[jdonn]

I don't feel like analyzing 50 but the constraints still have problems.

2 responder doesn't have a slam try, slam is possible but you are short on values and can never get there accurately with xxxxx of trumps.

3 opener doesn't have a super-acceptance.

8 opener is too good to rebid 2NT.

11 opener doesn't have a super-acceptance.

I stopped after 12 (but down the page on 34 I see responder with a spade void, huh?) Plus it's not like opener bids slam 100% of the time over the 5♦ bid, he has already shown 23-24 and responder isn't forcing to slam so doesn't he ever sign off? That seems to have not even been considered, as though 5♦ is a transfer to 6♥. Anyway it seems to me the only thing you are really close to proving is that a diamond is the best non-spade lead.

[karlson]

Effervesce: So basically you think that 4♥ denies two fast losers in any other suit. Not sure I agree, but that's fine, we can include that. I also took out any east hands with a void. The rest of the hands, where you don't think north would bid 5d, seem to me to be judged quite subjectively. If you want to come up with some objective criteria, I'm happy to rerun it. But for now, more hands with your constraints here. Btw the reason that there were so many hands where opener had no diamond control is simply that in order to beat it, we need partner to have a trick somewhere, and usually that's in diamonds. He has a much better chance of having a trick there if opener has weak diamonds.

Jdonn: You can always pick a few hands and say that you wouldn't have bid them this way and declare the whole exercise moot. I think some hands are probably a little too weak, and some a little too strong, but on average they are ok. What I don't see is any reason why any of this biases against a spade lead. What I've observed is that if you just keep making the south (or north) hand stronger, what goes up or down quite a bit is the chance of you being able to beat the hand at all, but the relative frequencies of the successful lead stay quite similar.

What I also want to point out is that in all of the simulations, the probability of beating slam at all is quite low. We're only going to beat this if they've misjudged, or if the hands unfortunately don't fit well. So it's not particularly surprising that on the hands where slam is beatable at all, either opener or responder has somewhat marginal values.

Btw, why is a diamond the best non-spade? Seems to me that a club is the big winer. Also, why can't responder have a spade void if 5d is natural?

[jdeegan]

OleBerg, on Mar 7 2010, 08:17 AM, said:

effervesce, on Mar 7 2010, 02:03 PM, said:

 

Nice posts.

They might even help jdeegan to "to gauge the quality of commentators."

It really has. One reponder wondered about the meaning of the 5♦ bid. I specifically stated that it was a Q bid showing the ♦A or KQ. Remember, south was Jimmy Cayne and north was Chuck or Susie Berger. Personally, despite my liking for old-fashioned bidding, I favor Q bidding the unsupported ♦K if it looks to be the best bid, but I certainly wouldn't do it sitting opposite JEC, and neither, I expect, would the burgers. I pays to know your opponents.

The other issue I haven't heard mentioned is that more than 2/3 of the time a ♠ lead will give up either a trick (very often) or at least a tempo - and, you know this from the bidding - no simulations needed. When you start from that deep a hole, how good can a ♠ lead be? Just wait on your ♠ trick (you are behind the ♠A), and hope pard's 3-6 HCP turn into a trick. If declarer is missing the ♥Q with nine, he will normally misguess.

Time now to crank up my ancient MS DOS copy of RPG and do my own simulations.

[jdonn]

karlson, on Mar 7 2010, 05:56 PM, said:

Jdonn: You can always pick a few hands and say that you wouldn't have bid them this way and declare the whole exercise moot. I think some hands are probably a little too weak, and some a little too strong, but on average they are ok.

One point, it's a lot more than just a few hands. The person who analyzed them closely feels it is 59 of them. I would estimate I would have a problem with at least 40 of them. But I don't think the exercise is moot. I just think, with all due respect since you are doing something I'm not at all inclined to do, the conditions being run are simply not realistic.

Second point it really means nothing to say on average they are ok. Just like you and I are 6'2 on average but that won't help either one of us in any endeavor where it's an advantage to be exactly 6'2. On top of that, it's the ones where either hand is too weak that are throwing it off in particular, if one hand is somehow too strong then there is probably no hope to set almost always.

Quote

What I don't see is any reason why any of this biases against a spade lead. What I've observed is that if you just keep making the south (or north) hand stronger, what goes up or down quite a bit is the chance of you being able to beat the hand at all, but the relative frequencies of the successful lead stay quite similar.

