Dark Widow

Isn't it a shame that someone did not come up with something similar to this at any point. Perhaps adding a modern twist (Control points) to a well-established method for handling such hands might be worth investigating.

It is pretty important in a relay system to have standard rules to apply and not to have special rules for every other auction. It does not matter so much what those rules are, only that they are consistent and apply (nearly) all of the time. 1. Have a rule with a limit as to how far above "standard" you are willing to keep relays on. If you have extra space then you can usually get back to standard by including the range ask at the beginning instead of at the end. If you have extra space and a limited hand describing then you probably need to assign the hand types more efficiently. 2. No problem, just omit the range ask at the end of the shape relays. Several systems do this anyway for passed hands. Make sure you agree whether zoom is still on in this case though. 3. If you have included the pre-ask for range then you can include these awkward hand types into the middle step. If not then you need to have a rule for fixing this. I think the simplest solution is to include this with your 5-5 hands and then dedicate the first step to 5440/5404. This might depend on how you have constructed your relays though. 4. If you use a symmetric-style scheme then once you have shown 2 suits naturally you can use the first step for 5+ in the second suit and higher bids for 4 cards in the second suit. If 5+ is not possible then you would just skip this step. 5. Since you specifically mention 1NT rebids, one variable which can be very helpful is to use an impossible suit as a transfer to a possible suit. In this way you can often use a standard NT structure with minimal changes. For example, if Responder has denied both majors and Opener rebids 2NT as a ganeral balanced hand with whatever range then you could play 3♣ to ask for a 5 card major or for 4-4 minor fits, with 3♦ as a transfer to clubs and 3♥ as a transfer to diamonds. That kind of approach moves away from pure relay but might actually fit better with what you are trying to accomplish.

I have played all 3 methods in a 4 card major context. As a beginner I learned not to include 5M332 ever. Later I started including some 5M332 hands and now I pretty much open any 5332 hand in range with 1NT. I have also sometimes played a 5 card major system where there is absolutely no rebid whatsoever after a 1M opening with a 5332 hand so these have to be opened 1NT if in range. In this style 2M is simply a bucket response denying the ability to rebid anything more helpful. If partner responds 1♠ then you do have the option to raise to 2♠ if you have 3 of them and it is a part of your agreements. There are 2 ways of handling 5-3 fits - either you include some sort of checkback for 5 card majors (usually based on Puppet Stayman) or you simply ignore them. Again I have played both ways with a preference for including Puppet. Even without I think it is advantageous to the rest of your system to open these hands 1NT most of the time though. If you do include Puppet there are 2 mainstream methods, either using a 3♣ response for this or using the traditional 2♣ response. The latter means giving up on Crawling Stayman while the former finds fewer fits (because it requires a GF hand). Opening 1NT tends to bury fits for both sides much more than opening 1 of a suit. This is true for any range. Usually opening a weak NT leaves you better placed than the opponents in the ensuing part-score battle but certainly not always. One of the qualities necessary for playing the style is to be able to live with the odd bad, or even very bad, result. If you cannot then it is (probably) better to play a 14-16 range and get enhanced constructive bidding in other auctions. If you look at the first answer you will see I mentioned the possibility of rebidding 2♠ with a 3532 or 3523 hand after 1♥ - 1♠. Of course you can also rebid 3♥ after 1♠ - 2♥ with 5332 or 5323. When I played this style I opened 1NT with 2533 and 5233 and 1M otherwise. I think this is a natural progression for a beginner to try out before committing to the "always" approach.

Perhaps you can cater to problem #1 by inverting pass and redouble. Then pass alerts Opener that they can double any advance or they redouble if 3rd seat passes. As others have said the disadvantage is not being able to play 1NTX as well as giving 3rd seat 2 ways of running. Most pairs include other hand types into this forcing pass too but there is no need to do so when you want a call available that unequivocally shows strength. Bidding 4-4s with a very weak hand is not a problem since you are very likely to find some kind of fit opposite unless you open alot of offshape 1NTs. Opener can still show a 5+ card suit by bidding something other than the pass or correct response too. Remember though that Opener has no guarantee of finding even a doubleton opposite. One last thing - if you do play a scheme like this then you need to think a little bit about how you want to deal with weak 4333 hands. The main options are to treat the hand as 1-suited, to upgrade a 3 card suit to 4, or to use one or more artificial redoubles. The last of these is arguably the best option but the second is probably more common.

Is there something to be said for giving up the natural 3♠ call over a 2♥ opening in order to have multiple levels for both spade 1-suiters and 2-suiters without giving up the stopper ask? So 3♥ stopper ask, 3♠ spades and a minor (limited), 4♣/4♦ spades and the minor (GF), 4♥ spades (strong), 4NT clubs and diamonds (GF).

You could also use (1♦) - 2♦ - (P) - 2NT as an asking relay, then (for example) 3♣ = min with equal or better hearts, 3♦ = min with better spades, 3♥ = max with equal or better hearts, 3♠ = max with better spades, 3NT = strong variant with no void, 4m = strong variant with void in m. Another option would be to use this 2NT response as a puppet to 3♣, either to play there or bid on, say 3♦ asking for better major and 3M setting trumps to ask for cue bids. 3♣ is now free for the good heart raise.

In most versions of Puppet this 3♥ rebid denies 4 hearts so there is no need to make a slam try in hearts now. A simple solution is to play that 4♣ over 3♦ is a slam try with both majors. With a hand that wants to make a slam try in clubs you start with 3♥ and then bid 4♣ next time around. You lose this as a cue bid for spades of course. Over 4♣ there are several possibilities, to name but one example: 4M as natural declining the slam try, 4♦ as accepting the slam try and either agreeing spades or with a double fit, and 4NT (and up) as key card responses accepting the slam try and agreeing hearts. Over the 4♦, 4♥ asks and then 4♠ agrees spades and 4NT (and up) are 6 key card responses with the double fit. If you play a version of Puppet where 3♦ does not promise a 4 card major then you obviously need follow-ups to cater to this. In that case it is probably best to use 4♦ as a general no fit flag, something like 4♦ = no 4 card major, 4M = fit + decline slam try, 4NT = spades and accept slam try, 5♣ (and up) = key cards with hearts and accepting the slam try. To make slam tries in the 4-4 fit when Responder does not hold both majors is not a problem at all in Puppet. For spades: 2NT - 3♣ - 3♦ - 3♥ - 3♠ - 4♥ is surely unambiguously a slam try, while for hearts Opener can preemptively signal readiness over 3♠ (= hearts) by bidding 4♣, while a simple 4♥ would be a decline of the (imaginary) slam try.

If I were playing this (and I sometimes do) then I would like to agree that the direct 2♥ response showed a max of around 7 and the delayed 2♥ can then be used for about 8-10. If that agreement were in place then North can afford to be more aggressive here. Perhaps something like 1♦ - 1♥ - 2♣ - 2♥ - 2♠ - 3♣ - 3NT - 4♥ would now be possible.

How does the popular 4.5/3/1.5/1 with 5/3/1 method fare? What about if a queen counts 1.5 when unsupported but 2 with a higher honour? Are you making any adjustments for honours in short suits, especially singletons? Are the correlations better if you subtract somewhere between a half point and a whole point for such singleton honours?