The Half Empty Overcall

Farmer John: Did you hear about Old Tom’s good fortune?

Farmer Bill: No. Did his daughter run away with the hired hand?

Farmer John: No, he fell off his horse and broke his arm.

Farmer Bill: Where’s the good luck in that?

Farmer John: Well, he could’ve broke his neck.

Competitive bidding is like that – often just avoiding disaster can be considered lucky.

Ulf Nilsson, the well-known Swedish bridge expert, has observed that contrary to orthodoxy overcalling on a suit lacking in the normal complement of top honors is often advantageous. His idea is that if you have AKxxxx in a suit, partner is less likely to have an honor in that suit than if you had overcalled on KJxxx. He refers to this as ‘the suit quality paradox.’ His conclusion is that one should overcall freely with a weak suit within a good hand, and his experience tells him this is a frequent winning action. ‘Less is more’ is a concept worth considering in detail. After all, it is a finite world, so the less we have the more partner can have. And if partner has nothing? ‘Down 500? Not to worry, they can make game’ – we’ve all heard that.

To take matters to the logical extreme, imagine if instead of bidding what we have, we bid what we don’t have. Say we open 1♠ to tell partner, ‘don’t expect anything in this suit.’ In fact, there are lots of useful bids like that, a splinter bid being the obvious one. The Precision 2♦ is such an opening bid, showing a hand 11-15 HCP with shortage in diamonds. It says, ‘Not diamonds, partner.’ It is notoriously difficult under standard procedures to show a hand with 4-4-4-1 shape – one has to bid 3 suits, and even then partners may not catch on. So here we have it in one bid. Also leading a suit in which you have nothing is more likely to find a suit in which partner is well endowed. Sometimes that works, but too often partner forgets how brilliant you are.

Bidding bad suits doesn’t always pay off when the alternative is a double. Can partner tell? Here is an example from yesterday’s game where I put an opponent to the test.

 

Dealer: West Vul: NS		North
		♠	752
		♥	1075
		♦	A73
		♣	AQ43
West				East
♠	A63			♠	QJ4
♥	983			♥	KJ642
♦	KQ92			♦	—
♣	985			♣	J10762
		South
		♠	K1098
		♥	AQ
		♦	J108654
		♣	K

 

West	North	East	South
Pass	Pass	2 ♥	3 ♦
3 ♥	4 ♦	Pass	Pass
Dbl	All Pass

 

Horrid preempts, nonvul vs vul, third seat, are part of the 2/1 system. Actually, I like preempting in a suit one above my shortage. South should have been warned by the presence of the ♥AQ in hand, but being optimistic, and having a long suit, she decided that partner could very well come up with the kind of support that an expert gets from a partner. In fact, her husband had good support. 3♦ makes, but not 4♦ , so there was really nothing to be done after the ‘excellent’ raise by North. However, an overcall of 3♦ hardly seems to be the best approach when 3NT or 4♠ should be kept in the picture.

Lest we think this was merely a bit of Friday night folly, here is an example from the German Bridge Team Trophy match between Turkey and Hungary, May 16, 2010, a match that, History suggests, was taken seriously by all concerned.

 

Dealer: South Vul: EW		Nikolits
		♠	J10
		♥	QJ832
		♦	AK4
		♣	A104
Assael				Aslan
♠	AK852			♠	43
♥	K106			♥	9754
♦	10			♦	J97
♣	Q765			♣	K832
		Lakatos
		♠	Q976
		♥	A
		♦	Q86532
		♣	J9

 

Assael	Nikolits	Aslan	Lakatos
1 ♠	2 ♥	Pass	Pass
Pass
.
Winkler	Kubac	Szilagyi	Zorlu
1 ♠	Dbl	Pass	3 ♦
Pass	3 ♠	Pass	3 NT
All Pass

 

North had the type of hand on which most feel it is worthwhile to take some action, and, indeed, partner held a top honor in his 5-card major suit. The Hungarian North overcalled in hearts and Lakatos could deduce his partner had an opening bid without a great heart suit, but he felt he could not make an encouraging 3♦ bid without the risk of going overboard. If he had thought like Ulf, he might have tried 3♦ on a weak suit, found partner with great support, and reached a making 3NT.

