Half-Astro

This is a simple defence to 1NT that I've been playing recently. It's my own invention, though given how simple it is, it would be surprising if no-one had tried it before. It works like this:

2C = both majors
2D = spades and a minor
2H/2S/3C/3D = natural

Double is normally for penalties. (A possible variation by a passed hand, or perhaps against a strong NT opening, is to play double as showing hearts and a minor.) And a 2NT bid would probably show the minors, though there are other possibilities for this as well.

The advantage compared to other Astro variants is that it is easier to play. In particular, overcaller's partner is is a better position when one of the two-suited bids comes up. For example, we can compare it to "Asptro" where 2C shows hearts and another suit and 2D shows spades and another suit, showing the weaker suit with both majors. Asptro is good at finding the right major to play in when overcaller has both majors (unlike "Astro" or "Aspro"). The problem is that for both 2C and 2D there are three different possibilities for the second suit, and it is not always easy to cater for all three. It is particularly difficult over 2D, where overcaller's partner can be faced with problems like these:

• Holding 3 spades and 3 or 4 hearts, and no interest in game. Here we want to play in 2H if overcaller has hearts. But bidding 2H invites partner to bid 3 of his minor if he has a 5-card minor. This may not be what we want: we would often prefer to play in 2S in that case (for example with 3=3=5=2 shape).
• Holding a decent hand with 4+ good hearts, where we are interested in game if partner has hearts as his second suit but want to stop in a part-score otherwise. This is impossible to handle because if we bid 2H (or 2S) then partner will pass when he has hearts, which is not what we want.

Playing Half-Astro there are not so many hand-types to worry about: instead of three possible two-suiters in each bid, we have only one (for 2C) or two (for 2D). And after either of these bids, overcaller's partner can use the next step to ask which suit is longer.

Of course, what Half-Astro does not have is a way to show hearts and a minor. With these hands we have to bid naturally (or pass if we do not have a suit good enough to overcall). But this is not such a huge disadvantage, since showing two-suited hands without spades is less important - sometimes when we could find a heart fit opponents might be able to bid spades over it.

I play this defence because it is effective and yet very simple. You could agree it with a new partner and expect not to have any mishaps. It's one of the very few conventions I've come up with which has been taken up by people who aren't my partners!

Polish Club: Responding to 1C

Since I posted the link to the Polish Club system notes last week a few of us have been practising bidding on BBO. Some parts of the notes have been clarified, and we've made some slight changes to the auctions after 1C : 1D , 1M. I expect there will be a few more details changed while we get used to playing the system. The latest version is on the webpage.

This post is about the response structure that we've chosen to use.

The responses are built around a 1D negative; 1H and 1S are natural with 6+ HCP. This is not the only possible way of doing things, but it is by far the simplest, and it's not obviously worse than anything else. When playing Polish Club, a 1D negative is very useful so that all the other responses can promise enough strength for game when opener has the strong hand. With more natural systems many people have started to use transfer responses to 1C, but if these were incorporated into Polish Club you would have to worry about how to show a strong hand while leaving open the possibility of playing in a part-score. While it is interesting to investigate how you might get this to work, I don't feel it offers any improvement on the simple negative. Another possibility is to keep the 1D negative but invert 1H and 1S; this seems to be more trouble than it's worth, forcing the bidding to 2S when opener has clubs and spades, which is worse than forcing to 2H when he has clubs and hearts.

So, our 1D, 1H and 1S responses are essentially the same as in most Polish Club variants. The other responses are perhaps a little more unusual.

In WJ05, the 2C response is forcing (with at least invitational strength), and weaker hands with clubs have to start with 1D. We prefer to bid an immediate 2C on the hands which want to play in 2C when partner has a weak NT. So 2C shows about 6-10 HCP with a 5+ suit. This works particularly well if the opponents interfere: then we are pleased to have got the club suit in immediately, whereas if we had started with 1D we would be worried about missing a club fit and do not have a strong enough hand to compete at high levels without help from opener. Even if the auction is uncontested, the 2C response is an excellent way to start when opener has a strong hand, because responder has shown a suit and limited his hand, and opener's rebids are all very easy, forcing to game with 18+.

