Saturday, October 17, 2009

Understanding natural bidding

When this hand came up in a national championship, here are three natural auctions that ensued.

Auction 1:
1H 1S
2D 3C
3D 4NT
5D 6D

Auction 2 :
1H 1S
2D 4D
4H 4S
5D P

Auction 3 :
1H 1S
2D 3C
3D 4D (rkc)
4H 4S (q ask)
5H 6D

The second auction (at my table) went off the rails in the ambiguity between the attempt at showing extra values and lack of a club control in the right hand.
But both the other two auctions involved one hand bidding a slam on their own once partner showed 5-5 in the red suits. Fair enough, for when we start giving opener a hand with ten red cards, slam usually has reasonable play and is often cold.

Could South's assessment change if North bid a slow 3D ?

What would be North's choice with a hand like
xx
KQxxx
AQxx
xx

Or a hand like
x
KQxxx
AQxx
Jxx

I prefer a 3S bid on the first hand above and 3NT on the second but is that universal ?

13 comments:

  1. Lawl at those auctions. I don't get it. When at some point South shows the Queen of trumps, North must find a way to tell South that all keycards are present because South has not even shown any potential life. When North closes with 6D who is to argue with South since 1KC could be missing. North should obviously recognise the Ace of hearts is HUGE, and when South hears that all keycards are present, all systems go!

    I strongly believe in South bidding 3D with any troubling 5-4 hands. 1H-1S-2D-3C is the worst standard auction and the maximal amount of space is required for responder to express what kind of GF he is up to. This is especially useful if North wanted to agree hearts (1H-1S-2D-3C-3D-3H is hot!) or if North wanted to self bid spades (1H-1S-2D-3C-3NT-...err?). I'd prefer to give up the 3D "promising" 5-5 because since when has responder gone "ah cool, opener has 5D so let's agree on our 5-3 diamond fit"? And if a true 5-5 is given up, there is usually space for the partnership to bid out naturally in search of a 5-3diamond fit after a 3NT bid by opener (3C-3D-3S-3NT-4D) or mayber responder will be able to bid 5NT pick a slam.

    I also remember Kokish highly advocates the 3D as well...

    ReplyDelete
  2. I think I would have produced auction 3. To be honest, I hadn't thought a lot about the implications of the 3D rebid, and probably would have assumed it showed 5 cards. But when I consider andyh's comment, I think he's persuaded me. Hopefully, I would push to slam anyway.

    ReplyDelete
  3. whilst you might simply be changing problems some of the time i think a 5422 is balanced

    when you bid 3 spades with the first hand responder is going to be commonly tortured by the possibility that you have a 3541

    ReplyDelete
  4. I'd assert 3D being neutral as opposed to 3D being 55 as much better, not least of which because GF hands with 3 card support can then rebid 3H. I wonder if I ever forced andy to do that. Interestingly, parallelism of thought is often an indication of a fair idea.

    Abetting this. another idea is to play 3C as invite with D OR GF and 3D as GF with D. It is usueful idea to seperate out GF 2nd suit raises early. Especially if you do I do now and (transfer) raise 1st suits so you are much less likely to have a fit there.

    david

    ReplyDelete
  5. I think that 3D shows a 5-5 in principle, although I might also rebid 3D on eg x AKxxx AKQx xxx.

    It helps to have your other auctions defined. For me, 3S over 3C can't be a minimum 3541, as I would have raised to 2S.

    With the example 1543 11-count, I would have rebid 1NT (playing a 15-17 NT). Otherwise it becomes too hard to handle auctions like these. Although it could be wrong, I would also rebid 1NT with the 5242 with no club stopper. It's harder playing a weak NT.

    ReplyDelete
  6. I bid the following hands a long time ago with Avi Kanetkar:

    Axxx
    xx
    KJxx
    KQx


    x
    AKxxx
    AQxx
    Axx

    1H 1S
    2D 3C
    3H 3NT
    4NT 6D

    Having 3D promise 5D makes it a lot easier to move beyond 3NT with a minimum GF and only 4-card support.

    ReplyDelete
  7. I don't particularly care for the second auction - the 4D bid seems wrong with such poor diamonds when there may be two other strains worth investigating. After all, couldn't opener hold KQx Kxxxxx AQJ x (7S very good)or KQ KQ10xxx AJxx x (7H very good). 4D is just too unilateral for such a flexible hand.

