Sunday, December 31, 2006

Worst bridge of 2006

Since we're all celabrating, I thought it was time for a laugh and post the worst bridge hand I've encountered in 2006. It happened quite recently and it's unbelievable what errors happened. If you know the site poorbridge.com (the name speaks for itself), this is one which is even worse than what you get there!

Here's the hand, I'm sitting South as usual:

Dealer:South
Vul:EW
Scoring:MP
x
QJ8x
976
T9864
Tx
K9xx
Kxx
AJxx
QJ97xx
Tx
QTx
Kx
AK8x
Axx
AJ8x
Qx

It looks pretty innocent, but wait until you hear what happened with it.

It all begins with the auction:
1! - pass - 1! - 1
1NT - 2 - 3 - pass
3NT - all pass

I have an easy opening, since we're playing a strong system. My partner shows 0-6HCP and RHO intervenes with . I have no problem with a rebid, and here we get our first two blunders: LHO supports with Tx while Vulnerable!!! Quite amazing since you know all finesses will be wrongsided and you don't even have an 8 card fit, but some people will do anything. Next mistake is the 3 bid from my partner. Since we play Lebensohl here, 3 shows a good hand while he intended just to fight the partscore battle. Dbl or 2NT would've done the job a bit better. Now obviously I like my Qx and raise to 3NT since I'm quite confident that I'll make it when s come in.

Now you better imagine (or put up) some music from horror movies.

LHO leads the T and I duck. It's a hopeless contract, but why not try to make the best out of it. A is continued and I take my Ace. Next step was imo to try and cash some s so I play A followed by a small . LHO hesitates, telling me where the K is, and decides to play low. This is the only thing that's done right, but includes the error of giving the position of the K. I'm in dummy and let 6 run for LHO's King.

Lets make it a bit of a quiz: can you figure out what card West played now?
For some reason she decides to return a diamond right into my finesse! Thanks, but I'm still nowhere. What else is there than to play another ? LHO takes her K.

Here comes blunder number 5, can you guess it?
She returns a heart to my J in dummy! Dummy was completely dead, so that's another trick for the good guys. I discard a and RHO has discarded 2 s by now. Still not home, I play my 2 remaining s and RHO gets it in her mind to discard another ! Probably she trusted her partner's raise too much, but this meant my 8 got promoted to my 9th trick.

Unbelievable: it took 1 mistake from partner to get me to a hopeless contract, and it took 4 and a half blunders from the opponents to actually make it! Only 1 thing was done right, but if LHO made the mistake of taking K immediately she wouldn't have had the opportunity to give dummy an unreachable trick. I prefer the latter for this article. Also my play was proven to be poor by GIB's double dummy analyzer.

Lets see how it went trickwise:
- If I duck T I have a possibility to go -2. Otherwise I can go -5 according to GIB!
- 2nd trick has to be taken, what I did. That's 2 things right, but now it starts...
- My only chance for -2 is to go for the s which I didn't do. This costed me 1 trick.
- to my Q was another trick I threw away.
- Now I got back a trick with almost every card I played.
So basically I should've gone -4 thanks to my poor play as well. However, taking the right line of play to go -2 would never bring my contract home...

All other tables made 1NT-1, I made 3NT= in the same direction.

That's it for 2006, I hope you've enjoyed my articles (why doensn't anyone sign my guestbook?), I'll probably keep encountering lots of interesting material to post on my blog next year, and you all have my best wishes for 2007!

Thursday, December 28, 2006

Newest MOSCITO verions

MOSCITO is a completely artificial system, but it tends to show suits whenever it can. It's imo more "natural" than precision, because after every limited opening a 4+ card suit is known (precision 1 is usually 2+, but in some variations it can be even shorter). Australia's Paul Marston is the main ambassador of the system, his current CC can be found here

