[kugw]

How to go about designing a bridge system for teams scoring.

[Cyberyeti]

1. decide what sort of system you want (approach forcing, big club/diamond/NT etc)
2. decide on the balance between ease of memory and accuracy
3. have a look at what people have done before, you don't have to do most of this starting from nothing

I've only designed one from scratch and heavily modified one which I still play.

The one from scratch was weird and for a specific purpose (to be very playable and fun, but abuse the permitted system regs of the time as far as possible), but elements were borrowed from several other systems.

[hrothgar]

kugw, on 2018-December-30, 20:25, said:

How to go about designing a bridge system for teams scoring.

1. Borrow time on alpha go or some other deep learning system.

2. Explain bridge scoring to it as well as bidding rules

3. Show it video of some bridge matches

4. Have it play bridge against itself for a day

[dokoko]

Cyberyeti, on 2018-December-31, 07:14, said:

1. decide what sort of system you want (approach forcing, big club/diamond/NT etc)
2. decide on the balance between ease of memory and accuracy
3. have a look at what people have done before, you don't have to do most of this starting from nothing

I've only designed one from scratch and heavily modified one which I still play.

The one from scratch was weird and for a specific purpose (to be very playable and fun, but abuse the permitted system regs of the time as far as possible), but elements were borrowed from several other systems.

1. Do it with a partner and practice it!
2. Don't forget competitive bidding!
3. Practice your system with an open mind!
4. Not everything you design will work.

If you face a problem at the table there might be a systemical solution aroud. Perhaps you have to redesign parts of the system.

Once the basic structure is there, find detailed description of other systems and look for things you may borrow.

Be detailed! When looking at a given bidding situation I found it helpful to assign a description to all reasonable bids, then look whether there was a bid available for any hand I could reasonably have. If more than one bid was available for a particular hand, I asked myself whether I would like to have a choice or would prefer to narrow the definition of one of the bids involved.

BTW, I found this interesting: http://dcrcbridge.blogspot.com

[fromageGB]

2/1 GF with strong NT that may include a 5 card major IS designed for teams scoring. If you have a matchpoint system that is similar in some ways then it is easy to take 2/1 standard as a simple start. Add Gazzilli, and concentrate on what you do after a 2♣ open - ie show/deny fit, controls, additional strength beneath game level. Similarly work on your style of non-serious and cue bids after major agreement, to ensure you can find the best slam when there is one.

If designing from scratch, put emphasis in getting to games that are there, rather than best part scores, but if thinking of taking a mixture of features from different systems, remember that they may not gel as systems should be holistic. If your memory is up to it, relays can be better than "natural", but then this is the wrong forum.