Wednesday, April 28, 2010

The Perfect Bid

Experienced bridge wags are fond of the saying "Bridge is an easy game."  But there is one situation where bridge really is an easy game.  It's when you're dealt a hand which is perfectly in the range of one particular bid, and not suitable for any other bid.  There would be no problem, right?  Just make the bid.

So why do we seem to go out of our way to make some other bid?  I don't know.  Sometimes, we just have a blind spot.  But I've seen quite a few examples lately. Take this one, for example.  ♠92 AQ96 AK62 ♣K73.  Red on white.  Partner passes, RHO bids 1♣ ("could be short").  There's one bid (and only one) into which your hand falls.  In fact, your hand is screaming "1NT, 1NT!".  We made 2NT.  Unfortunately, our side was on defense and our +100 did not score well when everyone else was +120.  I won't give any details, but the call chosen by the holder of this hand was, obviously, not 1NT.

Here's a similar one.  ♠AQ952 A8 K6 ♣AJT9.  Red on white, again.  LHO and partner both pass.  RHO opens 2.  Again, you have the perfect bid: 2NT.  This is especially true if you and partner play that 3♣ is looking for a five- or four-card major in this situation (e.g. Puppet).  Despite our 24 hcp, we ended up -800 on this board (in 3♠X).  2NT was in fact our par contract, although 3♣ also makes.

Sometimes, of course, we benefit from our opponents not seeing the wood for the trees.  ♠Q7 JT95 J964 ♣Q54 at favorable vulnerability.  Partner opens 1 and RHO bids 1.  Again, the right bid is staring you in the face: 2.  When I was the 1 bidder, this hand passed.  At Kim's table, this hand bid 1NT!

I hesitate to mention the next case, but in my role as intrepid reporter seeking out the truth, I feel duty bound.  ♠AKJ9 K965 T986 ♣2.  Nobody is vulnerable.  Partner deals and opens 1 (either a hand with distribution including 4+ diamonds, or a balanced hand with 15+ hcp).  RHO, somewhat surprisingly, bids 1♠.  While a trap pass could be right (although I think I'd rather be at favorable vulnerability for that action), I think there is one standout call here: double.  This typically shows four hearts and at least 8 hcp.*  That's exactly what we've got here, right?  If partner rebids 2♣, we can safely bid 2NT showing our spade stoppers, around 11 hcp, and the four hearts we've already promised.  If partner rebids 2, we have a more awkward decision.  In fact, partner has four hearts so we would presumably end up in 4.  Again, no details, but let's just say we never found our heart fit.

* in our system, we will occasionally have a good hand with a good suit of our own when we double.

Partnership Bridge: Redouble or pass?

Another thought-provoking article by the Granovetters in the May 2010 Bridge Bulletin. By chance, it is an excellent exemplar for my earlier blog Redoubling with three trumps.  Here's the hand in question (only we are vulnerable): ♠Q74 QT86 82 ♣AKQ3.  Partner deals and opens 1♠ and RHO doubles.  Now, what are our options?  2♠? no – too good.  3♠? no – ditto.  4♠? no – it gets us to the right spot but partner will think we have a weaker, more distributional hand.  2NT? not enough trumps.  Redouble?  no! – we have three trumps so, given the vulnerability, this doesn't look like a situation where we are enthusiastic about penalizing the opponents.  Pass?  Yes.  What can go wrong?  Nothing. Especially when we hold the boss suit.  Whatever they do next, we will bid 3♠ (or 4♠ if we think our hand is good enough) at our next turn.  If they bid 4 or 4♣ first, we can change tack and double.  Partner will know that we have a balanced hand with limit-raise strength and exactly three spades.

If by some miracle, they pass it out (they almost certainly won't) we will score 560 if we make 9 tricks or 760 if we can make game (even non-vulnerable, 1♠X+3 outscores 4♠).  Note, however, that on this hand it is just possible that the double will be passed out, as doubler's partner has ♠KJ985 942 T ♣T954.

