Game Tries

I was reviewing my system notes for what I play with Elwood, and it occurred to me I hadn't mentioned game tries.

These days, it seems like the “default” game try is what is known as a help-suit try. I have to say I have been distinctly unimpressed by this concept, and generally just don't understand them. When I was learning to play, 40-odd years ago, people used long-suit tries or short-suit tries. In either case, you weren't asking for help: you were describing your distribution, and hoping that the additional information assisted partner in re-evaluating his hand. Help-suit tries seem to me to fail to convey any useful information, and how Responder is supposed to judge what is “help” seems (at best) murky.

With that little rant out of the way, I can confess that I do play some game tries with Elwood, but it's a scheme he produced which he calls modified Kokish game tries. To the cognoscenti, that should ring alarm bells, because as well-known a theorist as Kokish is, he is also well-known for producing some of the most complicated schemes you'll find anywhere. (If you don't believe me, check out his “simple” defence to the Multi – it runs about ten dense pages!) But what Elwood put forward is not too bad.

The cunning ploy at the heart of the scheme is that (after a major-suit opening and simple raise) the next step bid is a relay asking Responder to name the cheapest suit in which he would have accepted a help-suit game try. If Responder retreats to 3M (declining all tries) or jumps to 4M (accepting all tries), Opener's hand remains concealed. This also leaves other bids free: there is room for short-suit tries, and also to mention a 4-card holding in the other major (to suggest an alternative trump suit).

The scheme is:

1S-2S,

2NT* asking bid for help-suit tries

3C*/3D* short-suit tries

3H natural, 4 hearts

3S* short-suit try in hearts

1H-2H,

2S* asking bid for help-suit tries

2NT* shows 4 spades

3C*/3D* short-suit tries

3H* short-suit try in spades

I don't know if this arrangement would meet Kokish's approval (I don't actually know what his scheme is), but I like that it seems to be pretty comprehensive.

These bids are off over interference, but Responder being a passed hand isn't a problem. They are also in play after Drury, for example after pass-1S; 2C-2D; 2S-?

HelLo

Elwood has expressed his dissatisfaction with Cappelletti (which we were playing against weak notrumps) several times over the past couple of weeks. I'm no fan of it myself, so I was looking for a replacement gadget that would be acceptable to us both. Elwood isn't too interested in Lionel, which is probably better suited against strong no-trumps anyway. Eventually, I came across HelLo, the improved version of Capp that Jerry Helms has invented. So we're playing that now, against weak no-trumps.

The scheme is:

double = penalty, usually balanced 15+, could be fair strength with a good suit as a source of tricks.

2C = either diamonds or a major-minor 2-suiter. Advancer must accept the transfer by bidding 2D, and Overcaller can pass or raise with diamonds or bid his major holding a 2-suiter. As with Capp, Advancer can then use 2NT to ask for the minor.

2D = hearts

2H = both majors

2S = spades

2NT = clubs

3C = both minors

This arrangement keeps a traditional penalty double, which is pretty much de rigeur against a weak no-trump, allows Overcaller to show all varieties of 2-suiter, and introduces a transfer element that Cappelletti doesn't have. Using 2H to show both majors isn't any better than using 2D as in Capp, but at least with this arrangement you get to show a major single-suiter straight away, and you get the partial transfer style.

I'm not entirely sold on the value of the transfer business. I wonder if keeping 2D as both majors and using both 2H and 2S as natural might not be just as good? Still, we'll keep it as specified for now.

Of course, since making this decision, we haven't encountered a weak no-trump, but there are some around. I'll post any interesting results.

Gadgets

I've been playing with Elwood a lot more regularly, usually once a week. Conversely, our system tinkering and email volume has dwindled, after a couple of big spasms to get some complicated stuff in place. So, I think it's time I put some documentation out there for what we're playing.

The basic set-up of 2/1 with a weak no-trump is still there, of course. A couple of gadgets have been added to the responses. The most complex item is Muppet Stayman.

We had been playing a relatively simple version of Puppet Stayman over the 2NT opening all along. Muppet Stayman involves switching Opener's 3H and 3NT rebids. The purpose of that is to handle Responder hands that include 5 spades and 4 hearts. The other way round, Responder can transfer to hearts and then bid 3S. Playing Muppet, with 5 spades and 4 hearts, Responder can start with 3C, and if Opener bids 3H (denying a 4- or 5-card major) Responder has room to continue with 3S. Of course, that loses the transfer effect if Opener chooses to play in spades, so Elwood suggested a further reversal, swapping 3S and 3NT from Responder in this sequence. Needless to say, when you try to map out the ramifications of all this, not forgetting that, as Responder, you also want to know what to do with a 4-major 5-minor type of hand, you finish up with some sequences that are disasters waiting to happen, and a severe headache from trying to remember it all.

To make the effort worthwhile, we decided to play the exact same thing over 1NT. So while regular Stayman is still in place and handling a lot of stuff, with a game-going hand Responder also has 1NT – 3C available as Muppet. This fits with our revised philosophy for opening 1NT. I started out with the traditional British attitude, that 1NT shouldn't include a 5-card major, except for odd cases. But influenced by results (and by Fantunes, I admit), I have let the pendulum swing towards the opposite extreme. We now explicitly open all 5-3-3-2s and most 5-4-2-2s (especially with 4-major and 5-minor) 1NT if they are in range. As a result, checking for a 5-card major is very much a good idea. For example, if Responder has a game-going hand 4-3 in the majors, the right choice is Muppet rather than regular Stayman. That way, he can find a 5-3 fit as well as a 4-4 or 5-4 fit. The one time we had a memory lapse that should have resulted in disaster, we actually dodged the bullet by both having the same failure. We both forgot the second inversion, and I passed Elwood's 3NT bid that should have shown 5 spades to find that he actually only had 2. If I had remembered, I probably would have corrected to 4S, to play in a 3-2 “fit”.

We are still playing 4-suit transfers over 1NT, and they can handle single-suited and wildly 2-suited hands. To complete the system of responses, we wanted to provide some tools for less extreme hands with both minors and a shortage in a major. Using 3H and 3S as the usual splinter responses, with a 3-suited hand and exactly 4 in the other major, takes care of quite a few. And we have defined the 3D response as showing both minors in a hand with a 3-card major. Then 3H asks, and Responder bids his shortage (3NT with short hearts). Of course, there are follow-up sequences, and the end result is (hopefully) that we can choose between 3NT, 4major and 5minor (or 6minor) on some sensible basis. This 3D response is quite cute (and original, I think), but rare. So far it has only come up once, steering us to 4H somewhat anti-climactically when Elwood showed short spades and I had opened with a 5-card heart suit. The real payoff for that bid will be when we get to a nice Moysian game, or to 5- or 6-minor, when no-trump doesn't make.

The other toy we have adopted is a mini-Roman 2D. This shows a 3-suited hand, 4-4-4-1 or 5-4-4-0. Commonly, this is played without much more definition, with some strength range around 11-15hcp. Some strong club pairs play a version where the the short suit is always diamonds, to handle a variety of hand that is difficult for their system. I started out somewhat skeptical about this bid's worth, feeling that it might be OK as a pre-empt, but would be difficult to use constructively, while if you finished up defending, declarer would have a blueprint for the play. Having run into it a few times, I changed my mind, conceding that it is annoying and sometimes quite awkward to defend against. So I decided to look into adopting it, and I produced some original ideas that have been run through Elwood's brain, and I think the result is quite interesting.

The ACBL's GCC insists on a minimum of 10hcp, so we are playing the opening as 10-14, which (as with Fantunes 2-openings) is weak enough to be quite pre-emptive, but strong enough to have some punch. To provide a constructive framework, I suggested some things to make our approach very major-suit oriented. Making the shortage always a minor is the first change. This does some good things to your response structure, but carries the penalty of reducing the frequency of occurrence. To help with that, and also just why not, Elwood suggested allowing 5-card hearts but not 5-card spades. That includes some “Flannery” hands, without screwing up our relays. Next, the usual set-up is to have 2NT as the strong response, whence Opener shows his shortage. Instead, I have Opener bid the minor suit he holds, and Responder can pass that or place the contract wherever he likes. Next, the 3C and 3D responses are not usually well-defined, as far as I can discover. So I suggested using them as proxies for hearts and spades. Elwood came up with the game-forcing relay sequences to follow, whereby Opener defines his strength and exact distribution. And then we have the 3H/3S responses as natural and invitational, and the 4H/4S responses as natural and to play (either weak or strong, like after a weak two opening). Almost all the time, when we play in a major, Responder is declarer. This is policy, since Opener's hand is well-known from the bidding (so if he becomes declarer, the defence can play double-dummy).

