semeai

Helene seems very nice, so I'm sure this is just an honest/humble post, but it comes off as an awesome thinly-veiled insult!

I don't see why. Any 4-6-0-3, 4-7-0-2, or 3-7-0-3 with ♠K, ♥AQJ, ♣A seems like an easy accept. Unfortunately (assuming you can set him in the problem) he's off 4 top tricks if he gives up the heart. For example Kx AQJxxx xx Axx. He needs the heart finesse.

Okay, my "certainly" was quite an overbid and some of the hands didn't make sense. It still seems to be a reasonable line with the singleton diamond hands, though. With Kxxx AQJxxx x Ax, declarer has several possible lines: 1) Play as he did, hoping the opponents don't lead diamonds and either the spade Ace is on or the opponents don't get to lead spades from that side 2) Cash two diamonds, pitching a club, and then lead a spade up, trying for a spade ruff. This works if the space Ace is onside, or if the opponents don't lead trumps twice if the spade Ace is offside, or if diamonds are 3-3 and the heart King is offside (dummy gets in with ♥10 to try the third diamond) 3) Try to cash three diamonds, then lead a spade up. This requires diamonds 3-3 and either the spade Ace onside or the spade Ace offside and hearts not xx-Kxx (or any 4-1 or 5-0 except stiff K) 4) Try to cash three diamonds, then lead a heart up. This requires diamonds 3-3 and the heart finesse on (or maybe some endplay chances) It seems to be between 1 & 2. Both work on this layout! There are variants with or without ♠J and ♥J. With Kx-AQJxxx-xx-Axx etc I will admit declarer would have to take the heart finesse.

It's easy enough to navigate jassem.pl with google translate. He doesn't have details about the system, but you can buy it as a (Polish) e-book. Google translate now works with pdf's. For example, he has a free excerpt on slam bidding from the WJ 2010 book here. Put it in as a webpage to google translate and it's mostly understandable. It loses the suit symbols, but you can follow along with the original Polish version for those. I didn't attempt buying the book and seeing if I could get google translate to work with it (I guess there would be multiple hurdles to overcome for that to be successful). One other somewhat useful thing is that you can take the WJ 2010 poll here. You have to sign up (standard sort of free sign up with e-mail deal) first, but then there's the full list of poll questions. It's in polish, but copying and pasting the text (not website as you need to be signed in) to google translate makes it mostly understandable. You can see how people voted and read comments after voting. Unfortunately, it seems you have to vote to see the poll numbers and comments. I only did one since I felt bad about skewing their poll.

Nice one. Of course, later in that 2+2 thread, Justin explains that he passes with a weak hand with no 5 card suit. Maybe we should split the difference and call it cards?

Some good reasons, thanks. Justin & Joe do have listed what Wyman's quote describes (where is this from?). The other two pairs play multi Landy. I was curious why this and not Woolsey. The bit about declaring instead of defending with your own good hand opposite a weak hand with 5 cards is good, and then there's the more obvious double & bid your major type hands. It would be nice to check on frequency of this versus frequency of 4M-5C (plus how much better it is to act instead of not acting with these two types, of course).

Bridgewinners posted the convention cards of USA 2, and I notice that all three pairs play penalty doubles of strong (or weak) 1NT in direct seat. I understand this is at imps so it's more attractive than it would be at matchpoints, but it still seems surprising and very neoclassical. Is there new thinking on this? Why use penalty doubles of strong 1NT? For comparison, none of USA 1 use penalty doubles of strong 1NT.

Thanks, but actually I still don't know what that is. Is what I added above (simultaneous to when you were replying) correct?

What is this? I know what a jettison squeeze is, but what does it mean when you spell it backwards? Added: Does it mean Ax opposite K instead of (for jettison) Kx opposite A? (Sorry for being off-topic.)

The bridgematters.com Countering Vexing Bids page seems pretty good for simple generic defenses. A 2H bid weak in one major was one of the examples. This isn't exactly that, but it's very close. This is basically what Free says (treat it as hearts) but with double being "PTO," i.e. passable takeout (no void or nice distribution with a singleton; see the link for more & follow-ups). Then 3C contains the hands you don't want partner to pass on but which are not too strong (and 4C contains stronger ones). I suppose the logic is that partner should pass with various hands with, say, 4 decent hearts to expose opener's hand, but that you don't want to be doing that opposite a void (or certain hands with a singleton). Added: I see Glen posted about this in the thread gwnn links to.

4♥ is wide ranging. and an immediate 5♥ would likely invite 6 in some specific manner (no preempting a preempt), so I don't see what this player not bidding 5 immediately has to do with anything. There are situations where that sort of logic applies, but I don't think this is one. 5♥ seems clear to me, with 8 trumps, a void, and no spades. It's definitely a two-way shot, and we could even have both making. (Of course, it's not what I think but what a poll of peers would return.)

To clarify this further: I believe in the idea that in addition to depending on the player, each auction also has its own expected tempo. In this auction some consideration on dealer's part is expected (a fast pass would be out of tempo). Dealer gets whatever time is expected (for him!) due to that, but no more. As I said in my other post, I don't really know anything good to do about/to the fast bidders.

