[rogerclee]

AKx xxx AKx K8xx

(P) 1N (2C) ?

2C = C+M

Give partner 14 points with a 5 card suit and 5332 shape, any balanced hand with 15 points, or any 16 point hand with exactly 4333 shape.
Give RHO 5+C, 5+M, and 5+ points.

Can you tell me how often we make 6N?

Thanks.

[nigel_k]

From a sample of 10,000 hands, it made 3962 times.

Also, slams make more often double dummy than at the table, as declarer has most of the decisions. Though it is easier when the defence have given you a blueprint.

[rogerclee]

nigel_k, on 2013-June-02, 04:13, said:

From a sample of 10,000 hands, it made 3962 times.

Also, slams make more often double dummy than at the table, as declarer has most of the decisions. Though it is easier when the defence have given you a blueprint.

Thanks, I thought 6N would make on more than 50% of these types of hands, but apparently this is not the case.

[barmar]

Suits aren't breaking, and you have a pair of balanced hands so you don't have an extra source of tricks. Overcaller probably has ♣A and a ♥ honor, so the latter finesse will be off.

[rogerclee]

barmar, on 2013-June-03, 09:33, said:

Suits aren't breaking, and you have a pair of balanced hands so you don't have an extra source of tricks. Overcaller probably has ♣A and a ♥ honor, so the latter finesse will be off.

Yeah I bid 6N thinking it would usually make on a strip squeeze or heart/club squeeze.

[Cascade]

I got significantly better results than Nigel from my sample of 10000 hands

    0          0
    1          0
    2          0
    3          0
    4          0
    5          0
    6          0
    7          6
    8         24
    9         43
   10        654
   11       3724
   12       4816
   13        733

Here is my dealer code:

predeal	north SAK2, H432, DAK2, CK832

thishand = 

(
  (
    hcp(south)==14 and shape(south,any 5332)
  ) or

  (
    hcp(south)==15 and shape(south, any 4333 + any 4432 + any 5332)
  ) or
 
  ( 
    hcp(south)==16 and shape(south, any 4333)
  )

) and

hcp(west)>=5 and 
clubs(west)>=5 and 
(hearts(west)>=5 or spades(west)>=5)

condition thishand

generate 1000000000

produce 10000

action

frequency(tricks(south,notrumps),0,13),

[rogerclee]

Thanks for the additional sim, hopefully nigel will follow up with a guess as to why the sims show very different results.

[gnasher]

rogerclee, on 2013-June-04, 01:06, said:

Yeah I bid 6N thinking it would usually make on a strip squeeze or heart/club squeeze.

I'd think that too. Our club length makes it likely that partner has a doubleton there, so he will probably have length in at least one of the pointed suits.The overcall suggests that partner's strength is in those suits too. That makes our two AKx holdings look rather good. For example, if partner has QJxx KQx QJxx Ax, we have only eleven top tricks, but RHO can't hold onto both suits.

[nigel_k]

Sorry, my logic was faulty and I managed to exclude all 5332 shapes for partner which obviously affects it a lot. After fixing and rerunning I have it making 5536/10000 times which is almost the same as what Wayne got.

[gnasher]

Is it correct to assume that the overcaller is 5-5? If 5-4s are allowed, that reduces our squeeze chances, as well as making it harder to play the hand (though the double-dummy analysis won't take account of that).

[PhilKing]

gnasher, on 2013-June-06, 03:32, said:

Is it correct to assume that the overcaller is 5-5? If 5-4s are allowed, that reduces our squeeze chances, as well as making it harder to play the hand (though the double-dummy analysis won't take account of that).

RHO has a max of nine quacky points, so 5-4 seems rather far-fetched (maybe KQJTx QJ9x?). And if they are vul, almost inconceivable. Maybe they just guarantee 5-5 if weak.

[gnasher]

PhilKing, on 2013-June-06, 03:39, said:

RHO has a max of nine quacky points, so 5-4 seems rather far-fetched (maybe KQJTx QJ9x?). And if they are vul, almost inconceivable. Maybe they just guarantee 5-5 if weak.

How do you know his points are quacky? Have you secretly bid Gerber and found out that partner has two aces?

I agree that to bid with 5-4 he needs concentrated values, some spot cards, an 8- or 9-count, and to be non-vulnerable, but 5-4s are about five times more common than 5-5s. Wouldn't you bid with xxx AK10x x J10xxx ?

[PhilKing]

gnasher, on 2013-June-06, 04:25, said:

How do you know his points are quacky? Have you secretly bid Gerber and found out that partner has two aces?

I agree that to bid with 5-4 he needs concentrated values, some spot cards, an 8- or 9-count, and to be non-vulnerable, but 5-4s are about five times more common than 5-5s. Wouldn't you bid with xxx AK10x x J10xxx ?

Quacky hands are 4.7 times more likely based on my hand.

I would not bid vulnerable on the hand you give.

[gnasher]

PhilKing, on 2013-June-06, 04:32, said:

I would not bid vulnerable on the hand you give.

No, nor would I. As I said, to bid with 5-4 he'd have to be non-vulnerable.

[PhilKing]

gnasher, on 2013-June-06, 04:41, said:

No, nor would I. As I said, to bid with 5-4 he'd have to be non-vulnerable.

Well I'm going to go out on a limb and say Roger knew what he was doing when he rejected the possibility of a 5-4 shape.

[Cascade]

Allowing 5-4 shapes into the simulation reduced the number of slams as follows:

 0          0
 1          0
 2          0
 3          0
 4          0
 5          0
 6          0
 7          4
 8         19
 9         55
10        791
11       4211
12       4356
13        564

The ammended lines of code was:

clubs(west)>=4 and (hearts(west)>=4 or spades(west)>=4) and

(
  (clubs(west)+hearts(west)>=9) or 
  (clubs(west)+spades(west)>=9)
)

I suppose if 5-4 hands are allowed then the double dummy result lies somewhere between the previous 55% and this 49% given that not all 5-4 hands will bid.

[rogerclee]

It's possible they would bid with 5-4 pretty routinely but they were facing a PH partner at w/w, so surely 5-5 was much more likely and I decided it was more accurate for "quick sim" purposes to just ignore it. I'm not claiming that my sim conditions match the set of hands my RHO would bid on, just the one he actually bid on (JT9xx void QJT QT9xx).

I'm also not sure about my "reject a quant invite" conditions, but they seem fine to me. We had not discussed what 2S (range/clubs) then 4N over 2N meant so I discarded it, but if the sim numbers are roughly accurate, it's the best option. I regret not doing this in retrospect at the table, seems like partner will pretty much always field it, and it's basically what we have.

Partner had Qx AKQ2 98xx Axx and went down 1 with essentially no play.

[Phil]

rogerclee, on 2013-June-06, 08:28, said:

It's possible they would bid with 5-4 pretty routinely but they were facing a PH partner at w/w, so surely 5-5 was much more likely and I decided it was more accurate for "quick sim" purposes to just ignore it. I'm not claiming that my sim conditions match the set of hands my RHO would bid on, just the one he actually bid on (JT9xx void QJT QT9xx).

Partner had Qxx AKQ2 98xx Axx and went down 1 with essentially no play.

14 cards...

[rogerclee]

Phil, on 2013-June-06, 08:32, said:

14 cards...

fixed, also he obviously did have play, but not after he found out that his major was spades.