Would you lead a diamond against 3NT?

[calm01]

http://tinyurl.com/72wzlme

[cloa513]

Sounds eminanently sensible given that partner has at most 1 HCP- whatever Jack it is almost doesn't matter. Though that K Spade lead later is a waste- just put through another diamond.

[barmar]

GIB tends to avoid leading away from honors, especially tenaces, against NT -- it's not a big believer in "4th from longest and strongest". That's why it frequently leads from 3 small. Very often it's right -- partner gets in and switches to its suit, and declarer is finessed. It's also wrong many times, because of the tempo that's given up.

 

This is why it doesn't always return the suit when it wins the opening lead -- if you don't necessarily lead your strongest suit, it's not automatic to continue it.

 

It doesn't look like it matters too much on this hand, every lead gives up a free finesse. But move a few of NS's cards around, and this passive lead is very likely to be the winner (or at least cut down the overtricks).

 

And once it sees dummy, it knows that all the finesses are working. So ♠K is no worse than anything else.

[calm01]

Thankyou Barmar.

 

What you describe fits GIB's behaviour very well both in choice of initial lead and follow-up.

 

I also agree about implied loss of tempo. Since the tempo of the opening lead is the main advantage of the defence over a NT declarer who can see the partnership resources, giving up this tempo advantage seems a losing strategy for the defence in the long run.

[barmar]

I haven't kept track of wins versus losses, but I know it's gotten the better of me many times with that strategy.

[cloa513]

GIB tends to avoid leading away from honors, especially tenaces, against NT -- it's not a big believer in "4th from longest and strongest". That's why it frequently leads from 3 small. Very often it's right -- partner gets in and switches to its suit, and declarer is finessed. It's also wrong many times, because of the tempo that's given up.

 

This is why it doesn't always return the suit when it wins the opening lead -- if you don't necessarily lead your strongest suit, it's not automatic to continue it.

 

It doesn't look like it matters too much on this hand, every lead gives up a free finesse. But move a few of NS's cards around, and this passive lead is very likely to be the winner (or at least cut down the overtricks).

 

And once it sees dummy, it knows that all the finesses are working. So ♠K is no worse than anything else.

I don't what you talking about- diamond is definitely best- it definitely costs no tricks- note if declarer had played two or more rounds of diamonds then it would be more difficult- Jhearts stands out then. KS costs one trick because declarer can cash all three spade tricks. If GIB is going by simulations then he would put partner with almost no HCP- declarer is supposed to have 25-26HCP for game. Obviously GIB is ignoring simulations.

[barmar]

GIB can't "ignore simulations" -- it's the only strategy it has on defense (declarer has an alternate algorithm called GIBson, which takes over at trick 3).

 

East could have ♠Jxx. If South has ♠J, then he can always get 3 spade tricks by finessing the 9. Since GIB uses DD analysis in its simulations, it assumes declarer will do this if necessary, so it's not giving up anything by playing the K.

 

If you click on the GIB button after West wins the ♦J, every card is labeled "+2", so with DD play there's really nothing West can do. And since West can tell almost the whole hand, it effectively knows this.

[barmar]

This was a basic bot in the MBC, wasn't it? If we make any improvements to play, we're going to be focused on the advanced bots. The basic bots are so time-constrained that dumb plays are likely. When I ran this hand with advanced bot settings, it open led 4th best ♠.

 

I'm not trying to gouge you, but if you're going to play with bots a lot, and want a less frustrating game, rent the advanced bots.