Of course we have to find somewhere else to put the strong hands with clubs, and we use the 2H and 2S responses for this. These bids are not particularly useful as natural bids. In WJ they are natural and show strong hands, but this does not come up very often, and if opener has a weak NT (as usually happens) we have so much space available after 1C : 1M , 1NT that there is no problem showing the strong hands there. In a natural system I like to use 1C : 2M as weakish (maybe 4-8 HCP), but the main advantage of this is that 1C : 1M , 2C : 2M can then be an invitational hand (rather than having to jump to 3M). This is not needed in Polish Club because the 2C rebid promises extra strength. Using 1C : 2M to show an even weaker hand would just pre-empt partner in the very likely event that he has the strong type. So none of the normal, natural meanings for 2M make much sense in my opinion. This, then, is the perfect place to put some strong hands with a minor suit. In fact we use 2H to show precisely 5 clubs and 2S to show 6 or more.

Next we need to think about the 1NT response. It's actually rather difficult to find a suitable meaning for 1NT. Here are two possibilities that we considered:

• 7-10(11) HCP balanced, no 4-card major. This has the advantage of preventing LHO from bidding a major at the 1-level. It's also a good start to the auction if opponents do compete. On the other hand, playing 1NT this way means that the sequence 1C : 1D , 1M : 1NT is underused and gives away too much information to the opponents (they could deduce that they have half the deck).
• Showing 5+ diamonds (6+ diamonds if less than a game force). This would probably be used in conjuction with a 2D response showing 5 diamonds and 4 clubs. This has the advantage that all unbalanced hands with positive values can show their longest suit immediately over 1C, putting us in an excellent position in competitive auctions. The problem with using 1NT this way is that if the partnership has the values for game, it has probably wrong-sided the potential 3NT contract. We like to play methods where responder can show shortage when he has an unbalanced hand, but it is not so good to point out this weakness to the defence if this hand is going to be declarer.

In the end we have chosen to do neither of those things: our 1NT response is natural and shows invitational values opposite the weak type (a good 10 or 11 HCP). This is one of those bids which is a bit infrequent but is great when it comes up; it avoids having to go to 2NT to make an invite. In fact we might use this bid even on some hands with a 4-card major. We can compare this to the 1NT response in WJ05, which shows 9-11 HCP. This seems an odd choice: it is not a genuine invite, and if you allow opener to bid 2NT on a maximum weak NT you might play an unnecessary 2NT with 14 opposite 9. If you're happy bidding 1C : 1D , 1M : 1NT on an 8-count, it seems much more sensible to do this on 9-10 as well, so that 1C : 1NT can be a real invite.

Our 2D response shows 6-10 HCP with 6+ diamonds, similar to the 2C response showing clubs. This means that our 2-level responses look rather similar to those in some Swedish Club systems (2C/2D natural and not forcing, 2H/2S show strong minor-oriented hands). However, we cannot bid 2D on hands with a bad 5-card diamond suit because we frequently have shortage in diamonds for our 1C opening. So hands with 5 diamonds in this range have to go through the negative instead - these hands will either be balanced or have 4+ clubs, since we would bid 1M with a 4-card major. We also have to start with the negative on strong hands with primary diamonds.

Our 1D "negative" in fact has four possibilities:

• 0-6 HCP, any shape.
• Up to 10 HCP with no 4-card major: balanced or 5D-4C.
• Game force with 5+ diamonds, or invitational with 6+ diamonds.
• Balanced game force with no 4-card major, not wanting to be declarer in no-trumps.

As with most Polish Club variants, this means that if the auction begins 1C : 1D , 1M both partners could still have a very wide range of hands. But it is very easy to sort things out: responder bids 1NT if he wants to play there opposite a weak NT type, and now opener can make a very descriptive rebid if he has a stronger hand. We use responder's 2C and 2D rebids artificially: 2C shows the strong hand with diamonds, and 2D shows the 5D-4C type with 6-10 HCP. We do not need to worry too much about interference in this auction because both opponents have passed at least once already.

Latest Developments in EBUland

For the bidding theorists amongst us, this minute from the latest meeting of the Tournament Committee is of interest:

The committee considered a letter from Bob Rowlands, addressed to various parties, including the Laws & Ethics Committee. The issue raised was the use of a system of transfer openings and responses [presumably moscito or something similar - DC] by a pair competing in the National Pairs Final. The pair was also the subject of previous similar correspondence regarding the National Inter-Club Knockout.

Mr Rowlands felt it was totally inappropriate that such a system should be allowed in events involving short rounds, when opponents have little chance to prepare themselves, and also in events such as the NICKO, which should be used to encourage club players to participate in national tournaments. If the EBU should stand by its decision to run all events at level 4, then this system should not be allowed at that level.

