jdonn, on May 5 2008, 07:55 PM, said:
I echo his comment, you can have your sims and I'll take stayman thanks. Seems more useful to find out if there is a major suit fit instead of calculating the odds of there being one when you are making decisions.
Your arguments about upgrading the hands are true as general statements but where is the cutoff? What if slam was making on a point less, is inviting with 9 and a bunch of tens ok, or accepting an invite (showing a max!) with 18 and a 6 card suit? I still find your auction implausible, neither player had any idea there could be a 9 card fit or that 100% of the high cards would be needed for slam to be so good (yes, every single one was working overtime). I mean no one has shown a single suit, could your north not be 4414 with 11?
I bet there might be a lot of different auctions presented if north had xx KQx in the minors instead of ATx xx.
I really do not understand this view.
I assume you play some sort of standard Stayman. You can have it. I concede it is a reasonable approach and might even be better than Puppet Stayman - I don't really know the answer to this and I suspect that you do not either. It certainly will be better for some hands and an alternative method may be better on other specific hands.
I don't have my eyes closed on the value of simulations and their limitations. However they have some major advantages over thought experiments about how the hand will play opposite some few example hands chosen by a biased analyst. Not least of which is that a 1000 or more hands chosen at random will on average be better (more representative) than one or a few hands chosen by the biased analyst.
In the end I am happy to strive towards bidding slams that are likely to be significantly better than 50% based on a large simulation.
In this case the example that you gave xx KQx in the minors is hardly representative. I suppose given our six diamonds and two clubs that 2=3 is more likely than 3=2 . However xx is much less likely than honour-x. Given we have KQ98xx there are four small cards and three honours (10 included which improves our chances in the suit). There are 4C2 = 6 ways of choosing two small cards out of 7C2 = 21 doubletons. It is much more likely that we have an honour (or two). And this analysis does not take into account that partner has shown more than a fair share of the outstanding high cards increasing the liklihood that the diamond holding is honour-small rather than two small. On average partner will have slightly more than two diamonds. In my simulations two small diamonds were held somewhere between 5% and 10%.
Its not surprising that when you choose a poor holding in your example for a marginal slam that slam will no longer be good. In making our bidding judgements it would be wrong to assume partner has the worst possible holding. A better example would be a holding like Jx and maybe the reality is that an average holding in this case is more like 1/4 Ax and 3/4 Jx or similar.
I am sure you would make the same sort of argument if I justified the bid based on partner holding an unlikely three or four diamonds. A simulation on the other hand takes account of both possibilities in approximately their relative frequencies.
Yes 4=4=1=4 is possible in our system. Even 4=4=0=5 might be possible although partner might choose something else with that hand. Somewhere between 25% and 30% of hands had one or no diamonds. Even so slam was still reasonable on average (double dummy). I am sure if our methods were better suited to this particular hand so that 4NT guaranteed a "balanced" hand then slam would be an even better proposition.
The cut-off for a bidding decision is where we on average get a better score by making the decision. In my experience it is best not to be prejudiced in where this cut-off is.
Help
. Yeah I think that's what I meant to say.
Suggest 1♦ - 1♥
2♣ - 3NT
4♦ - 4♥
6♦ - pass