I think this hand is an okay 5H bid even after the hesitation. As for the second scenario, I agree with gwnn. If it was only out of tempo with respect to the other players' immediate calls but not for dealer's normal tempo, then there's no BIT. If it is out of tempo for dealer, I don't think he can claim to be owed the 10 second pauses that the other players didn't use, and so it would be a BIT. Maybe the players who bid immediately can be given procedural penalties, but I think it would be unusual to do so.

You left out the underlined bit, which caused me to reread the sentence an embarrassing number of times. It is a tough decision between pass and 3NT over a 3♠ balance. KJ-sixth plus a side card is likely a trick short of 3NT, but partner would probably stay out of things with that. KJ-seventh plus a side card probably makes for a good 3NT (depending on the side card). J-seventh plus two side cards is also possible, and wouldn't be so good for anything above 3♠. Added: What about 2♥-P-3♥-P;P-4♠? Certainly sounds funny, at least. Also, as you said for the direct version, kind of random.

I just get to 4♠, by a texas transfer if you play them or otherwise by 3♥ and then 4♠. I'm not interested in looking for a diamond fit for a perfecto slam. Even looking for a spade slam is too rich. If I had a way to show a club splinter below 4♠ (not a diamond suit!) I would at least think about it but then still decide not to invite slam. I wouldn't be happy if I held a 1-count with 5 spades and partner bid 3NT over my transfer just to show a doubleton. I usually play any transfer break as a superaccept. I could see allowing 3NT as "I have a doubleton spade and a potentially running 6 card minor" instead if you want to have that agreement, but I really wouldn't want it to be bid on just any hand with a doubleton spade.

If you open with all balanced hands with no 5M, half of 5M332's, half of 5m422's, and no 6m322's, I get 54.8%. As Free points out it's really the 6m322's that hurt the percentage, so you're just getting some sort of average of the first two in my list based on opening 5M332's half the time, say. In terms of ending up in bad diamond fits with Stayman, using the above model for a 1NT opener (all balanced with no 5M, half of 5M332's, half of 5m422's, and no 6m322's) I get you ending in a 4-2 diamond fit 5.1% of the time, a 4-3 diamond fit 17.5% of the time, a 4-4 diamond fit 16.0% of the time, and a 4-5 diamond fit 6.7% of the time. (Of course these add up, other than an error of 0.1 due to rounding, to 100-54.8.) So in terms of what to do over 1N-2C;2D, if you pass you'll have the above percentages. If you bid 2H, you'll get to a 4-2 fit the 2.6% of the time that partner had 2-2-5-4 or 2-2-4-5, and the remaining 42.6% of the time you'll get to a 4-3 fit.

This is probably better suited to a simulation, but I tried it analytically but ignoring the confounding factor of hcp. If you open any balanced hand including 5M332 and neither 5m422 nor 6m322, I get 56.9%. If you open with any balanced hand including 5M332 plus any 5m422 and any 6m322, I get 53.2% If you open any balanced hand except never 5M-332 (and also neither 5m422 nor 6m322), I get 51.3%. If you open any balanced hand except never 5M-332 but do open all 5m422 and all 6m322, I get 48.4% (Before posting, I double checked and found I'd left off 3-3-4-3. I think I've used every distribution now, though.)

[quote name='[i]someName'[/i]]theQuote[/quote] is what did that, or the same with b instead of i for bold

And now they're all in italics too?

You don't have to do it based on how you pick up your cards, which could be influenced by the previous table's shuffling or lack thereof. You can shuffle them before you look at them. Of course if you're doing this every board (so as not to give away when your hand requires a random number to help decide your bid) it might get tiresome.

Added spoiler tags. Hadn't noticed this was in B/I.

We're discussing two types of randomization methods here: 1) Those due to shuffling your hand. For example, gombo121's first mention of this with red = 0, black = 1 (which suffered from being correlated with hand type), my suggestion using an absolute order of the cards to read off a permutation of the 13 cards, and hrothgar's suggestion of using transitions to get bits. The latter two have no correlation with hand type/strength. (I happen to prefer mine, as it contains much more info as well as being better suited to getting probabilities with denominators not a power of two, but the bits from transitions should be okay too. Really it's just a special case that records a manageable amount of data.) 2) Deterministic ones based on the cards you hold, such as gwnn's suggestion of adding up the spots and taking the remainder modulo something and lamford's suggestion of using whether you hold a 2, whether you hold a 3, etc. Is there a regulatory issue with either one? It seems unlikely, but (1) could be ruled as using an outside aid (though it seems just to use the cards) or as being impossible to disclose (though I think it should be fine just to tell what the method is). It also seems unlikely, but (2) could be disallowed for being encrypted (though partner doesn't have the key). If there's no regulatory issue, I don't know why you wouldn't use one of the uncorrelated methods from (1). Also, methods that use shuffling but have a correlation with hand type/strength seem strictly worse than either uncorrelated ones from (1) or (minimally) correlated ones from (2).

Partner's 2D bid was completely artificial.

For humans, this bidding sequence would show maybe a good 5 to mediocre 6 card heart suit, with some tolerance for another strain, whereas a direct 4H would be a hand more certain hearts is the correct strain. On the above hand I'd definitely bid 4H directly.