The latest changes are the following:
- 2/1 responses have always been a problem. He went from NF to GF and now he uses transfers (from 1NT and up). These are handy particular in constructive auctions where you want to show a side suit before supporting, or when you want a natural auction. It also gives the possibility of 2 kinds of 2-level raises: transfer to the Major is 9-11 and constructive, 2M is 6-9. Relays are great for slam investigation, but for games they don't have an edge. You can't find stoppers for 3NT, so sometimes you get to wrong games. This is imo a big improvement, because it still keeps the possibility to signoff (although I think he uses them only with good hands).
- He now uses the step-relay over 1 and 1 openings as 2-way: either GF relay or a normal 1NT response. He uses a natural relay scheme, which allows him to play this way. With an artificial relay scheme (like I play) it's nearly impossible to handle. What can responder do after 1-1-2 (showing a single suited hand with 6+) when you know 2 is a relay? With natural responses it's easy most of the time. The relay structure has lost some efficiency imo, but the tradeoff is probably worth it.
- 1-1 as any GF (except 5440's) and 1-1 as double negative, other responses are semi-positive. I always liked this structure, because you get lots of interference over 1 openings. When responder shows a negative, RHO can ruin the auction and cripple our abilities to fight partscore battles. When you know that opposite a 15+HCP 1 opening, responder has a semi-positive hand most of the time, and rarely a double negative, then it pays off to get those semi-positives in right away (I think it's GF 30%, SP 60%, DN 10% or similar). There's hardly a tradeoff, it's a pure improvement: the relays for GF auctions go 1 step up, but those for semi-positives go down at least 1 step. So percentagewise the relay structure improves.
- 4M-6m hands are now always opened in the minor. Before, you had a choice of opening the Major, usually it depended on the suit quality of the Major. I still like the MAFIA approach, but handling competition is probably one of the reasons why he changed. Missing a nice 6-3 minor fit to play in a 4-3M fit isn't always the best choice.

Friday, December 22, 2006

Kickback Turbo

Since Fantoni-Nunes use Turbo for their slam investigations, many people wonder what it is exactly, and how it works. Short: it is a cuebidding method, where skipping the 4NT bid means you have an odd number of keycards, while bidding 4NT shows an even number of keycards (similar principle like serious 3NT). There are a lot of different situations where Turbo can be bid lower (like when a fit is established at the 2-level, 3NT will usually be the "Turbo-bid") but this doesn't change much to the method. Usually it's clear how many keycards your partner has exactly because of the auction. Finding the trump Queen is done in several ways. Usually you can just continue cuebidding and suddenly stop at 5-level. Now your partner knows you were looking for the Queen, otherwise you'd just signoff immediately. Another way is to bid 5NT, asking partner to bid grand slam if he holds the trump Q.

I've noticed there are some problems with the normal method, and I'm not sure how Fantunes cope with that. For example, if you set as trump with 4, you can't figure out if partner has the Q when you are at the Turbo-stage. A simple auction, typical for their system:
1 - 3
4 - 4!
4! - 4NT*
???

3 shows 10-14 with 5+ and 5+
4 sets trump
! is cuebid
* is an even number of keycards

Opener can't know for sure to go past 5 with a keycard missing and not holding the Q. Responder can't go past 5 because they might be missing 2 keycards.

The problem is typical for fits, and sometimes comes up with fits as well, if you can't bid a cue in between.

Me and one of my f2f partners have found a solution to this problem: we play Kickback Turbo. It uses the same principle but other bids to skip:
- With a fit: 4
- With a fit: 4
- With a fit: 4
- With a fit: 4NT
4NT becomes the cuebid in whatever suit the Turbo-bid was. So in our auction from above, we get:
1 - 3
4 - 4*
4! - 4!
4NT! - ?
Now, since opener kept cuebidding after knowing the number of keycards, responder should go past 5 if he holds the Q, because opener was still looking for slam. If responder bids 5 he denies the Q and opener can make the right decision.

We use some other slam tools as well. Once we have set trumps, we have 3 main stages:
- serious or no serious slam interest
- number of keycards (odd / even) = Turbo
- looking for trump Q
Whenever we're at a pretty low level, all these bids go down a bit. We use following bids whenever they are available: 2NT, 3 trumps, 3NT, Kickback-Turbo-bid, 5NT. So for example, if you set trumps with 2:
- 2NT is frivolous, no serious slam interest
- 3 = Turbo
- 3NT denies the Q
if you set trumps with 3:
- 3NT is frivolous
- 4NT is Turbo
- 5NT asks the Q

Kickback Turbo doesn't have an immediate disadvantage imo. It is more difficult sometimes, because you have to think about the cuebids in another order (if partner skips the Turbo-bid, it doesn't mean he doesn't have a cue in that suit, you have to wait until you hear 4NT or not). Once you get used to it, you don't make mistakes against that anymore.
The advantage is clear: you show your number of keycards lower, and can more easily investigate about the trump Q. By using this method, you don't even lose any cuebids before the critical level!

