[TimG]

inquiry, on Jun 20 2008, 07:46 PM, said:

[*]South might should pass redouble to hope to find best spot, but this partner will take that as to play

I don't think passing the redouble is to play. But, this is a situation where special agreements are useful in landing the intervening side in their longest fit.

[sfbp]

huh?

[inquiry]

sfbp, on Jun 21 2008, 07:40 PM, said:

NickRW, on Jun 21 2008, 05:33 PM, said:

Sometimes I do a crossword puzzle.  One of those ones that has normal and cryptic clues.  I get the answer off the normal clue and then I look at the cryptic clue - and I still don't get it even when I know the answer.

I would sure like to see why NOT taking the first ♥ with the Ace gets that contract down 1.

it allows two trump entries to south hand so that the SPADE TEN cover king, ACE, can be followed up twice with low spades towards the QJ6 through the 875... QED. If you duck, any spade play (TEN COVER) allows the heart ace to be won and partner give a ruff.

[pigpenz]

Jlall, on Jun 21 2008, 10:18 PM, said:

han, on Jun 21 2008, 05:09 PM, said:

I ran a simulation and double by north is a clear winner.

i LOLed

yes after seeing this i ran a simulation with about 1500 hands then used DF analysis on them.
2♣ 27%
2♦69%
2♥69%
1NT 65%

so double does have enough merit to it, though maybe its not evident.

[han]

It was a JOKE! It is virtually impossible to say anything about the merits of a double by running a simulation. A double dummy analysis tells you nothing because you need to take into account how the auction will develop if you double and how the auction will develop if you don't. Also, if your partner knows you double with such shapes then your results on other hands where you double (and where you do have a normal TO double) will also change. Anybody who thinks their simple computer analysis tells them anything about the merits of a double here should refrain from posting on such topics because they simply have no clue. In my opinion.

[helene_t]

I think you exaggerate, Han. I can imagine dealing 1000 deals with the actual North hand and the West hand consistent with a standard 1♠ opening, and then let the Jack software (or some such) play out the deal after North having passed, doubled or bid 1NT respectively. Of course it wouldn't settle the issue but I wouldn't dismiss such a result as meaningless either. The overall results may be artifacts of the software's idiosyncrasies, but maybe the detailed results could reveal something interesting. Or maybe not.

Btw I am not sure if the issue is whether it is a good idea to agree to dbl with this kind of hands, or whether is is a good idea to double with this hand assuming some "standard" meaning of that double. In the latter case the answer probably is that one should adhere to whatever agreements one has, but suppose this hand is not covered by any agreement, and we have to chose the smallest lie.

[sfbp]

han, on Jun 23 2008, 04:08 AM, said:

It is virtually impossible to say anything about the merits of a double by running a simulation.

Isn't that the concept at the heart of my software?

[matmat]

sfbp, on Jun 23 2008, 08:56 AM, said:

han, on Jun 23 2008, 04:08 AM, said:

It is virtually impossible to say anything about the merits of a double by running a simulation.

Isn't that the concept at the heart of my software?

your software runs simulations?

[sfbp]

matmat, on Jun 23 2008, 08:03 AM, said:

sfbp, on Jun 23 2008, 08:56 AM, said:

han, on Jun 23 2008, 04:08 AM, said:

It is virtually impossible to say anything about the merits of a double by running a simulation.

Isn't that the concept at the heart of my software?

your software runs simulations?

More irony?

Where a comparison can be made between double dummy/ simulation and real results, it's been shown that there is a very good correlation.

There are many situations where simulation doesn't work, or cannot be done. That's exactly why determining the outcome of different actions in a given situation can be quite helpful to choosing an action. Measuring and comparing real outcomes is what the software does. Some people (here) apparently prefer to measure/experiment by playing one hand at a time, and wholly reject the concept of developing rules. They would rather spend 20 years becoming pro's than looking at the results of other people's actions, I think.

I wonder whether, given your allegiances, you needed me to tell you this.....