So far, it seems like the mini-Roman is occurring about once every other session. That doesn't sound like a lot, but I think it's comparable to the frequency of the weak two that we have given up – that didn't seem to come up all that often, either. And we've had at least a couple of good results from it.

Support doubles

I have been playing support doubles with Elwood all along, but I've started playing them with Agent 99 as well. Looking back over this blog, I don't see them being mentioned, so a new entry seems to be in order.

Support doubles and redoubles arise in auctions that begin
1suit – (something) – 1major – (call < 2major);
That is, Advancer bids over Responder's 1major, and it is still possible for Opener to raise to 2major. The commonest situation is an auction that starts 1minor – (pass) – 1major – (something), but the convention still applies if overcaller bid something other than pass as long as the basic framework is still there. And it applies to auctions 1H – (pass or double) – 1S as well.

The idea is that in this situation, Opener's rebid of double (or redouble, if Advancer doubled) shows exactly three card support for Responder's major. Raising the major explicitly therefore shows four-card support.

I am not a fan of The Law, and usually don't worry too much about whether a raise shows three or four cards. But in this particular situation, the support double turns out to be a useful bid when Opener might not have anything else really clear-cut to say. Systemically, Opener might have been planning to rebid 1NT or 2NT with three-card support for Responder. After the intervention, no-trumps may or may not still be attractive, but the double conveys good information without raising the level, keeping everything in play, so to speak.

A question I asked Elwood was whether support doubles only applied when Responder's major might be four cards. The dubious auctions would begin 1m – (1H) – 1S, where Responder is promising 5+cards. A counter-question is, if not support, what might a double show? There are various answers, but Elwood favors ignoring the complications and retaining all the doubles and redouble as support. I'm going along with that.

Another question was whether support doubles were mandatory. Elwood's answer was no, but if you don't have a good descriptive or temporizing call available, you should go ahead and double.

So that's a quick run-down of support doubles. There are other ways to play, but I would recommend giving them a try. I was skeptical at first, but I find that they do come up and are useful. You can't ask more of a convention than that.

Blackwood with a void

I forget sometimes (OK, most of the time) just how little competitive experience Agent 99 has. Not that I'm so terrifically well-seasoned, but I have been playing a long time, and I've lost to some internationally renowned players. And I spend an inordinate amount of time on the internet making myself feel closer to famous players, too. So there's actually a fair amount of stuff that is familiar to me that is probably a complete blank to her.

One of the more basic blind-spots cropped up the other day. What do you say when partner wheels out Blackwood, and you have a void?

Apparently, Agent 99 had never met this little conundrum before. Her improvised response was to treat the void as a key-card, which rapidly generated a zero as I put her in seven, and the defence found their ace. We were having a bad game anyway, and just to rub it in, some other pair bid and made seven on the wrong opening lead.

But there is an accepted way to respond. I suppose I thought Agent 99 must know it, but there's no reason why anybody would, unless they've been shown it. The basic scheme is to answer 5NT holding 0 or 2 keycards and a void, or bid 6 of the void holding 1 or 3 keycards and a void. There are some caveats and wrinkles, of course (there always are).

First of all, showing a void when you hold 0 keycards is hardly ever done. The only time Eddie Kantar (the RKC authority these days) would countenance it is when partner already knows you have a terrible hand, and is bidding RKC anyway. The sort of example Kantar gives might involve a 2C opening and relatively weak responding sequence, or something.

Second, only show useful voids. A void in partner's first-bid suit when you have agreed his second suit as trumps is almost certainly not an asset, and not something to brag about.

Third, if your void is in spades and hearts are trumps, and you have an odd number of keycards, you can't jump to 6S because that takes you past 6H. Jump to 6H instead.

This takes care of hearts and spades as trumps, when 4NT is the RKC asking bid. For us, it gets a bit more confusing when a minor suit is trumps, and we use Redwood, but it works just as well. Instead of 5NT to start showing voids, if you count the steps you see you get to 5D when the asking bid was 4D, or 5H when the asking bid was 4H.

Let's say clubs are agreed, and 4D is bid as 1430. Now,
4H = 1 or 4
4S = 0 or 3
4NT = 2 without the QC
5C = 2 with the QC
5D = 2 with a void
5H = 1 or 3 with a void H
5S = 1 or 3 with a void S
5NT = 1 or 3 with a void D
6C = I can't remember all this crap

Similarly, when diamonds are trumps and 4H is 1430:
4S = 1 or 4
4NT = 0 or 3
5C = 2 without the QD
5D = 2 with the QD
5H = 2 with a void
5S = 1 or 3 with a void S
5NT = 1 or 3 with a void H
6C = 1 or 3 with a void C
6D = I can't remember all this crap

As you can see, 5NT has become a “spare” bid, and can be used to show a void in the suit immediately higher-ranking than trumps. All the other voids can be named explicitly when you have an odd number of keycards.

Concealed Splinters And 4-card Raises

Splinter bids are bids that (by agreement) show partner a trump fit, a raise to game, and shortness in a side suit. This is terrifically useful information for assessing how well the two hands fit together, and splinters should be used at almost every opportunity.

There are a couple of tweaks that can be made, though. One observation is that usually, a splinter could be either a singleton or a void. Singletons are much more common, of course, but if you do happen to get a void, it would be nice to be able to distinguish it. Another thought is that the “traditional” splinter uses three responses to distinguish shortages in three suits. This is easy to use, but a tad inefficient. If you use one response to indicate that there is a shortage somewhere, a relay can be used to identify exactly where, if partner wants to know. And if he doesn't ask, you can bid game without revealing where the shortage is.

Another leaf on the tree of Niggle is the matter of jump-shift responses to major-suit openings, and using them for different flavors of raise. I don't like Bergen Raises, but the 4-card raises in Shep's 2/1 make sense to me. In addition, it's fair to say that if you don't need or want to use jump-shift responses as natural and strong, you need to find a good use for them, and that's not quite as easy as it sounds. Clearly distinguishing an assortment of 4-card raises seems like a very useful idea.

Putting all this together, you find that everything fits quite neatly. We include 2NT in the pantheon of jump-raises, and we include 3NT in the range of double-jump raises. Now, the cheapest jump raise is a concealed splinter. So 1H-2S and 1S-2NT are raises to game that include 4 trumps and a singleton somewhere. If Opener wants to find out where (ie usually), he bids the next step as a relay, and Responder bids the suit of the shortage, or bids the trump suit for the one suit he can't mention below 3. That is:
1H-2S; 2NT (relay) - 3C/3D/3H with a singleton C/D/S respectively. 1S-2NT; 3C (relay) - 3D/3H/3S with a singleton D/H/C respectively. This all leaves plenty of room for Opener to start q-bidding if slam is starting to look attractive. Note that if Responder q-bids the splinter suit subsequently, he is indicating the singleton ace.

We have only used one response so far, the cheapest jump-shift. There are 4 other single-jump responses, and these are used in reverse order of strength to indicate a 4+card raise (this idea being lifted from Shep's 2/1).
1H-2S 1S-2N Splinter (above)
1H-2N 1S-3C game-forcing raise without shortage (equivalent to Jacoby 2NT)
1H-3C 1S-3D 4-card limit raise
1H-3D 1S-3H 4-card mixed raise (weak but includes a defensive trick, an A or K outside trumps)
1H-3H 1S-3S pre-emptive, no defensive trick

Next we move up to the double-jump raises. The cheapest one is a concealed void splinter, showing a raise to game with 4 trumps and a void somewhere. As with the singleton splinters a level lower, a relay is used to locate the actual void. So,
1H-3S; 3N (relay) -4C/4D/4H shows a void C/D/S respectively.
1S-3N; 4C (relay) -4D/4H/4S shows a void D/H/C respectively.

