Wednesday, May 26, 2010

Who's overcalling?

Would you overcall in this situation (playing imps)?

red/white

pass - (1S) - ?

A4
94
AQT74
KJ84

Partner is a passed hand and you are vulnerable (the opps not).
Stepping up or not?

I'd say the mainstream 'expert view' is pass (WTP?).

Anyone who's been reading my previous post should know a thing or two about my views on the subject of 2-level overcalls, "the suit quality paradox" especially. Here we have above average strength and a less than robust suit, still I'd overcall for sure.

This is a hand from the Bermuda Bowl final 2009 in Sao Paolo. A couple of the most successful players in ACBL-land both overcalled. Eric Rodwell and Lynn Deas both came in with 2D, an action not duplicated at the other table in either match.

Was this a desperate attempt to create a swing? I doubt it. This was board 22 (of 128 deals in Open final, 96 deals in Women) and after the first stanza, USA led Italy 54-19 in the Open but trailed China in the Womens with 27-68.

What was the outcome of this 'frivolous' action? Not much actually. It was a partscore deal with your side having the highest partscore, double-dummy, in 4D and opps cold for 3S. Everyone sold out to 2 or 3S (overcall or no overcall) making 8 or 9 tricks. Partner had K72/872/K95/T932.

Which probably begs the question: What's my point then?

My point is that a couple of players some people maybe would consider a notch above "mainstream experts" seems (to at least some extent) share my view that this is a winning strategy. How about that?

Wednesday, August 26, 2009

Partnership defence?

Continuing where we left off, we could say that if declarer has 2-1, i.e. K of diamonds is the game-going trick, we will not beat it. Then our aim would be to make sure we won't fail to beat it when declarer is 1-2. If declarer's diamond is the J, partner could face a nasty decision. And if it's a low one, partner might have losing options as well. He doesn't know everything we does. Make it easy. Discard the Q of diamonds (the T would work also with the 9 visible in dummy). When declarer leads a diamond, partner will surely grab the ace and shift to something and we will know exactly how to cash out.

We could also say that we don't like to give up when declarer is 2-1. Can we trick declarer into going down, and minimize the risk of letting a beatable contract through if declarer's minor suit shape is reversed?

Our only option then is the 2 of diamonds (playing UDCA). By showing strength with the 2, partner will know we have the Q so if declarer leads a low, he'll take the ace. If declarer has Jx and believes you have the ace, he might try a 'chinese finesse' and let the J ride if not covered. Slam! The other variant is if partner gains the lead in clubs and shifts to a low diamond and declarer misguesses. Partner would never lead away from a holding including the J, so that might work. But, a suspicious declarer would probably get it right as with this particular dummy there is no hurry for the defenders to attack diamonds. So the only reason for doing that is 'trickery'.

Also, if we signal with the 2 and partner ducks a stiff J, declarer may put up the K and leave you regretting that one. On the other hand, maybe partner can deduce that you'd only try this from a 5-card suit, i.e. he'd duck the J from a Axx but win the ace from Axxx...?

Summing up: Going legit - help partner by ditching the Q. Possible con - signal with the 2.

This time either option would have worked as declarer held a glorious 9-count without the J of D.

KJTxxx
ATxx
x
Jx

As astout readers can gather from previous posts, the contract made at the table when the 5 was discarded and partner ducked the diamond.

Anyone wanting to add to this analysis in the 'search for truth' are welcome to post comments. We often learn the most from disasters.

Exploring 'the other side'

Sometimes trying to beat a contract is a lot of hard work. Some actual clues, some assumptions (which may be worth nothing in the end) and some imagination. Here's a chance to defend 4S from the other side.

_____AQ84
_____87
_____K97
_____T754
___________7
___________KQ9
___________QT532
___________K932
1S - 3H (artificial limit raise)
4S all pass

Lead is 2 of H, showing an honour 4th (or the lowest from xx, Polish style). Declarer captures our Q with the A and quickly leads a trump to dummy and a low towards his hand. We have to find a discard. What will it be?

For starters, what do we know? If partner has led from xx, declarer has at least 5-6 and he's always making. So, put partner down for Jxxx and declarer Axxx. Winning the ace of H immediately and draw trumps is really strange with late heart losers unless there's a sure trick source in dummy (to discard H's) or an abundant trump suit where late ruffs are always assured. Vs this particular dummy, this line of play therefore strongly implies that declarer has 6 spades.

