BBO Discussion Forums: GIB's interpretation of a double as takeout is not sensible - BBO Discussion Forums

Jump to content

Page 1 of 1
  • You cannot start a new topic
  • You cannot reply to this topic

GIB's interpretation of a double as takeout is not sensible

#1 User is offline   zzxjoanw 

  • PipPip
  • Group: Members
  • Posts: 15
  • Joined: 2017-June-13

Posted 2017-July-19, 16:52



Robot correctly annotates the double of East's Michaels cuebid as penalty. As far as I know, after such a double of Michaels, any further doubles of either of the opponents' two suits are conventionally interpreted as penalty. In any case, north's diamond holding is certainly not "rebiddable". Is there any sensible call for North other than "pass"?
0

#2 User is offline   steve2005 

  • PipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 3,162
  • Joined: 2010-April-22
  • Gender:Male
  • Location:Hamilton, Canada
  • Interests:Bridge duh!

Posted 2017-July-20, 06:15

Yes make no sense tohave 2X as penalty then 2X as takeout. Pass should be forcing here is cant pen.
Sarcasm is a state of mind
0

#3 User is offline   Bermy 

  • PipPipPipPip
  • Group: Full Members
  • Posts: 220
  • Joined: 2017-April-11
  • Gender:Male
  • Location:South Africa
  • Interests:Bidding Theory

Posted 2017-August-04, 03:51

View Poststeve2005, on 2017-July-20, 06:15, said:

Yes make no sense tohave 2X as penalty then 2X as takeout. Pass should be forcing here is cant pen.


So if everybody agrees that Gib interpretation of double is not sensible......why do these programmers insist on letting the robot bid this way? Why isn;'t someone doing something about it. Surely all it needs is a decent expert bidder to design a proper working bidding algorithm. Its clear to me that the current programmers have their own version or their own agenda on how to bid, and its very very poor.

1000's of people are now paying good money to play with these robots every day, and while I not saying they are not fun, why do we have to endure such rubbish bidding.

New players are learning the game all the time, and robots are such a good learning tool. Why teach them this rubbish version of 2/1?
0

#4 User is offline   Stephen Tu 

  • PipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 4,097
  • Joined: 2003-May-14

Posted 2017-August-04, 12:12

They don't insist on it. It is just that the number of competitive auctions is immense, and it's a Herculean task to fix them all. It's not that the programmers don't know how to bid. It's that it is hard to transfer their knowledge to a computer, which is extremely dumb, and can only follow rules to the letter. A computer has no common sense. It can't extrapolate an idea from one auction to a similar but slightly different auction, and do something sensible like humans can. So you almost have to write down rules for every possible auction. This is really hard even for an expert bidder+ programmer.

Imagine writing down a set of rules so a 6 year old who never played bridge can bid correctly in all competitive auctions, and every day has forgotten and is starting from ground zero.

They are working on it, it's just progress is slower than we'd like.
0

#5 User is offline   Bermy 

  • PipPipPipPip
  • Group: Full Members
  • Posts: 220
  • Joined: 2017-April-11
  • Gender:Male
  • Location:South Africa
  • Interests:Bidding Theory

Posted 2017-August-05, 22:11

View PostStephen Tu, on 2017-August-04, 12:12, said:

They don't insist on it. It is just that the number of competitive auctions is immense, and it's a Herculean task to fix them all. It's not that the programmers don't know how to bid. It's that it is hard to transfer their knowledge to a computer, which is extremely dumb, and can only follow rules to the letter. A computer has no common sense. It can't extrapolate an idea from one auction to a similar but slightly different auction, and do something sensible like humans can. So you almost have to write down rules for every possible auction. This is really hard even for an expert bidder+ programmer.

Imagine writing down a set of rules so a 6 year old who never played bridge can bid correctly in all competitive auctions, and every day has forgotten and is starting from ground zero.

They are working on it, it's just progress is slower than we'd like.


Well then is rather appears that they are setting the "wrong" set of rules then. It not the bidding logic that is rotten its the basic bidding system. These are rules that quite frankly don't work at the bridge table. The parameters for finding 4-4 major suit fits are altogether wrong, their 2 and 3 level bids are wrong, take out doubles are wrong and most importantly penalty doubles are wrong. Cue bids are altogether wrong and simply cannot be trusted. To make matters even worse explanations are usually wrong. How hard can it be to program a robot to give the "correct" explanation for a bid, and then respond properly to the information given. The rule of 2 and 4 belongs to us and 3 and 5 belong to he opps do not exits here, so try push the bidding one more level and you can rely on the overbid? Partner plays against you......not with you, surely improvement is needed urgently.
0

#6 User is offline   cloa513 

  • PipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 1,529
  • Joined: 2008-December-02

Posted 2017-August-06, 05:33

When they set 2 double as penalty then the programmers should have fix all reasonable continuations- it is laziness as usual. It is one thing to say handling all bidding sequences is hard and then you should say why they find the time to set 2 double as penalty shouldn't have had the time for that.
0

#7 User is offline   jdonn 

  • - - T98765432 AQT8
  • PipPipPipPipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 15,085
  • Joined: 2005-June-23
  • Gender:Male
  • Location:Las Vegas, NV

Posted 2017-August-08, 10:31

View PostBermy, on 2017-August-05, 22:11, said:

How hard can it be to program a robot to give the "correct" explanation for a bid, and then respond properly to the information given.

For "a bid"? Not that hard, although much harder than you seem to think (it can be anywhere from a matter of hours to months). For all the bids? Incredibly hard and time consuming.
Please let me know about any questions or interest or bug reports about GIB.
0

#8 User is offline   Bermy 

  • PipPipPipPip
  • Group: Full Members
  • Posts: 220
  • Joined: 2017-April-11
  • Gender:Male
  • Location:South Africa
  • Interests:Bidding Theory

Posted 2017-August-08, 19:19

View Postjdonn, on 2017-August-08, 10:31, said:

For "a bid"? Not that hard, although much harder than you seem to think (it can be anywhere from a matter of hours to months). For all the bids? Incredibly hard and time consuming.

More excuses?

OK so Im not qualified as a programmer, however that should't disqualify me from commenting here. 25 years ago I owned a little hand held bridge "robot" made by Saitek that could not only bid and play better than Gib, I also seem to remember that I could decide which bidding system I would prefer (2/1 sayc acol precision or Polish club) and it also allowed me to switch on or off whatever convention I wanted or didn't want. The robot bidding and play was far more accurate than modern Gib. Gib has been around for maybe 15 years now and we still have to endure this. And still making basic mistakes that should have been rectified in the last century. 10 years ago I remember reading reports and complaints on how bad the bidding system was and as all can observe......very little progress has been made. All complaints seem to generate nothing but excuses.
0

#9 User is offline   iandayre 

  • PipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 1,114
  • Joined: 2013-December-23
  • Gender:Male

Posted 2017-August-09, 11:15

Let's just say that this particular bidding situation needs to be prioritized for improvement. The OP and others are correct that there is no conceivable call for North on the third round of bidding other than pass.
0

Page 1 of 1
  • You cannot start a new topic
  • You cannot reply to this topic

1 User(s) are reading this topic
0 members, 1 guests, 0 anonymous users