At this point, we can look back and see that we have an array of raises with 4-card support, and can distinguish singleton and void splinters, which is a lot more than most people have in their bag of tricks. [infomercial voice]But wait, there's more.[/infomercial]

We actually have a few bids spare now. The double-jump raises, that most people use for splinters, are available for some different meaning. There is a type of splinter used by Opener called an Auto-splinter. For example, the sequence 1C – 1S; 4C! would indicate 4-card support for spades and a raise to game that includes 6 solid clubs (and a shortage somewhere else, therefore). This is off the main track of splinters, as the most important part of the message isn't actually the shortage but rather the honking great source of tricks. But it's still a good idea, and I think it makes sense to make that type of raise available to Responder also. So, for example, the sequence 1S – 4C! would show a hand like KJxx/xx/x/AKQT9x. It's true that you could start to bid this hand by bidding clubs, or by showing a diamond splinter. But I don't think either of those routes is really going to convey the combination of a very good suit with 4-card trump support.

Now we're talking. All I have to do is sell Agent 99 and Elwood on the idea.

Opening 2NT And Responses

 

The opening bid of 2NT sits in the twilight zone between ordinary opening bids and forcing strong openings. It has a reputation for being a “slam-killer”, in that it can be difficult to identify slam potential when the optimal contract depends upon fit. Responder knows that opener has a big hand, but 21 hcp or so can have more than one weakness, especially in a balanced hand, so he can’t go hog-wild. And there is very little room for investigation before you find yourself past game and heading into dangerous territory.

An additional complicating factor is that the distributional requirements for the 2NT opening tend to be somewhat lax. These days, the 1NT opening gets used on semi-balanced hands or maybe even with a singleton, and that sort of thing would have been heresy 30 or 40 years ago (and still is frowned upon in some quarters). But even back then, it was routine to bend the “balanced hand” rules for the 2NT opening, because the grey area between opening 1suit and opening 2C needs the help in Standard American and similar systems. (For a while, one of the cornerstone advantages of Acol as a system was its intermediate 2-bids, because they specifically addressed this region of hand-strength. But as people were seduced by weak 2-bids, that advantage was surrendered by many Acol players.) So whatever responding scheme you choose, you have to be aware that both opener and responder may not exactly be balanced.

Nevertheless, the usual responses are similar to those used over the 1NT opening. Stayman, transfers, and think of something to do with the 3S response and some way to handle minor-suit responding hands. Puppet Stayman addresses the possibility that opener has a 5-card major suit, as well as making the stronger hand declarer, so that sees quite a lot of use over 2NT. I play it with Elwood, but not with Agent 99. To be honest, I’m not as thrilled with it as I might be. I just can’t recall any hand I’ve played where we finished up in the wrong contract and I said to myself “if only we’d been playing Puppet…”. The last time I opened 2NT with a 5-card spade suit, playing with Agent 99, she answered 3H (transfer)! I bid 4S, just to make sure we got there, and we played quietly in game like everybody else. There is a memory-burden attached to Puppet, not a large one, but still, it was something that Agent 99 didn’t want to add to her repertoire when we were starting out, so we play just regular 4-card Stayman.

There are other possible approaches. I came across an old scheme of Jeremy Flint the other day. After 2NT – 3C; opener rebids:

3D = no major OR flat major OR 4-4 with diamonds

3H = hearts and clubs

3S = spades and clubs

3NT = 3334 or 3325

4C = majors, minimum

4D = majors, good controls (7+)

4H = majors, maximum points, <7>

Notice that all bids from 3H to 3NT show C, and all bids over 3NT show both majors. 3D is basically natural.

2NT – 3C – 3D – 3H (natural)

3S = 4333, 4243, 4342

3NT = no major

4C = H and D **

4H = 3433

2NT – 3C – 3D – 3S (natural)

4C = S and D **

** Note that the 4C bid shows support and D, allowing partner to bid 4D control ask (4H=5).

2NT – 3C – 3D - 3NT = 54xx.

This is kind of interesting. The opening has limited opener’s high-card strength, and the 3C inquiry seeks elucidation of his distribution. Although it looks a bit intimidating at first glance, a lot of the bidding is actually pretty natural, and the provision of control counts or inquiries for slam investigation is quite economical. The biggest weakness I see straight off is that it doesn’t seem to take account of off-shape openers, not even a 5-card major. So I don’t think anybody will be taking this up these days.

On the other hand, the other part of Flint’s responding scheme uses the 3S response to account for various minor-suit hands. Opener has to make the puppet bid of 3NT, and then responder has

4C/D = invitational

4H = xx45

4S = xx54

4NT = good 5-5 minors. 

In the original, 2NT – 3NT showed a moderate hand 5-5 in the minors, so to simply raise to 3NT you would make the 3S response, and then pass the forced 3NT rebid. This multi-purpose 3S response is still used today in some circles. At the Manhattan, the version that is sometimes taught has 2NT – 3NT as natural. The 3S response followed by 4NT shows 5-5 in the minors but is non-forcing, and 3S followed by 5C is 5-5 minors forcing to slam. Depending on who is doing the teaching, the 4C and 4D bids may be natural or may indicate the opposite minor (that is, 4C shows diamonds and 4D shows clubs). The advantage of leaving the bids as natural is that it’s one less thing to remember, and opener may be able to use Redwood or something conveniently. The idea of not bidding your suit is, of course, to allow the strong hand to be declarer. This whole thing isn’t that great, but it provides a relatively simple “one-stop shopping” solution for various hands, so it’s definitely better than nothing.

In his columns written for OKBridge, Marc Smith advocated using 3S as Minor Suit Stayman. In conjunction with this, he suggested 4C/4D as Texas transfers (for H/S, respectively), thus freeing the 4H and 4S responses to show single-suited hands with clubs and diamonds, respectively. I haven’t tried that arrangement, but if he suggests it, it must be somewhat playable.

To be honest, the best solution for the 2NT opening may be to stop using it. In ACBL territory, one of the best uses of the 2D opening may be to replace the 2NT opening. The Mexican 2D is a known convention, originally part of the Romex system. This opening shows a big balanced hand, usually 18-19 hcp, but there’s no reason why you couldn’t make that 20-21 or 21-22, or whatever your usual 2NT range is. The extra couple of bids available to responder may not seem like much gain, at first glance, but if you check out Martin Johnson’s ideas at Bobby Knows Bridge (http://www.freewebs.com/bobbybridge/), you can see how to use (in effect) four-suit transfer responses. You actually gain quite a lot of flexibility, responder being able to stop in lower part-scores when he has a bust, and also able to show more below 3NT when he has some distribution. It seems to me that the 2D opening looks much better than 2NT.

So then what do you do with the 2NT opening? Well, Alder pre-empts look good, but the ACBL would have the vapors. Transfer pre-empts are relatively innocuous, but even they are listed as a mid-chart convention. Or you could just go with 2NT showing a minor two-suiter – that actually made it to the General Convention Chart.

lebensohl over 1NT

I can’t believe I haven’t yet posted anything about lebensohl. I have a very ambivalent, or at least confused, attitude to this convention. You really have to play it or something equivalent: just relying on natural bidding and common sense doesn’t work well enough in these situations. In principle, the convention is simple and should be easy to use. And yet, when it comes to actually using it at the table, I find that it drives the new user crazy (speaking both as a once-new user myself, and as a teacher who has had the dubious pleasure of trying to teach this thing). Perhaps the real difficulty is that while the problem situations occur often enough for you to need the convention, they don’t occur often enough for you to become comfortable with the different sequences just in at-the-table play. Some dedicated practice with customized deals is in order. Anyway, here is a quick overview.

When the opponents interfere over 1NT, responder may have difficulty describing his hand. Given that he has a suit worth mentioning, there are at least three levels of strength that he may hold: weak, invitational, and strong. And generally, he only has two bids available below 3NT, mentioning his suit at the two-level or jumping to the three-level. The lebensohl convention is a way to give him more flexibility.

