[steve2005]

hnad
Look at clubs led towards end declarer has Ax
opps have K opposite Q
dummy has JT etc
so opps lead club and Q doesn't cover board so no club loser.
daylong tournament

[jdonn]

Since you made a takeout double, GIB thought you had 3+ clubs. This doesn't change when you show a big hand later. If you have 3+ clubs then covering can't help (or hurt).

[pescetom]

As often happens, I don't understand GIB's bidding.
After the takeout double, why does he force with 2♠ as if he had opening strength and no clear strain, instead of an honest 3♥?

[steve2005]

jdonn, on 2018-February-12, 08:52, said:

Since you made a takeout double, GIB thought you had 3+ clubs. This doesn't change when you show a big hand later. If you have 3+ clubs then covering can't help (or hurt).

I can still have a balanced hand too good for NT with 2♣ or even a hand too good for 2♥.
Gib when it thinks it doesn't matter seems to play random cards.
There is still a correct bridge card to play usually even if it probably doesn't matter.

[johnu]

jdonn, on 2018-February-12, 08:52, said:

Since you made a takeout double, GIB thought you had 3+ clubs. This doesn't change when you show a big hand later. If you have 3+ clubs then covering can't help (or hurt).

Certainly continuing to promise support for the unbid suits is one of the "features" of making strong, off-shape takeout doubles in Gibberish and why doing so is usually dangerous to the doubler's side. To paraphrase Jean-Luc Picard, "Don't make it so".

[jdonn]

steve2005, on 2018-February-12, 12:15, said:

I can still have a balanced hand too good for NT with 2♣ or even a hand too good for 2♥.
Gib when it thinks it doesn't matter seems to play random cards.
There is still a correct bridge card to play usually even if it probably doesn't matter.

You can't have a balanced hand too good for NT any more when you bid 4H. But when you double and bid a suit GIB thinks it shows 3+ in the other suits. That's just how it is, for now. From what I am told, this is a harder problem to fix than I would have otherwise thought. I recommend to players to avoid very offshape takeout doubles (not that this was one, but just saying).

[steve2005]

Gib often has bidding problems with off-shape good doubles too.

But in any case, playing a random card if the simulation indicates the card doesn't matter is a huge flaw. Early chess programs had similar flaws.
With equal cards Gib should be making a bridge decision on what card to play. Often Gib plays an unnecessarily high card which makes no difference double dummy but allows mortals to claim on hands they could go down. On this hand even a beginner bridge player would recognize that the Q is right card to play and has almost no possible downside that late in the hand.

Even with Gib bidding all 4 hands you just can't be sure you can rely on the bidding accuracy to produce hands for simulation. Gib is often off on pts and distribution.