simulation request

[scoob]

1) can someone run the following simulation for me?

 

- 13-18hcp

- 5 card major

- 5332 / 5422 / 6322 shapes

 

a ) how often will this hand come up?

b ) out of those hands, how often will one (or both) of the doubletons contain neither an ace OR king?

 

2) can anyone recommend good simul software that can be purchased online?

 

thanks in advance!

[mrsock]

1) According to my simulation of 10,000,000 deals,

 

a) South will get a hand fitting your description on approximately 3.9% of deals.

 

:( South will have an A or K in one or both of the doubletons on 64% of the deals that fit your description*. So I think the probability you are interested in is 36%.

 

*(49% of the 5332 hands, 78% of the 5422 hands, and 80% of the 6322 hands)

 

 

 

2) get the dealer.exe program for free. It's quite simple to use.

 

See http://www.dombo.org/henk/dealer.html

[scoob]

thanks!

 

hmm, that's an even rarer occurence than i had imagined. which is good!

 

i have tried to use dealer but was unable to get it working correcty on my windows machine. i found an online interface to an older version that i've been using a little bit, but it's limited to 5000 deals i believe, and i'm not even sure how you get % results with it

 

i'd be interested to see the input coding you used if you happen to have it handy :)

[Guest Jlall]

49 % of 5332 hands with 13-18 HCP have Ax or Kx? does this surprise anyone else?

[Blofeld]

I should point out that mrsock has answered a question slightly different from the one posed. In the case of 5332s the answer is doubtless correct, but with two doubletons scoob asked how often at least one doubleton will have neither ace or king, and mrsock gave figures that say how often both doubletons will have neither (assuming that he means what he said).

 

Justin, that figure will presumably also include AK, AQ, AJ, KQ or KJ holdings. In particular, I don't know whether the simulator will count these at full face value, so you may find that hands with KJ tight are getting counted as 13HCP even if you wouldn't assign that value to them, and boosting the numbers for this sort of thing.

[MickyB]

A priori, a doubleton having A or K happens

 

2/13+(11/13)*2/12 = just under 0.30

 

So I'm surprised it is quite as large as 0.49 for 13-18 point hands. Instinctively it felt like it should be a fair bit less.

[luke warm]

i'm not a mathmatician, but intuitively it seems wrong that in a 5422 hand, within the hcp constraints posted, the aces and kings are as or more likely to be in the doubletons than the long suits

[scoob]

ok i finally got dealer working under windows (thanks cygwin!)

 

this is my input file:

 

n_hcp = (hcp(north) >= 13 && hcp(north) <= 18)
n_maj = (hearts(north) >= 5 || spades(north) >= 5)

s_dbl = (spades(north) == 2 && not hascard(north,AS) && not hascard(north,KS))
h_dbl = (hearts(north) == 2 && not hascard(north,AH) && not hascard(north,KH))
d_dbl = (diamonds(north) == 2 && not hascard(north,AD) && not hascard(north,KD))
c_dbl = (clubs(north) == 2 && not hascard(north,AC) && not hascard(north,KC))

n_ud = h_dbl || s_dbl || c_dbl || d_dbl
n_sv = (shape(north, any 0xxx + any 1xxx))

n_ubl = n_ud || n_sv

condition
 n_hcp && n_maj && n_ubl

generate
 1000000

action
 average "Opener" (hcp(north)),
 average "Responder" (hcp(south)),
 frequency (n_ud,0,1)

 

netting results like this:

 

Opener: 14.6311

Responder: 8.46342

Frequency :

    0      36177

    1      33718

Generated 1000000 hands

Produced 69895 hands

Initial random seed 1117826063

Time needed    2.525 sec

 

Opener: 14.6311

Responder: 8.46342

Frequency :

    0      36177

    1      33718

Generated 1000000 hands

Produced 69895 hands

Initial random seed 1117826063

Time needed    2.525 sec

 

Opener: 14.6311

Responder: 8.46342

Frequency :

    0      36177

    1      33718

Generated 1000000 hands

Produced 69895 hands

Initial random seed 1117826063

Time needed    2.525 sec

 

which seems to be inline with what mrsock came up with...

[Fluffy]

49 % of 5332 hands with 13-18 HCP have Ax or Kx? does this surprise anyone else?

Without the 13-18 HCP assertion a doubletion will have either Ace or King or both 1- (11/13*10/12) times wich my calculator says 29.5%, having 2-8 HCP above average doesn't seem worth extra 20%.

 

 

Woops! Micky already said so :).

[DenisO]

 

 

2) get the dealer.exe program for free.  It's quite simple to use.

 

See http://www.dombo.org/henk/dealer.html

This seems to be for Unix machines - Is there a version that works for Windows XP - without need for any specialised knowledge of compilers etc.

 

Denis

[DenisO]

Found it - ie the Paul Baxter version :)

[Zoltan]

Nice code scoob!

 

But...

 

n_hcp = (hcp(north) >= 13 && hcp(north) <= 18)

 

Why not

 

n_hcp = (hcp(north) > 12 && hcp(north) < 19)

 

?

 

:)

 

Zoltan

[Fluffy]

cos it is clearer, wich is a very important fact if your codeis gonna be used by other people.

[Zoltan]

Thx, was a little joke.

 

Btw, can I get some codes for dealer program? I'l be :)

 

You can find my registered email address in my profile.

 

Thx, nice people are on BBO and Forum!

 

Zoltan

[pigpenz]

Bridge Browser can let you look up million of hands played on Bridge Base or OK Bridge and I am not sure if you can enter in constraints or not....the best simulator I have ever seen is by Fred Gittelman and it was in Bridge BASE III, but he says he doesnt have any copies anymore, or atleast that he can get at very easily....Maybe if we all pull on his strings alittle bit we can get him to start selling it again.

[scoob]

i should have mentioned that i do in fact own a copy of BrBr already, and it's been endlessly useful to me. now that i've gotten henk's dealer working for me, i've all my needs covered :P

[vang]

another good and free dealer is

http://bridge.thomasoandrews.com/deal/

[inquiry]

i should have mentioned that i do in fact own a copy of BrBr already, and it's been endlessly useful to me. now that i've gotten henk's dealer working for me, i've all my needs covered :)

You most certainly can add constraints to bridgebrowser, but you will not necessarily get 1000's of hands that meet those constraints. So a good simulator still has uses. But Bridgebrowser does put you in the relm of real world hands and has a functional windows interface.

[scoob]

i should have mentioned that i do in fact own a copy of BrBr already, and it's been endlessly useful to me.  now that i've gotten henk's dealer working for me, i've all my needs covered :rolleyes:

You most certainly can add constraints to bridgebrowser, but you will not necessarily get 1000's of hands that meet those constraints. So a good simulator still has uses. But Bridgebrowser does put you in the relm of real world hands and has a functional windows interface.

this was the main reason i started looking for a simulator to accompany my copy of BrBr. also, i've been able to constrain much more complicated hands via dealer than i can with BrBr - but that could easily just be that i don't know how to untap BrBr's full power.

 

i intend to pick the author's brain thoroughly next week when i see him :blink: