eyhung

Answer here:

Also, I don't see how "lying indicates the Queen" matters here. The suit is KNOWN to be exactly 3 on the left, 2 on the right. Whoever has the queen by definition has told the truth.

The actual spots were : KJ92 AT73 Two, four, ACE, eight Three, five, ? Also, in real life, I knew these opponents, and leading the jack would have told me nothing, but I'm curious about the general case when you are sitting down against someone you know nothing about.

I was playing 2 notrump at matchpoints and had KJ9x of clubs in dummy, ATxx in hand. I got a count on the hand and knew LHO held exactly 3 clubs. So I played a club to the ace, and LHO played the 8 (upside-down). I led a club up and LHO completed a peter. So what was going on? Had LHO played up the line I would have finessed without a second thought, but why is LHO going out of his way to show he has 3 clubs with Qxx of clubs and that club holding in dummy? So, my question is : If this were the only relevant suit (say all the other suits are stopped and the only remaining guess was this), and assuming the opponent were "unknown but doesn't look incompetent", would you play for the drop or finesse?

Unfortunately, in real life it's impossible to take 9 tricks if the defense doesn't screw up, because hearts are 3-6 and responder holds the spade ace, club queen, and four diamonds to the ten (OP mistakenly gave West the DT -- he had the D9).

yeah but -200 is going to be a good score? OP didn't mention that on normal play declarer will probably take 7 tricks in notrump. -100 would be average-minus, -200 was a near-bottom, and -300 was a cold bottom, so perhaps declarer took an anti-percentage play in an attempt to get back to -100.

I blame the partnership for not being on the same page with their pre-balancing style. If West can double on an aceless 9 with a bad queen (not my cup of tea at this vul), then East can't bid 3NT -- but if East expects a sounder sandwich double then West is to blame for doubling. I don't have a strong opinion on the merits of an aggressive vs. sound pre-balancing approach, but it's important to be on the same page! Re: the first-round pass, I actually held this hand as East in the same game. Doubling on the first round felt terrible to me -- you have no shape, nothing in spades, soft values in hearts, and the vul is the worst. The opponents actually went 1NT P 2H P P so partner didn't even think about doubling, and we took all our tricks to beat 2H a trick for average-plus.

I actually think, if you had a system, it seems simple enough to have a policy that would satisfy all of the constraints you mentioned. Show people ratings for potential partners/opponents when joining a table with an empty seat, but don't show people ratings for themselves. Allow people to "opt out" of broadcasting their rating, but make the default option to show the rating. And if a player really wants to know how they're rated, they don't opt out, and can ask their friends -- or they can "opt in" by clicking on "show me my rating" somewhere in an options menu. I think the majority of users don't bother to change options, so that way you can cater to the desires of the majority without hurting the desires of the minority. You can claim that the vast majority will just turn off broadcasting their ratings, but I don't think so. From my experiences on okbridge with the Lehman rating in its infancy, a lot of people did not, and turning it off also sends an important data point -- that the owner is not happy to broadcast his rating. :lol:

Unfortunately, I didn't witness the hand in question, so I couldn't provide any of the three factors that you are asking for, but I thought the problem was interesting enough to relate. The "obvious" line of ruffing 2 hearts in dummy is indicated, but fails with 4-1 trump. I think your line, leaving open the possibility of a cross-ruff, is superior, because it can even pick up some hands where both suits break 4-1, but I was wondering if there was an even better line that I was missing.

[hv=n=saq432hdak32cq832&s=sk95hkt832d4cakj9]133|200|Scoring: IMP[/hv] I was given this hand by a friend. You reach the contract of 6♣ with no opposing bidding, and apparently the declarer never supported spades. The lead is ♠J (standard leads). How would you play this at teams?

I didn't realize that we were only talking about the 1♦ - 1♥ - 1♠ auction. Most people don't really want to differentiate between that auction and 1♣ - 1♥ - 1♠, even if it would be technically correct to do so. (And yes, I agree with all your technical points.) From a practical standpoint, I feel you're making an unfair comparison between unbalanced spade + simple followups vs. balanced spade + XYZ + symmetric relays. I believe the majority of standard system partnerships are not interested in having different methods over 1♣ and 1♦, or into complicated 3rd-round followups such as a symmetric relay after the XYZ 2♦. That obviously doesn't apply to some of the regulars on these Forums, of course. :) But if you're just interested in being on the same wavelength as partner, playing an unbalanced spade + simple followups is hard to mess up, and I think that's an underappreciated virtue of a system for casual partnerships.