Let's assume that LHO bids 2♣.  Partner rebids 2 and RHO raises to 3♣.  Now, having passed earlier, we enter the auction with 3♠, showing a three-card limit raise.  Partner will know exactly what to do.  He will pass.  Yes, on this particular occasion we give up on a penalty of 800, but 140 is still a reasonable plus score, at least at matchpoints.  On a really good day our LHO will now double 3♠, giving us 730.

What happened at the table, according to the article, was that our hand redoubled.  The bidding continued as above.  Over 3♣, our hand doubled.  Partner who had opened with a fine, if minimum, distributional hand ♠AT632 K AJ9764 ♣2, pulled our penalty double to 3 (a questionable action, as Pam points out in the article, since he's already shown a weakish distributional hand when he rebid 2) and we bid 3NT, never having shown our excellent spade support.  As it turned out 3NT cannot make from our side of the table, nor can 4♠ as it happens, each contract failing by a trick.  This is primarily due to a foul distribution of the missing spades (5-0).  But, generally speaking, when we have such good support and 13 hcp, we would normally expect to be in 4♠.  The fact that it doesn't make on this layout is neither here nor there.

I was surprised that neither Matthew nor Pam criticized the redouble.  If you've read my previous blog (referenced above) you'll know that I feel that redouble should guarantee no fit! Qxx opposite a major suit opener is a fit!  Partner can never know whether to penalize the opponents if he thinks that we might have three of his suit.

Monday, April 26, 2010

Opening Bidder's Compact

In this blog, I expand on one my earliest blogs The "no undo" principle, in particular as it relates to making marginal opening bids in first or second seat.

We've all opened a marginal hand at times and thought better of it as the auction climbs up to the stratosphere.  But, when you open in 1st or 2nd seat, you make a compact with your partner.  The implicit terms of this contract are (as far as I know this is the first statement of these terms):
  • I promise never to pass a forcing bid;
  • I promise that, having limited my hand, I will respond appropriately when partner requests a control;
  • I promise always to tell the truth about the number of key cards I hold;
  • I promise to support partner if I can do so at the two-level.
Of course, this all sounds pretty straightforward, doesn't it.  You would never pass any forcing bid whether you had opened or not.  But other people do it.  And of course you understand your obligations when it comes to control showing bids.  And you wouldn't dream of lying about your key cards!  But I'm betting that you haven't given a lot of consideration to the final point.

Here's a case in point.  Matchpoints, white on red.  You deal yourself this following hand: ♠K984 Q5 Q8 ♣AJ963.  Playing strong notrumps, you open this hand 1♣ and your left hand opponent bids 1.  Partner doubles, which at this point could be a normal negative double or a good hand with a good suit, to be introduced later.  RHO bids 2.  Your call.

If partner has a "normal" negative double, he will have both majors.  If he has some other good hand he might just have one major.  If he's chosen to double with only 3-4 in the majors, he will have a good balanced hand.  In any case, now is the time to show your four spades.  You're still able to show support for partner's presumed spades at the two level.  Do it.  Here's what might happen if you don't.  Partner reopens with another double (two-way, this time) and this time you show your spades.  LHO now bids 3.  Here's what's going through partner's mind.  You didn't bid 2♠ over 2 so you can't have four spades.  You didn't rebid clubs so you probably don't have six.  So presumably your hand is something like 3235 shape and probably 13 or 14 hcp because you probably wouldn't open a 3235 hand with 12 hcp unless your clubs were strong enough to rebid.  Partner still feels it's our hand but doesn't want to bid 3♠ on a 4-3 fit.  So partner doubles again.  By your agreements, this double is intended as penalty (based on the three-strikes-you're-out rule).  At this point, you really have no reason to pull the double, even though you have an uneasy feeling that your hand is more distributional than you'd like.  The opponents take 10 tricks and we record the lovely score of -870.  3♠ would be down 1 so that would be -100 assuming it's doubled.  If the opponents bid on to 4, our side would feel that it had done everything it could have and happily pass.