The basic idea is that after a 2-level overcall, a natural bid of 2NT is very unlikely to be required. Using lebensohl, a bid of 2NT by responder is a relay, requiring opener to bid 3C. Then if responder passes (showing clubs) or bids a new suit at the 3 level, the bid is competitive or invitational. If the responder could have mentioned his suit at the two level, his three-level bid after the relay is invitational. If responder didn’t have a chance to bid his suit at the two-level, his bid after lebensohl is just competing, and opener is expected to pass. Conversely, if responder doesn’t go through the relay but bids a new suit at the 3 level directly over the overcall, that is strong and forcing. And if responder can get his suit in at the 2-level over the overcall, that is just weak.

So for example, if the bidding goes:

1NT – (2H) – 2S that is natural and weak, just competing

1NT – (2H) – 3S that is natural and strong, responder wants to be in game

1NT – (2H) – 2NT! – (Pass) – 3C! – (Pass) – 3S that is not forcing, invitational

1NT – (2H) – 3D that is forcing with at least invitational strength

1NT – (2H) – 2NT! – (Pass) – 3C! – (Pass) – 3D that is just competing

So far, so good, but there’s more. Another common convention, without a name that I know, is that responder can cue-bid to indicate game-going values and four cards in one or both majors (a way to make up for Stayman being lost). For example:

• 1NT – (2H) – 3H! means I’m strong enough for game, and have four spades
• 1NT – (2H) – 3NT means I want to be in game and I don’t have spades

The lebensohl relay can also be used in both these sequences, and the usual meaning attached is related to whether or not responder has a stopper in the enemy suit. The two possible methods are generally referred to as “fast denies” and “slow denies”. If you play “slow denies”, then the above mentioned two sequences also show a stopper in the enemy suit, hearts in the examples above. Going through the relay denies having a stopper in the enemy suit. So

• 1NT – (2H) – 2NT! – (Pass) – 3C! – (Pass) – 3H game-forcing with four spades, but no heart stopper
• 1NT – (2H) – 2NT! – (Pass) – 3C! – (Pass) – 3NT game-forcing without either spades or a heart stopper

If you play “fast denies”, the meanings are reversed: jumping to 3NT immediately over intervention actually denies having a stopper in the enemy suit. (Opener therefore has to stay awake and bid something if he too lacks a stopper). There may be an advantage to playing “fast denies” if the opposition wants to keep bidding. The immediate jumps tend to shut out advancer, and the times when responder doesn’t have a stopper are perhaps more likely to be the times you really want to do that. So, while it would seem that “slow denies” has the edge in naturalness (if there is such a word), in practice, almost everyone plays fast denies. Just for the hell of it, with Agent 99 I play “slow denies”, and so far we haven’t had any problem.

 

1NT Forcing and Bart

If you play 5-card majors, it is best to play the response of 1NT as forcing for one round. While this prevents you from playing in that most admirable contract, it provides an extra round of bidding to find an accurate part-score. While most people manage without too much further elaboration, there is a case for an additional layer of artificiality. For one thing, opener’s “natural” rebids of 2C and 2D can only promise a 3-card suit at best. And after 1H – 1NT opener may be forced to bid 2C on a doubleton in a 4=5=2=2 hand that isn’t strong enough to reverse. Also, responder’s hand may be balanced and weak to invitational in strength, or it may be unbalanced, and range from weak to just under the strength needed for a 2 over 1 response (and if you play 2/1 GF, “just under” is really not a bad hand). After 1S – 1NT or 1H – 1NT, opener will make a jump rebid with a maximum opening, generally 18+hcp. There is therefore a potential problem when opener is 15-17hcp and/or responder is near his maximum, say 9-11hcp - game needs to be considered in these cases. And yet another situation arises after 1S – 1NT when responder has a 5-card heart suit and a doubleton spade – it can be difficult or impossible to identify the best part-score (two hearts or two spades).

There is more than one way to address some of these issues. Perhaps one of the simpler ideas is a convention called Bart that addresses sequences after 1S – 1NT; 2C. (The situation after 1H – 1NT isn’t quite so bad, but while a similar treatment can be applied, there are differences, and I’ll stick with the spade opening for this description). There’s more than one version of Bart floating around as well, but this version seems fairly straightforward to me.

First, the agreement can be made that opener’s 2H and 2D rebids promise at least 4 cards. The 2C rebid is then redefined as not meeting the requirements for any other bid, making it alertable as not really natural any more. Second, bidding following opener’s 2C rebid is elaborated by making responder’s 2D at this point an artificial relay (also alertable).

We need to consider how opener will bid, and also what hands responder is trying to show. First, opener’s first rebid:

1S – 1NT;

• 2S shows 6+cards, 11-16hcp
• 3S shows 6+cards, 17-18hcp
• 4S shows 6+cards, 19-20hcp
• 2NT shows 18-19hcp
• 3C/3D/3H are natural, 4+cards, 18+hcp, forcing
• 2D/2H are natural, 4+cards, 11-17hcp
• 2C catches everything else, generally either clubs or a balanced hand in the 11-17hcp range

Now consider responder’s hand. Generally, the 1NT response is made on hands that are balanced, or that are unbalanced but not strong enough for an immediate 2 over 1 response. The 2D relay will be used for:

• Balanced 10-12hcp including 4-card club support
• Strong (9-11hcp) with clubs
• Strong (9-11hcp) with diamonds
• 3-card limit raise that includes 4-card club support
• hands with 5 hearts and 2 spades

Direct bids (not using the relay) will then show

• Balanced 10-12hcp without 4-card club support
• Weak (5-8hcp) with clubs
• Weak (5-8hcp) with diamonds
• 3-card limit raise without 4-card club support 

(The exclamation is used to indicate a bid that should be alerted).

After 1S – 1NT; 2C! responder bids

• Pass – obviously weak, should be 5+clubs and at most 1 spade
• 2D! – artificial and forcing, generally 9+hcp but can be weak with 5 hearts and 2 spades
• 2H – natural, weak, 6+cards, opener should not return to spades
• 2S – natural, weak, usually a doubleton
• 2NT – natural, invitational 10-12hcp, denies 4 clubs or 3 spades
• 3C – natural, 5-8hcp, 5+cards, principally a courtesy raise to shut out the opponents
• 3D – natural, 5-8hcp, to play. Weak diamond hands have to play in 3D rather than 2D, the only serious drawback to the convention.
• 3S – limit raise with 3-card spade support, denies 4 clubs
• 3NT – natural, 13-15hcp, denies 4 clubs or 3 spades

After 1S – 1NT; 2C! – 2D! opener defines his hand a little more:

• 3H – 3-card hearts, 15-17hcp
• 2NT – natural, 15-17hcp, less than 3 hearts
• 2S – 11-15hcp, less than 3 hearts
• 2H – 11-15hcp, 3-card hearts

If opener has bid 2NT or 3H, responder should be able to place the contract. Otherwise, responder completes his description:

After 1S – 1NT; 2C! – 2D!; 2H

• Pass – natural, 5-card hearts
• 2S – usually a doubleton honor with 8-10hcp
• 2NT – natural 10-12hcp, but with 4-card club support
• 3C – good club raise, 9-11hcp
• 3D – invitational with good diamonds, 9-11hcp, 6+cards
• 3H – invitational, 8-11hcp, 5+cards
• 3S – limit raise with 3-card spades and 4-card clubs
• 3NT – natural, 13-15hcp, includes 4-card club support
• 4H – natural, distributional

and after that, opener should know where to go.

Laid out like this, it all looks rather intimidating. But actually, it’s quite straightforward and pretty natural, just difficult to describe concisely.

Reverse Flannery Revisited

A couple of posts ago, I had a fairly incoherent ramble about jump-shift responses, and I said I thought Reverse Flannery sounded like a fair-to-good idea. Well, Agent 99 and I are now playing it, but in a very abbreviated form. Essentially, 1minor - 2H shows 5 spades and 4 hearts, 6-8hcp. 1minor - 2S shows the same but with 9-11hcp. Neither bid is forcing, and opener is more or less expected to set the contract at his second bid. The only forcing rebid from opener would be 2NT, but we haven't yet defined how responder would reply to that, which sounds a bit odd - how can we play like that? - but the truth is, a detailed description of distribution is not entirely necessary, and the 3-point strength ranges are tight enough that narrowing the range seems a bit superfluous. I guess we'll go with something like:

3C - 5=4=1=3 or 5=4=0=4

3D - 5=4=3=1 or 5=4=4=0

3H - 5=5-2-1 or 5=5-3-0

3S - 6=4-2-1 or 6=4-3-0

3NT - 5=4=2=2

That should be easy enough to remember. And if the 2NT bid implies a fit with one or both majors, plus game interest, that should be the most critical information needed for picking the best game.