Monday, December 18, 2006

Declare or defend?

A little quiz today. Look at following hand:

Dealer:South
Vul:NS
Scoring:imps
KT765
97
J8
9632
-
A32
QT976543
87
AJ983
J4
K2
QT54
Q42
KQT865
A
AKJ

Short question: when you're West, looking at all hands, do you want to declare 5* or defend 4, and why do you think you're right?

[edit:19/12]
Ok, the quiz has been solved. You want to declare 5* because 4 is laydown. In the comments there was suggested to endplay East. There are several ways to do this (basicly after you've played you'll run trumps and endplay East):
- you can start with a small to the King. Then you have 2 ways of endplaying him: either with or with .
- you can also start with a small and let it run! Later you play Q and East can't make a good decision.
The only thing you shouldn't do is start with Q, since East can hold up and there's no way to endplay him anymore.

Tuesday, December 12, 2006

Lead directing psych

Yesterday I was playing with some friends and at one point I got the following hand:
A97
Q987653
-
J52
We're NV vs V, RHO opens in 1st seat with 2NT (20-21). What do you bid and why?

I thought opponents might have 4 or a slam in , so why not mess up their auction. If they bid to 4, the contract will be wrongsided and partner can immediately give me a ruff. Perhaps I can even reach his hand for another ruff for -1. So I bid 4 which turned out to be a success. If opponents don't double (unlikely) I'll be happy to go -9, otherwise I can safely return to 4H. It's always exciting to bid natural with a void, especially if you expect partner to have lots of support and he should have the discipline not to raise...

The entire auction went:
2NT - 4! - Dbl - pass
pass - 4 - Dbl - pass
pass - pass

Now the psych has been exposed, but opponents can't make things right. South did a great job not to bid 4 at any time, and take his money on 4x down as much as possible.

The full hand:

Dealer:North
Vul:NS
Scoring:imps
85
AK
AQT9
AK973
KJ
J2
J654
QT864
A97
Q987653
-
J52
QT6432
T4
K8732

As you can see, when is played by South a lead will defeat the contract if I can manage to play low when declarer plays (I should be able to find that, but we'll never know for sure). 4 by North is laydown. 4x-2 cost us 300, which is way better than 4=. Some however got to 6 or 6, I've even seen a 6NT!

Monday, December 11, 2006

Guard squeeze

This weekend I had a bad day in competition. Half the match I was asleep, got a game handed over to me but didn't accept their offer, and left lots of overtricks to opponents. We still managed to win 15-15 (0-3 in imps) but that no consolation price... Anyway, I still had some good moments now and then: a suicide criss-cross squeeze for -1, and here a guard squeeze, also for -1:

Dealer:South
Vul:NS
Scoring:imps
J4
Q8
QJ52
T9764
K97
T9743
A
KJ83
Q8652
652
K9763
-
AT3
AKJ
T84
AQ52

I end up in 3NT by South. LHO leads a so I like it pretty much if behave. I take the first trick with Q and immediately play a small , but RHO shows out! Quite a disappointment, since I don't see an immediate way to 9 tricks now. 3 + 1-2 + 3 + 1 = 8-9 tricks, but losing 2, 2 and at least one in another suit will get their trick total to 5 more quickly...

I play the Queen to keep control which LHO takes the K, and he continues T. Now I'm in my hand and decide to make it a bit hard on LHO: I play a small . GIB says I can make only 7 tricks if he takes the J, so here's a little mistake from LHO: he plays small! He didn't have much information and it looked like I needed 3 tricks at that time so he wanted to block my suit. Now however he has to guard the Jx... I switch to to the Ten and Ace. LHO again plays a high . So I continue to RHO's King (LHO discarding a ) who plays a small . I already know everything I need to know: LHO has a 3-5-1-4 distribution, discarded a after asking them, so he has the K. I can easily squeeze or endplay him, so I take A and now it's time for some magic: the 3rd squeezes LHO. He has:
K
73
-
J8
Here are his options:
- If he parts with a , I'll endplay him in to bring me the final 2 s.
- If he parts with a I can take my tricks.
- So the only alternative left is discarding K and hoping his partner has QT behind the J.

