Thursday, January 31, 2008

Crackpot?

It amuses me that the theory of not overcalling in a lower ranking suit at the 2-level with top honor fourth in enemy suit was deemed 'crackpot' over at BBO-forums, one of my hang-around places. Oh well - I'm crushed ;-)

But, I'll expound on the reasons, feeling a bit more motivated today. First a couple of qoutes: "I hate experts who give opinions without any valid theory behind it." and "... just consider the number of saves/preemptions that this approach misses. After all, on the hands he passes, partner rates to be short in opener's suit and so may well have a fit for ours... and we are staying out????" and from the same post "...it strikes me as almost certainly a long-term loser, in a big, big way."

Well, a post like that, not giving logical reasons, may have been a bit provocative but also a sort of a challenge. If you've read my blog and keep coming back, maybe you think I know something about bridge and would think about why there could be any truth in my statement. Over at BBO, most struck out :-)

I first made this observation in 1998 playing an international high-level team event in Holland. I've been keeping track ever since then and I'm no stranger to top-level bridge and I'd say I also analyse more matches than most living people (PS-bridge/BB-records etc, owning most World Championship books since 1958). Am I clueless or how can this overcall be wrong?

The reasoning (theory) behind the approach goes like this:

1) When we have length and defensive strength in enemy suit, that decreases the chances of them having a fit and IF they do then our holding is a liability for them, decreasing the chance of them making anything on a higher level (especially if our partner has some values).

2) When they have the values to be bidding and their honors aren't in their suit (i.e we have at least one top honor), that means that they are located in our suits or sidesuits; potential defensive tricks if we are playing (and we have points wasted not pulling any weight offensively). So, their offensive potential has decreased and our offensive potential has decreased.

3) When partner raises, as we have 9+ cards (often 10 as in 6-4) in 2 suits, we're less likely to find useful honors with partner. Only trump honors and aces/kings in sidesuits may help as we're short (31/13/22/12 etc) in the remaining suits. Queens or KQ in our stiff suit may be a little late when they cashed out the 5 first tricks against say 3C.

4) When they have the higher ranking suit, we have to contract for more tricks then they do if we want to declare. This makes it even less likely that we can do so profitable and it will always be tempting for partner, who will then have shortness.

5) With 4-cards in their suit, they don't have a fit in that suit that often (haven't calculated by how much). This means that the hand is quite often a misfit, with points fairly even divided between the sides.

6) Do you want to declare or defend under those circumstances?

7) What happens most of the time is that we go down after an overcall instead of them.

8) Could still work out, but we want to take percentage actions, right?

On that initial hand in Holland almost 10 years ago however, partner was short and sacrificed in 5C over 4S. It turned out to be a phantom which isn't surprising at all. But this sac scenario isn't making up the majority of the losses I've seen. Getting caught or just buying the contract when neither side has a paying fit/contract and going minus declaring instead of plus defending, doubled or not, that's the biggie!

The way I gained in the deal in my previous post, by opps missing a fit, is very rare but it was a nice bonus for the correct approach. The one from Teneriffe was getting caught for -300 with teammates going down 2 vulnerable for -200 at the other table.

Theory is one thing, empiric observations over 10 years is another.

Cracking, or taking the pot?


Edit: To clarify, the cut-off point to me seems to be TOP honour. Jxxx hasn't rendered bad results in the deals I've seen so far.

Wednesday, January 30, 2008

Different view

My views on 2-level overcalls in lower ranking suits differs from the expert community at large. For one, I like 5-card overcalls and I think diffently about suit quality requirements (and more). That's for another entry or article.

One of my 'theories' is that you shouldn't overcall 2x over 1M with 4-cards headed by a top honour in their suit. Not even with a good 6-card suit of your own. It's not that it might not work out fine. It's that the odds of it not working out is so much greater that it's no contest. None. Zip.

I'm not in the mood right now for laying down the reasons why. Either accept it, reject it or think about it. As ususal, you decide for yourself.

(Btw don't confuse this with the writings of Mike Lawrence where length in enemy suit is considered a positive factor when overcalling. He writes about 1-level overcalls in a higher ranking suit.)