The last point (right-siding the notrump) can be addressed by raising to 2♥ instead of rebidding 1NT on hands with clearly anti-positional holdings in the 4th suit. The biggest problem with a balanced spade rebid + XYZ structure is that you cannot play in 2♣ after opener starts with 1♣ -- responder can have a hand that is a clear weak preference to clubs if 1♠ promises an unbalanced hand. I would never play XYZ over a 1♣ opening with the unbalanced spade style. Obviously if you do play the (standard) balanced spade style, XYZ + symmetric relays have much to recommend them, but that is a significant amount of complexity that I feel most partnerships don't have the inclination to adopt.

I wouldn't have said anything if there were at most one poster who passed, but I feel this is a clear opening in the KT9x suit 3rd favorable if you consider yourself to be a better team than your opponents. Sometimes passing out is right, but it feels very random. Meanwhile, bidding is right when : 1) They misjudge the bidding 2) You and partner judge well in the bidding 3) They misjudge the play 4) You and partner judge well in the play It's wrong when the converse happens, but that's less likely since 1) you + partner are unlikely to misjudge the future auction/play since you have what is expected for your bid (a light opener with a lead-director) and you are the better team 2) They are the worse team. I remember a great tip from a Grant Baze article : when you are up against weak opponents, do not lose the board with an early non-obvious unilateral action. Keep things flexible with a flexible hand and give them opportunities to lose the board. They bid worse, they play worse, they defend worse: that's what weak opponents do. Passing is such a unilateral random decision: you may be staking everything on passout being right when you don't know much about the hand.

As an adherent of the 1NT > 1S on balanced hands style, you will note that I did not rebid 1NT but 2♥. Responder has the hand type I was worried about: weak hand, mediocre (and unrebiddable) hearts, and a potential problem in clubs. It turns out the clubs are worth 3 tricks more on the defense vs. notrump, so the difference is down 3 vs. down 1 (hearts are a level higher). So I think this is less about your system and more about a misapplication of judgement. Re: hearts vs. spades, both strains take only 7 tricks on best defense. Partner did not have a third club to make use of the ruffing value, and yet either major still plays better than notrump. One advantage of spades is that the North hand can pass 1♠. But then the East hand has a clear reopening double and then they can find the E/W par of 2♣ making 2. -50 in 2♥ or 2♠ seems to be practical par at IMPs.

If you bid 1S with this hand you are not planning to bid 2H next. I see no reason to think that the combination of 1N/2D/2H/2S that you will get to after bidding 1S will be worse than playing in 2H opposite all of those hands. Right, I would not be planning to bid 2♥ next, but just like bidding 2♥ directly will cause us to miss good spade partials, bidding 1♠ may cause us to miss good heart partials where partner has a weak hand with 5 mediocre hearts (more likely given your KTx). Even if partner only has 4 hearts in a bad hand, the worthless club doubleton and two heart honors also gives hope that even the 4-3 heart fit may play better than 1NT or 1♠. However, I am not saying that 2♥ is +EV vs. 1♠ on this particular hand. I think it's actually -EV, and if I were only dealt 4342s for my 1♦ openers, I'd want to be playing a system where I could rebid 1♠ on any 4342 hand. But I think in the context of my preferred rebidding style, I believe the +EV I gain from 1♠ promising an unbalanced hand when those hand-types occur outweighs the -EV action I have to take by raising to 2♥ on this hand, because I think -EV is small given the reasons I advanced. Note that everything had to be right for me to raise to 2♥ on 4342 -- either one of the changes mentioned by the OP causes me to rebid 1NT instead.