Placing declarer with 6-4 in the majors, how about the minors? With 30/03 nothing matters, he's always making/going down, so we have to assume 21/12. With declarer assured of 9 tricks (six spades in hand, ace of hearts and two ruffs in dummy), we need the rest. A heart (check!) and 3 minor suit tricks which means that partner must have both minor suit aces.

If declarer has a singleton diamond and a doubleton in clubs it's easy, ace of diamonds and A-K of clubs. If declarer has two diamonds and a stiff club, we need to con declarer from winning a trick with dummy's K. Can we do that with this holding? Are we even willing to go for a possible con if this means that we risk letting it through when declarer has a singleton?

Tuesday, August 25, 2009

Conclusion

Anyways, our partnership agreement is that style 2 is the general methodology. So, declarer should have the Q. But, part of the attraction is the intellectual challenge/stimulation, right? Let's say we had chosen to adopt style 1 and take a closer look.

What can we deduce about the heart suit? Partner could always afford a heart with five of them, so declarer is very likely to have at least three. If declarer has at least three hearts, then game always makes if he's got the K as well, unless we take the ace and can cash three club tricks, i.e declarer having

KJTxxx
AKx
4
Jxx

That would give partner...

x
QTxx
QT5xx
K9x

... and I can't see the 5 of diamonds discard coming from this holding. I'd let go of my lowest diamond then to inform partner I have a high card there. That should help him see that ducking a diamond is useless and that setting tricks must come from other suit(s). The problem then is that partner wouldn't know which K to play me for. But that dilemma is solved by our second diamond play. We'd follow low with club interest and high with the K of hearts. This information is more important than length at this point.

So, after that diamond discard let's place the K of H with partner and it seems reasonable to place declarer with either the Q of diamonds or the K of clubs because otherwise he'd only have 9 hcp at most. (Sidequiz: would a heart discard indicate original length or show/deny another honour?)

Doesn't that mean we can now fly the ace of diamonds (anyway) and put partner in with K of hearts for a club through and collect the setting tricks?

The answer is maybe. What if declarer has a stiff club? Could he have a singleton? With a stiff K, declarer would probably have led a club from dummy at trick 3, or partner might have discarded a club from Jxxxx. But what about...

KJTxxx
ATx
Qxx
J

Now we have to duck.

Is 6-3-3-1 then a possible shape for declarer? I could see partner letting go of a diamond from a short holding with both K's.

x
KQxx
T5x
K9xxx

I'd discard the 5 of diamonds from this holding, not wanting to emphazise either sidesuit.

Edit: It has been suggested elsewhere that one would never discard from a three-card holding here, rendering this layout impossible. I say one should consider letting a diamond go from 532 as well. If you always discard from your longest suit(s), you'd be an easy read for a competent declarer. Here, if declarer has QTx (vs K97) this diamond discard might lead him astray, playing you for an original Jxxxx. Don't be too rigid in your approach here.

Summing up: The only time the setting trick could go away, by ducking, is when declarer has a stiff diamond and partner has discarded the 5 from QT532. Otherwise we just wait for declarer going down, getting one heart trick and either two diamonds and a club or a diamond and two clubs (catering to either layout). I can't construct a reasonable layout where partner would have QT532. And, as stated above, is that really possible with the actual partnership agreement? So, ducking must be right.

Anyone with a differing opinion?

Discarding - partnership signalling philosophy

The most common signalling agreement when discarding is 'discourage/encourage' or 'show/deny strength'. This can be done 'directly', i.e. by playing a card in that suit, or 'indirectly', i.e. by discouraging one suit we may indicate strength in another because otherwise we might have chosen to discourage that suit instead. That's the easy definition. In real life many other things may factor in. Does dummy has a trick source? What did the lead reveal? How many cards does dummy have in a particular suit, can I afford to discard one? And so forth.

A very important aspect is what general signalling philosophy that your partnership applies. The main 'sides' are:
1) Obvious-shift style where the key word is INTENTION.
2) Actual holding (what do I have), where the key word is POSSESSION.