On one of the s RHO discarded a , so after I play A and small to his Q, he still had to bring a last .

3NT-1 still won us some imps, unlucky my squeezes were for -1 every time...

Tuesday, December 05, 2006

Quite an interesting holding

Here's one from this weekend. Dummy holds:
QT8x

You hold:
K9x

Because of the auction you know that LHO (= declarer) has 5-6 s. He leads Q from dummy (fixed typo tnx to Binkley). What should you do?

I had a split second to decide what to do without giving away too much information, and played low because it just felt right. Afterwards I thought about the holding, and realized low was indeed the best. There are a few interesting holdings partner can have:
- x or void: it doesn't matter, you won't make a trick.
- A: if you cover, you give away a trick.
- J: if you cover, you'll make 1 trick. If you don't cover, you'll make 1 trick as well, unless LHO decides to play the Ace dropping partner's Jack (but why would he play like that?).
So it only matters when partner has the A stiff.

What about when LHO only has 4-5 s? Then partner can be singleton or doubleton.
- xx or x: doesn't matter, you won't make tricks.
- Jx or J: doesn't matter, you'll make 1 trick.
- Ax: again it doesn't matter.
- A: if you cover you give away a trick.
So again, the only situation it matters is when partner has the singleton Ace, and you still shouldn't cover... Nothing changes!

There are even extra (small) advantages of not covering: in some rare occasions, you'll be able promote your 9.

Sunday, December 03, 2006

One of those which can win matches

Playing a team match, you pick up this lovely collection:
A54
Ax
x
AT98xxx
You're in 3rd seat, V vs NV, and you're playing a light opening system. The bidding starts:
pass - 1 - ?

I guess most will bid a simple 2. Bidding any higher with 3 Aces and such distribution is quite useless. So we continue:
pass - 1 - 2 - pass
2 - 3 - ?

Now what?
I know partner has less than 10HCP, and I thought partner had a good holding, and probably some tollerance as well. Since I have such control over all the suits, I decided to just bid 4, my partner can bring home tight contracts.

Here's the full hand:
QJT9x
KQxx
QTx
x

A54
Ax
x
AT98xxx

The play was quite easy, and we made 11 tricks. The other table didn't find 4 so we won bigtime with this board.

My partner decided not to open 1 with 10HCP although we usually open 5 card Majors with 10-14HCP with very few exceptions. I guess he didn't like his slow values, although he has both Majors and no bidding problems can occur. After a 1 opening I think I'll start slow but we'll still end up in 4.
I don't know why opponents didn't find it...

Monday, November 27, 2006

HEEMAN after 1NT

After a 1NT opening, many players play Stayman and transfers (usually 4 suit transfers). There are some discussions if you should use 2NT as invite without Majors or not, do you have to go via Stayman to invite if you play 4 suit transfers,... What do you do with invites with 4M and 5+m, 5M and 4+m,... And what with GF hands? Do you include 5 card Majors in your 1NT opening, and how do you search for them? Can you show a weak hand with BOTH minors? How about both Majors? How can you have a decent slam approach?

Many more questions arise, and no system is perfect. One day I found HEEMAN and I immediately like it! You can find a full writeup in English at http://www.jackbridge.com/pdf/eheeman.pdf. There's a Dutch writeup here: http://www.jackbridge.com/pdf/dheeman.pdf.

The main advantages imo are:
- it has lots of possibilities and is VERY easy to remember!
- opener usually plays and there's not much known about his distribution.
- many invitational sequences are possible.
- you can find weaknesses which will cause 3NT fail. So you'll be able to play a better contract.
- easy for slam auctions, and there are many ways to show slam interest.
- you can show 5-5M below 3NT.
- because responder shows his hand, opener can deviate from the normal NT distribution.

The biggest (and only) disadvantage imo: you usually can't find a 5 card Major in opener's hand. The only possible time is when responder shows an invitational hand with the other Major. Now opener can bid 3M with a maximum hand only.
This however is not that bad, but sometimes you go down in 3NT while you had 4M, and sometimes you make 3NT while 4M goes down...

