[bd71]

---

How does this happen? GIB doubles 7N when on lead and with an ace in hand...and then leads another suit.

This post has been edited by Gerardo: 2010-November-02, 07:27
Reason for edit: Changed diagram size using HV instead of direct link

[cloa513]

Obviously no simulation involved- GIB always picks what it thinks is the worst lead. I equally love GIB for taking the contract from a near solid 7H to force it into only robot makeable 7NT. Why I'd never play with a GIB!

[Phil]

LOL this is one of my regular partners. He will never live this down.

[bb79]

4 ♦ showed spade support, so spades are set as trump, that's the reason of correction from 7♥ to 7 ♠.

[cloa513]

This all describes how badly designed the GIB bidding system is. North is only worth a 2♠ rebid- Clubs can hardly be excluded as a contract and spades are strong nor the whole hand.

[bucky]

cloa513, on 2010-October-27, 16:56, said:

This all describes how badly designed the GIB bidding system is. North is only worth a 2♠ rebid- Clubs can hardly be excluded as a contract and spades are strong nor the whole hand.

Except that 2♠ will be insufficient -- South jump-shifted to 3♥, so the computer disallowed 2♠ bid.

[Bbradley62]

bucky, on 2010-October-27, 20:14, said:

Except that 2♠ will be insufficient -- South jump-shifted to 3♥, so the computer disallowed 2♠ bid.

Oh... let's not let the facts get in the way...

[barmar]

GIB believed the auction. Since NS supposedly have a ♠ fit, it was practically impossible for the ♠A trick to disappear. It was trying to do even better. Since N looked for slam opposite the splinter, it probably doesn't have wasted values there, so the ♦ lead may set up an additional defensive trick, which they can take when W wins the ♠A.

[rogerclee]

FWIW this has happened to me before also, GIB was on lead against 6N (undoubled) with AKxxx of spades and led a side suit.

[bucky]

Actually, when playing with GIB, before you make a bid, you can get the meaning of the bid (as interpreted by GIB) by moving your mouse cursor over the bid. Here 4♦ is known as diamond shortness in support of spades (BTW that is a standard interpretation of Soloway jump shift). Given the bidding, the diamond lead is correct if the goal is to maximize the expected number of defensive tricks. Unfortunately GIB probably isn't programmed to analyze how good the contract is (compared with other normal tables), so GIB didn't realize that leading ♠A to beat the contract one trick was as good as beating it 10 tricks in MP.

[TylerE]

How hard would it be to add logic such that if GIB on opening lead has a 100% cashout for at least -1 vs. a NT contract, to always force that play.

[bucky]

TylerE, on 2010-October-28, 20:19, said:

How hard would it be to add logic such that if GIB on opening lead has a 100% cashout for at least -1 vs. a NT contract, to always force that play.

This would backfire too. Sometimes beating a contract by 1 trick just ain't enough to get a good score, for example if opponents sacrificed.

[manudude03]

Sacrificing in NT seems like a rare occurance. Sure, when it does happen, then cashing out is probably going to be a bad score but...