But how often does it really work out well? And how often do we end up saying "we were just too high when we started looking for a fit." Let's think about what usually goes wrong.
Balanced hands are no problem. We have 22 or more points and rebid 2NT over partner's usual 2♦ bid – everyone knows where we stand. Single-suited hands too, normally.
Let's take a look at a recent example: ♠KQT8742 ♥A ♦KQ7 ♣A2. Is there a danger that partner will let this hand be passed out even though we can make game? What do we need to make game, anyway? ♠Ax ♥xxxx ♦xxx ♣xxxx would be enough but maybe partner would find a response with this hand. ♠xx ♥xxxx ♦xxx ♣Kxxx would be sufficient also and partner will definitely be passing with that. Are the opponents likely to go quietly when we have a seven-card suit? They might if they are vulnerable and each have a balanced hand with 10 or 11 hcp each. When this hand came up in a robot tournament, 40% of players – including myself – with this hand chose to open 2♣ (the rest opened 1♠). As it happens, not one of us made our contract. Most went down one in 4♠. Partner's hand was ♠5 ♥T8743 ♦AJT8 ♣J86. Most of the spade openers made their 4♠ contract when LHO did not lead a club. I was in 6♦ personally after partner showed a red two-suiter. It never occurred to me that he would show his two suits with such scant values. Wasn't I already showing a good spade suit when I rebid 2♠? Wasn't I showing even longer spades when I rebid 3♠ over 3♥?
It got me thinking that there is so much ambiguity after a 2♣ opening that responder doesn't really know how good our suit is, although in this case with me bidding it twice, the message should have got across.
The biggest problems arise when we have a strong hand that is neither balanced nor single-suited. Two-suiters aren't too bad, although it's amazing how often partner is 1165 when we are 6511. There's just about room for opener to make two suit bids. And, at the slight risk of the opening bid being passed out, there's the possibility of a jump shift (or reverse) to show a big two-suited hand that doesn't quite qualify for a 2♣ opener. But three-suited hands are appallingly difficult to bid even when starting at the one-level. Let's say you pick up this nice hand: ♠AKQ8 ♥3 ♦KQ7 ♣AKQ62. You might be able to make game with as little as ♠xxx ♥QJTxx ♦Jxxx ♣Jx opposite. And the danger of being passed out now is quite significant (the opponents only have 12 points between them and probably no long suit).
So, what can we do in this situation? You could open 2♣ and rebid 3♣ over the expected 2♦. But the chances of stopping in time are not great given that we started describing our three-suited hand at the three-level!
I started thinking about using the three-level to show a self-splinter: i.e. rebid 3♥ over 2♦, to show a three-suited hand with short hearts. Sure, it prevents us from using 3♥ to show a solid heart suit but, given that we now will always be showing a goodish suit with our first rebid, there's less to be gained by showing the solid suit. We should probably insist that we have at least four of the other major. It still doesn't help however with distinguishing between one- and two-suited hands.
So, let me propose the following somewhat artificial scheme:
2♣ – 2♦ (waiting) –
3♣* (short clubs) – ditto
3♦* (short diamonds) – ditto
3♥* (short hearts) – ditto
3♣, 3♦, 3♥, 3♠: single-suited hand with a good 6+ suit (not necessarily solid)
3NT (25-27 balanced)
Note that in this system, opener will normally end up as declarer, with a one- or two-suited hand as the partnership will generally not play in one of responder's suits. That is not true, however, for approximately half of the three-suited hand situations (spades, diamonds and notrump).
It is possible of course for responder to make a positive suit response to the 2♣ call. However, the price of breaking the relay is quite high (all of these follow-ups are missed) so a suit should only be bid by responder with a very good suit of his own, typically six cards with two of the top three honors and good intermediates. In the case of a two-suited hand (the 2♠ rebid), there may be a chance to show a good suit by bypassing the next higher suit, i.e. if it is the next step up. Otherwise, it is better for responder to bid the relay as this conserves most space.
I would really appreciate your comments on this scheme. Is it workable? Is it legal (ACBL rules are so arbitrary)? Is it too complicated to remember?