There is a possible semi-exception to this rule.  Suppose that, again playing strong no trumps, you open with a balanced 12 count and you have the opportunity to make a support double (showing three of partner's suit).  Should you feel obliged to double?  Or could you pass?  I think that it's acceptable with a minimum hand and only three card support to pass, at least for this round of bidding.  After all, the "Law" says that we may not be safe at the two-level with a seven-card fit.  And, partner has another chance to bid.

The alternative strategy of passing at your second turn any time you have a minimum hand, even with a known 8-card fit, is losing bridge, I believe.  How many times have we heard the mantra "support with support"?  Thus, supporting partner's suit (or implied suit) at the two level does not promise any extras.

Wednesday, April 14, 2010

A hidden advantage of the weak no trump?

I've always been a fan of the "weakie".  Usually, that's the 12-14 range (and sometimes a good 11).  I'm also a fan of the ranges that I play when playing precision: 10-12 not vulnerable, 1st/2nd seat; 14-16 otherwise.  The usual advantages espoused for the weak no trump are:
  • 1NT is a very descriptive bid and you get to use it more often then when it shows a stronger hand (at matchpoints, it's often the first side to 1NT, especially not-vulnerable, that "wins" the auction);
  • Minor suit openings are either good distributional hands or balanced hands with 15+, so we aren't so afraid to open a minor in case it makes it too easy for the opponents to overcall a major;
  • The corollary is that 1NT is semi-preemptive, using up the entire one level, thus making it more difficult to intervene.
There are of course a few disadvantages:
  • sometimes we "wrong-side" the notrump contract, that's to say we play it the opposite way to most of the other pairs -- this can generate tops, but it also generates bottoms;
  • opening 1NT vulnerable and finding partner with a poor hand can be troublesome: we might go down 2 undoubled and still get a zero.
Then there is the de facto advantage at the local bridge club: many opposing pairs either don't know how to handle the weak no trump or they have forgotten to discuss it.  A common error is to think that you can intervene over the weak no trump with weaker hands than you might over a strong notrump.  In fact, the exact opposite is the case.

Today, I found a new advantage.  You can use it for a semi-preemptive 4th-seat opening, in the same way that you might open 2 or 2♠ with a six-card suit and a minimum opening bid: you plan to make your contract and you are gambling on the opponents not coming in to the auction.

Clearly, 1NT isn't as good a 4th seat preempt as 2 of a major, but it can be effective.  Take this hand as an example (I held this hand earlier this evening): ♠J3 73 AQ954 ♣KJ97.  We were not-vulnerable versus vulnerable and there were three passes to me.  I can tell you now that passing the hand out would have guaranteed a very good board.  But I wanted to open just in case partner had passed in 2nd seat with 10 or 11 hcp.  Naturally, I would want a diamond lead if the opponents get the bid, but I wasn't planning on letting them get into the auction (I'm short in both majors).  So I decided that a preemptive 1NT was in order.  Call me crazy if you like (yes I know it doesn't have 15 Cansino/Pearson points).  But the local club is a great place to experiment, surely?

The lead was 3 and my partner laid down a dummy that was, let's say, a disappointment (in fact, about as bad as it could be on the auction): ♠985 QJ65 J62 ♣QT2.  The bad news was that I was contracted to take 7 tricks in notrump with a K less than average (compared to their K greater than average).  There were actually two good bits of news.  First, with eight spades and 23 hcp between them, the opponents could probably make 3 or 4 spades.  Second, the lead had given me a crucial tempo.

I won RHO's 7 with the 9 and proceeded to establish my clubs.  Perhaps the defender with the ♣A would continue the attack on diamonds.  No such luck.  RHO won the trick and of course didn't have any more diamonds.  A spade was led and they fortunately split 4-4.  I pitched my two hearts.  A club was then continued.  I won and played towards dummy's J.  LHO went up with the K and returned... a diamond!  At this point I was able to claim the rest, making my contract.  Had they cashed their two hearts for -100, I think I would still have had a very good board, as they can make 4H or 4S their way.