I play this with my Fantunes-partner, where 1NT (11-14) can be all 4333/4432/5332/4441/5422/6m322 shapes! I haven't had much problems, but sometimes it's not easy to know what bids will work out best.

We've made one little adjustment:
1NT - 2
2 - 2
???
(responder shows an invite+ hand with at least 4-4 in the Majors)
Here we show immediately if we have fit or not:
2NT = minimum without fit
3 = maximum without fit
3 = maximum with fit
3 = minimum with fit
3 = minimum with fit
3NT = maximum with fit
This makes it easier for responder to try for slam or not.

We also don't like the super accepts, because I don't see the reason why you have to play at least 3M when you have a minimum hand with a good 4 card support, or a 4333. We only super accept with maximum hands and show our shortness (so no super accepts with 4333's).

Tuesday, November 21, 2006

Respond to asking bids!

A disaster from last week I want to share with you. Playing MOSCITO, the bidding starts:
1! - pass - 1! - 2
2NT - pass - 3 - pass
3NT - all pass

1 = 9-15HCP, 4+, can have longer m
1 = relay with invite or better hand
2NT shows stopper and maximum hand, probably pretty balanced

3NT down a few, 4 went only 1 down I think, but both contracts can be made if you find the stiff K...

First of all, my hand:
x
Qxx
QJxx
AQ9xx
Partner blamed me for not bidding 3 instead of 3, to show a 3 card support. Imo he knows nothing about my hand, so he just has to respond to my question, "I'm the captain of this auction, I ask, you respond", end of story. What are the possible meanings for a 3 bid? I guess there's only one sensible one (since we're in a GF auction): asking for a double stopper. So bid 3NT with a double stopper, bid something else without it.

Now partner's hand:
A9xx
AKJxx
x
Jxx
Clearly no double stopper , so a clear 4 response (maybe even 4!). Partner said he can't know that I have a 3 card support, but he doesn't need to know this to respond to the question. I know why I ask the question, he doesn't.

Perhaps you're interested why I don't find 3 a good bid. Well basicly, I think 3 should show a 4+ card with slam interest! With invite/GF hands 4+ card support and no slam interest we immediatly start with a 2 response, so what else can 3 be after such an auction? Opener probably has a 4-4M, so why search for a 5-3 fit if he can hardly have a 5 card ? The only thing that interests me is a double stopper, without it I'll signoff in 4 which I thought would be a 4-3 fit (which is pretty ok with singleton in the short hand).

Sunday, November 19, 2006

Close one

Here's one from a club evening friday evening:

Dealer:North
Vul:EW
Scoring:MP
x
KJxxx
QJxx
Axx
QT9x
xxx
KTxx
xx
A8xx
Ax
x
KQTxxx
KJxx
Qxx
Axxx
Jx

Playing f2f with my MOSCITO partner, he opens the bidding with 1 (showing 9-15HCP, 4+, can have a longer minor) and RHO intervenes 2. I have an easy takeout double, LHO passes and partner bids 2 showing 5-4 or 4-5. I bid 2 showing an invitational hand with 3 support (with weaker I would've just supported immediately, even with 4 - it's the system). LHO and partner pass, and RHO balances with 2. Now what? Well, I decided to just double, since I wasn't sure about a good fit. Partner decided to pass it in. I disagree with this decision, because we have a double fit and it seems logical that opps will also have a double fit. Playing 2 anything will probably be easy, and they also know where to find the s. Even though we're NV vs V and may have an easy +200 it's too risky imo.

Anyway, the full auction:
1! - 2 - Dbl - pass
2 - pass - 2 - pass
pass - 2 - Dbl - pass
pass - pass

I start with A and don't like what I see in dummy! My J will be caught in a finesse and I just made K high (and chance is that RHO doesn't even has one!). RHO seems still to have one, so I switch to . RHO takes his Ace and plays a small . It's imo quite clear that he wants to try to discard a on his K so I take K and play Q and another . RHO ruffs and plays K which partner takes. He returns a which RHO takes with the Q catching my J and he plays for some reason small to his Queen!!! Because we cut off declarer's communication, I think he thought he could only make it if s would split 3-2, which is virtually impossible. The winning line of play is to just run and make me ruff... Now he's completely screwed. He cashes K and ruffs a with trump Ace and plays T. This is what's left in dummy:
T9
T
And I still have:
Jx
x
Partner still holding QJ we still have to make 2 tricks. If I discard my I'll only make 1 more trump trick, if I ruff with the Jack dummy discards his losing , so the only way to make 2 tricks is to ruff low!

