The Mulberry convention is designed to distinguish sign-offs, slam tries and Keycard auctions where space is limited and many possible trump suits are in play. The classic example is the following auction:
Responder may wish to play in any of the four suits and, depending on other responses to the 2♦ opening, may be wishing to invite game, bid game or make slam tries of various strengths. The auction does not permit all of that at comfortable levels using traditional methods.
The Mulberry convention solves the problem by setting aside bids above the level of 3NT for particular purposes. In particular:
I have been playing this convention for a little while, and it has a few features which I do not particularly like.
Firstly, as noted, the mild slam try bids are easy to forget. To some extent, that is a price of many more complex methods, so it was never a deal-breaker for me. Of course, these ones usually look like they are sign-offs in game, making memory lapses more likely.
Secondly, I have never been particularly fond of Keycard. My preferred method of slam investigation is cue-bidding and Turbo. Mulberry is not really conducive to that approach.
Thirdly, it requires one hand (the responder in my example auction) to make a decision about whether the hand is a sign-off, a mild slam try or a slam drive (presuming the requisite number of Keycards are held) at a time when often opener's shape is the only information that responder possesses.
It occurred to me that a better way of solving Responder's problems in auctions of this nature would be to identify the suit first and ascertain slam interest (if any) later. In order to permit that information to be conveyed, at least "two under" transfers are needed. Thus Mulleyberry was born.
In the example auction given (repeated here for convenience):
Responder gets to then sign off in game or proceed towards slam if (s)he wants to via whatever method is favoured. For me, that would be cue-bids with Turbo. For others, that might be Kickback.
It is possible to remove "impossible" suits from this structure.For example:
Spades is not a possible trump suit on this auction, so 4♦ can show clubs and 4♥ can show diamonds to give additional cuebidding space. 4♣ remains as showing hearts.
This particular auction also illustrates that this method might expand the possibilities for Mulberry-type auctions. In my methods, both opener and responder are unlimited on this auction. This method enables opener to show useful extras for slam along the way once responder has indicated that (s)he knows in which denomination the contract will be played.
Furthermore, it means that the available bids below 3NT can be used to unequivocally probe for strain. In the auction above, opener can bid 3♥ to promise a doubleton and use 3♠ (4th suit forcing) to deny holding that many hearts. With 3-card heart support, responder would bid 4♣, as (s)he knows that hearts is the correct strain.
I welcome any feedback on any of this - it is late night here and this came in a moment of inspiration. I have learned that such inspiration is often flawed. I would also welcome naming suggestions - the title of this article was very much tongue in cheek. In fact, I would not be surprised to hear that such thoughts had already occurred to others long ago.
Plus... it's free!