The Turkish North made a more flexible call of double, so South was forced to encourage with his half-empty diamond suit. No problem, it seems, and 10 IMPs to Turkey. Kubac’s double was heaped with praise by the BBO commentators, but I am not so sure they are correct. Traditionally overcalls are meant to show a hand that is playable in the suit named. That is too narrow a view, especially if one is to follow Nilsson’s advice and overcall with values largely outside the suit, in which case, advancer should be able to bid freely. The use of transfer responses to overcalls can be of some help.

Note that if the clubs and hearts were interchanged, North could have overcalled 2♣ and South could have bid 2♥ easily with a happy result. Thus, an overcall especially of 2♣ can be taken as a first step in a competitive process in which playing in clubs is not necessarily the one and only aim. There is a takeout aspect to the call.

Doubling in America traditionally is meant to show a hand playable in more than one suit, so if over a double by North, South had bid 3♣ , the doubler could hardly have bid 3♥ to bring the major suit back into play – that would show a much more powerful hand. Doubling on a 1-suited hand hoping later to be able to bid the suit conveniently is one of the worst aspects of the American double. Uncertainty is risky.

Here is a hand reported on BBO on May 20, 2010. Roy Welland and Zia were engaged in a practice match against up-and-comers from the Netherlands, the practice being intended primarily for the youngsters. Welland passed in first seat only to enter the auction at the 3-level after the opponents had limited their resources, a bad approach.

 

Welland	Zia	Pass 1 ♠ Pass 2 ♠
♠ KQ65	♠ J	3 ♦ ! Dbl* All Pass
♥ J7	♥ K9832
♦ KJ976	♦ 103	* penalty double
♣ 73	♣ KQ952	-500 against 140 making

 

We can see Welland’s logic: the opponents have a spade fit, so we have one, too. Diamonds are poor, but that merely increases the odds that partner has support! Wrong this time, as this is a rare 5=7=7=7 division of sides. Unlucky? Not really. The ♠ KQxx represent a very bad omen, to which Zia’s ♠ J adds more doom and gloom. However, Welland’s logic is not entirely wrong. Look at Zia’s shape. Over 1♠ should he not be taking action, such as a 2♠ cuebid? If Welland thinks the aggressive Zia would have acted with 5-5, then Welland’s outrageous 3♦ bid doesn’t seem as mad as at first glance.

This is another example of how competitive bidding depends on the partnership stance. If the aim is always get into the action, then more latitude has to permitted partner which allows for (or even demands) light actions. An absence of action is in itself informative. This applies to action taken over light opening bids, such as employed by Meckwell. Nilsson has given us an example that illustrates the urge of all to get in early.

 

Dealer: West Vul: Both		Nilsson
		♠	J3
		♥	K9
		♦	A643
		♣	AJ652
Meckstroth				Rodwell
♠	A64			♠	K9752
♥	A65432			♥	QJ
♦	Q87			♦	K109
♣	4			♣	1085
		Wrang
		♠	Q108
		♥	1087
		♦	J52
		♣	KQ93

 

Meckstroth	Nilsson	Rodwell	Wrang
1 ♥	2 ♣	Dbl	3 ♣
All Pass

 

The NS division of sides is 5=5=7=9, so there are 17 Total Trumps. EW teammates made +140 in 2♥ when the American North didn’t see his way to overcalling, vulnerable, on a half-empty suit. Nilsson took 9 tricks in clubs for a gain of 6 IMPs.

As a first cut at analyzing results, I ask myself, ‘what would have happened if the East and South hands were interchanged?’ Deep Finesse tells me, ‘nothing much changes.’ NS will bid to 3♥ on their 9-card fit and make 140 on a club lead. The NS division of sides is 7=4=7=8, and the Total Trumps are still 17. Thus, Nilsson was neither lucky nor unlucky in that respect. If 3♣ gets doubled and stays doubled, it goes down 1 on a heart lead for a loss of 2 IMPs, but it is unlikely that NS will be tempted to try to collect.

Remarks

The question of the advantages one obtains from overcalling at the 2-level on half-empty suits can’t be answered with a few examples. Nor can statistics help greatly, for, as we have seen, the results will depend on the methods adopted by the players involved. One might say that overcalling on a weak suit is more likely to have a bad affect on partner’s game than on the opponents’. Why? Because the information content of the overcall will be greatly reduced without the restriction on suit quality and length. In many cases the opponents will bid again, and partner may be in doubt as to whether to compete further. Once the opponents have limited their hands it becomes much more dangerous to push without good trumps. Of course, the opponents may misread the situation and allow themselves to be pushed too high. Part of a winning strategy is to generate plus scores on defense. (There is a cure for this: cooperative doubles. Not many partnerships have the confidence to adopt this strategy.)