Jump-shift responses

I'm at a bit of a down-spike at the moment. Real life is interfering with Agent 99's availability for playing bridge, and Elwood actually has a real job, so his availability is always limited. I don't care to fill my days with meaningless games as a house player or whatever, so the end result is I'm not playing so often myself. Bummer.

My usual pastime in this sort of situation is to trawl the internet for interesting bridge ideas. Sartaj Hans (The Imp Chimp) is looking for best practices to do with training, and related stuff. Having no great ideas myself, I'm waiting to see what everybody else comes up with, so I can steal it. Cathy Chua seems to be taking a few days off. So I'm reduced to either working on class notes to give to beginners, finding material for lectures to give near-beginners (which is way too much like work), or digging for the odd bidding convention that I might be able to persuade Agent 99 or Elwood to play.

An area that has some scope for Agent 99 and myself is the jump-shift response. We don't play 2/1 responses as game-forcing, so there is an argument for using strong jump-shift responses to identify certain strong hand types immediately and facilitate slam investigation. (I don't believe that the old-fashioned approach of using a jump-shift on any hand past a certain point-count is productive. If you are going to use them at all, restrict them to one of two sorts of hands: either a self-sufficient suit or a sort of fit-jump with good support for opener's suit. Which of the two it is gets clarified very quickly and you can proceed with the slam investigation without feeling the effects of the lost round of bidding.) But the truth is, neither of us is greatly enamored of strong jump-shifts, and so mostly we play pre-emptive.

After minor suit openings, we play inverted raises, where the simple raise is 10+hcp and the jump raise is 4-6hcp with 4+ (probably 5+) card support. The criss-cross raises (1D-3C and 1C-2D) are used for pre-emptive raises in the 7-9hcp range. This seems to be a slightly unusual use of the criss-cross raises - I made it up one day when I became dissatisfied with the wide (possibly) range of the pre-emptive raise. Since we play a weak no-trump, opener may be looking at a decent balanced hand (a strong no-trump opening), and he really would like to know if a pre-emptive raise is on the kamikaze edge or is maybe close to a forcing raise.

At the moment, all other jump-shift responses are simple pre-empts, showing 6+cards and 0-5hcp. (I don't believe in pre-empting partner unless my hand is really, really bad.)

An idea that might work is Reverse Flannery. After a minor suit opening, if responder has 5+cards in spades and 4+cards in hearts, he is not well placed for getting his distribution described accurately, especially when he is not too strong. Reverse Flannery uses the jump-shift responses of 2H and 2S to cover these hands, 2H being (say) 6-9hcp and 2S being 10-12 (or maybe just 10+) . 2NT can be used as a relay/distribution inquiry, similarly as in the original Flannery convention, if opener wants the details of what responder has. I have no experience with this, and I don't know anyone who plays it, but it sounds like a very reasonable idea, and a solution to a genuine (if not all that common) problem. Of course weak responding hands frequently cannot describe their distribution, but when they hold both majors, we should try and find a way for them to manage it. There are some details to be worked out before we try and play this. Maybe we should make the ranges 6-8 and 9-11, and make both bids non-forcing, on the grounds that if you're 12+hcp you'll manage anyway.

Adopting some version of that would take care of all the jump-shifts after minor-suit openings. There's still the question of what to do after major-suit openings. I can't believe simple pre-empts are such a great idea - I can't even remember ever bidding one, or hearing one from Agent 99.

At the moment, my plan is to talk her into using them as fit-jumps - essentially, limit raises with a decent side-suit. That's not too sexy, but I'm not interested in Bergen raises (or variations thereof)  because I don't believe in the Law (I'm an LTC fan, myself). Elwood persuaded me that mini-splinters weren't a good idea in practice (although they have some attraction in theory). And I haven't come across any other particularly great ideas yet. Maybe strong jump-shifts by unpassed hands and fit-jumps by passed hands? That might work.

Reverse Drury

With Elwood I play 2/1, and Drury is one of his favorite conventions.

Systems that use sound first and second seat openings (particularly 2/1) often use light opening bids in third and fourth seat. It is not uncommon these days to see third seat openings on 8 counts, but usually a light opening is in the 10-12 HCP range. The expert style is to always open in a real suit in third seat if the hand is light, so four-card major openings are common. In responding to these light openings, you need to have a way of showing a near opener (good 9 to bad 12 HCP hand) with support for opener’s suit without going past the two level. The Drury response of 2C fills this need. In the original version of Drury, a rebid of 2D showed a minimum opening, other rebids showed a full opener (12+ HCP). Using Reverse Drury, opener rebids his suit to show a light opening (he may have to rebid a four card suit) or rebids 2D with a full opener, but perhaps only four cards in the suit opened. After the 2D rebid the responder may bid up to three of opener’s suit with four card support, after the rebid of the major he is asked to pass except with an exceptionally good passed hand.

Responses by a passed hand

After pass – 1S; ?
• 1NT - semi-forcing, up to 11-12 HCP. Opener passes if light or 5332 shape with 12-13 HCP
• 2D/H - not forcing, decent 5 card suits, 9-11 HCP. Deny three spades
• 2S - normal single raise, not constructive, limited to a bad 9 count
• 2NT - 4 card spade support and 9-11 HCP with a singleton somewhere. Opener relays to discover the singleton
• 3C - shows 6 clubs and about 10-11 HCP. The hand type is a flawed initial preempt (two side cards or bad suit are possibilities).
• 3D/H - a flower bid (4 card support for spades, decent 5 card side suit, near opener).
• 3S - this is preemptive, something like SJxxxx Hx Dxx CQxxxx

• 2C - any hand of 10-12 HCP with three plus spades, could be a good 9 count with 4 trumps and a ruffing value (Reverse Drury)

Responses after a 1H opening are essentially the same, except that with 4+ spades and only 3 hearts you should generally prefer a 1S response to a Drury response.

All 2C responses are Drury, even after intervention. For example,
Pass – (pass) – 1S – (1NT); 2C is Drury

Drury sequences: Pass – 1S; 2C - ?

• 2D - full opener, may be a suit also, but presumed to be a flattish 12-14 hand. Responder continues:
     o 2H - natural, 5 card suit, usually only 3 spades, not forcing, suggests hearts as alternate trumps
     o 2S - usual rebid, 3 or four spades, not forcing
     o 2NT - natural, 11-12 HCP, only 3 spades
     o 3S - strong game invite, 4+ good spades
     o 3 bids - short suit game tries, singleton or small doubleton 

• 2H - natural, does not guarantee a full opener, could be a light 55 hand for example
• 2S - a light opener, less than 12 HCP
• 2NT - 15-17 HCP with 5332 shape
• 3C/D/H/S - slam tries, new suits are natural
• 3NT - balanced 18-20
• 4x - short suit slam tries
• 4S - a common rebid, no slam interest but enough extra to play game.

Lionel

I was somewhat upset when I realized that Multi-Landy is a Mid-Chart convention per the ACBL. And when I checked, they won’t let me and Agent 99 play it at the Manhattan (not in the afternoons, anyway – we might be able to get away with it in the evenings, because the usual evening crowd is more sophisticated). Apparently, the 2D overcall to show an unspecified major suit is more than the average little old lady should be able to defend herself against. I think that’s bullshit, but I think the blame lies with the ACBL, not the Manhattan. If the Multi 2D had been legal in the US for the last 40 years, neither it nor Multi-Landy would be raising any eyebrows anywhere in the US.

I want to use decent tools. If I have to follow ACBL strictures, fine, but we’re going with Lionel.