Monday, April 12, 2010


I'm a fan of the number 113.  First of all, it's a prime number and all primes are just a little bit special.  Second, it's the denominator of one of the better approximations to pi: 355/113 which is remarkably close (and a whole lot better than 22/7).

So, I was moderately pleased to get 113 victory points in yesterday's flight B Swiss at Watertown, with teammates Jan, Steve and Leo.  I think this is the highest total I've had for an 8-round Swiss.  The only fly in the ointment, as it were, was the fact that another team got 125 vps!  Under the (very) old rules, I think we might have won the event: each team had 7 wins, but we beat them when we played head-to-head.

But we had a lot of fun at our table.  The cards were definitely running our way (the flip-side was that our teammates were generally on defense most of the day).  At our table, we played 33 out of the 56 boards (there were no passouts).  On those 33 hands, of which no fewer than 23 were game contracts, we won 62 imps (approximately 2 imps per board).  When we defended, we won 31 imps (approximately 1.5 imps per board).

The one area that we didn't do so well, as a team, was in bidding slams.  Our teammates missed two slams, while we also missed two our way (though we also bid and made two) and played what should have been a grand slam as a small slam.  These five hands contributed a net loss of 13 imps, but if we'd all really been on the ball, would have given us 49 imps in our direction!  But neither of our pairs is a really established partnership and I think slam bidding is an area where being really in tune with each other is very important.

All in all, a satisfying and very pleasant day.  Perhaps it's time to take a ride up to Dunstable and check out those two magnificent copper beeches... along Route 113.

Saturday, April 10, 2010

Defensive Strategy: Follow partner's lead

At love all, you deal yourself ♠T K983 9864 ♣AK96 and pass.  LHO bids 2♠ and there are two passes to you.  You decide to double and everyone passes.  Partner leads the A and dummy is ♠5 Q762 AJ7 ♣Q5432.  A wish trick is completed (your 3 being an encouraging signal).  Partner continues with 5, obviously his last heart.  What's the right strategy here?  How do you know if partner is looking for quick ruffs (he has several small trumps) or whether we can take our time and develop side suit tricks (partner has good intermediate trumps).

Duh!  It's obvious.  If partner had good intermediates, he'd be leading a long suit.  The fact that he led from a doubleton indicates that he wants ruffs quickly.

On this hand, I didn't see the hurry for giving partner his ruffs (after all he did pass 2SX so he must have good trumps, I reasoned).  But partner set our side's defensive strategy with his lead and presumably he knows how good his trumps are.  Unfortunately, we only defeated the contract one trick because of my poor logic (57%) instead of two tricks (93%).

How often do we see opening leader make a good lead only to find his partner countermanding the plan and switching to a different strategy?  It's a question of trust.  Good partners assume that their partner has made the right lead and continue the program (assuming there isn't some glaringly obvious alternative).

Friday, April 9, 2010

DSIP rule summary

  1. Unless explicitly rendered otherwise by one of the events listed below, all doubles are either pure takeout or cooperative, that is to say that you expect partner to take the double out, but you won't be unhappy if he doesn't; level is immaterial – even if the last call by the opponents is 7.
  2. If either partner on our side does one of the following, doubles become penalty:
    1. bids notrump "to play" (generally this is at the 2-level or above or at the one level when showing a stopper);
    2. makes a penalty double (e.g. a lead-directing double of an artificial or cue bid) or passes partner's double, or makes the third double for our side;
    3. redoubles (of any sort);
    4. shows a two-suited hand (except that if one of the suits is unknown, it's best to play partner's first subsequent double as asking which is the suit);
    5. makes a cuebid or bids the last remaining unbid suit (or bids the "fourth" suit as a force);
    6. jumps the bidding (this last item covers the gamut of preemptive bids and fit-showing bids -- because they tell partner so much information in one bid, he/she is now in a much better position to make a penalty double -- note however that a subsequent double by jumper is not typically 100% penalty).
  3. If we pass over a suit bid on our right in a competitive auction and subsequently double that same suit, thus exposing the pass as a trap;
  4. If their side does any of the first four of the above actions (they can jump all they want, it's only when we jump that penalty doubles are turned on).
For more background and discussion please see my earlier post: Double!