+200 was a nice top, and both me and my partner had a bit of excitement because the downtrick seemed very far away. All decisions turned out to be great: the switch was necessary to make declarer ruff, I had to take K immediatly because the loser would've gone away, it was very good from partner to take A immediately and continue destroying all communications, and I found a way in the end to make 2 tricks by letting declarer overruff.

Tuesday, November 14, 2006

Heavy preemption handling

This weekend there was the Belgian Junior Championship, the last time I could compete since I'm too old now. It wasn't a success, but we had an amusing day. The boards (computer dealt) weren't great imo, since too many finesses were onside (I think about all of them), too many good splits, and whenever trumps split 4-0 it didn't even matter. So basicly there weren't many playing skills necessary. On the other hand, they were still fun because you get some distribution and competitive action. Here's one of the most extreme ones (I like this auction the most out of 60 boards):
NV vs V, South dealer:
JT
-
AQJ654
AQT76

A
AQ764
K32
J985

Not that difficult to find a minor slam if you have silent opponents. However, after 1 round the bidding was at 4!

1 - 3 - Dbl - 4
???
My 1 opening shows 14+HCP unlimited (we played a modified version of Fantunes). Dbl is obviously for takeout.

I thought that Dbl from me now would be takeout but this wouldn't show such great holding in . So I bid 4NT, showing some hand with tolerance for 3 suits, implying a control as well. Now my partner found a great bid: 5NT! Bid a slam he says, so I bid 6 since I'm still a minimum.

The full auction:
1 - 3 - Dbl - 4
4NT - pass - 5NT - pass
6 - all pass

If I was a bit stronger I would have to make a decision to bid grand or not, but luckily I got spared of that one. We probably would've missed grand if I had K instead of Q, so you can't say that preempting doesn't work.
Btw, as I mentioned above, there was 7NT laydown because K was singleton onside! Where's the challenge...

There were only 2 pairs who found the minor slam, so this was a good board for us.

Congratulations to Willem and Rien, my f2f MOSCITO buddy and a friend, who won the championship for the 2nd time!

Tuesday, November 07, 2006

Nicely defended

Here's one from competition this weekend (rotated for convenience):

Dealer:East
Vul:None
Scoring:imps
QTx
Axxx
Txxx
xx
Kx
KQTx
Jxxx
Jxx
Jxxxxx
Jx
AQ
ATx
Ax
xxx
Kxx
KQxxx

The bidding is quick and simple: - all pass. 2 shows 12-14HCP with 5+.

I thought my best chances for defeating this contract were in leading a and either set up an extra trick or give partner a ruff. So I lead K, partner shows an even number of s and RHO takes the Ace. Declarer probably was afraid of a bad split or K behind him, and decided to try to set up some s first... I show an odd number of s and partner takes the second with the Ace. He now returns to my Queen and I give partner a ruff, obviously showing prefference. Partner nicely returns and it's all over for declarer. He takes the Ace and plays a low to the K. I now know that he held 5-6 s, exactly 2 s (why else would partner take the A the 2nd and not the first or 3rd time), exactly 3 s, so 2-3 s. I close the screen and ask declarer more information about his opening: can it be balanced? He says they usually open 1NT with balanced hands, so now I know his exact distribution unless he gave false info: 6-2-2-3. The only chance of defeating this contract is by giving partner trump promotion by playing in a triple void. Before doing this however, I need to cash K first, otherwise declarer just ruffs in dummy and can discard his losing from his hand.

2-1 is the result.

Declarer made the biggest error by being too pessimistic about trumps. Thanks to the K lead he only has 1 loser, 1, 1 and shouldn't be that afraid of a ruff (it's probably the 3 card trump that will ruff), so he'll have maximum 2 losers for the contract.

Saturday, November 04, 2006

Thin grand

This one came up during partnership bidding, but it's an interesting one for declarer play:
Axx
AJxx
AKxx
xx

K
KQxx
xxx
AKxxx
We ended up in 7 by South, because North was a bit agressive.