I think this is why the work on better rating systems like the one at http://www.coloradospringsbridge.com/pr.htm deserves wider recognition. It is not perfect, but it does generate conclusions that feel far more representative of recent performance than other systems. (Unfortunately, although it has full access to the ACBL tournament database, it does not have access to many club game results, so data that would help rate more top players is missing -- but it does do a good job of measuring top partnerships.) From my experience in analyzing baseball players for fantasy baseball, it's easy to misrate people based on a limited sample of experience -- going to an objective ratings system smooths out observer bias. Edit: link didn't work right, hope this fixes it.

Fine, I'll be the spokesman for the 2♥ camp. I choose 2♥ over 1NT because of the suit-oriented nature of my hand and the small doubleton worries me, and my followups can cater for a 3-trump raise. Yes, there are definitely downsides to bypassing spades, but my style is to only rebid 1♠ with an unbalanced hand. I certainly don't think my hand is worth bidding 1♦ ... 1♠ ... 2♥, I think that should show extras. When I do have an unbalanced hand, I have an advantage in being able to describe my two suits and unbalanced nature immediately, and when I don't have that hand, I don't lose as often as one might think -- you may be able to checkback over 1NT into your 4-4 spade fit, or partner may not have 4 spades. I'd bid 1NT with either change.

So you're asking how to continue after T1: ♥J, ♥K T2: ♣ to LHO's Ace T3: ♥9 exit, obviously win ♥Ace. Now the H8 is a menace against RHO. Oddly, LHO could have led a low heart to preserve the H9, so he has already misdefended (if we are assuming that LHO is honest, then he won't have T96). So now I would cross in diamonds, draw trump pitching a heart and spades from the dummy. Then I'd cash the DA and ruff a diamond, picking up any Q-third of diamonds (more likely than K-doubleton of spades). Then I'd run the rest of the trumps with menaces H8, DJ, and SQ. The 3-card end position would be AQ void J void T 8 void x In this ending, lead the last club and discard the DJ. Assuming the HT has not appeared, now I play a spade to the ace -- this wins whenever the person with the HT holds the SK, or whenever LHO holds the DQ and SK. If the carding is "honest", East holds the HT, so this line wins when East has the SK (simple major-suit squeeze) or when West holds the DQ (double squeeze). It also wins when East holds the DQ-third. The failing case is if East holds the HT, DQ-fourth and West holds the SK -- in which case their tempo of discards may lead me to take the spade finesse (for example, if East is a poor player who will never bare a king quickly, but discards a spade without pause at trick 10). But against an expert East defender who is capable of baring his king in tempo, I would reject the finesse and play for the drop.

Not everyone leads 4th best vs. suit contracts, and second, defenders are far more likely to falsecard on lead vs. a slam. I actually think 6 from 6x is most likely but Qx6(x) can't be ruled out if the leads are 3/5. The OP didn't state what the leads were so I assume he didn't note it and so I'm not going to analyze the HJ play. It's certainly not unreasonable -- clears up one of the questions so you can focus on the spades/diamonds. And yes, I chose to leave diamonds as the remote threat because the defenders can theoretically knock out the spade threat when they win the club ace.

This may be with the benefit of hindsight, but consider that if you could ruff out the spade king with the extra entry provided by the club king, the opponents should have taken the trump ace to deny you that entry. And it's not like that it's a particularly hard defense to find after you ruff a spade in your hand at trick 3. Maybe a better line is to play a club to the king at trick 2, ruff a spade back to hand, and lead another trump. If the club ace wins and exits any side suit, that will increase your chances to pick up that suit (a heart exit will isolate the menace in hearts). Assuming the club ace exits with a neutral trump, you can then draw trump, cross to a heart, ruff another spade, then trumps coming down to Q -- AJ3 -- --- 8 K9 x If both the HT9 are still out, I'd play the last club throwing spade queen from the dummy, then I'd probably take the diamond finesse (East is unlikely to have been squeezed). If the heart 8 is now second in rank, I'd probably use the information provided by the spade ruffs, club break, and discards to decide whether to try to drop the diamond queen or finesse.

True, I suppose that was in response to the "robot reward" tournaments, where there is a significant luck factor on how many distributionally strong hands GIB gets to play against you. In Robot Duplicates that should be ok.