In 2005 I played the Open Europeans Champoinships in Teneriffe with Fredrik Nyström and told him about it. He was sceptical, of course, and the second day he couldn't contain himself and overcalled 2H over 1S with length in spades. The cost was 11 imps and that's one of the few times I've smiled while recording a double-digit imp loss ;-)

Anyways, this weekend I pulled up, playing imps:

Q743
J93
AKJ973
----

Partner passed and Cecilia Rimstedt opened 1S (11-16 5+) and I passed of course. How did this end up?

This was the full deal:

_____J86
_____T85
_____QT8
_____A754
5_________AKT92
KQ7642____A
5_________642
KQ632_____JT98
_____Q743
_____J93
_____AKJ973
________

They subsided in 2S after Pia Andersson responded 2H and passed a 2S-rebid. Cecilia made it while our teammates (Berteau-Nyström) played 4H making for +420 (the board was played 24 times, making a game 16 times). See how much easier to manage the auction after a 2D-overcall. Strange, right?

Some imps won feels better than others. Yet, I did nothing.

Sunday, January 20, 2008

Mirage revisited

No one is immune to the 'bridge mirage' syndrome. I repeat, NO ONE. Here's a deal from the last segment of the 2007 Bermuda Bowl final in Shanghai between USA1 and Norway where the mighty Zia fell victim. He had:

J73
T82
AT
97642

I watched on BBO vugraph when the bidding started 2C by Helness, pass by Zia, 3C positive by Helgemon and and partner (Rosenberg) overcalled 3S at all red. The bidding continued 4H to the right, 4S by Zia and 6C from LHO which was corrected to 6H, all passing.

N_______E______S_______W
________2C______p_______3C
3S______4H_____4S_______6C
p_______6H_____p________p
p

What would you lead?

I can't know what was going through Zia's mind, and I wasn't up to the task of asking him when we talked in San Francisco. A list of lead options might go like this:

1) Ace of diamonds. Unbid suit, we cash it, hopefully and see dummy.
2) A spade. Partner bid the suit viulnerable and should have something useful.
3) A club. Maybe declarer has a stiff and we can cut him off from the suit if he lacks an entry.
4) No choice.

Yet, Zia lead a heart. Why??

I can only speculate (imagine) how Zia's mind was racing off. Why no Blackwood? Dummy must have a void. Maybe void in diamonds as he would bid 5S with a spade void but can't show void in diamonds below 6C. Declarer removed to 6H, why? Maybe a club void over there. Then maybe partner can stop the club suit with the A/K there and we should lead a trump to remove a ruff/entry on the go.

Hrm. Yes, that's it. Let's lead the T of hearts in case dummy hits with xx and partner only has a lower stiff x (records say he led the 8 of H, but as I recall it when watching, his choice was the T).

_______AQT964
_______73
_______9843
_______5
K82_________5
____________AKQJ9654
7652________KQJ
AKQJT8______3
_______J73
_______T82
_______AT
_______97642

Not the best shot on the actual layout and it lost big against 6H going down at the other table after a more mundane lead (and no intervention). This is a prime example of the 'mirage', making far too much of too little hard evidence.

Well, knowledge of imminent danger (this syndrome can occur when you least expect it) can prepare us to overcome it, so:

Praemonitus praemunitus ("forewarned is forearmed") !

On a sidenote: As most probably knows, Zia is replacing Soloway on the Nickell team from the summer NABC on. So who is Hamman playing with in the Vanderbilt and the trials? What might be news is that Chris Compton, Dallas buddy and expert player, is the 'chosen one'.

Saturday, January 19, 2008

Mirage

A mirage (from the Latin mirare, meaning 'to appear, to seem') is something I can suffer from at occasion, and it stems from reading to much into the bidding and/or the opponent's choices and general quirks. Some sort of mutant form of table presence going wrong, diverting you from the main track. It sets your mind off on erroneous ways, causing you to stray from more straight logic or 'truth' (for lack of better word). This malady hit me on this hand from last Sunday when I had:

K32
10875432
7
kn4

Matchpoints, all white and LHO opens a weak 1NT (12-14), partner doubles and the next hand pauses, and then passes. Hrm.

