[fromageGB]

Assume you play a constructive raise so that 1♠ 2♠ shows 3 card support and a typical 7 to 10 count. If opener wants to invite game he traditionally makes a trial bid, such as 3♦. Responder is asked to bid game with a maximum, sign off in 3♠ with a minimum, and if midrange to base his decision on his diamond holding.

To make this judgement, he needs to know what sort of trial bid it is. It may be "short suit" (eg x); a "gappy long suit" (eg KJx/KJxx ); or a "rubbish suit" (eg xxx). For the first and last, a holding of Qxx is wasted values, for the middle it is a plus feature. What sort of trial bid do you make?

It seems to me that any of these gives away unnecessary information that will probably help the defensive opening lead, and there is no gain if responder is minimum or maximum anyway. So a better method is to have an "unspecified trial bid" of 2NT. Responder bids 3♠ or 4♠ as before, but if uncertain now he makes a trial bid of 3♣/♦/♥ for opener to judge whether that is a positive or negative holding. So you have avoided helping the opening lead in two thirds of the cases, and for the other third you are giving information only when it is needed.

The question now is "what sort of trial bid should responder make?" My feeling is that a short suit (x or xx) trial is least helpful to the opposition, but I would welcome your thoughts on the matter.

If the trump suit was hearts, with 1♥ 2♥ then 2♠ is the unspecified trial, and if responder has a midrange hand with a spade trial he bids 2NT.

[han]

We play the cheapest step is ambiguous and then responder can bid a suit in which he has secondary values. For example with xxx xxx Axxx KJx responder bids 3C, saying that he would "not accept a short suit game try in clubs". Of course opener is not required to have shortness.

[kenrexford]

After 1M-P-2M, semi-constructive or constructive, I use a fairl;y simple, but fairly important IMO, structure. 2NT is a general GT (and can result in a contract of 3NT); new suits are natural and either game or slam tries.

The importance of the natural game tries is in the recognition of the power of secondary fits. You have established a 5-3 fit. If Opener shows a second suit and Responder has a better fit for the second suit, amazing things are possible. For instance, imagine a simple 5-3 fit initially established and then finding a 4-4 fit. First, you now know that if you convert to the 4-4 fit, you will probably take an extra trick beyond that initially expected, not just because of a ruff but because the ruff is now known to be likely. Second, your initial fit now becomes a trick source.

Imagine, for example, 5431 pattern across from 3433 pattern. If we have all four top honors in both fit suits, probably overkill, we will expect to win 10 tricks in those suits if we have control over the other two suits. 20 points yields 10 tricks. Thus, if you add in the AK in Opener's fragment, or both side Aces, 27-28 HCP yields slam.

For me, then 1M-P-2M-P-new-P-3NT is used to re-focus trumps to the new suit and show slam interest (little is needed for slam now). 4♣ asks questions.

All of that may be a bit much for most people. Fine, but still the real prospect of a slam is nonetheless present anyway. Give opener a wild-shape 4-loser hand, and the right three covers does the trick. So, 1M-P-2M may often induce slam thinking.

Because of this, the style of GT's should either mirror the normal action for slan tries or should enable catering calls when Opener is slammish. Natural bids for me seem easiest, as Opener's new suit call if natural can easily handle most game try scenarios and slam try scenarios equally well, so long as Responder accepts game tries with intelligent slam try responses.

I think you are right, though, in the GP that nondisclosure in game try scenarios is good. My partner once commented, concerning the upside to game bash bidding, that 1M-P-2M-P-2NT is usually his game try (if not simply blast-and-pray), such that a new suit bid was probably more likely to be a slam move than a game move.

That seems a tad much, but the idea seems to be valid even if hyperbole.

[sfbp]

I've liked the following structure for a long time. Note that it may work better in the context of limited (precision) openers. There are other flavours such as making the first bid after 1♥-2♥ different from that after 1♠-2♠ but it seems like they just place a burden on partnership wanting to adopt - you can always refine. I learned it originally in Acol , so it works fine with 4-card majors (ie 4+4 raise) as well. NS = New Suit.

==
1M - 2M -

7-10 points, always some ruffing value (small doubleton or singleton) and can be alerted as "will accept at least one game try".

==
1M - 2M -
NS - 3M -

No thanks (usually means that the shortness is opposite now-wasted high cards)

==
1M - 2M -
NS - 4M -

Done!

==
1M - 2M -
NS1 - NS2 -

I don't know, what do you think (pass the buck)

==
1M - 2M -
2N -

I don't have useful shortness but still interested if you can bid shortness now (assuming you cannot bid game)

==
1M - 2M -
3M -

Usually a 6th card in the major

==
There's more, such as defining carefully what hands to go via Forcing NT to 2M, and via FNT to 3M; but I assume most using these methods already have that sorted out already.

The minimum biddable game (admittedly very lucky) I ever saw was 11 opener opposite 7 responder. And the biggest responding hand (that doesn't qualify for FNT and jump to 3M) is 10 points and a doubleton, and that hand gets "upgraded" if the doubleton is small.

In the context of standard a game try can be a concealed slam try, but rather unlikely with limited openers. I did manufacture an example once.

Some of the biggest advantages to the whole scheme seem to be where opener is at the top or bottom of the range. In the former case he blasts 4, in the latter he passes (and, xyz-like, you're not in 3 going down 1), and opponents don't have much information.