    ReplyDelete
  8. While 1H-1S-2D-3C is notable for leaving the partnership less room than any other 4SF auction after a one-level response, the issues it raises are mostly applicable to all 4SF auctions:
    -- Which of O(pener)'s rebids should be pure?
    -- Which one(s) can be fudged? Does it make sense to always follow the same principle or to vary between auctions?
    -- Can some of the problems in third- (and later-) round auctions be solved by better definitions of earlier bids?

    In some partnerships I have had the agreement that, with fudge hands after 4SF, O makes the cheapest bid (so, here, 3D is ambiguous). In others I have played that it is O's first suit (here 3H) that can be ambiguous. Another variation is to play that, if O rebids 1S, then 2S is the fudge bid (after, say, 1C-1H-1S-2D). These principles can be extended to other auctions: playing that one should make the cheapest of equally attractive (or unattractive) bids helps O choose a rebid with AQxxxx Kx Qxxx x after 1S-2H-2S-3D.

    Any of these agreements can cost: if 3D is ambiguous then O will often be unable to show that he is 5-5; if 3H does not promise a six-card suit then R(esponder) will choose 3N when 4H is better (change Khokan's example to AQxxx xx KJx Kxx to see why this matters); if 3S can be a doubleton then how are 5-3 spade fits to be found?

    Part of the answer lies in better definition of O's second-round actions. I think that -- even if one does not regularly raise major-suit responses on three-card support -- 1H-1S-2S should be routine on 3=5=(41) hands unless O has extra strength. And, even if playing a weak notrump, one should play 1H-1S-1N as weak (as KS wisely does). This should be O's preferred choice with minimum 1=5=(43)s and 2=5=(42)s. (A useful adjunct is to play that R's 2C inquiry asks for O's spade length: 2D=(0)1, 2H=2, 2S=3.) Minimum 6-4 hands should rebid 2H rather than 2D or 2C. Thus a rebid of 2D (or 2C) shows extra values (could be 16-17 with 5332) or extra shape. With these underpinnings the partnership is on firmer ground. And it now makes sense to play 3H as the fudge bid: if O has 64 he will have extra values and can usually afford to rebid 4H if R chooses 3N after 1H-1S-2D-3C-3H. (And, if you play modified Bergen raises where all GI hands with three-card support bid 3C directly over 1H then R's 3H rebid after 1H-1S-2D shows a GF so further reducing the dangers of using 3H as the fudge bid. Otherwise R has no natural way to show slam interest in H below game.)

    (Another useful agreement -- one that would have helped Sartaj in his auction -- is to play that a jump in the fourth suit above 3N is neither natural nor a splinter [which would require R to have length in O's first suit as well as his second] but a control-showing raise of O's second suit. Thus 1H-1S-2D-4D denies a C control, as well as showing slam interest and setting the suit. Either way -- after R's 4C or 4D rebid -- the partnership is well placed.)

    David

    ReplyDelete
  9. I should have added that, while a fudge, 3H in the methods I propose does promise extra values (else rebid 1N or show 55), denies three spades (so can't be 3=5=3=2 16-17) and five diamonds and a C stopper. So, O has something like 2=5=4=2 or 1=5=4=3 15+ with no C stopper. Not bad for a fudge!

    David

    ReplyDelete
  10. I play 3C as INV+ (still 4th suitish), and other INV-looking bids as GF.

    So 1H-1S-2D-2NT is a naturalish GF. Now -3D (definitely 5-5)-keycard-etc

    With hand 1 (xx KQxxx AQxx xx), just bid 3NT over 2NT. I would still try with 4C next - clearly a cue for D because I would have agreed H with 3H over 2D if that was where I lived. A few cues and the auction dies in 5D.
    With hand 2 (x KQxxx AQxx Jxx), go the same way. Not quite good enough to pattern out with 3C.

    ReplyDelete
  11. Can't see the linked hand with any browser -- chrome, ie, firefox all show a blank page that give the board number only.

    ReplyDelete
  12. Works in internet explorer. Dont know about the others.

    ReplyDelete
  13. Works in Firefox too.

    I liked David Morgan's suggestions and Ben Thompson's use of inverted meanings..

    I toyed with using 3C as a base for inv+ transfers, but you run into serious problems when opener has extras with 3S.

    ReplyDelete