BBO Discussion Forums: robot nt defense - BBO Discussion Forums

Jump to content

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

robot nt defense

#1 User is offline   cybourgh 

  • Pip
  • Group: Members
  • Posts: 3
  • Joined: 2012-April-24

Posted 2013-April-18, 14:16

I posted this in the general forum and then realized this was the correct forum for this subject.

I am using the cheap gib so maybe the expensive one plays better. But the gib's NT defense is horrible. It almost always leads from its shortest suit even if I have never bid and even if the opponents have bid that suit. Rarely do I see the conventional 4th from longest and strongest lead even when it would have been a setting lead. It doesn't lead the top of a 3 sequence either. It hardly ever returns my lead so me leading 4th from longest and strongest won't work. This needs fixing.
0

#2 User is offline   barmar 

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • Group: Admin
  • Posts: 22,109
  • Joined: 2004-August-21
  • Gender:Male

Posted 2013-April-19, 02:40

View Postcybourgh, on 2013-April-18, 14:16, said:

I posted this in the general forum and then realized this was the correct forum for this subject.

I moved that thead here, then I noticed that you reposted it yourself....

I've posted about this a number of times, and even included information in the GIB System Notes. Read the book "Winning Notrump Leads" by David Bird and Taf Anthias, and you'll understand why GIB defends as it does. Passive leads often have the best double dummy expectation. Leading away from honors often gives away a trick. You want to get to partner's hand, so he can lead through declarer to your strong suit.

And if you get the book "Winning Suit Leads", you'll notice that GIB's leads there are also very similar to those recommended. It almost always leads a singleton if it has one, frequently leads doubletons, and often cashes aces (isn't it annoying how often he then gives his partner a ruff?).

#3 User is offline   cloa513 

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

Posted 2013-April-20, 06:40

View Postbarmar, on 2013-April-19, 02:40, said:

I moved that thead here, then I noticed that you reposted it yourself....

I've posted about this a number of times, and even included information in the GIB System Notes. Read the book "Winning Notrump Leads" by David Bird and Taf Anthias, and you'll understand why GIB defends as it does. Passive leads often have the best double dummy expectation. Leading away from honors often gives away a trick. You want to get to partner's hand, so he can lead through declarer to your strong suit.

And if you get the book "Winning Suit Leads", you'll notice that GIB's leads there are also very similar to those recommended. It almost always leads a singleton if it has one, frequently leads doubletons, and often cashes aces (isn't it annoying how often he then gives his partner a ruff?).

GIB doesn't do that either he had three suits with Hxx and 1098x and which suit did he lead from Qxx- giving me one trick- its just random suit choice.
0

#4 User is offline   barmar 

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • Group: Admin
  • Posts: 22,109
  • Joined: 2004-August-21
  • Gender:Male

Posted 2013-April-20, 10:22

It's not random, the auction is also a factor.

GIB defense is very pretty simple: it deals out a bunch of hands consistent with the auction, performs performs double dummy analysis after each possible lead, and chooses the lead with the best average result across the hands. The only time it picks randomly is if there are multiple best choices.

GIB does NOT use rules of thumb, like "lead 4th best from your longest and strongest". It knows about standard leads, so if it decides to lead one of the high cards from 1098x, it will lead the top of the sequence. But the sequence does not automatically make that suit a good lead, that has to come from the DD analysis.

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