Total Tricks and Overcalls

Nilsson considers he was somewhat unlucky to find partner with a 4-3-3-3 shape. It is easy to analyze the situation with regard to what is most probable, and to see that 4-3-3-3 is not that unexpected. First, Meckstrorth’s 1♥ opening bid is most likely to be with a 5-3-3-2 shape. We can combine that with Nilsson’s shape to obtain the most likely distributions of the South and West hands, as follows. A and B designate the hands held by East and South without specifying in which direction they belong. The directions are equally likely.

 

North	West		NW	SE		A	B	N & A	N & B
♠ 2	♠ 3		♠ 5	♠ 8		♠ 4	– 4	6	6
♥ 2	♥ 5		♥ 7	♥ 6		♥ 3	– 3	5	5
♦ 4	♦ 3		♦ 7	♦ 6		♦ 3	– 3	7	7
♣ 5	♣ 2		♣ 7	♣ 6		♣ 3	– 3	8	8
						Total	Trumps	16	16

 

If North has A as his partner, the division of sides is denoted under N&A; if B, under N&B. In this situation the most likely division of sides is the same for each potential partner, thus so is the total number of trumps. On that basis it makes sense for both sides to compete at the 2-level. The danger is that South may compete to the 3-level holding ♣ xxx if he expects North to hold 6 clubs. Note that a 4-3-3-3 shape in the South hand is not unlucky, but most likely. Is it unlucky to find Meckstroth with 5-3-3-2 shape?

Some expert writers claim that the more hearts North holds, the more likely it is that he has a fit with South’s hand. We can test this under the assumption that North held 4 hearts and 2 diamonds rather than the other way around.

 

North	West		NW	SE		A	B	N & A	N & B
♠ 2	♠ 3		♠ 5	♠ 8		♠ 4	– 4	6	6
♥ 4	♥ 5		♥ 9	♥ 4		♥ 2	– 2	6	6
♦ 2	♦ 3		♦ 5	♦ 8		♦ 4	– 4	6	6
♣ 5	♣ 2		♣ 7	♣ 6		♣ 3	– 3	8	8
						Total	Trumps	15	15

 

Unexpectedly, the most likely number of clubs held in combination remains at 8. The ominous portent is that the most likely number of total trumps is reduced to 15, and it becomes more dangerous in theory to compete in clubs to the 3-level. Note that South will hold a doubleton heart, so over a negative double from East he may be inclined to bid 3♣ on what is essentially a misfit deal. He shouldn’t, because his 4 spades tells a story, however, he may wish to prevent West from rebidding his hearts cheaply. Ideally West or East should be able to double cooperatively for penalty a venturesome 3♣ excursion by South, but if they can’t, the cost of a bad bid is greatly reduced. That is the confused state in which we presently find ourselves, where competitive methods are based on what used to be true, but no longer is.

For more directly from Ulf Nilsson, himself, visit his website.

Partner A and Partner B

We don’t want to leave the impression that it doesn’t matter theoretically whether Player A or Player B becomes North’s partner. The above cases are special in that the SE division of sides consists of all even numbers. Here is a case where the division of sides is a mixture of odds and evens so that the Total Trumps vary.

 

North	West		NW	SE		A	B	N & A	N & B
♠ 2	♠ 3		♠ 5	♠ 8		♠ 4	– 4	6	6
♥ 2	♥ 6		♥ 8	♥ 5		♥ 3	– 2	5	4
♦ 4	♦ 3		♦ 7	♦ 6		♦ 3	– 3	7	7
♣ 5	♣ 1		♣ 6	♣ 7		♣ 3	– 4	8	9
						Total	Trumps	16	18

 

The Total Trumps can be 16 or 18 depending on whether Hand A or Hand B is dealt to East, a 50-50 proposition. If East holds Hand A with 3 hearts, he will raise hearts and South will raise clubs. North is likely to declare in 3♥ . This is correct when the Total Trumps are 18. If East hold Hand B he will be inclined to try for a penalty double of 3♣ . If the red suits are exchanged in the North hand, the situation can be worse.