I think you could do one of two things: 1) Change the "cooking" so that the "best hand" is the hand with the greatest distribution (longest two suits) + HCP. That makes it harder to draw inferences about the other hands, and I believe it would also make it more enjoyable for the human player. 2) Don't cook the deals at all, but allow the human player to switch seats and declare when they are supposed to be dummy. This means that the human would declare approximately 50% of hands, which would alleviate the "boredom" that Josh mentioned.

If you say so. As far as I can tell, there are only 4 relevant holdings, and the last time I checked, each is equally likely. K♦ - J♥ K♦J♥ - neither J♥ - K♦ Neither - K♦J♥ With the finesse you win in 2/4 cases, diamond King with West (50%) and the other way, you still win in 2/4 cases (KJ with East or West), same 50%. Excuse my math, but I am too dumb to understand anything else. That's like saying that if you buy a lottery ticket, one of two outcomes must occur: 1) you will win the lottery 2) you will not win the lottery Under your reasoning, there is a 50% chance that you will win the lottery. This reasoning is clearly flawed because the chance of the second case is not equal to the chance of the first case. Similarly, saying that KJ void K J J K void KJ are equally likely is false. The 1-1 breaks are slightly more likely. Say you knew that the King was held by West, and that was the only information you had. The chance of the jack being in the same hand is now 12/25 because there are 12 chances for West to hold the jack, but 13 for East. So the 1-1 breaks occur 26% of the time, and the 2-0 breaks occur 24% of the time. KJ void 24% K J 26% J K 26% void KJ 24% There are four cases but each case is not equally likely, so the 1-1 break is 52% and the 2-0 break is 48%. And that's why it's mathematically odds-on to drop a king instead of finessing with 11 cards. The logic is the same when deciding whether to drop or finesse a queen with 4 missing, or to drop or finesse a jack with 6 missing. You cash your extra high honors and lead up to the tenace. Your 2nd-hand opponent plays the last small card. At this point, it's 52% to drop, 50% to finesse. So barring any extra information, it's correct to play for the drop.

Yes, to confirm, I am using total bridge points, so the original sim says passing is about as good as bidding if both sides play par bridge. I have tweaked the simulation to disallow some of East's 10-11 counts so North's Average HCP is going to be somewhat higher than his expected 9.33. I also used the true spots -- the original post only specified x's, while a followup post mentioned the actual spots (QJ3 J75 K2 KQ754) Hands = 1000 Average North HCP = 9.734 Average NS score = 35.75 points N/S plus = 606 N/S minus = 394 160 [score "NS -110"] 26 [score "NS -120"] 15 [score "NS -130"] 85 [score "NS -140"] 3 [score "NS -200"] 8 [score "NS -400"] 20 [score "NS -420"] 1 [score "NS -450"] 10 [score "NS -80"] 65 [score "NS -90"] 1 [score "NS -920"] 154 [score "NS 100"] 99 [score "NS 110"] 91 [score "NS 120"] 11 [score "NS 130"] 98 [score "NS 140"] 11 [score "NS 300"] 16 [score "NS 500"] 17 [score "NS 600"] 9 [score "NS 620"] 1 [score "NS 630"] 3 [score "NS 650"] 12 [score "NS 80"] 84 [score "NS 90"] All the +100s and -200s are from doubled contracts, so it's easy to adjust. The total score on the 1000 hands was 35750. Double gains 50 points 154 times, or 7700 points, and loses 100 points 3 times for 300 points. So assuming no matchpoint-style doubles, the adjusted total score would be 35750 - 7400 = 28350, giving N/S a more realistic par average of 28.35 points from opening. So with the East adjustment, it's clear to open in 4th chair. Also, the -920 involved A7 T6 AT87654 J6 opposite KT842 A984 Q93 8, but it only influences the par average by 1 point. I admit my preempt / weak two function is not robust -- I'd be happy to solicit suggestions to make my future sims better.

You are right that my sim ignores light 3rd openings, so the true expectation is higher than +4.2. But how does opening a good suit give us an advantage over double-dummy par? It would give us an advantage over real-life outcomes, but I'm not modeling a real-life outcome. In fact, if the sim came out slightly negative par for N/S I would still conclude that it's right to open given the factors Justin mentioned, and the 4-5% declarer's advantage that part-score single-dummy contracts have.