[ArtK78]

Played this hand in an ACBL Instant Tournament this afternoon:

---

Whether you agree or disagree with my opening bid of 2NT is not relevant.

North is going to bid 7NT. Why not ask for aces first? North has 16 HCP. South's opening bid showed 20-21 HCP. It is possible that 4 HCP in the form of an ace are missing.

Even if North had 17 HCP so that an ace is not supposed to be missing, what does it cost to ask?

On this hand, it would have avoided a very silly result (which was duplicated 5 times in the 15 times the hand was played).

[barmar]

ArtK78, on 2016-July-04, 14:26, said:

Even if North had 17 HCP so that an ace is not supposed to be missing, what does it cost to ask?

And even if it has all the aces, it could be missing a king -- you don't want to bid a grand on a finesse.

[johnu]

I can see this hand making 7 after GIB ducks the ♠A while going for a 2 trick set.

[robert2734]

Story time. Partner opens a precision 1C showing 16+ points and I had 21 points 5-3-3-2 so I respond 7NT. Buddy asked me why I didn't ask for kings and I said I thought I could make 7NT off a king. My five card suit, partner's four card suit, three card suit and the ace in the suit we are missing a king. There are a couple of flaws with that argument but the point is 7NT is at worst on a finesse.

If you take the toppost hand and replace the ♠ king with the ♠ ace, 7NT isn't on a finesse. It's cold.

[lycier]

This is my simulation hand.

---

Responding 7nt seems to be a buggy bidding.

[ArtK78]

robert2734, on 2016-July-04, 20:59, said:

If you take the toppost hand and replace the ♠ king with the ♠ ace, 7NT isn't on a finesse. It's cold.

Yes, but that is not the point. I ask again, what does it cost to ask? If you not off an ace, then go ahead - bid 7NT. Just check first.

[Stefan_O]

ArtK78, on 2016-July-05, 10:54, said:

Yes, but that is not the point. I ask again, what does it cost to ask? If you not off an ace, then go ahead - bid 7NT. Just check first.

1. Yes, hard to find an example in this specific situation where Blackwood can lose you points, but that's just not how Gib works (i.e it does not exhaust all the possible actions and checks "what cannot cost?").

From the description, it seems there is a rule in the database, suggesting: with 16+hcp bid 7NT.
On top of that, it probably makes a simulation, and sees that (double-dummy) the contract makes in, say, 97% of all cases (assuming South has 20+hcp), which is considered good enough to just place the final bid.

Something like that, I'd guess, is how the algorithm works.

2. Additionally, I do not think in Gibs system 2NT-4NT is Blackwood -- when there is no potential trump-suit, Gib usually (always?) interprets 4NT as quantitative.
So it would then need to start with Stayman or something else, which gives opps options to Double for lead or bid a suit (which theoretically CAN cost something).

[johnu]

Stefan_O, on 2016-July-05, 13:05, said:

2. Additionally, I do not think in Gibs system 2NT-4NT is Blackwood -- when there is no potential trump-suit, Gib usually (always?) interprets 4NT as quantitative.
So it would then need to start with Stayman or something else, which gives opps options to Double for lead or bid a suit (which theoretically CAN cost something).

IIRC, 4♣ is Gerber. Of course, opponents can double or bid a suit (but one opponent is a passed hand, and the other couldn't bid over 2NT and would be on lead against a NT contract) but GIB gives 0% priority to that possible drawback.

Besides missing an ace which will probably be fatal in 7NT, or a king which makes 7NT a 50% contract most of the time, you could be missing a crucial queen. If GIB does do a simulation, double dummy analysis will let the contract make much more than a 50% guess, because a 2 way finesse is 100% double dummy, and if there is a possible squeeze, the contract will make. I don't know what the threshold is for bidding 7NT based on simulations, but it should be a lot closer to 100% than what is probably is now.

[Stephen Tu]

Stefan_O, on 2016-July-05, 13:05, said:

1. Yes, hard to find an example in this specific situation where Blackwood can lose you points, but that's just not how Gib works (i.e it does not exhaust all the possible actions and checks "what cannot cost?").

2. Additionally, I do not think in Gibs system 2NT-4NT is Blackwood -- when there is no potential trump-suit, Gib usually (always?) interprets 4NT as quantitative.

GIB plays Gerber as do most players in SA-based systems. So it can check for aces with 4♣, and here without a long suit source of tricks as North it's also appropriate to check for kings with 5♣.

Gerber is a vastly overused convention by weaker players, and rightfully derided, they often use it in spots where 4♣ is much better played as natural, or a cue bid, or a splinter. Most of them would be better off on average never having learned the convention. But on very rare occasions it is actually the right call, and this happens to be one of them. So it's just a matter of programming the rules for GIB to favor the safety check above blasting 7nt directly.

[Stephen Tu]

johnu, on 2016-July-05, 13:46, said:

Besides missing an ace which will probably be fatal in 7NT, or a king which makes 7NT a 50% contract most of the time, you could be missing a crucial queen.