The above was published April 9th 2010. It's now Oct 18th 2014 and much water has flowed under the bridge. The triggers (rules) mentioned above have not changed significantly since then, but there have been revisions. It is probably time to write up a new summary, especially rule 3 which essentially [should] cover[s] all of the exceptions. If you would like to read all of the relevant posts on this general subject, then follow this link: all DSIP posts.

    Thursday, April 8, 2010

    Doubling to say "don't lead my suit"

    One of my favorite bridge books is George Rosenkranz' Tips for Tops.  As well as being the inventor of the contraceptive pill and a fine bridge player, he's a real gentleman.  And his book is full of well-reasoned bridge logic that you don't find in most other books.

    One of the agreements he recommends is a slight change from the normal precedence for leads against 3NT when your partner has doubled.  Most people play that if they have bid a suit and then they double 3NT, it says "damn well lead my suit: and if you're void, then at least apologize."  But what's the value in this silly agreement?  Assuming that you aren't actually void, weren't you most probably going to lead partner's suit anyway?  Remember that we don't use lead-directing doubles of freely bid games and slams to increase the penalty.  We use them to create a penalty where otherwise the contract was going to make.

    So whenever possible I get my partners to agree to use a lead-directing double of 3NT to say "don't lead my suit."  It's possibly I had that agreement with my partner on this hand from BBO but, if so, I forgot (we don't play very often) and I apologize!

    My hand was ♠A6 K632 J742 ♣Q54 in second seat (none vulnerable).  RHO and I passed, then LHO opened 1♣ and partner passed.  RHO raised to 2♣ (no alert).  This was passed around to partner who doubled.  RHO raised again (3♣) and I felt that my hand was worth a 3 call.  Two passes followed and now RHO bids 3♠!  LHO bid 3NT and it came around to me.  Should I double?

    Obviously (so it seemed) 3NT was going down and I didn't particularly want a heart lead (not unless partner had an obvious heart lead himself).  Playing the way I prefer, I should double to suggest a different lead (a spade for example).  But I didn't double because I was worried that that would guarantee a heart lead.

    As it turns out, no lead sets the contract!  However, against this particular declarer (I was allowed to win my ♣Q), a spade lead probably would have set it.  Or, if upon winning my ♣Q, I switch to A and a low spade, we would get 1♣, 1, 3♠ and maybe a (dummy's "spades" were 9752, while partner had QJT3).  I could even have set it by switching to a low diamond (or even continuing clubs) too.  Just so long as I don't woodenly continue hearts.  But naturally I thought partner's hearts were better, given that I didn't double.  You see how important these agreements can be.

    Wednesday, April 7, 2010

    The squeeze that wasn't

    When you are thinking about running a squeeze, do you analyze minutely all of the possible layouts, threats, entry positions, etc.?  Or do you rectify the count and have at it?  I'm afraid I tend to do the latter.  75% of the time it works out just fine – the squeeze either works or it doesn't and nothing is lost (that's the great beauty of the squeeze: you never end up worse than you might have been).  I don't usually stop to think of isolating the menace, transferring the menace, etc.  I should.

