Bridge in Oz

The Problem

Consider the following hands (which we will call "Case 1") for an opening bid of one spade, using five-card majors:

• AQxxx Kxx Kx xxx
• AQxxx Kxx xxx Kx
• AQxxx Kx Kxx xxx
• AQxxx Kx xxx Kxx
• AQxxx xxx Kx Kxx
• AQxxx xxx Kxx Kx

and the following holdings by responder:

• Kxxx QJxx Qxxx x
• Kxxx QJxx x Qxxx
• Kxxx Qxxx QJxx x
• Kxxx Qxxx x QJxx
• Kxxx x QJxx Qxxx
• Kxxx x Qxxx QJxx

Notice that in one-third of the possible combinations of the above, game is excellent, while in the remaining situations it is virtually hopeless. The only difference is the degree of fit in the side suits.

Let's look at some other possibilities, this time with opener holding the singleton ("Case 2"):

• AQxxx Kxxx Kxx x
• AQxxx Kxxx x Kxx
• AQxxx Kxx Kxxx x
• AQxxx Kxx x Kxxx
• AQxxx x Kxxx Kxx
• AQxxx x Kxx Kxxx

and responder's hands:

• Kxxx QJx Qxx xxx
• Kxxx QJx xxx Qxx
• Kxxx Qxx QJx xxx
• Kxxx Qxx xxx QJx
• Kxxx xxx QJx Qxx
• Kxxx xxx Qxx QJx

Clearly, the same observation applies as above.

In all of these situations opener holds 12 high card points (HCP) and responder holds 8 HCP. Another interesting point here is that if you combine the Case 1 opening hands with the Case 2 responding hands then game is always hopeless, and in two thirds of those combinations it will usually fail by two tricks.

These examples illustrate two valuable principles:

1. In spite of the tendency by many players to evaluate a hand's strength only in terms of HCP, distribution matters a great deal. A responding hand with four trumps, 8 HCP and a side singleton should be evaluated as 11 points in support and is in the "limit raise" category. Similarly, an opening hand with a singleton should be upgraded by two or three points once partner shows support, and by considerably more if the cards fit well.

2. Regardless of whether a hand is worth a minimum, limit or game-forcing raise, bidding methods should allow for exploring the degree of fit and should consider the possibility that either opener or responder may hold the key singleton or other distributional feature.

Today's popular methods do not adequately address these issues. And most partnerships give little or no attention to game tries, perhaps agreeing on "help suit" tries without any real understanding of what that means.

The Solution

The system I recommend is simple and covers all of the example cases shown above, and many more. It works as follows:

• The cheapest jump response to a major suit opening (1♥ - 2♠ or 1♠ - 2NT) is a limit raise or better, promising four or more trumps and ten or more points in support (as already mentioned, this includes distributional points).

• The direct jump raise (1♥ - 3♥ or 1♠ - 3♠) is preemptive and shows four or more trumps and 0-5 points in support.

• Other single jump responses show four or more trumps and 6-9 points in support.

Incorporated into this system is the concept of "fragment" game tries. The general idea is that when making a game try, you bid the cheapest suit in which you hold one or more high cards (2NT shows spades), and therefore skipping a suit indicates only small cards (or a singleton or void) in that suit. If partner is still unsure after hearing your fragment bid, then he may bid a fragment of his own and pass the decision back to you.

Development of the Auction

Here is an example:



The 2NT response showed at least a limit raise. Opener's 3♣ shows something in clubs. Responder likes the club fit but is not sure about the red suits, so tries 3♥. Opener, encouraged by the heart fit and knowing that if partner's hand includes a singleton it is in diamonds, does not hesitate to bid the game, despite only 20 HCP in the combined hands and no wild distribution.

There are two other very significant aspects to this auction. One is that we have done little to aid the defense; the opponents do not know for sure that responder has a singleton, and a knee-jerk lead of the "unbid suit" will work poorly. The other is that we have avoided bidding diamonds, which could allow the opponents to double and find a profitable sacrifice.

Similarly, the 6-9 point raises (3♣ or 3♦ or 3♥ in response to 1♠, or 2NT or 3♣ or 3♦ in response to 1♥) are "premature" game tries; i.e., they immediately show the cheapest fragment (1♥-2NT shows spades), and as before, partner may choose to explore with his own fragment bid if he sees the possibility of game.

We have already discussed opener's option to show a feature after 1♥ - 2♠ or 1♠ - 2NT when exploring for game. This might also be done when opener has strong slam interest, with 18 or more points (including distributional values), in which case his intentions will be clarified on the next round. Other possible rebids by opener are:

• Three of the major: bad hand with scattered values, signoff
• 3NT: about 16-17 points, scattered values
• Four of the major: accepting, but little to spare
• New suit above three of the major: about 16-17 points, shows a fragment

Here's a variation of the above auction in which opener has slam interest:



Here opener clarifies his slam interest with a 4♣ cue-bid, and responder, who previously denied a diamond card, confirms a singleton by cue-bidding that suit. Opener checks for key-cards and bids the excellent slam, with only 26 HCP in the two hands.