[Stefan_O]

---

At trick 8, West leads C7, and East with AQ behind declarer's King plays the Ace and hands declarer the contract.

This occurred in: "Free Daylong Tournament (IMP) - 2016-11-16"

[cloa513]

Stefan_O, on 2016-November-17, 16:17, said:

---

At trick 8, West leads C7, and East with AQ behind declarer's King plays the Ace and hands declarer the contract.

This occurred in: "Free Daylong Tournament (IMP) - 2016-11-16"

It had to be consistant. It was giving tricks to declarer from trick 3 onwards. GIB:I wasn't awake earlier otherwise I would have dropped the J and Q of ♠ so I just dropped of Q♠- now what more can do to help. Helpful robot syndrome.

[Stefan_O]

cloa513, on 2016-November-17, 16:53, said:

It had to be consistant. It was giving tricks to declarer from trick 3 onwards. GIB:I wasn't awake earlier otherwise I would have dropped the J and Q of ♠ so I just dropped of Q♠- now what more can do to help. Helpful robot syndrome.

Right... apparently, it switched into "reverse mode" on this deal.

[jdonn]

I'm not sure about the spade play, but the club play was definitely because he thinks declarer would be too strong for his bid with the king of clubs.

[Bbradley62]

jdonn, on 2016-November-18, 21:02, said:

I'm not sure about the spade play, but the club play was definitely because he thinks declarer would be too strong for his bid with the king of clubs.

Since you brought up what GIB expects 4♠ to show... is it reasonable to include "11- HCP" in that expectation?

[iandayre]

jdonn, on 2016-November-18, 21:02, said:

I'm not sure about the spade play, but the club play was definitely because he thinks declarer would be too strong for his bid with the king of clubs.

OK, let me ask you this. It seems GIB invariably makes the "dramatic" play in these situations. If it believes neither play can cost, I doesn't take the play that can't possibly cost, but prefers the play that could be wrong if its assumption about declarer's hand is incorrect. I have a hard time believing that this preference has not been programmed in.

[jdonn]

iandayre, on 2016-November-27, 17:22, said:

OK, let me ask you this. It seems GIB invariably makes the "dramatic" play in these situations. If it believes neither play can cost, I doesn't take the play that can't possibly cost, but prefers the play that could be wrong if its assumption about declarer's hand is incorrect. I have a hard time believing that this preference has not been programmed in.

Not to my knowledge. I believe it chooses randomly between the plays, but since 50% is quite a lot of the time for a human who would never play the club ace, that is the play that sticks in our memories. When it plays the club queen, nothing even registers.

[iandayre]

jdonn, on 2016-November-28, 08:19, said:

Not to my knowledge. I believe it chooses randomly between the plays, but since 50% is quite a lot of the time for a human who would never play the club ace, that is the play that sticks in our memories. When it plays the club queen, nothing even registers.

OK you may be right. But one thing I am sure of. GIB does NOT play equals randomly. It will throw away the higher ones to keep the lower 100% of the time. Seems like there could be a connection.

[smerriman]

iandayre, on 2016-November-28, 12:30, said:

OK you may be right. But one thing I am sure of. GIB does NOT play equals randomly. It will throw away the higher ones to keep the lower 100% of the time. Seems like there could be a connection.

I'm not sure about throwing the higher of equals, but I definitely agree there is a consistent lack of randomness in certain circumstances. I've noticed if Gib is guaranteed to lose one of the final 2-3 tricks, it *always* opts to lose it as early as possible (eg, with a trump and a loser, it always discards the loser rather than trumping in then leading the loser).

[johnu]

jdonn, on 2016-November-18, 21:02, said:

I'm not sure about the spade play, but the club play was definitely because he thinks declarer would be too strong for his bid with the king of clubs.

After declarer has shown up with 16 HCP already, shouldn't GIB W decide that the description of 4♠ was not correct and not rely on a bad description? Just a rhetorical question.

What are the chances BBO will abandon GIB and lease/buy a new bridge program?

[jdonn]

GIB doesn't think that way, and 0%

[iandayre]

smerriman, on 2016-November-28, 13:30, said:

I'm not sure about throwing the higher of equals, but I definitely agree there is a consistent lack of randomness in certain circumstances. I've noticed if Gib is guaranteed to lose one of the final 2-3 tricks, it *always* opts to lose it as early as possible (eg, with a trump and a loser, it always discards the loser rather than trumping in then leading the loser).

You are right. GIB defends as if there were a bonus for winning the last trick.