Help
What I guess ruins the game for most who have tried it, and prevents it from ever becoming a serious game, is that it's virtually impossible to determine whether a player is (consciously or subconsciously) acting on the UI he (always) has.
So here's an idea I have for a form of UI free two-player bridge:
Auction phase:
Suppose the two players, call them A and B, control the NS and EW hands, respectively, and suppose "North" (or A, controlling the North hand) is "dealer" (although the dealing can be assumed to be done automatically, as on BBO).
Then the auction phase will cycle through the following events,
The cards are redealt face down consistent with the vulnerability** and auction so far.
A, looking only at the North hand, makes a call and explains*** it.
B, looking only at the East hand, makes a call and explains it.
The cards are redealt face down consistent with the vulnerability and auction so far.
A, looking only at the (new) South hand, makes a call and explains it.
B, looking only at the (new) West hand, makes a call and explains it.
until the auction ends the way it does in four-player bridge, i.e. with a call being followed by three consecutive passes.
Play phase:
Suppose "West" (or B, controlling the West hand) becomes "declarer". (The other cases where the hand isn't passed out are similar.) Then the following will take place:
The cards are redealt face down consistent with the vulnerability and auction so far.
A, looking only at the (new) North hand, plays a card from the North hand. (This corresponds to the (blind) opening lead in regular bridge.)
The North and South hands are made permanently visible to A and permanently invisible to B.
The East and West hands are made permanently visible to B and permanently invisible to A. (The East hand would be visible to all (i.e. be Dummy) in regular bridge.)
The play then proceeds as in regular bridge, except with A and B controlling two hands each.
Some observations:
Playing the game with a real deck of cards would be very cumbersome, so let's assume it is played in an electronic environment (on bizarro-BBO, perhaps).
Every bidding or opening lead problem has an exact counterpart in regular (four-player) bridge.
Unlike in regular bridge, there is no dummy visible to all (both) players. But if we imagine the two players occupying what corresponds to declarer's and opening leader's seats in regular bridge, then the opposite hand from each, which is permanent only after the opening lead has been made, may be thought of as a pseudo-dummy visible only to that player (the "declarer" or "defender").
"Declaring" will be just like declaring in regular bridge except that deception will sometimes work differently. For example, leading a low card early from "hand" towards KJ in "dummy" may not have the same effect on "LHO" who now sees "partner's hand" (his own pseudo-dummy) instead of "dummy" ("declarer's" pseudo-dummy).
"Defending" will be just like "declaring" except on the first move (the opening lead) where only one hand is visible to the "defender".
The duplicate format is not very suitable for this game, because even if game starts the same way at two tables, the players will soon almost certainly be bidding and playing very different deals. Instead the table result could be compared against par and each player e.g. awarded 0,1 or 1/2 points according to whether they lost, won or tied the (two-player, one-board) match, like in chess (or BAM).
Since this is a two-player game in every sense of the word, it might be easier to implement an ELO-type rating system for individual players than in regular bridge.
It seems hard to cheat at this game, at least by collusion or self-kibitzing. So maybe this is what the world of bridge needs right now?
* Here I mean 'player' in the common sense of the word, not in the game-theoretic sense where arguably each partnership is a player (and regular bridge already a two-player game).
** The meaning of a call may depend on the vulnerability.
*** The explanation might be translatable to code that can be be used as input for the (re)dealing software.
