kfgauss

I agree with this basically (though I'm not certain what the precise ruling/result should be). One should further note that North's double is quite reasonable, as South seems to have a really big hand: he doesn't have running diamonds, and he really can't have running clubs when East appears to be 5-6. East hasn't yet shown that he'd bid 5♠ over 5♣, and we can allow South to bid 5♣ directly over 1♣. I do think that 5♠ over 5♣ will probably be allowed then, but West, even though he competed to 5♠, has no ability to claim that he'd bid 6♠, so an adjustment to 5♠ seems warranted. Andy

Mid-range (16-18) point hands like this are much more frequent than the GF ones (and I think the game/partscore decision on those is more important than the distinction you're making here). 2NT isn't forcing in sayc (remember that you just told us you don't know sayc). --------- As for the actual hand, 3NT does (by inference) show a solid 6+ card suit (as mike suggested) in sayc. I'm worried about the diamonds, though (perhaps I shouldn't be as much as I am). I'd like to try 3C, but at the table I might chicken out and bid 4H or 3NT. Andy

I agree. Too bad we don't yet have a quality single dummy hand evaluator (read: a solution to the play of the hand at bridge :P), though. All this DF analysis will lead us to make our bidding systems treat A10x opposite KJx as 3 tricks. (Yeah, yeah, it's a very small effect -- I'm mostly joking.) Andy

Yes, my chat message went to a real user too. Andy

I sent a chat message to someone just as they were leaving and it went to someone totally random (I hadn't seen their name before and their name wasn't anywhere on my screen beforehand). Wouldn't know how to attempt to reproduce this (it's just happened once). I'm using version 4.4.1. Andy

Thanks for the response, Ben. This is a nice example of what to do if you're going to give up 2NT -- of course I'm still looking for the potentially impossible combination of our two methods (and while we're dreaming, we might as well hope for actual short suit & long suit game tries too). Defining more precisely what happens after certain 3X bids in my methods (and/or making some of those artificial to some extent) seems promising, but I haven't done this in a satisfactory way. That said, I found this quite amusing. Couldn't quite tell whether inquiry just meant "ask" or whether you'd named it after yourself :P. Andy

I haven't played this style seriously. I just recently played with a partner who likes rebidding 1N on 1444's too (as opposed to playing that lovely 4-2 diamond fit). Does anyone have recommended response schemes here -- e.g. ways one should modify 2-way checkback to account for these possibilities (I haven't thought about it much, so any observations would be useful)? ---------- As for the original post, I tend to open 1C on these hands and rebid 2C when partner bids my stiff with most partners (the ones with whom I don't rebid 1N with 1444). I also basically always raise if pard bids my 3 card suit (there's a discussion of this going on in this same forum right now). If I have 10xxxx or something, I might open 1D, but the club suit would have to be pretty bad. Andy

With one partner, where we raise on 3 cards basically anytime we have a worthless doubleton and a balanced hand (maybe the 3 card suit needs an honor), in addition to various unbal hands (though if we like our rebid we usually won't -- e.g. if we can rebid a nice 6 card suit -- but note that 3154's without extras almost always raise), after 1C-1H type auctions (except when we have a spade suit to bid), and we use the following rebid scheme: 1m-1M;2M: 2N = inv NF 3m = inv NF (4+m, exactly 4M) 3X = inv forcing, usually exactly 4M (often just a stopper) 3M = invite almost always with 5 and we tend to bid 3N on most GF hands with 4 cards in the suit when we've got all suits vaguely stopped (we'd bid 3X in some suit we've got stopped if we're worried about some particular suit) I'm not so happy with this setup, but it works reasonably and we haven't come up with anything we like better. We don't really want to give up the ability to play in 2N. The game tries are rather unspecific, but that seems to be the trade-off. Note that we like matchpoints and so I'd rather not hear disparaging comments towards 2N contracts :). I suppose I'd be willing to hear an "imps only" response structure, but I'm not overly convinced that one should just toss away these partscores at imps. If you're wondering why we bid like this when we then try so hard to wriggle out of playing a 4-3, it's because we're happy with the sort of 4-3 fits we get to when playing only 2M and because we like the way it helps out other rebids by opener. Does anyone have suggestions -- I'm willing to hear your thoughts on why I should give up the ability to play one of the partscores (as well as rebid schemes you prefer that do this), but I'd be very interested in ideas that allow me to keep the ability to play 2N/3m. If you care, this is in the context of a 2/1-type system. Andy