I think the normal bid on this hand is 3H, showing a rather weak hand with long hearts and sufficient playing strength. Bidding 2H doesn't show the potential of the hand and could be just crap. Passing is way off for even contemplating, although it could be a winner.

And yet, 2H was what I bid. Why? Because I read to much into the pause without any knowledge of that person's skill level and ways. I asked about pass, which was just neutral and a suggestion to play 1NT. I got thinking that RHO probably had something like a 5-card diamond suit and some overall strength because otherwise, it would be better to run.

If RHO had some strength, this was maybe a partscore deal and partner was likely minimum for double and I therefore decided to just bid the nothing bid of 2H. See how wrong this thing went in my head? This was faulty thinking for more reasons than one. If my take on RHO is correct, a 3D bid may be forthcoming and I need to bid 3H anyways, but now maybe attracting a double if wrong.

So, I broke one of my own cardinal rules, "when in doubt bid the normal bid", here 3H, and deservedly paid a price when this was the layout:

______D10976
______ED9
______K10
______ED3
kn854______E
Kkn________6
E63________Dkn98542
K876_______10952
______K32
______10875432
______7
______kn4

I have no idea what the heck RHO was contemplating but it can't have been anything bridgerelated ;-) Pass was a really bad bid that worked fine here with me messing up with an probably equally bad bid and there is no excuse for me racing my mind in the wrong direction.

Result was a missed game - partner would have had an easy raise over a direct 3H.

Don't fall prey to the bridge 'mirage'.

Friday, January 18, 2008

Survivor

Everyone is familiar with the expression: "the operation was succesful but the patient died". My partner, Magnus Eriksson managed to get it right the other way around: "the operation was a failure, but it was only way to save the patient"!

First look upon these hands as a play problem:

AQ43
AJ
AKJ87
T3

K972
Q62
54
AK92

After you open 1C (red/white), the next hand jump overcalls 3H and there's no stopping your partner after you admit to a spade suit over a takeout double. She puts you in a grandslam in spades after checking the 'vitals'; how would you play it on the ten of spades lead?

I think the best way, taking the preempt into consideration, is to cash two rounds of trumps and go about establishing the diamond suit in dummy. With the last trump with RHO, you're fine. A, K and another diamond ruffed, heart finesse and another diamond and finally enter dummy with a club ruff to draw the last trump and enjoy your grand, taking the final trick with the 'beer card', the mighty 7 of diamonds! Ain't life great?

Not so fast! Magnus was dealt:

865
9
QT632
J865

... and decided to 'fake' a heart void by doubling 7S! He of course hoped to scare the opponents into 7NT instead, making them think a heart lead would then give him an immediate ruff and he liked his chances much better in 7NT with those minor suit holdings.

Some people can't be bluffed (as many poker players probably have learnt the hard way) and everyone sat for it without any particular deliberation. I was about to lead a spade but now of course led a heart away from my K empty-seventh instead.

Oh, yes. You are right. This deprived dummy of an entry prematurely and he was belly up. Amusing? Yes. Deserved? Maybe less so..

Sometimes being too clever for your own good, is exactly what's best for you!

Wednesday, January 16, 2008

Shifting gears

Temporarily reviving an old partnership, I played a 2-day pair event over the weekend with Magnus Eriksson. Neither exactly remembered the old relay club system that we used to get to the Spingold semi back then and we went with old familiar stuff, the Swedish 2-way 1C. Midway through the event, I picked up:

Ax
xx
xx
JT9xxxx

Partner opened 1C (11-13 bal or any 17+) and I bid 1D negative. Now I had a choice when partner rebid 1NT showing 17-19. Should I pass, maneuver to play in 3C, invite 3NT with 6+C or just blast 3NT?

People playing 4-way transfer would probably start by doing just that, but we weren't playing anything that sophisticated (dropped the relay club, remember ;-).

I opted for the straightforward 3C-bid; invitational. This triggered a slightly surprising 4C bid in return! What!?