    Having a general idea of entry positions, potential blockages etc. is usually enough.  But last evening at the club I failed to extract full advantage of a situation.
    Dealer South. None Vulnerable.
    ♠ A Q 10 9 8 7 2
    J 6
    ♣ 8 5 4
    ♠ K
    10 7 4 2
    A K 5
    ♣ K J 10 9 3
    Bridge deal ♠ 6 5 4
    8 5 3

    J 7 4 3
    ♣ Q 7 2
    ♠ J 3
    A K Q 9
    Q 10 9 6 2
    ♣ A 6

    Here's the auction:
    W Me

    We play 1NT as 12-14 which is why the South hand opened 1 thus allowing me to end up as declarer.  Otherwise, there'd be no story.  Note that my 2♠ was non-forcing.  The lead was a trump.  If the lead had been a diamond (or even a club), again there'd be no story.

    I won the lead and played four more rounds of trump.  Somewhere along the line, E shed a second club, which gave me my opportunity.  This is how things stood:
    Dealer South. None Vulnerable.
    ♠ 7 2
    J 6
    ♣ 8 5 4

    10 7 4
    A K
    ♣ K J 10
    Bridge deal ♠ 
    8 5 3

    J 7 4 3
    ♣ Q
    A K Q 9
    Q 10
    ♣ A 6

    At this point, I knew that I needed a trump squeeze because I had no other way back to my hand.  But I thought it was a Vienna coup situation where I had to unblock the ♣A first.  On the contrary, I needed that card intact as a potential late entry to the dummy for the last diamond, should that become good.  So I played another spade, pitching a diamond (shame!) and now W was able to pitch one top diamond and hang on to the last club.

    If I actually stop for a moment to think about my menaces against W, they must be a diamond in dummy and a club in my hand.  Therefore, I must pitch dummy's little club.  This blocks the clubs, but that's the magic of the trump squeeze: when and if the club threat materializes, I can unblock the A and return to hand with a trump to enjoy the last club.

    As much as one might read about squeezes on paper, there's nothing like running one and failing to ram the appropriate rules into that thick skull!

    Sunday, April 4, 2010

    More discipline

    There are two distinct types of discipline breach.  One is a uni-lateral breach (also known, when it works, as "operating").  The other is known as hanging partner out to dry.

    Let's look at a couple of examples. First, hanging partner out to dry. It's the first board of the National Swiss (i.e. IMPs) and you hold ♠532 T83 Q965 ♣974.  Both sides are vulnerable. Partner deals and opens 1.  RHO bids 1♠.  Your call?  Pass is the disciplined call.  2 risks hanging partner out to dry. Why?  Partner is still very much in this auction.  The opponents (one of the best teams in the event) bid to 4♠ and partner with a balanced 18-count and three sure tricks and a probable fourth (the K) doubles (in other words, he believes you rather than the opponents).  Then, with every expectation of collecting 500, he misdefends allowing declarer's singleton Q to score an early trick, thus kissing goodbye to one of the three sure tricks.  Declarer draws trump and claims.  So instead of defending 4♠ for down one (+6 imps), we actually allow them to make 790 (-12 imps).  That little stunt cost 18 imps!

    Now, what about a uni-lateral breach?   It's match-points and you pick up a reasonable hand (to be honest, I don't remember the hand exactly): ♠AK KQ82 QT84 ♣A54.  RHO opens 1♠ and you bid 1NT.  LHO raises the ante to 2♠ and partner bids 2NT (Lebensohl).  You dutifully bid 3♣ and partner completes the picture with 3.  Everyone at the table knows that this is to play.  Partner is not inviting game.  But partner might have KJxxxx and still bid this way.  You take a unilateral decision: 3NT.  Partner is done and won't be taking any further action unless it is to double a 4♠ call (which you won't mind).  Unfortunately, partner's hand is something like ♠87 J94 J97652 ♣Q62.  You can't set up the diamonds and a heart trick before they get four spades and the A.  Was this such a terrible action?  Yes, it was undisciplined (partner requested you to pass and you overruled her).  But there was a realistic expectation of making 3NT within the parameters of partner's actions.

    My contention is that these two types of disciplinary breach are not the same.  The first is a felony, the second a misdemeanor.  That's my story and I'm sticking to it :)