 

North	West		NW	SE		A	B	N & A	N & B
♠ 2	♠ 3		♠ 5	♠ 8		♠ 4	– 4	6	6
♥ 4	♥ 6		♥ 10	♥ 3		♥ 2	– 1	6	5
♦ 2	♦ 3		♦ 5	♦ 8		♦ 4	– 4	6	6
♣ 5	♣ 1		♣ 6	♣ 7		♣ 3	– 4	8	9
						Total	Trumps	15	17

 

When the Total Trumps are 15, North is in danger of being doubled for penalty as East will hold a hand with shape 4=1=4=4. Matters will be made much worse if South gives a courtesy raise to 3♣ with 3 low clubs. Of course, if East holds the 4=2=4=3 hand, he will make a negative double and South will raise clubs with 4. According to the Law of Total Tricks this is the correct action, and the contract may devolve to NS without further action by either side. This is what happened in the Meckwell hand discussed above.

This analysis involves only the most likely splits so we are dealing with the most even splits possible. A full analysis requires a computer program that takes into account all possibilities. At the table the extreme cases will take care of themselves. This analysis points to the need for methods that deal with what is most probable, by which we mean the development of cooperative doubling techniques based on shape information.

A Mike Lawrence Overcall

In his 1979 classic on overcalls Mike Lawrence suggested the following hand as a 1♠ overcall of a 1♥ opening bid: ♠ QJ97 ♥ 86542 ♦ A ♣ AJT. He considers this an automatic call at matchpoints and a minimum call at IMPs. So here we have another version of the half-empty overcall. Lawrence argues that holding 5 hearts is an inducement to bid, as it increases one’s chance of finding a fit. The overcaller can make game opposite ♠ K8642  ♥ J ♦ T863 ♣ Q93. Well, anything is possible, but I have not had great success when I have tried it. It seems that the LHO holds that hand more often than partner, but that is unlucky – it should happen only half the time.

We can look at the hand shapes to see if some insight emerges from the consideration of the most likely division of sides. Assume West has opened 1♥ and North has overcalled on 4=5=1=3 shape. What shapes are the most likely for West, East, and South?

 

	North	West	A	B	N&A	N&B
Spades	4	2	3	4	7	8
Hearts	5	5	2	1	7	6
Diamonds	1	3	5	4	6	5
Clubs	3	3	3	4	6	7
				Total Trumps	14	16

 

One should place West with 5 hearts for his opening bid. It is most likely that the other missing cards are divided as evenly as possible between the 3 hands. Although the best fit for North-South is likely to be in spades, the number of total trumps doesn’t encourage vigorous action by South even when holding a 4=1=4=4 shape (Hand B). A jump raise is not recommended under these circumstances. Holding Hand A any red-blooded South will give partner a single raise, but the portent couldn’t be worse. This is the half-empty aspect of Lawrence’s overcall. Let’s look at the half-full aspect and assume Player B holds the favorable shape of 5=1=4=3. The following layout is the most probable.

 

	North	West	A	B	N & A	N & B
Spades	4	2	2	5	6	9
Hearts	5	5	2	1	7	6
Diamonds	1	3	5	4	6	5
Clubs	3	3	4	3	7	6
				Total Trumps	14	17

 

If North finds Player B sitting opposite, prospects are bright, but if he finds Player B on his left, prospects are bleak. East passes and South bids 2♦ , West passes, and the doubling begins. In this situation the glass is not half-empty or half-full, it is more like ¼ full and ¾ empty. The reward may be great, but the risk is also great, which makes sense. Acceptable risk is a matter of probability: the expected gain versus the expected loss.

In both situations considered above, the downside is represented by a Total Trump count of 14, the worst possible situation. How likely is it that the Total Trumps reach 17? We can compare the numbers of combinations. The ratio of the combinations for more favorable 17 to the less favorable 16 is 3/5. We then come up with this comparison:

 

Conditions		Weight	Percentage
Favourable	(16 Total Trumps)	3	19%
Neutral	(16 Total Trumps)	5	31%
Unfavourable	(14 Total Trumps)	8	50%

 

North may escape dire consequences if EW can’t double for penalty, but the bad results will outnumber the good results – I would guess in the ratio of 5:3. Taking action when one holds length in opener’s suit doesn’t look like a good bet to me, even when balancing.