If you assume all small-card leads vs NT are from 4 cards (ie if you assume that West has 4 clubs), then you always make (with his line). Indeed, his analysis begins "[t]he contract is assured" and the fact that you duck was part of the setup of the problem. Also, Fluffy, he doesn't exclude Whereagles line because "it's a squeeze book" -- indeed he makes a point of telling you when you shouldn't be squeezing and has a few hands sprinkled throughout to keep you on your toes -- it's just because he makes this "textbook" assumption. Let's analyze this, though (assuming diamonds break): by ducking the club, you'll make when (you have a choice between 1 & 2 corresponding to whether, when you test one suit and find east stopping it, you guess the other suit is 3-3 or 4-2): 1) clubs 3-3, hearts 3-3, west has both clubs & hearts 2) hearts 3-3, west has both clubs & hearts, east has hearts & west has clubs [could swap hearts & clubs here, but let's ignore the lead inferences for the moment] By winning the first club and playing a spade to the Queen, you'll make when (should prolly rattle off 4 diamonds before leading a spade to the Q, but this has nothing to do with the double dummy odds): 3) West has Spade K, Hearts 3-3, West has Spade J & stops Hearts (and then minor chances like either hand having 5 clubs and stopping hearts, which we haven't ruled out because we didn't duck the club -- though I guess this means West led small from J10973 or led a stiff. Also, this line has better chances when diamonds don't split than a ducking the club line.) [note that it's technically an equivalent line, but I suppose you should finesse the 10 and thus swap SK, SJ in the above because occasionally an opponent will hop SK] I haven't done the math, but it seems pretty clear that (3) is the winner, so nice catch (and it involves a squeeze as one of the chances :)). Andy

This sort of reasoning always strikes me as sort of funny. Perhaps it is in fact true that good players don't lead singletons against slams when they have a side ace, but shouldn't they, assuming they have no lead that is actually "good"? This would confuse the issue of who has the Ace when a short suit is led (or if the rest of your peers are not leading stiffs when they have Aces, it'll lead the opps to guess wrong when you do). Maybe the point is that the difference between leading the singleton and leading something else when you don't expect to get a ruff is larger than the confusion you could cause. Is this true (let's assume you have no other lead that "suggests itself" particularly)? Because you're finessing partner probably and/or giving away distrib info (regarding your stiff)?

Sorry for unearthing an ancient thread :), but I was rereading this chapter in Love's book recently and redid this problem and realize now that it's much worse than I originally thought. The squeeze chances interfere with finding 3-3 splits. Say you cash hearts and find righty with 4. Then you play for the double squeeze, but you have to decide whether to toss your club winner or your spade exit in North on the trick where you've squeezed West. So you need to guess whether West has 3 or 4 clubs. So you have to decide whether to play West for: One of: J73-Q109 973-Q107 [edit: oops, I mean 973-QJ10] or J973-Q10 Both of the 3-3 splits are rather reasonable (ie considering the play of the opponents), so I'd really have to go with that. (Combined with my above stated view that nobody decent *ever* leads from J-fourth against 6NT -- again, thinking of this as a "real" problem instead of a "textbook" problem). [Note that the fact that finding the 3-3 splits interferes with the operations of the double squeeze(s) makes choosing which suit -- hearts or clubs -- to attack first either technically irrelevant (if you always choose to go for the 3-3 split when you find righty with 4 of the suit you attacked first... though not really irrelevant as you'll be watching lefty to see if he squirms) or rather odd if you'll play for 4-2: it's not immediately obvious which one to choose to attack if you, say, expect one suit to be more likely to be longer with one opp (I'll leave out the detailed analysis for now).] This is closer to a real "erratum" for the book, as Love doesn't mention this at all, and presumably if you were going to lead a 3 card suit in his day, you'd lead small (at least from J73). For textbook purposes, I guess every lead is supposed to be 4th (after all, you've gotta lead 4th from your longest & strongest against NT :)), which I guess I'm sort of ok with here. That said, does anyone have a squeeze problem book/collection (or just general declarer play, but including many squeeze hands) they know of that isn't devoted to "test driving" various squeezes but rather counting and drawing inferences in order to choose the correct squeeze or guess correctly which suit to discard/etc? I guess it seems hard to make such a collection, as you'd need to discuss the defenders' discards but wouldn't want to give away the setup of the squeeze. A two-page approach might work, where you flip the page and you've got a partial solution and all the discards listed and are at the critical juncture. Andy

None; a light bulb will always remain a light bulb. Dunno about that. The filament breaks and it no longer produces light: seems a fundamentally different object :). But really, the answer is: It takes one dummy! [meant to be spoken. insert your own punctuation.. I recommend , or :] Andy