The committee unanimously agreed with Mr Rowlands’ sentiments, and wished to strongly recommend to the L&E that they reconsider their stance.

Despite playing a Level 4 system myself, I have to agree with this - in fact I said so in an earlier post. It just reinforces my belief that the "correct" level for general tournament play is somewhere "between" Level 3 and Level 4. It will be interesting to see what the L&E has to say about this, since they previously decided that it was "a Tournament Committee matter" ...

Polish Club System Notes

One of the reasons I haven't posted much on my blog recently is that I've been writing up my system notes for Polish Club. This is a system that I've been playing recently with Mike Bell, and we've tried to come up with a version which is an improvement on standard Polish Club variants such as WJ.

The notes can now be downloaded from this page.

The notes focus only on the 1C opening, and are very detailed, with a particularly extensive section on dealing with interference.

A few things that make our version different from other Polish Club variants such as WJ05 are:

• All balanced hands without a 5-card suit in the 12-14 HCP range are opened 1C.
• Not all 18+ HCP hands are opened 1C: some are opened 1D or 1H instead.
• The 2C and 2D responses to 1C are not forcing; we have artificial sequences to deal with game-forcing minor-suit hands.
• The 1NT response to 1C is invitational.
• After an overcall, we use a mixture of natural bids and transfers by responder.
• There are many other artificial sequences, both in an uncontested auction and in competition, including frequent use of opener's diamond rebids as artificial.

Some of the reasons for these things have already been discussed in this blog. I might write about a few of the other ideas at some point.

Doubles Webpage

I've created a new page to explain why I believe the EBU rules need to be changed.

Alerting of Doubles

Hopefully this means I'll be able to avoid cluttering up this blog with further rants on the subject. :)

Invitational Jump-Shifts

This method is quite popular amongst players of 2/1 GF: after a major-suit opening, a jump to 3 of a lower suit is invitational and shows a good suit (6+ cards). Personally I think this is a truly dreadful idea.

The problems aren't too difficult to spot. It's a very space-consuming response. This means that it will make subsequent bidding difficult, and you can't afford to make the bid on too wide a variety of hands.

In my experience the invitational jump-shift is almost guaranteed to put opener in a difficult position. Let's try these very ordinary hands, after the bidding starts 1S : 3D -

S AJ974
H J3
D Q5
C AKT4

Here perhaps partner can stop the hearts and you can run lots of tricks in NT. Or perhaps he can't. How are you going to find out, with no bid available below 3NT?

S KQJ52
H AQJ83
D 3
C 84

Here if partner has heart support you might belong in 4H, but alternatively the hand could be a complete misfit not making anything higher than 3D. Make the hand a little stronger and you know you should be in game, but will you find 4H if that's the best spot? How will partner know you have five hearts?

S AK965
H A53
D K842
C T

A lovely hand in support of diamonds, but how do you distinguish this hand from all the other different shapes that might want to raise?

Of course, if you play a 2/1 as absolutely forcing to game, you do need somewhere to put these invitational hands. But not all invitational hands with a good suit are suitable for an IJS. It is dangerous to make an IJS on hands with tolerance for parnter's major, or with four cards in the other major, because of the possibility of missing a major-suit game. So playing IJS does not really solve the problem of what to do on these hands.

Much better, at least when one of these invitational hands comes up, is to be playing a system where a 2/1 is not absolutely forcing to game, with responder's rebid of his suit showing the invitational type. Starting at the lower level gives you so much more flexibility: not only does opener have an extra chance to describe his hand, but responder can make the bid on hands which are not such pure single-suiters, because he is not committed to showing the IJS type. For example, over 1S, hands with four hearts and a six-card minor are no longer a problem, because after a 2/1 response, responder can afford to raise a heart bid to game, and will only rebid his minor to show an IJS hand if opener does not have hearts.

Naturally this would make life more difficult when you do actually have a game-forcing single-suiter. But personally I think it is relatively easy to find a way to bid these hands. Even if constrained to play natural methods, I would much prefer to be playing 2/1 "GF except rebid" than absolute game-force. And with a bit of artificiality there is plenty of room in most cases to distinguish invitational from game-forcing hands without having to invent suits or NT bids. Certainly in a natural GF system, responder's 2-level rebids (when available) tend to be underused, and can be redefined to include the game-forcing single-suited type. Really I think that using a cheap response like 2C solely for natural game-forcing hands is a serious waste of space.