Double shows 11+hcp, a two- (or possibly three-) suited hand that includes 4 or 5 spades. The defined minimum is 4-4 in two suits, one of which is spades, but against a strong no-trump I think we should stick to at least 5-4 (although spades might be the 4-carder) or extra high cards. Against a weak no-trump, you can be as adventurous as you like. If responder passes, advancer with 11+hcp can (and probably should) pass also, converting the double for penalties. (This implies that overcaller shouldn’t shade his point-count, only the distribution.) If that doesn’t look attractive but he likes spades, advancer can bid spades at what seems like the best level, or can bid a new suit at the 3-level as a splinter raise of spades (game-try). If he doesn’t like spades, with a weakish hand he should bid his cheapest 4-card suit and we’ll try to find somewhere to play at a low level. With a really good hand, he can bid 2NT, suggesting “game somewhere”, and overcaller should show his second suit.

Overcalling 2C shows clubs and hearts, overcalling 2D shows diamonds and hearts, as mentioned above promising only 4+4+ cards but usually better. Neither overcall is forcing. A 2NT overcall is the Unusual No Trump, showing both minors 5+5+ cards. And that takes care of all the two-suited hands. All of those bids, including the double, need to be alerted, but they are all legal per the ACBL General Convention Chart.

And all the other overcalls are natural. 2H and 2S could be 5 cards against a weak no-trump, but will usually show a 6+card suit. Three-level overcalls would be natural and 6+cards.

The method doesn’t change with the strength of the opening 1NT. But the weaker 1NT is, the more aggressive the overcalls can be. I have the feeling that this could be much more destructive than Multi-Landy, but that’s the way the ACBL wants to play.

Splinters

I think one of my lectures sometime soon is going to be about splinter bids. There are a number of the low-ranked players who show up for one or more of the 299er games who don’t seem to know anything about them (not especially surprising, since I’m sure splinters don’t feature in the beginner or intermediate classes). But some splinters should be simple enough for them to manage without too many disasters. I’m thinking those from paragraph 1 below, and maybe paragraph 2.

Splinter bids are bids that (by agreement) show shortness. There are a large number of possible bidding sequences where a splinter can usefully be defined. Some are commonly known and used, others are less common and typically only used by regular partnerships. In general, it pays to take a little time to define a fairly wide variety of splinters, and use them as often as possible, for a couple of reasons. First, the overhead cost is low: the bids defined to be splinters typically have no other particularly good meaning, and the memory-burden is generally small (for that reason). Second, the payoff can be quite large. Knowledge of a singleton or void in partner’s hand allows a much more focused assessment of your values, which can lead to games and slams on relatively low point-count when assets are working, and also to good stops when combined high point-counts are not fitting so well.

All splinters are alertable, whether or not most people play them.

1) The splinter bids that everyone knows are 4-level responses to a major-suit opening. For example, 1H – 4D! shows a game-raise of hearts that includes a singleton diamond. To give this better definition, we should note that this sort of raise is laying some groundwork for slam bidding, so the raise to game needs to be full-strength: 10+hcp, no more than 7 losers on the LTC, and including 4-card trump support. That doesn’t make it excessively strong, and with many hands, opener will merely return to game and that will be the end of the auction. But even with a near-minimum, opener may want to investigate if it looks like the hands fit well. The Losing Trick Count is a better guide, but a rule for high card points is that slam will usually be a good bet if you can count 27 points in the combined hands counting only the ace in the splinter suit, or with 24 points and no high cards in the splinter suit. (This rule guarantees that you will not be off two cashable aces or the ace and king of a suit.) The one sequence I always mention to a new partner is 1S – 4H. I like to play that as a splinter raise of spades, some people like to play it as natural. I can live with it either way (for one session), but I really hate playing 4H in a 3-1 “fit”.

2) A similar sort of splinter happens the other way round, in sequences that begin 1minor – 1major. For example, in the sequence
1C – 1H; 4D!
opener is showing a game-raise of hearts that includes a singleton diamond. Typically, opener’s hand will not only be in the upper range (16+hcp) but will also include length in his first-bid suit. Responder should be aware of the possibility of a double-fit, if he has some support. A special case is sometimes called an auto-splinter: this where opener “splinters” in his own suit, for example
1C – 1H; 4C!
This sort of sequence should show a solid 6-card suit, 4-card support for responder, and an unspecified singleton in one of the other two suits.

3) Some bids of a new suit at the 3-level can also be defined as splinters, similar to those just mentioned but forcing only to 3 of responder’s major. For example,
1C – 1H; 3D!
doesn’t have an obvious natural meaning, since 2D would be a reverse and so opener doesn’t need to jump to show strength. I recommend that this and similar sequences be played as splinters supporting the major. Some attention has to be paid, since some sequences are needed for natural bidding, for example 1D – 1H; 3C looks similar, superficially, but opener is just showing clubs and a good hand. I don’t think there is much real risk of a misunderstanding, because the general rule is that if a bid might be natural, it is. It’s only the “jump reverses” and other such oddities that become splinters.

4) Any unusual jumps (often a jump in a sequence that is already forcing) should probably be treated as a splinter. For instance, playing inverted minors,
1C – 2C; 3H!
should be a splinter where opener is 16+ and is suggesting game or slam in clubs, and alerting responder to only try 3NT if he has the hearts well-covered. 
1NT – 2D; 2H – 4C!
should imply a 6-card heart suit and be a mild slam try, since just bidding 3C would show a club suit and be forcing. This rule of thumb can be applied in competitive auctions, also, although I’m sure some partnership discussion is needed. But whether you are the opening side or the overcalling side, there are a lot of sequences where a jump or double-jump in a new suit makes no sense normally and can be used as a splinter raise of partner.

5) Splinters typically show singletons. They can also show voids, which can cause problems. A solution for the 1major – 4x type of splinter (paragraph 1) is to set aside the cheapest triple-jump bid as a void splinter, showing a void in an unspecified suit. The 3NT response (currently unused) comes into play for this. So
1H – 3S and 1S – 3NT
are the void splinters. Opener bids the next step as a relay for responder to show where the void is (bidding 4 of the trump suit to show the suit that isn’t available to bid below game level):
1H – 3S; 3NT – 4C/D/H void in clubs, diamonds or spades respectively
1S – 3NT; 4C – 4D/H/S void in diamonds, hearts or clubs respectively
The only singleton splinter that is compromised by this scheme is 1H – 3S. That is fixed by using 1H – 3NT as (specifically) a splinter showing a singleton spade.

This idea has a little bit more memory-burden than basic splinters, but it does clear up an (admittedly rare) issue using only otherwise-unused bids.

I play all of this stuff with Elwood. Agent 99 wouldn't go for the void splinters, but Elwood has a mischievous streak.

Crawling Stayman

Here’s a hand that came up in a pairs session with Elwood.

N-S Vulnerable, dealer West
♠ K Q J 4
♥ J 10
♦ A K J 9
♣ K 10 5

  []

♠ 10 8 6 5
♥ 9 7 4 2
♦ 4 3
♣ 8 6 3

W    N    E    S
1♣   1NT  Pass Pass
Pass

There was no way for Elwood to make his contract, despite his high point count. -100 didn’t get very many matchpoints at all, though, because several pairs playing our way made impossible contracts, and several others registered small pluses as E-W “won” the contract but found they couldn’t make much either.

Elwood wasn’t happy, and I was very embarrassed. With my other partners, I play the usual Garbage Stayman. On this hand, I considered the bid, but rejected it, on the grounds that if the answer from North was 2♦, I wouldn’t have anywhere to go. But at Elwood’s insistence, many months ago, I had agreed to play a variation he calls Crawling Stayman. In this version, a bid of 2♥ over the 2♦ reply doesn’t promise more than a 4-card suit. Opener (or overcaller, in this case) is expected to correct from 2♥ to 2♠ in the event that he holds a doubleton heart. Of course, this isn’t something that comes up very often, but there’s very little downside, and it just so happens to work very well on this particular hand. The crawling mechanism doesn’t come into play, but because it is available, South can bid 2♣. The answer of 2♠ finds the one contract that N-S can make against any defence.

And I just forgot all about it. Completely. Never even entered my mind.

I hate it when that happens.