Most play that passing 2Dx shows diamonds and partner will usually pass. This increases safety (helps you stay out of 6-0 fits and the like) and perhaps puts more pressure on the opps (they have fewer options & must act now). Some sort of lead direction being built in instead seems like an ok idea. Note that bidding 2!S would say you like hearts (but only want to play 2!S opposite spades), so 2!H being lead directing for hearts is a bit funny. If something is going to be lead directing, I'd recommend pass: then you still play the hand in the more likely event that you're not lead directing, and also advancer is more likely to play the contract when you are (since must bid now [edit -- this doesn't make any sense]). As for whether it should be lead directing for hearts or spades, dunno. Actually, a thought occurs to me: if you lead direct for a suit presumably you've got something like Qx. What if this is not pard's suit? Then you've located a card for the opponents (e.g. give pard xxx ... I suppose this isn't so terrible since usually they'll finesse you for stuff, but it's still something and what if they count out the hand and find pard is 3-6-2-2 and decide the Queen is more likely to be with the length). Something to keep in mind. Andy

I tend to play at a "deliberate" pace (read: moderate to slow) and must admit I'm a big fan of slow play. When playing quickly, I can form a reasonable plan, but I enjoy the game much less because there's less subtelty and actually thought-out deduction. Perhaps you mean you take a good chunk of time at trick one and then have most of the board worked out, but from your "slow play epidemic" comment, I'd guess/assume you only take a little time at the start and don't pause for thought for very long when an interesting point or tough decision specifically comes up. I haven't seen you play, but I doubt your play is at the level of Lauria and Versace (don't take this the wrong way: clearly almost nobody's is) and you'd probably pick up on more stuff if you took a little longer occasionally. Try playing more deliberately (this isn't meant at all to sound patronizing: we presumably play the game with slightly different attitudes and this comment in no way is meant to imply that I think I play the cards better than you do) and see whether you pick up on more stuff and how you like it. As for playing quickly inducing mistakes by your opponents (because you're more used to it or better suited for it or whatever than they are), yes, this does happen. The defenders can/should keep their own normal pace, but it is sometimes hard and requires a bit of conscious thought. I'm not advocating not playing quickly sometimes to induce mistakes/etc, but rather taking more time initially and pausing when it's warranted and there's an interesting problem, and perhaps trying experimental sessions where you try to think a lot and don't worry about inducing mistakes (which might not be how you'd play if you were doing your utmost to win). Andy

Oops. Then I don't have a squeeze -- East just tosses both spades and I have to lose 3 tricks. That just means my terminology can only apply to the actual squeeze :P. As for the jokes which hint at the view that terminology is silly nonsense, I agree except that terminology helps with memory and with thinking about squeezes in a structured way (but surely you agree and were just having fun). Andy

heh...I did this on vugraph. It should happen more than people do it imo. Are you referring to hand 20 vs Canada in the semis where you played for Wolpert not to have made a takeout x of 1C on 109xx/Jxx/AQJxx/x, or is there a juicier one I didn't see? This one seems good (assuming Wolpert isn't known to take strange actions like this), probably well meeting the requirement of being (significantly) worse than 12:1 against. Since USA1 were the favorites, though, it probably wouldn't be worth it if it were only say 20:1 against or some such -- in addition to the variance, there's the psychological blow. (Of course, the latter depends on your psychology and whether you'll be needled for letting the one imp get away.) Anyone want to guess odds on the above hand having existed and Wolpert having doubled? Andy

Quite a pretty hand. There might be a better term/description, but this seems to be what happens when you combine a three suit (vulnerable stopper) strip squeeze with a winkle. "Vulnerable stopper" is just a fancy name for the heart setup here: East has a heart stopper, but it goes away if it's led from (hence "vulnerable"). A winkle would be the following setup: (I've cashed the DA and altered the hand somewhat, including swapping E/W) [hv=n=shk92dj8c&w=shq103dk3c&e=sh876dq9c&s=shaj5d10ca]399|300|[/hv] South cashes the CA, tossing H2 from dummy. West is squeezed: a heart obviously fails, a low diamond allows an endplay, and tossing the DK allows declarer to setup the DJ in dummy. This combined name "three suit (vulnerable stopper) strip squeeze winkle" (which is a bit odd, because a winkle really already is a weird form of a strip squeeze itself) isn't quite a perfect description, though, as (after we've cashed DA) East has "already" come down to the stiff DK or stiff Dx (East gets to choose which) and has no chance to keep DKx, but it still seems to be appropriate, especially given the fact that East still has this sort of choice. Adding CK to south and a low diamond to East (and cashing the DA) might fit the above terminology very slightly better. Andy