I can only give my general feeling, but if the opponents are on the strong side it's very important to lead aggressively so that you can set up and/or take your tricks before they vanish. If they are on the weak side it's very important to lead passively because they may have no hope to make except to be given a trick on lead. So that one or both hands are too weak in so many examples definitely biases against a spade lead, I feel.

Quote

Btw, why is a diamond the best non-spade? Seems to me that a club is the big winer. Also, why can't responder have a spade void if 5d is natural?

I may have misjudged on the diamond it was only a glance, it seemed to me to set most frequently leading through dummy's diamonds. And if 5♦ is natural and says nothing about black suit controls (in which case opener won't use them as the decision maker to bid slam) then ok, maybe we just have the AK of spades to take! Or the spade ace could even be in dummy then...

[jdonn]

jdeegan, on Mar 7 2010, 07:05 PM, said:

The other issue I haven't heard mentioned is that more than 2/3 of the time a ♠ lead will give up either a trick (very often) or at least a tempo - and, you know this from the bidding - no simulations needed. When you start from that deep a hole, how good can a ♠ lead be?

It can still quite easily be best. It only matters if it gives up exactly the 12th trick, and if you can't get that trick back elsewhere. Your argument is like saying exercise isn't healthy because it dramatically increases your chances of a sprained ankle.

Quote

Just wait on your ♠ trick (you are behind the ♠A), and hope pard's 3-6 HCP turn into a trick.  If declarer is missing the ♥Q with nine, he will normally misguess.

I like how the spade is just 100% to score if we don't lead it. As though opponents in 6♥ and lacking the king of spades and queen of hearts won't have a minor or two to throw any spade losers on.

[karlson]

Effervesce: I did the stiff ♦ sim you asked for. Spade did better on this than on most other possibilities, but it was still about a 1-2.5 underdog.Hands.

Jdonn: average was not a good word, I meant "most". I tried another sim with both opener and responder having much better hands (10hcp/7 losers for responder, 9 controls for opener), but I didn't find a single hand in 10k where 6 was beatable (lol). I think at this point, I've tried it with opener on the weaker side, and on the stronger side, opener without all controls, opener with all controls, responder on the weaker side and the stronger side, 5d as cue, as length, and as shortness, and through all of that, it's never been close. If there was some magical middle where a spade was clearly best, I think we would have seen at least part of the effect somewhere.

Anyway, since no one else seems to care anymore, I'm done with these for now (unless someone has a specific request I guess, since it's easy).

[jdonn]

karlson, on Mar 8 2010, 02:34 PM, said:

If there was some magical middle where a spade was clearly best, I think we would have seen at least part of the effect somewhere.

Like this perhaps?

effervesce, on Mar 7 2010, 06:42 AM, said:

That is, only 17 hands of the 50 qualify. Of the 17, 8 only a spade will beat it, 6 any non-spade, one only a club lead, 2 any lead.

[jdeegan]

jdonn, on Mar 7 2010, 07:26 PM, said:

jdeegan, on Mar 7 2010, 07:05 PM, said:

The other issue I haven't heard mentioned is that more than 2/3 of the time a ♠ lead will give up either a trick (very often) or at least a tempo - and, you know this from the bidding - no simulations needed. When you start from that deep a hole, how good can a ♠ lead be?

It can still quite easily be best. It only matters if it gives up exactly the 12th trick, and if you can't get that trick back elsewhere. Your argument is like saying exercise isn't healthy because it dramatically increases your chances of a sprained ankle.

Quote

Just wait on your ♠ trick (you are behind the ♠A), and hope pard's 3-6 HCP turn into a trick.  If declarer is missing the ♥Q with nine, he will normally misguess.

I like how the spade is just 100% to score if we don't lead it. As though opponents in 6♥ and lacking the king of spades and queen of hearts won't have a minor or two to throw any spade losers on.

Your last statement addresses the whole point. When I started constructing hands for RHO where 3-4-2-4 distribution let two spades go on diamonds (remember, Cayne is very unlikely to open 2NT with two doubletons), I saw that this made LHO at most 2-2 in the black suits, so 6♥ makes even if pard has the ♠Q and a ♦ or ♥ trick whenever dummy gets a spade pitch on a club. Essentially, when we gamble on finding pard with ♠Q, we still don't win probably more than half the time. As far as waiting on the ♠K, if we don't make a ♠ trick, we probably aren't going to beat the hand anyway.