Re: Invite or not?

On Nov 14, 8:42 am, ewleong...@xxxxxxxxxxx wrote:
On Nov 13, 11:09 am, henrysun...@xxxxxxxxx wrote:



On Nov 11, 5:06 pm, "Bud H" <budh9534REMOVES...@xxxxxxxxxxx> wrote:

K10
K7x
107xx
Q8xx

Partner non-vul at imps opens a 15-17 1NT. Do you pass or invite? How
about if vulnerable?

Bud H

In an attempt to provide some data, I ran two separate 1000 hand
simulations for the auction 1NT-2NT with this hand. The first gave
opener a 17 point balanced hand (i.e., a hand that would always accept
the invitation) and the second gave opener a 15 point hand (i.e., a
hand that would usually decline). I was interested in seeing the
total number of tricks taken over the two samples and to see how often
2NT went down when opener had a minimum and how often 3NT made when
opener had a maximum.

Here are the results.

First, when opener has a balanced 17 count:

<7 tricks = 28
7 = 91
8 = 328
9 = 420
10 = 118
11 = 15
12 = 0
13 = 0
total number of tricks = 8549

Second, when opener has a balanced 15 count:

<7 tricks = 146
7 = 279
8 = 393
9 = 166
10 = 16
11 = 0
12 = 0
13 = 0
total number of tricks = 7581

So by these figures, when opener is maximum 3NT can be considered to
be makeable around 50% of the time (553 successes), depending on
margin of error. When opener is minimum, 2NT is going down about 40%
of the time (425 failures), again depending on margin of error.

I interpret the simulation as showing that this is, in fact, a pretty
close decision, and not clear cut one way or the other. Perhaps the
best way to resolve it is to invite when red when the gain is greater
and not when white.

Henrysun909

Is this generousity to bunnies week?

If the worst that can happen is the situation is break even with the
assumption that your opponents are finding the double dummy lead and
double dummy defense

and that you are finding double-dummy play.


Also, you have
a chance to play skillfully to enhance your chances of making game
when the opponent make a slight mistake.

I see. So you think a human will play better than a double-dummy
solver?


One would think if your goal
is to be a winner you welcome the opportunity to give your opponents
every chance to make their God given right to make a mistake and
maximize your reward when they do make a mistake.

About the only situation where I can see where you wouldn't invite is
where you are playing against two world class opponents and your
partner as a card player is a cripple..

The simulation suggests that when NV there is very little wrong with
inviting. Of course there is very little right with it either. If the
simulation is correct then when NV passing and inviting will produce
almost exactly the same number of IMPs in the long run. So if you want
to invite, no one can criticize. Similarly if someone else does not,
there is no reason to criticize.

There is one factor which affects the decision worth considering. If
you are the type who stretches 1NT openings:
* down to 14s, then passing becomes clearly right.
* up to 18s, then inviting becomes right.


Andrew

.

Relevant Pages

  • Re: Stupid polarized capacitor tricks
    ... Fields simple analysis of a rectifier-capacitor filter. ... I make a living out of fixing messes created by people with simulation software and little or no understanding. ... but you consistently waffle, and essentially never back it up with theory - ascii schematics, maths etc. ... I would rather find a mistake of my own that I made and not pound ...
    (sci.electronics.design)
  • Re: Say you are asked to work on the "Deep Design" project
    ... It's exactly the same as perfect double-dummy play. ... because it doesn't investigate the results of opponents' losing ... but even 1 simulation at 1 depth (the least work you ...
    (rec.games.bridge)
  • Re: Statistical Pattern Recognition
    ... also tests exact hypotheses. ... do the simulation themselves. ... William Calvin, being a neurophysiologist, is automatically a "Good ... small mistake that is forgivable or not being found out, ...
    (talk.origins)
  • Re: How common is rolled up vs. rolled up?
    ... To restate, assuming that you are rolled up, the odds that one or more of your seven opponents is also rolled up is 54.3:1 against. ... The simulation assumes that the upcards of the opponents are unseen. ...
    (rec.gambling.poker)
  • Re: Pre-empted again
    ... I would not bid 5s - little to gain and a lot to lose. ... sacrifice by the opponents. ... Assuming unknown opponents (i.e., at a tournament of some kind and ... Perhaps you would be kind enough to run a simulation? ...
    (rec.games.bridge)
Loading