Alternatively, Richard Pavlicek's website www.rpbridge.net has a file format converter so you can change .lin to his file format .rbn and then he has a .rbn to .html display utility. This produces quite nice looking hands (see his "sample after conversion"). Andy

I just remembered something that seems relevant, given the DK lead on one of the above hands: Twice yesterday, after I'd "sent" another hand to try again, but hadn't yet posed that altered problem to GIB (I don't recall the precise circumstances), one of those GIB answer boxes (ie the box that appears when you ask for advice) popped up with "I'd lead the S10" or some such. (I may have been sitting around talking to someone for a minute or so before getting to posing the next problem -- I don't recall whether this happened immediately or not.) Andy

I tried it multiple times, changing the hands to other things that worked and then changing back. I set up an invisible teaching table in the main bridge club, sat at N, E, S and put GIB in the W seat, entered the hand as west and gave N the rest of the hearts and C32 and S the rest of the spades and D32 (using "x" to do all of this). Then I entered P P 3H and waited for GIB to bid. Trying this morning I did also have GIB bid 3NT. I then changed the clubs (by erasing the club suit, erasing East's lowest club, and entering a different club suit) to CQ864 and it stalled again. I've switched between these two twice now today and it's bid (3NT) each time on Q1086 and stalled each time on Q864. I just hit undo, entered 3H again, and it bid 3NT on Q864 in normal tempo (~3 seconds or so). I've changed it to J864 and it's stalled again (> 1 min) and then when I hit undo, it passes (~1 second). Now I changed it to J1086 and it thought for ~5 seconds and led (!) the DK (auction is still P P 3H ?). I hit undo (twice -- 1st one undid the lead) and bid 3H again and it bid 3NT (~3 seconds). I have been having occasional connection problems (not with my internet connection though) -- being disconnected and not being able to get back on for several minutes or until I restart my computer [which takes a minute or two, so maybe this isn't really doing anything] (this is not more than once an hour), but I haven't been having problems while doing this. Indeed, I was chatting with someone while it first happened. I just restarted my computer to see if it persisted, and it bid on Q864 immediately, but then I tried J1094 (a hand I hadn't tried before -- could be relevant?) and it stalled (& then after ~1 minute I undid and bid 3H again and it worked in ~3 seconds). As for the "hadn't tried before" theory, I just put back in J1086 and it's stalling again (several minutes -- I went and did something else & left it). It again bid 3NT after ~3 seconds when I undid & bid 3H again. (While trying today, I haven't changed the tempo from the default setting -- looks like 3/4 the way to fast.) Andy PS Are these 3NT bids in any way normal/good? I've been surprised by them all except perhaps the 1st one (the one with the DQ too) which seemed to me to be aggressive but possibly ok.

I have version 4.3.8 and gave GIB (at a teaching table) the problem: None vul. 103/A3/AK964/Q1086 P P 3H ? It stalled and wouldn't answer. I had been trying various example hands: 103/A3/AKQ96/QJ86, 103/A3/AK964/QJ86, both of which it bid 3N on, and I made the hand weaker to see how long it would keep bidding 3N, but when I changed the !CJ to !C10 (or !C4) it stalled (it's still running as I type -- has been many minutes). I tried this multiple times at various speeds and it stalled every time except when I put it all the way to "fast". When I weakened the hand further, to 103/A3/AK964/8654, it passed in normal tempo. Andy

The old instant matchpoint games posted on Richard Pavlicek's website may be the sort of think you're looking for. He has them in .rbn format, but his file converter will convert these to .lin I believe. Andy

