[Free]

whereagles, on Apr 14 2008, 03:11 PM, said:

Dude, you can get as picky as you want, but it won't change the fact that the true results probably won't differ much from the bugged ones, which took like 1 minute to code and run. You probably heard of Pareto's law, no?

First you need to invest 20% of the required time ofcourse

I think if you exclude 5332 distrubitions (and perhaps limit the number of ♥s to 0-4), you'll get a pretty accurate result

[TimG]

hrothgar, on Apr 14 2008, 08:44 AM, said:

Oh yes, from what i can tell your simulation doesn't permit a 1♦ opening on hands like

♠ K653
♥ K73
♦ A763
♣ K8

If I opened that hand 1♦, I think I'd raise partner's spades rather than rebid 2♦. Of course, I would also raise to 2♠ holding three spades with some frequency. And, rebid 1NT with many 5332 hands (especially the ones where my suit isn't so good as hinted by 2nd hand's trap pass).

I agree with Richard that it would be better to give at least a brief summary of the conditions you used.

[helene_t]

Nuno, I think you should set diamonds to 6+. Almost all hands with only 5 diamonds would bid differently. Maybe not Kx-xxxx-AKQJx-xx but you could ignore that exception.

[cherdano]

Whereagles, I don't mean this as a flamebait but your assumptions are so far removed from what I would consider the correct assumptions your results do seem worthless to me. You should really include "unbalanced" as a condition, and for example for me a 2♦ rebid almost never has less than 6 diamonds.

[hrothgar]

whereagles, on Apr 14 2008, 05:11 PM, said:

Dude, you can get as picky as you want, but it won't change the fact that the true results probably won't differ much from the bugged ones, which took like 1 minute to code and run. You probably heard of Pareto's law, no?

I'm well aware of Pareto's Law. However, in this case it gets trumped by one of the basic rules of computing. "Garbage In, Garbage Out"

I'm perfectly willing to believe that the code in question took you less than a minute to write. However, I'd hardly brag about this if you're trying to convince people that they should have any faith in the accuracy of your results or pay attention to your simulations.

Take a close look at what your simulation gets wrong:

1. Your simulation fails to exclude NT openers from your 1D opening

2. Your simulation doesn't exclude any hands with Hearts > Diamonds from your NT opening.

3. Your simulation doesn't recognize that there are hands with 4-4 in the minors that will get get opened 1D (thankfully these won't rebid 2♦)

4. Your simulation doesn't recognize that there are lots of hands that will rebid 1NT rather than 2♦ on balanced hands

Its hard to place much faith in a simulation that misbids on both balanced and unbalanced hands.

Given that you appear pretty new to this all, here's a bit of advice on methodology: Code reuse is your friend

Invest some real time and effort developing a set of scripts that model a bidding system.

Create a function that describes a 1♣ opening.
Create another function that describes a 1♦ opening
Work your way up (or down) to 3m or so...

If your feeling ambitious, you might even start working on first round responses to thevarious openings.

At this point in time, if you want to run some simulation you can slap together some code that is both quick to assemble and - get this - accurate...

[whereagles]

helene_t, on Apr 14 2008, 03:02 PM, said:

Nuno, I think you should set diamonds to 6+. Almost all hands with only 5 diamonds would bid differently. Maybe not Kx-xxxx-AKQJx-xx but you could ignore that exception.

You think so? Well, I'm used to rebid diamonds if 4♥-5♦ not strong enough to reverse, but ok. Setting diamonds to 6 (and dropping 5+ hearts and 5♦332 15 hcp hands from the simulation) yields

spades: 1.9778
hearts: 2.58765
diamonds: 6.35535
clubs: 2.0792
hcp: 12.755
hcp in diamonds: 6.03114
Generated 10000000 hands
Produced 408072 hands

Seems like pass is the winning action, despite lurking bad breaks. Even more if you usually support with 3 spades, of course.

[whereagles]

cherdano, on Apr 14 2008, 03:13 PM, said:

Whereagles, I don't mean this as a flamebait but your assumptions are so far removed from what I would consider the correct assumptions your results do seem worthless to me. You should really include "unbalanced" as a condition, and for example for me a 2♦ rebid almost never has less than 6 diamonds.

Roger. See above.

[whereagles]

hrothgar, on Apr 14 2008, 03:18 PM, said:

Invest some real time and effort developing a set of scripts that model a bidding system.

That's all fine.. if I had time. It's rather obvious I'd be paying more attention to stuff if I had time, don't you think?

[dcvetkov]

I would pass

If we try to reconstruct ppps hands, doubles probably is 4-4-1-4, or 4-4-0-5, and his partner decided to take his chance defending. (I would be very surpirsed if LHO has a trumps truck in front of the 2♦ bidder?) They probably judged well, depends who has the better spots cards in trumps, but anything else, tends to be worse, and of course you will play doubled.

[TimG]

dcvetkov, on Apr 14 2008, 11:54 AM, said:

I would pass

If we try to reconstruct ppps hands, doubles probably is 4-4-1-4, or 4-4-0-5, and his partner decided to take his chance defending. (I would be very surpirsed if LHO has a trumps truck in front of the 2♦ bidder?) They probably judged well, depends who has the better spots cards in trumps, but anything else, tends to be worse, and of course you will play doubled.

Of course LHO doesn't have a diamond stack -- he's the one that made the balancing takeout double. Couldn't he be 3424 or 3415 along with your perfect 4414 and 4405 guesses? RHO has diamonds (and probably the best hand at the table).

[Cascade]

cherdano, on Apr 13 2008, 06:15 PM, said:

SOS redoubles at the 2-level are overrated.

I agree. I don't think I have ever had the occasion to use one.

[Cascade]

whereagles, on Apr 15 2008, 12:44 AM, said:

Yes, there was a bug in the code about hearts and clubs. Here're the new figures.

spades: 2.09555
hearts: 3.1117
diamonds: 5.60782
clubs: 2.18493
hcp: 12.7525
hcp in diamonds: 5.38587
Generated 10000000 hands
Produced 942490 hands

So most likely shape is now 2362 and this argues for a pass.

Except that the opponents are much more likely to have the trumps to double you with when partner is some sort of 4=5 in the reds.