Well, this was great news. My hand improved by a whole lot and I cuebid 4S and Magnus jumped to 6C, making 7. His hand was:

Jxxx
AKJx
A
AQxx

He picked off a couple of offside honors enroute to all the tricks and a shared top. The 1NT-rebid may appear questionable in some people's eyes. But, it isn't that strange in the 2-way club setting with a stiff high honor, because a 1M-rebid is still 2-way and may be a 3-card suit with 11-13. 1NT gives partner a better picture although it comes with a price with both majors considering it was matchpoints; a better partscore could easily go begging now.

As it turned out, bidding the correct, value bid (3C) led to the optimal contract. When in doubt, that (i.e. the bid that strikes you as normal/correct) should always be your choice.

Thursday, January 10, 2008

'Creative' effort

Against Thomas Berg, former teammate and the sports agent for rising female tennis star Caroline Wozniacki (http://www.carolinewozniacki.dk/), we got outmanuevered in a shrewd operation. He had

Q862
863
T5
9632

Vul vs not, partner opened 1NT showing 12-14 and the next hand doubles showing a strong hand. Any last thoughts?

Thomas found the creative 2H which was a success when we happened to play doubles as penalty over 2m and negative over 2M! I had an opening hand with 5H and eagerly awaited a reopening double that never came when Håkan had too many hearts and the contract drifted 5 off for a 2 imp gain vs 4H making +420 at the other table!

When successful, bids like 2H is called 'creative' or 'imaginative' or 'brilliant' (if Zia makes them), when unsuccessful, they're 'idiotic', 'lunatic', 'madness' etc. As always it's the outcome that decides the label... ;-)

Lesson of the day: play penalty doubles over 2M as well (if you don't already).

Wednesday, January 9, 2008

Little contract

There's a reason written up hands seldom features the small partscores. There isn't often an early decision time that is suitable for that format. Both sides win tricks and the variation of continuations makes bad reading material.

Last match in Denmark I played a 2C-contract though after opening a 2-way 1C (11-13 bal or any 17+) and partner bid a nonforcing 2C after a 1S-overcall. I got the ten of diamonds lead and saw this:

8764
Q63
J
AK753

AQ5
T94
AQ54
864

After the first trick collected the J, K and A, I paused for reflection. How should we make this?

Clubs need to be 3-2, giving us 4 tricks and we have AQ of D and the ace of S. This brings the total up to 7. The final contract-winning trick could come from an endplay of West, giving us an extra spadetrick, from being able to ruff dummy's last spade in hand, the AKx(x) of hearts onside or HJx of hearts onside and a defensive slip with West failing to cover the 9 of hearts from hand or

The endplay seems pretty remote as this requires West to have AK of hearts so that East can't push a spade through, the same requirement as getting a trick with the queen of H unless West is bare AK. We can rule out AK onside as West would have led one of those puppies with that holding (not necessarily true 100% of the time but something to rely on when analysing the hand).

The best chance is therefore ruffing a spade in hand, but then we need West to be 5-3-3-2 and the last club with East and sever the defensive communication so that East can't remove our last club in hand as we'll have cash ace-king or suffer the loss of extra trump trick promotion by the defence.

Having reached this conclusion, we cash the queen of D pitching a H and lead a H to the queen and K. East shifts to his singleton spade and we win and continue hearts and are safe from harm, scoring +90.

______8764
______Q63
______J
______AK753
KJT93_______2
A76_________KJ82
T96_________K8732
QT__________J92
______AQ5
______T94
______AQ54
______864

Check for successful opposing layouts and play for the most (legit) likely one. Surprisingly often this gains; at least you can rest assured you gave it your best shot!

Tuesday, January 8, 2008

Greater reward

The same match I picked up:
8
T98
AK642
AQ42

All red and partner passes, my RHO opens 1D natural (4+) and I pass. LHO jumps to 2S weak and partner surprisingly comes to life with a takeout double! Next guy passes and it's your call. What do you choose?

What are the choices? We could bid a feeble 3C or probe with 3D/3S. Maybe an invitational 4C or just stab 5C. What does partner have? This is a somewhat dangerous position for partner to get into, passed hand or not, and he should have some useful shape, maybe 5-card hearts.