I think, in general, I’m pretty good about remembering the different systems I play with my various partners. I have two regular partners, and two fairly regular but somewhat less frequent ones (including Elwood). And then again I play occasional games with all sorts of people as I fill my roles as TD/house player. No two of them play exactly the same stuff, except maybe the 299ers when I have to fill in the movement in their game. But my recent experiences with Elwood are ruining my self-image, and maybe my reputation. He is one of the most intense and thoughtful players in the club (the biggest fault he ever gets accused of is over-thinking things, which isn’t a complaint you hear too often at the club level). I love that in a partner, and I’m more than willing to listen to what he wants in our bidding system. I know that anything he brings up will have been considered, usually in great detail, and often in some directions and situations that wouldn’t have occurred to me. I do get my ideas in, of course, and veto some of his, but the end result is that our bidding system is probably 60%-40% his ideas versus mine. I’m sure it’s very good, but the result of this mixture is that there’s a memory burden on me that I haven’t successfully addressed. With him, I’ve agreed to play stuff that I don’t play with anybody else – and I’m failing in my obligation to remember it all.

So this is coming at an awkward time, and it’s the wrong time of year anyway, but I’m making a resolution. I resolve to document our frigging system, and read the system summary before every session. I think I remember stuff better after I’ve written it down in my own words. And if that doesn’t work, I’ll blame Mad Cow Disease and English roast beef.

Bidding over opponent’s 1NT response

When the opponents start the auction 1D – 1NT, in Standard American auctions it is pretty much a given that the 1NT bidder will have a club suit. This observation leads to a couple of neat overcalling possibilities.

First, consider how you are placed when in the “sandwich” position, after (1D) – pass – (1NT) - ?
The BobbyBridge page about Multi-Landy observes that it works here. Since we won’t want to make a simple minor-suit overcall in this position, we can play that 2C shows both majors and 2D shows a weakish overcall in an unspecified major. That lets us specify that the 2H and 2S overcalls are full-value – opening bid strength – so we can intervene relatively cleanly in either (or both) major.

Second, consider where you stand in the re-opening position, after
(1D) – pass – (1NT) – pass; (pass) - ?
On Eddie Kantar’s website, he notes that a re-opening 2D here should be natural, because you may have been silenced by the opening bid (which might only be a 3-card suit). However, it is still true that you won’t want to reopen in clubs with a known suit behind you. So we can (again) use 2C as showing both majors, in this case not strong and possibly only 4-4.

Third, Kantar also remarks on
(1C) – pass – (1NT) – pass; (pass) - ?
He notes that a re-opening 2C here should be natural, because you may have been silenced by the opening bid (which might only be a 3-card suit). However, it is also true that a natural re-opening 2D bid is very unlikely, since any such hand would almost certainly have overcalled 1D in the first place. So we can use 2D, this time, as showing both majors, again in this case not strong and possibly only 4-4.

These may seem like little things to clutter-up your memory, possibly causing more trouble than they are worth. That remains to be seen. Being able to fight for part-scores at the two-level is important at any form of scoring. Allowing the opponents to settle in 1NT too easily is faulty strategy.

Opening 2C And Responses

I have described in an older post what Agent 99 and I play. I wanted to put out another post, because there are a couple of good alternatives floating around that I would be interested in trying, and in hearing how other people get on with them.

An idea I used to like is an artificial positive set-up, perhaps similar to:
• 2D: any 0-3, or 4-7 with a 5-card or longer major
• 2H: any 8+ unbalanced or 10+ balanced
• 2S: 4-6 balanced, or 4-7 with a 5-card or longer minor
• 2NT: 7-9 balanced
The reason why I consider this sort of scheme playable is that while it seems harmful or perhaps just not useful to condense all positive replies into 2H, I have looked at hundreds of deals where a 2C opening is faced by a positive response. In virtually every case, the final contract should be a slam. So it actually makes sense to use only one bid for those hands, and use two or three responses to categorize the "semi-positive" hands.

A similar sort of philosophy gets a very different implementation as described by Martin Johnson at
http://www.freewebs.com/bobbybridge/
(some interesting stuff there). He calls it "Two, Weak, Transfer".
• 2D: A positive hand with at least an ace or 6 working points, unsuitable for any of the special higher responses. All hands of 10+ HCP and most hands with 1.5 quick tricks use this response.
• 2H: A negative response, the auction may stop at 2NT, 3 of a major or 4 of a minor
• 2S: A medium strength TWO-suited hand. Minimum strength is any hand that does not intend to stop below game, maximum of two kings or 1 quick trick.
• 2NT: A WEAK single suited hand with no side entry, and a decent suit. The bid denies two top honors, but any seven card suit qualifies as "decent".
• 3C, 3D, 3H: A good single suited hand minimum suit KQJxxx or AQ10xxx with no side king or ace, suit shown is the next higher (TRANSFER response).
Once again, there is a positive response and a negative response. But this time, the other bids are used to describe particular types of responder hands that are difficult to get across otherwise. I like this idea, because while I appreciate that the ideal auction has the opener as captain (because his hand is so strong), I also feel that standard methods provide too few mechanisms for responder to describe whatever contribution he might make. Discovering point count or controls can be done later: identifying shape, or that there is a suit without an outside entry, is both more difficult and more useful. That makes this set-up far better than, say, step responses.

For something completely different, you can go to Dr. Chris Ryall’s website
http://chrisryall.net/bridge/two/clubs.htm#responses
for his description of Paradox responses.
• 2D: Can supply a trick in support of either hearts or spades.
• 2H: Cannot supply a trick for hearts, may or may not have a trick for spades.
• 2S: Cannot supply a trick for spades, but promises a trick for hearts.
Both the 2H and 2S responses are a form of negative, bidding the suit that responder doesn’t like (hence, Paradox). They are therefore not forcing, if opener doesn’t have game in his own hand. An auction like 2C – 2H; Pass would probably make the opponents blink, but it’s perfectly logical. Suppose you open 2C on a hand that is single-suited with hearts, and you estimate 8 or 9 tricks in hand. Partner responds 2H, saying he has no help. Obviously, the thing to do is pass. Of course, that means the requirements for the 2H and 2S responses must be pretty tight – opposite a typical 2C opening, you really don’t need much to contribute a trick. But the general idea is interesting, and may be worth exploring.

Of the three methods, Martin Johnson’s “Two, Weak, Transfer” system has me the most excited at the moment. I’m going to run up some random 2C openings and try to compare possible sequences.

Run-outs from 1NT Doubled

The only significant problem with opening a weak no-trump is the danger of getting caught for a penalty. Going for a large number really doesn’t happen very often at all, but if you are vulnerable, getting hit for 200 or 500 on what is essentially a part-score deal is a real possibility. But we don’t have to sit still for 1NT doubled if a suit fit looks to be better.

I am not suggesting that running is always the best policy. If responder has 6-8hcp in a more-or-less balanced hand, the points are split roughly equally between the two sides. The best choice will usually be to stand your ground and trust partner to bring home 1NT. If you are not vulnerable, even when very weak, responder may do best just to sit: if declarer can escape for one or two down, you may find that they have a good part-score or even game the other way.

But then there are the times when we need to escape. There are numerous methods for escaping from 1NT doubled. For better or worse, we have decided to go with Moscow run-outs, which are among the simplest methods. The advantages that I see are that the run-outs are easy to remember, and our constructive bidding doesn’t get damaged too much when the opponents are being silly. The biggest disadvantage is that we often lose the ability to play in 1NT doubled: we can play redoubled, but not doubled. That raises the stakes to an uncomfortable level when assets appear to be roughly equally divided. But to be fair, that’s a blade that cuts both ways. (For a much bigger discussion of the philosophy of competing over 1NT and different methods for competing and escaping, see Chris Ryall’s website:

http://chrisryall.net/bridge/1nt-complex.htm

I’m not saying I agree with everything he says, but I think I agree with most of it).

There are two basic situations, 1NT – Double and 1NT – Pass – Pass - Double.