[matmat]

sfbp, on Jun 23 2008, 10:30 AM, said:

matmat, on Jun 23 2008, 08:03 AM, said:

sfbp, on Jun 23 2008, 08:56 AM, said:

han, on Jun 23 2008, 04:08 AM, said:

It is virtually impossible to say anything about the merits of a double by running a simulation.

Isn't that the concept at the heart of my software?

your software runs simulations?

More irony?

what do you think?

there is a substantial difference between compiling a set of data (which your software does do) and running a simulation, which I don't believe it does.

[sfbp]

I will restate using as close to the given wording as possible.

The concept of my software is that given "It is virtually impossible to say anything about the merits of a double by running a simulation." it is however possible to measure outcomes of a given action.

Your sarcasm is not welcome.

You're right in one way: the software doesn't make conclusions. It allows the user to draw them. If you're not motivated to do that, nothing will help.

[matmat]

sfbp, on Jun 23 2008, 10:57 AM, said:

Your sarcasm is not welcome.

That's OK. your shameless self promotion isn't welcome either.

[sfbp]

matmat, on Jun 23 2008, 09:07 AM, said:

sfbp, on Jun 23 2008, 10:57 AM, said:

Your sarcasm is not welcome.

That's OK. your shameless self promotion isn't welcome either.

Classic. When you run out of arguments, make the attack personal.

[matmat]

sfbp, on Jun 23 2008, 11:11 AM, said:

Classic. When you run out of arguments, make the attack personal.

look, if you charge for software, you should have to pay for the advertising. inserting little snippets plugging your work in a "matter of fact way" should not be acceptable.

[sfbp]

matmat, on Jun 23 2008, 09:13 AM, said:

sfbp, on Jun 23 2008, 11:11 AM, said:

Classic. When you run out of arguments, make the attack personal.

look, if you charge for software, you should have to pay for the advertising. inserting little snippets plugging your work in a "matter of fact way" should not be acceptable.

Interesting point. I've never charged for the software. So if we are discussing software that's free, I guess I don't violate your "shameless" rule.

[TimG]

sfbp, on Jun 23 2008, 10:30 AM, said:

That's exactly why determining the outcome of different actions in a given situation can be quite helpful to choosing an action.

Suppose we look at all hands where dealer opens 1S and 2nd hand doubles with 4441 shape (within some point range). We have no way of knowing whether these doubles were by partnership agreement or were an improvisation. We don't really want to lump those two situations together when analyzing the results. Further, when trying to determine the merits of doubling with this shape/count combination, we need to also determine how this agreement affects other auctions -- both those that start with double and those that don't.

I'm not suggesting that a double-dummy simulation can address these points. But, I don't think this is one of those situations where looking at actual results will be particularly helpful.

[TimG]

sfbp, on Jun 23 2008, 11:25 AM, said:

Interesting point. I've never charged for the software. So if we are discussing software that's free, I guess I don't violate your "shameless" rule.

Is this just a technicality in that you charge for the data, but not the software?

[sfbp]

TimG, on Jun 23 2008, 10:16 AM, said:

sfbp, on Jun 23 2008, 11:25 AM, said:

Interesting point. I've never charged for the software. So if we are discussing software that's free, I guess I don't violate your "shameless" rule.

Is this just a technicality in that you charge for the data, but not the software?

No. You are welcome (and it is possible) to play around with the software with various small, free datasets.

Presumably you could figure out how to make your own, but the software is not designed to support that endeavour.

[beowulf]

I'm surprised that nobody has seen fit to mention the principle of equal level conversion in regard to this hand. If you and your partner have an agreement about ELC (which needs to be declared of course) then this hand can make a take-out double and correct 2C to 2D showing a red two-suiter. The 15 hcp somewhat make up for the poor shape. Without such an agreement, the double is a calculated risk. At white on red, it probably won't cost a lot -- if you partner has a good hand and clubs then 3NT may be on.

I don't think it is "horrendous", although I wouldn't have doubled with that hand myself (unless I had an ELC agreement).

The rest of the carpings about software, simulations, etc. are very hard to follow. Plain statements of opinion would be nice -- instead of sarcasm, sarcastic answers to supposed sarcasm, etc.