An alternative for people who want to keep their 2/1s as game-forcing is putting the invitational single-suiters into the 1NT response. This has several drawbacks. You don't get to show either the suit or the strength immediately, and if you later bid your long suit it might be difficult to distinguish this from a weaker hand. Also if 1NT is not forcing, you may well be missing a better contract if opener passes. It works better over a 1H opening than over 1S - particularly if playing Kaplan Inversion so that the "forcing NT" hand bids 1S rather than 1NT - because responder has a 2S rebid available as artificial to distinguish weak single-suiters from invitational ones. Even then, I still prefer making a two-over-one response, showing the suit immediately, if the system can be arranged to allow for it.

What does "Natural" mean?

Most bridge players understand roughly what to expect from a "natural" bid, but it's quite a difficult term to define precisely. This is unfortunate because, as well as being a useful word for describing what a bid means, it's also often used in system regulations and alerting regulations, where you really need a good definition.

The meaning also varies slightly depending on where you live. Let's look just at suit bids. In America, most people seem to consider a suit bid to be natural if it promises length in that suit, even if there is some additional information given. For example, a Muiderberg 2S opening (showing 5 spades and 4+ cards in a minor) would be described as natural. They would say it is "natural but conventional". The WBF, on the other hand, defines "natural" to be the opposite of "conventional". Of course, "conventional" is another term which is extremely difficult to define, but it is clear that the Muiderberg 2S is conventional, and so is not natural according to the WBF's definition.

In England our understanding is somewhere in between. The old Orange Book (pre-2006) summed it up quite well: a natural suit bid was

a bid of a suit which shows that suit and says nothing about any other suit. ...

So a Muiderberg 2S opening is not natural, because it shows length in another suit (even though that suit is unspecified). But, unlike with the WBF's definition, it is still possible for a natural bid to be conventional. We would say that the opposite of natural is artificial. An example of a bid which is conventional but not artificial is a 1D opening which promises an unbalanced hand. This certainly shows diamonds, and says nothing about any other suit, but there is an additional agreement about the hand as a whole which makes it conventional.

Because I learnt my bridge in England, this is the definition which feels most right to me (and not just because I'm an avid reader of Orange Books! - this really is what people mean by "natural" in England, give or take some of the details). Unfortunately, both of the conditions in that old OB definition are slightly faulty.

First of all, what does it mean to "show that suit"? The old OB definition continued:

The suit shown will be at least four cards before opener rebids but may be three cards from then on; exceptionally a bid of 2C in a 3=4=3=3 hand precisely in response to 1S is considered natural.

This definition appears to make bids such as 1H (pass) 2H artificial if they could be made on three cards. In practice, this oversight was just ignored - no-one was suggesting that these bids should be alerted - but it did look a bit silly. It was corrected in the new version of the OB, where the definition of a natural suit bid is:

A bid of a suit which shows that suit and does not show any other suit; the suit shown will be at least three cards long except that preference bids and raises may be on shorter suits. Note that in earlier rounds of bidding a natural suit bid usually shows at least four cards.

This is a much better definition of what it means to "show" the suit bid. It's also an improvement for the bids which show only three cards suit. According to the old OB defintion, a 3-card 1H opening was alertable because it was not considered to be "natural". But this doesn't seem right - it surely is a natural bid, but ought to be alerted because it is unexpected.

The second part of the definition is much harder to get right, and neither of the two EBU versions really works properly. According to the old regulation, a natural bid should "say nothing about any other suit". Presumably this was discarded when it was realised that nearly every natural bid shows something about the other suits, even if this is only from the negative inference that some other natural bid could have been made instead. And, for example, a pre-emptive opening bid which is played as denying a side 4-card major should still be considered to be natural. But the new version goes too far in the other direction, so that (for example) a 3H bid which shows heart length and club shortage is considered natural. This isn't a problem for the alerting regulations because club shortage is still unexpected, and therefore alertable, but it doesn't correspond to what we really think of as being "natural". I think the old definition is closer to the truth here, and they just needed to make an exception so that a bid could still be natural if it denied a certain amount of length in another suit (or suits).

Still, there isn't really any obvious way to define it perfectly, and it's interesting to see how various different authorities try (and fail) to do it. And that's without even considering what a "natural" no-trump bid should mean.