[gszeszycki]

i am curious as to why any 4333 shape was eliminated as an option for the responder hand:)

[bluecalm]

Quote

i am curious as to why any 4333 shape was eliminated as an option for the responder hand:)

English is not my native language so maybe I needed a coma there.
I wanted to say that responder has either (2-3♠ 2-3♥ without 8 card major) or he has any 4-3-3-3 shape. Any English native speaker care to comment on meaning of my original sentence ?

[gwnn]

You need some well-placed parentheses to make your criteria mathematically accurate

[phil_20686]

bluecalm, on Jun 23 2010, 06:47 AM, said:

i understood it as:

Quote

N has 9-14hcp and either 2-3♠, 2-3♥ without an 8 card minor, or any 4-3-3-3 shape.

[pooltuna]

bluecalm, on Jun 23 2010, 08:20 AM, said:

Quote

i am curious as to why any 4333 shape was eliminated as an option for the responder hand:)

English is not my native language so maybe I needed a coma there.
I wanted to say that responder has either 2-3♠, 2-3♥ without 8 card major or he has any 4-3-3-3 shape. Any English native speaker care to comment on meaning of my original sentence ?

You seem to have eliminated the following holdings from the N hand
1345
1354
3145
3154

[bluecalm]

Quote

You seem to have eliminated the following holdings from the N hand
1345
1354
3145
3154

Yes, as I believe most expert partnerships don't bash 3nt with those
I may be influenced by polish bridge culture though as everybody here plays jumps to 3♥/3♠ to show those hands.

[gszeszycki]

I do not know the limitations of your software but if possible maybe the best logical request would be 1n hand (any 5332) (any 4333) (any 5422)
(any 4432) 3262 3226 2362 2326
responder 9-14 is fine but limit side suit length to 6. If this partnership is using puppet stayman then we have a huge variety of hands that need
to be excluded from responder (they will generally have a max of 2 cards in either major) 2263 2236 2254 2245 <2164 2146 1264 1246
0265 0256 2065 2056> <> these are possible but will almost always use another method of bidding. if using regular stayman then also add 2335
2353 3225 3253 3334 3343 3244 2344 0364 0346 3064 3046 1345 1354 3145 3154 0355 3055 sigh i probably forgot something but this should cover it. We are also assuming the opps will search for major suit fit when able (an assumption that will much less guarantee the more hcp responder has). Im tired:)

[bluecalm]

Quote

We are also assuming the opps will search for major suit fit when able (an assumption that will much less guarantee the more hcp responder has). Im tired:)

Yeah, yeah a lot of depends on system. I can tell you though that it doesn't matter much. Most hands will fall into my range anyway. I don't see why I should limit suit length to 6. Responder is auto bidding 3NT on 2-2-7-2 for example.
If no pupper then my range is almost perfect. Puppet is too difficult because different people bid it on different hands.
Anyway it doesn't matter you can make small changes and results will be the same. I am sure of that because I did hundreds of those simuls. ♠ is the best and ♦ sucks no matter what you try

[jdonn]

Anyone have the actual hand?

[gwnn]

1 sec.

here you go

http://tinyurl.com/36e4qt2

[pooltuna]

bluecalm, on Jun 23 2010, 09:08 AM, said:

Quote

You seem to have eliminated the following holdings from the N hand
1345
1354
3145
3154

Yes, as I believe most expert partnerships don't bash 3nt with those
I may be influenced by polish bridge culture though as everybody here plays jumps to 3♥/3♠ to show those hands.

in SAYC I believe 1NT (p) 3♥/3♠ are slam trys in those suits not singletons consequently these hands have to bash 3NT.

[Cascade]

JDonn said:

I know people say you shouldn't lead from Axxx here but a heart really feels right to me. ...

and two posts later

JDonn said:

I would lead a spade all day, nothing else even comes close.

What is going on here?

LOL

[Cascade]

bluecalm, on Jun 23 2010, 11:47 PM, said:

Simulation (1000 hands):

S has 15-17balanced with possible 5M or 6m.
N has 9-14hcp 2-3♠, 2-3♥ without 8card minor or any 4-3-3-3 shape.

Number of hands given card is winning lead (at imps):

7♠ - 221
3♥ - 192
3♦ - 176
9♣ - 190

High score of 9♣ is certainly surprising but overall things are as expected:
♠ is best while ♦ sucks.

Not for this particular hand but I have over 10000 hands analysed for the opening lead single dummy on the auction 1NT (15-17) 3NT.

xxxx on average won 1.5 IMPs

ATxx on average won 0.4 IMPs

KJxx on average lost 1 IMP

x on average lost 1.2 IMPs

[jdonn]

Wayne glad you finally caught on. The truth is I think a diamond lead stands out.

[Cascade]

jdonn, on Jun 24 2010, 12:28 PM, said:

Wayne glad you finally caught on. The truth is I think a diamond lead stands out.

Obviously.

Why did you dismiss a club so quickly?

[hanp]

Cascade, on Jun 23 2010, 07:18 PM, said:

bluecalm, on Jun 23 2010, 11:47 PM, said:

Simulation (1000 hands):

S has 15-17balanced with possible 5M or 6m.
N has 9-14hcp 2-3♠, 2-3♥ without 8card minor or any 4-3-3-3 shape.

Number of hands given card is winning lead (at imps):

7♠ - 221
3♥ - 192
3♦ - 176
9♣ - 190

High score of 9♣ is certainly surprising but overall things are as expected:
♠ is best while ♦ sucks.

Not for this particular hand but I have over 10000 hands analysed for the opening lead single dummy on the auction 1NT (15-17) 3NT.

xxxx on average won 1.5 IMPs

ATxx on average won 0.4 IMPs

KJxx on average lost 1 IMP

x on average lost 1.2 IMPs

Thanks Wayne, I like these results.

In your analysis, was the suit led a major, a minor or either? Also, could you specify how the simulation was done? Did you force the computer to lead from any suit against each 1NT - 3NT? Or did you let it pick a lead?

On those 10000 hands, can you indicate on how many the lead was from one of these four holdings?

[Cascade]

10000 was an approximation.

On each hand I forced the computer to lead from each suit and then cross-IMPed the results. (The average IMPs I quoted may actually be cross-IMPs from three comparisons.)

So in fact each of the approximately 10000 hands was played four times with a different lead each time.

There were approximately 100 hands of each type as the 10000 hands were actually 100 hands with each honour combination. I included only hands with no void and no six or longer card suits. So x, T, J, Q, K, A, xx, Tx, Jx, Qx, Kx, Ax, JT, QT, KT, AT, QJ, KJ, AJ, KQ, AQ, AK etc.

So related to this problem there were 100 hands generated with KJxx and a low card from this suit was led. Similarly 100 hands with xxxx and 100 hands with ATxx. The statistics were just based on the 100 hands that is I did not adjust for the times where I was analyzing xxxx and the hand happened to contain ATxx or KJxx in another suit.

I do not have the numbers at my finger tips but there as a big premium on leading a major on this auction.

Obviously the simulation has some flaws but on the other hand it is a much larger sample than any of us will ever encounter at the table in a life time of playing bridge.

[hanp]

Thanks for the info Wayne, very interesting simulation.

[Phil]

Board 2 from the USBC:

(essentially) 2N - 3N at both tables.

♠ATxx versus ♥QTxx.....

Kamil, Diamond, Etter and Katz all led the heart.

The spade was better and would have led to an easy set.

After the heart lead the contract should still go down (interesting problem really) but was let in both times in the Weinstein / Diamond match.