I wouldn't be worried about missing Q with 16+ over a 2nt opener if all the aces and kings are held. It's probably impossible systemically to figure out which is the crucial Q. The contracts will tend to be either laydown or on some sort of finesse/squeeze guess, perhaps with some extra chances in another suit e.g. 3-3 break. With enough strength that the other table is in slam for sure, you only need > 50% at MP and > 57% at IMPs to bid grand, and "cold or on a hook" will easily clear that barrier.

[lycier]

I think there are some issues on Gib CC.

After opening 2nt. here is following Gib CC.

4♣=garber
4nt=Quantitative invite to 6nt---2-5♣,2-5♦,2-4♥,2-4♠,11hcp,12TPs.
5nt=11-Hcp,12-TPs.

So we have to think 5nt is undefined.Hping Gib programmers would improve the Gib CC system.

[ArtK78]

There is no doubt that the method for checking for aces over 2NT is 4♣. I am very surprised that there is a discussion of this.

All that I am saying is that, before GIB launches into 7NT, it should bid 4♣ Gerber to check that an ace is not missing. This should be programmed into GIB. I cannot imagine that this would be difficult.

[johnu]

Stephen Tu, on 2016-July-05, 13:55, said:

I wouldn't be worried about missing Q with 16+ over a 2nt opener if all the aces and kings are held. It's probably impossible systemically to figure out which is the crucial Q.

I agree it is impossible for GIB. George Rosenkranz invented something called SuperConfi which allowed you to find 7 of a minor with 2 balanced hands and one of the key points was trying to determine if a queen was missing in a 4-4 or 5-3 fit on the theory that you probably can't make a grand with "guessing" the queen. You could also adopt this when looking for 7NT with no modifications. Sometimes you could find out about another key queen. Relayers will say WTP, 8 rounds of relays and you'll know everything

[barmar]

Stefan_O, on 2016-July-05, 13:05, said:

From the description, it seems there is a rule in the database, suggesting: with 16+hcp bid 7NT.

Yes. There's a rule that says "After partner opens NT, with combined 36+ HCP and any shape, bid 7NT".

Quote

On top of that, it probably makes a simulation, and sees that (double-dummy) the contract makes in, say, 97% of all cases (assuming South has 20+hcp), which is considered good enough to just place the final bid.

Not in this case. The above rule disallows simulations. It seems more like the person who wrote the rule just calculated that the chance that the missing 4 points are an ace is low enough not to complicate the rule. And if the holder of the ace isn't on opening lead, there might be 13 tricks in the other suits.

[lycier]

barmar, on 2016-July-06, 08:57, said:

Yes. There's a rule that says "After partner opens NT, with combined 36+ HCP and any shape, bid 7NT".

I would guess you are a Gib programmer,right?

[iandayre]

barmar, on 2016-July-06, 08:57, said:

Yes. There's a rule that says "After partner opens NT, with combined 36+ HCP and any shape, bid 7NT".

Not in this case. The above rule disallows simulations. It seems more like the person who wrote the rule just calculated that the chance that the missing 4 points are an ace is low enough not to complicate the rule. And if the holder of the ace isn't on opening lead, there might be 13 tricks in the other suits.

Wouldn't you say Barmar that is a poor rule? Obviously you can be off an Ace with 36 HCP. And there is a simple and effective way to avoid that possibility by using Gerber. It is also accepted practice to upgrade hands with extra playing strength and good spot cards, and the GIB rule you quote ignores this. Mind you I realize that some Robot tourney players take this to excessive, sometimes nonsensical, levels, but Ace asking bids are still available. If they stop at 6 off an Ace and still go down, they get what they deserve.

[Bbradley62]

lycier, on 2016-July-07, 04:27, said:

I would guess you are a Gib programmer,right?

Your guess would be correct...

barmar, on 2012-January-11, 00:40, said:

Yeah, we never made an announcement about it (now that I think of it, I'm not even sure if we mentioned it in the "Meet the Yellow" I did), although it got out in the GIB forum.

I've been doing some programming, both on the site and behind the scenes (the first thing I did was some small changes to the BB$/MB$ purchase pages), and I've been doing some GIB-related work that I can't go into detail about. I'm a software jack-of-all-trades, so I expect over time I'll have my hands all over the place.

[Stefan_O]

barmar, on 2016-July-06, 08:57, said:

Yes. There's a rule that says "After partner opens NT, with combined 36+ HCP and any shape, bid 7NT".

Not in this case. The above rule disallows simulations. It seems more like the person who wrote the rule just calculated that the chance that the missing 4 points are an ace is low enough not to complicate the rule. And if the holder of the ace isn't on opening lead, there might be 13 tricks in the other suits.

OK. But how do you know this?
Have you been involved with the Gib source-code?

[Bbradley62]

Stefan_O, on 2016-July-10, 05:43, said:

OK. But how do you know this?
Have you been involved with the Gib source-code?

Yes, Barmar is a BBO employee who sometimes works with the innards of GIB, and when he makes a statement like this it should be accepted as fact.