There are layers of this, but an easy example of this might be when partner leads an ace vs a suit contract and I have a weak holding in that suit. Style 1 would encourage or discourage depending on the holding in the suit partner would be most likely to shift to if I discourage. Style 2 would discourage, as we have nothing to contribute, and leave it to partner to work out how to continue (with the information that I don't have any strength in this suit). There are advantages and disadvantages to both and exceptions, such as cases when you intentionally choose to mislead partner because you can see it's correct (by looking at your hand and dummy).

It might actually even be proper to have a different terminology for these styles. Perhaps the INTENTION-style is best described by 'encouraging/discouraging' and HOLDING-style should be coined 'strength/weakness'. That would also be better from a disclosure perspective as it would give the declaring side a better understanding of your signalling methodology, something he/she is entitled to.

Well, I digress. Back to the defence vs 4S. The reason for the above is that playing style 1, a high diamond (relatively, according to UDCA) here wouldn't necessarily say anything about the Q if he really want another suit/shift. It would only say that he's got no interest in you playing a diamond if you're on lead. Playing style 2, the [relatively] high diamond would (should) pretty much deny the Q. When looking at this particular dummy it's hard to find a case where an exception might apply.

Monday, August 24, 2009

What's in a signal?

Playing a top-bracket knockout at the Washington NABC's, red vs white, you look down at

65
J742
AJ86
AQ5

Playing a light, aggressive opening style, partner deals and passes and your RHO open 1S. First choice, pass or X?

Since you are outranked by the boss suit, you decide to pass and the bidding continues with 3H, alerted, on your left and 4S from opener. 3H is explained as being an artificial limit raise in spades and you decide to lead the 2 of hearts. Your agreement is 2nd/4th, Polish-style, so this promises an honour (4th) or is the lowest from xx. Dummy hits and this is the view:

_______AQ84
_______87
_______K97
_______T754
65
J742
AJ86
AQ5

The 2 draws the Q from partner and declarer wins with the A. Without further ado he draws trumps in two rounds ending in his hand. Partner follows low once and then discards the 5 of diamonds, playing upside-down count and attitude (UDCA). Declarer now leads the 4 of diamonds from his hand and we do what??

Initial analysis shows 8 hcp in the majors from declarer and a reflection that any late heart losers can be ruffed in dummy. 6 spade-tricks, HA, maybe a H ruff in dummy and the K of diamonds brings up the total to 9. If declarer has the Q of diamonds we can duck now without risk (not going anywhere) and get some more info. If partner has the Q, declarer might have a stiff and we need to grab the ace immediately.

What can we deduce from that diamond discard?

Friday, May 22, 2009

Fingerspitzgefühl

Nation's Cup is an excellent 2-day invitational tournament held each year in Bonn-Bad Godesberg. 2009 was the 24th occurence, all organized by expatriate Swede Göran Mattsson. Everything is top-notch and the format is national teams, Germany as the host providing two, divided in two groups meeting in each other in a round-robin with the winners from each group squaring off in a final. The other teams continue in a Swiss for the remaining prizes.

This year the field consisted of 16 teams, Canada and a bunch of Europeans, and Sweden was represented by Ulf Nilsson - BG Olofsson and PG Eliasson - Thomas Magnusson. Teammates played well but we didn't play up to par and a 8th place finish was a disappointment for sure.

Here's a fine display of "fingerspitzgefuhl" by PG in our 2nd match. First take a poke at the auction:

AT2
-
KT987643
KQ

All red and the auction starts (1S) - pass - (2C) to you with 2C showing 10+ with 3-card spade support. Yes, this is Europe and bids that doesn't carry the same meaning as in the club game are allowed. It seems pretty obvious to bring the diamond suit into the picture, but at what level?

PG opted for 'the full Monty' and jumped to 5D which was greeted by RHO with a red card. The lead was the K of hearts and a decent dummy came down.

K4
T4
J2
JT97654

AT2
-
KT987643
KQ

A surprisingly few number of hearts in dummy is the initial reaction. What can the opponents hands look like? We know spades to be 5-3 and hearts are very likely to be 5-6. If diamonds are 0-3 we're toast so assume 1-2. So this looks like 5-5-1-2 vs 3-6-2-2 with KQ of hearts to your left.

Ruff the lead and play the K of clubs to find out who has the ace. West wins and shifts to a spade. This pretty much guarantees that East has the ace of diamonds and PG played three rounds of spades, ruffing in dummy and a diamond to the king, dropping the offside queen for a gratifying +750 and 12 imps.

Full results are available here: