[fuburules3]

http://tinyurl.com/3egmtoj

5NT is explained as having 15-17, but the robot has 12. I suppose if robot has 17 and I open 2C, maybe it should just go to grand, but still.

[barmar]

The book bid is 4NT, but simulations tell it to try 5NT.

Look at the range it shows for your 3NT: 19-22. 19+17 is only 36, not enough to jump to grand. If you have a hand that was planning on bidding 2♣-2♦-2NT (22-24), you're supposed to rebid 4NT in this auction, and then it will go to 7 with 17. If you had bid that way with this hand, it would have just jumped to 6NT.

[Bbradley62]

barmar, on 2011-November-05, 20:41, said:

Look at the range it shows for your 3NT: 19-22.

Does the 3N bid imply a NT-oriented hand? If so, wouldn't 20-21HCP have opened 2N (with 19 opening 1m)?

[barmar]

It looks like 3NT is in there as a catch-all, just in case no other rules match -- if you can't find anything better, bid 3NT. I don't think it will normally bid it, except maybe in competitive auctions that have taken away its normal rebids.

I think the auction in the OP could never have happened if the bot were sitting South, but it needs a meaning in case a human bids it. And since the bid to invite slam is 4NT, 3NT has to be weaker than that.

[Siegmund]

Welll.... double-dummy sims say 6NT is >50% at 31HCP and 7NT at 34 (provided you have all 4 aces.) It may be an unavoidable flaw of the simulator for it to make these slam tries and raises more agressively than mere mortals do.

[Bbradley62]

barmar, on 2011-November-05, 22:07, said:

I think the auction in the OP could never have happened if the bot were sitting South...

Do you mean that SouthGIB would have bid his suit holding this particular hand, or that the auction wouldn't exist, regardless of the actual holdings? Why wouldn't this auction simply show the same South hand as 2♣-2♦-2NT?

[barmar]

GIB opens that South hand 2NT, not 2♣. Add a J, and it opens 2♣; then the auction would go either 2♣-2♦-2NT or 2♣-2NT-3♣.

[Bbradley62]

So, according to these explanations: if South makes a stronger 4NT rebid, NS stops in 6N, but if South makes a weaker 3NT rebid, they get to 7N. Seems inconsistent.

Let's try to bring this back to the original question. South chose to treat this hand as balanced 22HCP, which seems completely reasonable to me. Let us suppose that he instead held ♠A9, ♥AJ63, ♦A75, ♣AKQ9.

The auction begins 2♣-2N-3N... If GIB doesn't want South to rebid 3N with this hand, the given explanation shouldn't encompass it. North continues with 5NT, which instructs South to bid 7N with the top end of his previously-stated 19-22 range. So, the question is still:

Why does North think that his balanced 12HCP belongs in 7N opposite a balanced 22HCP?

Earlier, the response was

barmar, on 2011-November-05, 20:41, said:

The book bid is 4NT, but simulations tell it to try 5NT.

Did North really simulate that "most" or "many" balanced 22counts will made grand opposite his hand?

[barmar]

Bbradley62, on 2011-November-07, 08:48, said:

So, according to these explanations: if South makes a stronger 4NT rebid, NS stops in 6N, but if South makes a weaker 3NT rebid, they get to 7N. Seems inconsistent.

The difference seems to be that 4NT is well defined, but 3NT is a default.

Quote

Let's try to bring this back to the original question. South chose to treat this hand as balanced 22HCP, which seems completely reasonable to me. Let us suppose that he instead held ♠A9, ♥AJ63, ♦A75, ♣AKQ9.

The auction begins 2♣-2N-3N... If GIB doesn't want South to rebid 3N with this hand, the given explanation shouldn't encompass it. North continues with 5NT, which instructs South to bid 7N with the top end of his previously-stated 19-22 range. So, the question is still:

Why does North think that his balanced 12HCP belongs in 7N opposite a balanced 22HCP?

Earlier, the response was
Did North really simulate that "most" or "many" balanced 22counts will made grand opposite his hand?

What seems to be happening is that lots of the maximum North hands in the simulation have a 5-card spade, heart, or diamond suit that fits well with South's, resulting in many hands where grand is either cold or at worst on a finesse or squeeze. There are also a bunch of 5-5 hands.

You might argue "but those hands aren't consistent with the bidding, they're supposed to bid their suit." But if GIB only looks for hands where IT would bid 3NT, it won't find any -- with a 2♣ opener it would either bid a 5-card suit or 4NT to invite to 6NT.

[Bbradley62]

Again, you seem to be saying "we like what North-GIB did". Is that the case?

[barmar]

What I'm saying is that when you make a confusing bid, your partner (human or robot) gets confused, and it's hard to blame them if they go wrong. I don't "like" GIB's bid, but I'm not sure how to fix it.

I suppose there's one thing we could do: change the system. Why use 4NT as the invitation instead of 3NT, when 3NT is an idle bid? I'll discuss this with Uday and Georgi.

[Bbradley62]

barmar, on 2011-November-07, 10:50, said:

I suppose there's one thing we could do: change the system. Why use 4NT as the invitation instead of 3NT, when 3NT is an idle bid? I'll discuss this with Uday and Georgi.

Thank you.

[Bbradley62]

This may or may not be related.

---

If the 2N rebid shows a balanced hand, why does it include the 19-21HCP range? Also, shouldn't it deny significant heart support?

[barmar]

Some of the rules have something like this in them "Shows enough for game opposite the minimum partner has shown". It then subtracts what partner has shown from game values, and this is included in the explanation as our minimum. So don't read 19+ as "could be as little as 19", read it as "guarantees at least 19". Since it has earlier promised even more, this is trivially true. GIB doesn't know how to merge all the constraints to produce a less confusing description.

[Bbradley62]

Cool. Thanks.

[Lovera]

7 NT (at double dummy you can see) there are because on the same play on diamond impasse, W is squeezed ♠/♥ at 10 trick on last club.

[Bbradley62]

There you go again (in my best Ronald Reagan voice)... digging up an almost-four-year-old thread...

[Lovera]

Bbradley62, on 2015-July-18, 14:41, said:

There you go again (in my best Ronald Reagan voice)... digging up an almost-four-year-old thread...

Yes, it is so. But while i was searching for another aim i have found this thread and i wonder myself that anyone had seen and said what i have told as general information.(Lovera)

[iandayre]

barmar, on 2011-November-05, 20:41, said:

The book bid is 4NT, but simulations tell it to try 5NT.

Look at the range it shows for your 3NT: 19-22. 19+17 is only 36, not enough to jump to grand. If you have a hand that was planning on bidding 2♣-2♦-2NT (22-24), you're supposed to rebid 4NT in this auction, and then it will go to 7 with 17. If you had bid that way with this hand, it would have just jumped to 6NT.

Could you explain further? It seems very counter-intuitive to have to jump to 4NT to show a minimum balanced hand, having opened 2C. Also, if you had an unbalanced hand, why would you not just bid a suit over 2NT? Given GIB's parameters, please provide an example hand that would raise 2NT to 3NT. Or, as I suspect such a hand is not likely, consider changing the programming so that raising 2NT to 3NT shows 22-24 or perhaps 22-23 or even exactly 22.

I now see that this has been discussed and is under consideration. Thank you. But I can't help but wonder why any experienced bridge player would design a system where 2C-2N-3N is an idle bid! Perhaps it has something to do with that fact that the use of a direct 2NT response to 2C as a natural call is no longer common practice in North America.

This post has been edited by iandayre: 2015-July-19, 16:48