BBO Discussion Forums: LEAD OUR SUIT - BBO Discussion Forums

Jump to content

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

LEAD OUR SUIT what the heck is with these random garbage leads

#1 User is offline   gszes 

  • PipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 3,655
  • Joined: 2011-February-12

Posted 2018-June-09, 05:48

bot holding xxx xxx Kxx Kxxx I am not going to give full bidding because it matters not to the computer BUT I have bid diamonds (not a preempt) and I got a stinking heart lead I am TIRED of these irrational and stupid leads which constantly give away
tricks and worse ALWAYS seem to work in favor of the bots on defense when their p has the miraculous card they need. STOP THIS MADNESS force the computer to LEAD OUR SUIT UNLESS THERE IS A STRONG REASON TO BELIEVE ANOTHER SUIT IS BETTER (on both offense and defense). I stopped playing in middle of a challenge to type this because this type of stuff borders on cheating (or at least gives the really good appearance of it). I am tired of doing things with the computer to no avail and this is a SUPER EASY fix. How about it people can we all stand together and yell our lungs out in frustration??????????????
0

#2 User is offline   DozyDom 

  • PipPipPipPip
  • Group: Full Members
  • Posts: 191
  • Joined: 2017-November-05
  • Gender:Male
  • Location:UK

Posted 2018-June-09, 05:54

... maybe we shouldn't because forcing the bot to lead your suit would be an absolutely ridiculous idea?
0

#3 User is offline   wbartley 

  • PipPipPipPip
  • Group: Full Members
  • Posts: 289
  • Joined: 2010-July-23

Posted 2018-June-09, 15:01

To be fair, the request wasn't to force the bot to lead his/her suit. It was to force the bot to lead partner's suit unless there is a strong reason to believe another suit is better. The real question is, what constitutes a strong reason to believe another suit is better? GIB generates a number of hands that match the bidding. It then runs double dummy analysis to decide which card is the most effective. Since the generated auction is what GIB would bid and GIB doesn't concern itself much with suit quality when overcalling, that might be skewing the results toward not leading partner's suit. I wonder if GIB leads partner's suit when choosing from what it believes to be equally effective options. If it doesn't, that change alone might make GIB lead partner's suit a lot more often than it does.
0

#4 User is offline   Stephen Tu 

  • PipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 4,076
  • Joined: 2003-May-14

Posted 2018-June-09, 15:19

This is one area where GIB probably is better than humans on average. The GIB general approach is likely more accurate than human rules of thumb. We like to lead partner's suit too often because it's so easily defensible in the post mortem. Even if it turns out wrong, it's easy to shrug off, partner won't criticize you for leading his bid suit.

Now if the auction shows huge errors in the bidding database, so its estimate of what the typical hands for the auction are way off, causing a bad lead, that's a different story, and the DB should be fixed.
Against NT, GIB likes to lead passive a lot, which is often right, though it would do a ton better if it had some attitude component to the lead and understood attitude signals so it would know when partner wants to continue or not. So often it finds the right lead but then inexplicably stops continuing the suit when declarer is ducking out of desperation.
1

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