Forsaking science for practicality, and with the extra chance that RHO might sacrifice with 4S with some hands he checked with this round, I pretty quickly decided to bid a confident 4H on my 'nice' T98. For 5C to be right it has to be a 2-trick better contract (11 vs 9) and I should be able to ruff some spade losers.

This was the layout:

_______A962
_______AJ62
_______T
_______9865
KQT753_______J4
73___________KQ54
53___________QJ987
J73__________KT
_______8
_______T98
_______AK642
_______AQ42

After West failed to find the trump lead (confident 4H instead of wiggling into it, maybe helped), instead opting for partner's suit, the hand played itself for a crossruff and +620.

Imps is sometimes, bidding wise, a strange mix of taking chances, going scientific and being practical with a sense for the auction. The trick is knowing when to pick the right club out of the bag and then the 'big shuffler in the sky' has a propensity for rewarding you.

Or am I only rambling to cover up dumb luck...?

Monday, January 7, 2008

Satisfaction not guaranteed

Technique in card play has its own rewards. By trying to cater for various layouts you may, or may not, get the prize. Here's a hand from this weekend that came close to making a difference in the scoring.

K
Q97
AJ42
87652

AQ9653
AK54
9
A3

I played 4S after opening a 2-way Swedish 1C and getting a 1D-overcall to my left and got a K of clubs lead. How to play it?

It seems like a rather trivial deal with 5 likely spade tricks, 3 hearts and 2 aces. If spades or hearts break or a squeeze developes, we may get more. Is there any danger?

Well, if neither major breaks we may have problems. Say spades are 1-5 and hearts 4-2. So we immediately win the club, ruff a diamond, go back with a spade to the K and ruff another D. Now when I cashed a high spade, my LHO discarded a diamond. Ace of H, heart to the queen and the last diamond from dummy assured 10 tricks and an inner sense of satisfaction.

That feeling lost some luster when this was the whole layout:

______K
______Q97
______AJ42
______87652
7____________JT842
832__________JT6
KQ763________T85
KQJ9_________T4
______AQ9653
______AK54
______9
______A3

With hearts breaking, anyone could take 10 tricks and you could actually make 6 by playing for hearts 3-3 before leading the last diamond off dummy (the only declarer in slam, in another match, went 2 down). Just another wash at the comparison (don't know which line the other table in our match took though).

Just beat the drum. Next time...

Sunday, January 6, 2008

Perfect play

Once again playing the top division in Copenhagen with Håkan Nilsson this weekend, I had this collection:

AT7
765
A72
JT98

Partner opens 1NT (14-16) and next hand overcalls 2S constructive. What to do at all white when X is negative and 2NT is artificial?

I stabbed 3NT as a practical shot and Håkan got a low spade lead and this was what he had to work with:

AT7
765
A72
JT98

J6
AKT4
J95
KQ76

Not too shabby with 14 vs 9. Seems the contract hinges on finding an extra trick in hearts, either playing for H/Hx with overcaller (cashing AK and leading towards the T later) or hooking the T, playing for both QJ to be onside.

We can stall our decision a bit and after winning the J at trick 1, we play a high club which LHO wins and continues with the queen of spades and another one, RHO following 3 times. Håkan now cashed one high heart (dropping the 9 to his left) and played 2 rounds of clubs, ending in dummy. LHO followed and RHO tanked before discarding an encouraging low diamond. How do you tackle the hearts?

Håkan figured it was more likely that overcaller was 5-1-4-3 for his constructive overcall as he seemed to lack a high diamond H than for him to be 5-2-3-3 so he cashed the last club and hooked the heart for a wonderful +400. When teammates were +110 in 2S this translated to +11 imps.

______AT7
______765
______A72
______JT98
KQ542_____983
9_________QJ832
KT83______Q64
A53_______42
______J6
______AKT4
______J95
______KQ76

It wouldn't have helped East to insert a heart honour as we can just duck and still have the ace of D for a re-entry.

When faced with a choice (guess), try to postpone the decision as long as possible and hope to gather some clues along the way. Sometimes this actually works.