There's an article in "For Experts Only" (by the Granovetters) called "The Monty Hall Trap" by Phil Martin (which I haven't actually read, but have heard something of) which discusses the difference between "volunteered" information and information that you "happened upon [randomly]" and how you should use these various sorts of info in your use of vacant spaces (there's also something on this topic in the link above to Richard Pavlicek's website). This is not, as has been suggested above, simply because the difference is so small. If that were the case, then the second one would be correct but the first would be close. The reason is as follows: Suppose I have 9 diamonds. The chances they split 2-2, 3-1, and 4-0 resp. are: ("3-1" means either way) [x C y is "x choose y": the number of ways to choose y cards from a collection of x cards] 2-2: (22 C 11)(4 C 2)/(26 C 13) = 6*13*12*13*12/(23*24*25*26) 3-1: 2*(22 C 10)(4 C 1)/(26 C 13) = 2*4*13*12*11*13/(23*24*25*26) 4-0: 2*(22 C 9)(4 C 0)/(26 C 13) = 2*13*12*11*10/(23*24*25*26) Thus the odds of 2-2 instead of 3-1 (if we rule out 4-0 by playing a round) are: 2-2:3-1 is: 6*13*12*13*12:2*4*13*12*11*13 = 3*12:4*11 = 36:44 = 9:11 Suppose we first remove 2 diamonds from the deck: 1-1: (22 c 11)(2 c 1)/(24 c 12) = 2*12*12/23*24 2-0: 2*(22 C 10)(2 C 0)/(24 C 12) = 2*12*11/23*24 1-1:2-0 is: 2*12*12:2*12*11 = 12:11 After you draw a round of diamonds, it's not the case that 1-1 is now more likely than 2-0. You gained no new info except that they weren't 4-0 (again, assuming opps just play random diamonds), so it must still be the case that an original 3-1 split is more likely than an original 2-2 split. The (false) change in the odds (9:11 changing to 12:11) is basically due to there being more ways for them to play from an original 2-2 split (4 ways = 2*2) than from an original 3-1 split (3 ways = 3*1). You can work out the implications for other suits, but this shows that at least for the diamond suit itself, you shouldn't be counting "partial information" in that suit. Others have posted calculators I guess, but you can also just do this by hand using the (x C y) stuff as above. (If you're unfamiliar with how this works, I can explain it more thoroughly.) Hope this is useful, Andy

A (minor) bug report and a suggestion (maybe low priority): The bug: In partnership bidding tables, when opp bidding is active, you can lead if the opps win the contract (not the bug -- but something mildly surprising). I've found that when the host is not one of the players and the players lead, the host crashes (bbo freezes and one needs to "end task" with windows task manager). The (not really related) suggestion: It would be neat if one could specify hand types to work on at teaching tables like one can at partnership bidding tables. They're rather similar seeming entities, especially with opp bidding turned on and a non-playing host bidding the two opponents' hands. We were having fun leading when the opps ended up declaring (and wondering why the host kept crashing until we noticed the correlation) which is sort of where the idea came from. Thanks, Andy

Kokish's methods are "option 2" in my original post (unless you mean something different), and there are some problems with it here: 1) 13-15 is too wide a range to bid 2NT with, unless you think 13 is enough to be GF (When playing a standard system with a weak notrump, e.g., this is not an issue, as your notrump rebid is GF). 2) What do (41)44 (and (40)45 and, if you don't open 2C with them, (41)35 and maybe even (42)25's that aren't suitable for notrump) hands do? There isn't enough room after raising clubs to find your major-suit fit. These can be solved by having more of a "relay"-like setup and choosing different artificial meanings for 2H/2S (and maybe even 2NT). How best to do this is something interesting to discuss (though maybe this would be more complicated than Mauro wants), but I also think that "option 1" above is a nice simple "natural" method that doesn't have the problems above (it has different ones -- weak 4441's probably bid a NF 2NT or have to force game [or lie and claim 5 diamonds], and you can't make a NF club raise [a loss, but not a huge one]). Having said it'd be interesting to discuss such, I'll suggest one (probably not the best one, as I'm just making this up on the spot): 1D-2C;? 2D = 5+ diamonds, any 2H = Min bal or any 4441 (ie 4-4-4-1, 4-1-4-4, or 1-4-4-4). Also stick in (40)45 (treat like a 4441) and play another method if you could also have (41)35 or a (42)25 (that doesn't want to call itself bal) 2S = GF club raise 2N = GF bal (14+-15) 3C = NF club raise Over 2H, 2S asks, and: 2N = Min bal 3C = 4-4-4-1 3D = 4-1-4-4 3H = 1-4-4-4 (maybe there's a follow-up after 3D and 3H: bidding the stiff asks whether (41)44 or (40)45, plus range info or some such) Also over 2H, 2N tells opener to: pass with min bal or min 4-4-4-1, correct to 3C with min (41)4-4, and otherwise bid 3D (4-4-4-1), 3H (1-4-4-4), or 3S (4-1-4-4) (I'm not certain which order/bids for these three are best over 2S/2N). Higher bids (3C+) over 2H would be rather specific, but I'm not going to think about that now (let's say they're "natural" GF and try to work out on the spot how responder should bid with the various hand types). Andy