2/1 gives you a "system advantage"; that's also an advantage for "average" pairs who don't want to miss slam, but don't notice as much when they get out-competed for the part score.

It looks like I'm denigrating 2/1, and I'm not. It looks like I'm saying "it's a crutch for the flight B players" - which is in fact true (and a "stepup" for new/pickup partners, no matter their strength), but not the point. Experts with expert judgement will be able to play any system (including 2/1) that both players are compatible with, and will put the work into that system to iron out the wrinkles - so the choice there doesn't boil down to "what's better" as much.

* (edit to add missing footnote) "I know a lot of people are going to think, that's not 'homogenous', that's 'homogeneous'. But that isn't what I mean, I mean 'homogenous' as in milk." -- Anna Russell.

[Cascade]

han, on 2011-December-19, 08:33, said:

Are you trolling as usual, or is this a serious post? If you meant it seriously, could you give us a specific hand that is high-frequency?

Serious.

In answer to your question - No.

However in case you have not noticed Qxxxxx, AKJ, AKJ, x is not a specific hand. Yes the poster who I quoted posted a specific hand. My numbers were clearly not based on that specific hand but a generic one where the spot cards were not specified. That is I just looked at the more specific honour combinations and ignored the relatively insignificant spots.

My point wasn't that Q65432 AKJ AKJ 3 was an unlikely hand. I am fully aware that every specific hand is equally likely a priori. But that Qxxxxx AKJ AKJ x is a relatively unlikely hand and something like AKQxxx AQx Kxx x is more likely

With 18 hcp it is highly unlikely that your honour strength will be concentrated into two short (three card suits) and your long suit will be weak.

With 18 hcp and six spades in a 6331 hand the simulation delivered these frequencies (I can probably calculate the true values rather than simulate if I needed to - I think i have a spreadsheet somewhere for this):

xxxxxx	10
Jxxxxx	27
Qxxxxx	69
Kxxxxx	131
Axxxxx	280
QJxxxx	143
KJxxxx	287
AJxxxx	472
KQxxxx	537
AQxxxx	827
AKxxxx	1385
KQJxxx	521
AQJxxx	936
AKJxxx	1394
AKQxxx	1872
AKQJxx	1109

and for the three card suits (actually one of the three card suits but its symmetrical and i didnt bother to record the other one)

xxx	373
Jxx	336
Qxx	651
Kxx	1139
Axx	1985
QJx	288
KJx	478
AJx	701
KQx	774
AQx	1085
AKx	1438
KQJ	117
AQJ	173
AKJ	203
AKQ	259

The average hcp in each suit was approximately

spades 7.2
hearts/diamonds 4.6 each
clubs 1.7

(there are some rounding errors)

My point is that I think it is unlikely that big gains from 2/1 come from catering to strong hands with very weak long suits. These are very low frequency hands.

[Cascade]

These are the calculated not simulated expected values of high card points in each suit in an 18 hcp hand with 6331 distribution

7.188678192	4.523218063	4.523218063	1.764885682