1NT – Double

In this situation, we want to play “systems on”, so that our constructive bidding isn’t disrupted too much when the opponents are being silly. The escape bids are Pass, Redouble, and 2NT, all alertable.
Redouble requires opener to bid 2C. Responder can Pass with clubs, or bid 2D with diamonds.
2NT over the double shows both minors (pre-emptive), 5-5 or better.
Pass requires opener to redouble. Responder can then pass (to play in 1NT redoubled) or bid his lowest suit holding any two suits 4-4 or better. Opener passes with 3-card support or better, or bids the next suit up if he has a doubleton in responder’s first suit. Responder then passes or corrects to his second suit. We should therefore finish up in at worst a 4-3 fit at the two level. The opponents may find that harder to double.

1NT – Pass – Pass – Double – Pass – Pass

Note that with a very weak hand, after 1NT – Pass, as responder you should often try and bid something. The best hand at the table is on your left, and if you pass it will most likely double, announcing the situation to the world. If you can transfer to a suit or use garbage Stayman, you start to confuse the issue. A weak minor two-suiter can be handled by bidding 2S and either passing 3C from opener, or if he bids 2NT, correcting to 3D. That all being said, sometimes you pass, and a double comes back around to you.

In this situation, presumably responder either has enough to pass (which is how we can occasionally still play 1NT doubled), or has a one-suit run-out to a minor via redouble (same as above) or we’re in trouble. In the latter case, if responder has two suits at least 4-4, he bids the lower one. Opener passes with 3-card support or better, or bids the next suit up if he has a doubleton in responder’s first suit. Responder then passes or corrects to his second suit. As after the pass and compulsory redouble, we should therefore finish up in at worst a 4-3 fit at the two level. Since this means we can move to a suit contract, after 1NT – Pass, a responder looking at any weak hand with a mild two-suiter can elect to pass first and run later, if staying in 1NT looks plausible unless they double.

Two-suited Overcalls

When you have a genuinely two-suited hand, 5-5 or better, it is worth getting into the auction even when you are not strong. You have two chances of finding a fit with partner, and if you do catch one, the hand usually plays well. Of course, a problem is that if you finish up defending, declarer pretty much has a blueprint of how to play the hand. But that’s a trade-off that many people are willing to make.

There are various methods for showing a two-suiter with one bid. With all these overcalls, the high-card strength should be concentrated in the suits. Generally, the calls are most often used as pre-emptive weapons, and you don’t want to mislead partner into making a sacrifice that turns out to be a phantom because you have an unexpected ace or two showing up in your short suits. A common idea is to play the bids as weak or strong. With a weak hand, you make the overcall and then keep quiet. With a strong hand, you make the overcall and then bid again. If you have a middling sort of hand, you don’t use the two-suited overcall, just bid one of your suits and hope to get the other one mentioned later. Exactly what constitutes weak, middling and strong is a subject for debate. A common sort of arrangement would be 7-11hcp = weak, 16+hcp = strong.

The Unusual No-Trump is a 2NT overcall to show the minors. This is usually generalized to say it shows the two lowest unbid suits, so that after a 1D opening, for example, you can overcall 2NT to show clubs and hearts.

This seems easy, but the mistake a lot of people make is to overlook that 2NT is occasionally needed as a natural overcall. For starters, 2NT over an opponents weak two opening is similar to a 1NT overcall of a one-level opening. And even over a one-bid, in 4th seat 2NT is generally natural, for example:
(1H) – Pass – (Pass) - 2NT or
 (1H) – Pass – (2H) – 2NT
In both these cases, the 2NT overcall is natural, something similar to a 2NT opening bid, around 20 points with hearts stopped. It is also worth mentioning 1C openings. If the opponents play a “short club”, meaning it may only be a doubleton, you may want to discount that as being a bid suit and have your Unusual NT show both minors – at least it’s something to discuss. The same issue arises if they play a strong club system and a 1D opening may not have many diamonds.

The other really common two-suited overcall is the Michaels cue-bid. (1minor) – 2minor cue-bid shows both majors. (1major) – 2major cue-bid shows the other major plus an unspecified minor. Advancer can bid 2NT to ask for the minor. Some people play that 2NT is a constructive way to ask, while bidding 3C is simply “pass or correct” with no interest in going higher.

The advantage of Michaels is that it only uses the cue-bid, and covers a couple of the most important situations. The weakness is that your partner doesn’t always know both suits immediately (which may limit his ability to pre-empt), and that even with both UNT and Michaels in play, there are still some two-suiters not covered. For example, over a 1D opening, neither bid handles a spade-club two-suiter. If you want to be able to cover all combinations, you have to use a third bid, and the one usually sacrificed is the 3C jump overcall. For example, the method known as Ghestem (after its inventor, Pierre Ghestem) uses (in the original version)
2NT = the two lower unbid suits
3C = the two higher unbid suits
qbid = the highest plus the lowest
and of course there are variations on this.

One other set of overcalls worth mentioning is Leaping Michaels. Over an opponent’s weak two opening, jumps to 4 of a minor show two-suiters. Over 2H or 2S, a jump overcall of 4C or 4D shows the bid minor plus the other major, strong but not forcing. Over a weak or natural 2C or 2D (perhaps a Precision 2C that shows 11-15hcp with 5 or 6 clubs), a jump cue-bid shows both majors. A jump to 4 of the other minor shows the bid suit plus an unspecified major.

This hardly ever comes up, but it’s worth having in the back of your mind just in case. It uses bids that aren’t likely to be needed for anything else, and if you do happen to get the right hand for it, it could be a life-saver.

Gambling 3NT

The opening bid of 3NT is still played as showing a monster balanced hand by some die-hards. But the majority of people, including almost all experts, seek a better use for the bid. A very large balanced hand can be adequately handled through a 2C opening (or whatever is your conventional strong opening bid). And starting to try and bid constructively at 3NT is fairly hopeless, making the very old-fashioned opening ineffective as well as unnecessary. The commonest alternative is what is called the Gambling 3NT.

The Gambling 3NT was originally part of the Acol system. The opening showed a solid or semi-solid 7- or 8-card minor suit, with one or two stoppers outside. With some help, responder could pass, hoping that declarer could gain the lead and run off nine tricks before the defence could organize five. This might depend on a favorable opening lead, or an ace being on-side, or some such – hence the “gambling” designation.

The modern incarnation of this bid is modified from that original description. These days, a seven card suit is enough, but it is expected to be pretty solid (minimum AKQxxxx). Rather than having outside strength, the opener is expected to have little outside strength, usually defined to be no ace or king outside. Responder still has the option to gamble with a pass, or bid 4C or 5C for pass-or-correct to opener’s suit. There is also a 4D response to ask for a singleton (with no aces or kings, a singleton is the most useful feature opener might have outside his suit). Opener rebids:
After 3NT – 4D ;
4 major = singleton in the bid suit
4NT = no singleton
5 minor = singleton in the other minor
(The reason for reversing the minor suit calls is that it wouldn’t necessarily be a good idea for opener to go past 5C holding a singleton diamond – what if responder was really looking for a heart control?)

This incarnation of the bid is clearly more frequent in occurrence (though still pretty rare), and also fits in to a spectrum of pre-emptive bids that runs from very weak (3-minor pre-empts are often made on 6-card suits these days) to very strong (4- and 5-level pre-empts are usually 8- or 9-card suits). But one obvious weakness of the modern version is that the contract is wrong-sided. Since the responder is the one with the stoppers, it would be much better to have the opening lead running around to his hand. That is why the usual advice for the opening leader is to lay down an ace, so that you get a look at the dummy and can (hopefully) spot the flaw that will give you your defensive tricks. This also is why transfer pre-empts or Alder pre-empts or some such scheme are attractive for this case – if the opening is 3S , responder gets to be the declarer when he wants to gamble, which is much better.

With virtually all of opener’s assets concentrated in one long suit, there is scope for the opposition to have decent prospects of game in one of the other suits. But finding a fit when you have to start at the 4-level is bound to be somewhat problematic. The best idea looks to be to use a defence against 1NT, just a few levels higher. We would go with Multi-Landy:
4C = both majors
4D = one major, unspecified
4H = hearts and a minor (hopefully, the one opener doesn’t have)
4S = spades and a minor
Double can be used to show general values, but clearly you aren’t going to get fat on penalties most of the time. Opener will probably be able to round up at least eight or nine tricks, even without too much in the dummy. So probably it would be better to use an immediate double to show the other minor.