Perhaps an interesting question: What lead should you be afraid of? I'll come back on that later, because if you know how to play this hand, you'll know the answer to this question.

On this hand I think you'll get a trump lead, which is usually a good choice against grand slams. It doesn't matter much, but realise that some suits split nicely and some badly. Plan the play!

First thing we notice: trumps will need to split 3-2 or 3-3. You'll need to set up sooner or later to get rid of your losers in the North hand. First things first, test trumps. How? If trumps split 3-2 you need to be able to handle a 4-2 split. So you need to keep exactly 1 tophonour and a small trump in dummy. Wherever you end up, it seems trumps split 3-2. Now we need to develop the while there's still a trump out, since we need to handle a 4-2 split. AK and small, LHO still follows. For safety reasons you ruff with the tophonour you kept. Success and a problem at the same time: RHO shows out, so LHO still has a and there's still a trump out. Is all hope lost? No, there's still a possibility: if RHO started with 2 trumps, he won't be able to overruff, so you return to your hand with K to ruff another small. A very exciting situation when RHO decides what to do, and luckily he can't overruff! The situation is as follows (you're in dummy):
Ax
-
AKxx
-

-
Kx
xxx
x
The safest way now is to ruff a , draw trump, play the last and cash your remaining high cards in dummy.

Now back to my question earlier: what lead should you be afraid of?
A spade will kill the communication to your hand and you'll need 3 in LHO's hand to get to your hand. This however won't work, since he had a stiff...

Wednesday, November 01, 2006

Problem hand

Here's an interesting one from BBO. Playing MP, all Vulnerable, the auction starts at your left side:
1 - 2 - pass - pass
Dbl - pass - pass - ???

You hold the following monster:
97543
QJT953
-
Q6

What do you bid?

Personally I think passing is very dangerous, even though I trust my partner's overcalls. My hand is worthless in a contract and the trumps will split very badly.
So the alternatives are:
- RDbl: risks partner bidding 3, so what's next?
- 2: risks that partner has a stiff and you won't have any trump control, so the hand is again useless.
- 2: this was my choice at the table.

I think 2 should show a 2-suited hand with good and a poor second suit (otherwise you have another bid), so partner can scramble if he doesn't have tollerance for . The bidding continued:

1 - 2 - pass - pass
Dbl - pass - pass - 2
Dbl - pass - pass - ???

Now what? Should you bid again?
Obviously partner had an easy way out (RDbl or a bid), so pass accepted the contract. End of the auction.

Partner held:
Q
A4
A987654
J83

We found our 8 card fit (we knew how trumps split, but ok), and the best partscore was reached:
- 2 goes -3.
- 2 also -3.
- 2 only -1.

Thursday, October 26, 2006

Impossible slam?

I haven't had many interesting hands lately, except this one:

K7xxx
K9
AKQT9x
-

LHO starts the bidding, all Vulnerable (Matchpoints):
1 - 2 - pass - ?
With that partner in particular, I don't play fit jumps, and I didn't have enough trumps to splinter imo. So I just decided to jump to 4 and hope for the best. Partner made +2 thanks to a defensive blunder (AQ was behind the King, but somehow the guy decided it was a good idea to discard the Queen at some point), so there's only 5 available. Great one may say, but here's partner's hand:

xx
AJTxxxx
J
xxx

We had 6 laydown if it's played by me (K protected, Q stiff onside). The problem is: how to get there? It's quite a weak hand to jump to 3 at this vulnerability although you have 7 of them. Either way, you'll probably never end up in 6 with a 6-1 fit while you have a 7-2 fit...

Anyone who can give me a convincing auction, be my guest. :-)

Thursday, October 19, 2006

Development squeeze

This one I played yesterday (rotated for convenience). It's very easy to squeeze your opponents on this hand, but it just doesn't come up very often. According to a source on the bridgebase forums (= Ben) it's called a "development squeeze". Here's the hand:

Dealer:East
Vul:None
Scoring:imps
3
KQ83
AKT97
JT4
J54
J765
864
983
AQ9762
A2
QJ5
Q6
KT8
T94
32
AK752

The bidding went:
1 - 2 - pass - 2*
pass - 2NT - pass - 3NT
all pass

So I'm in 3NT, and LHO leads 4 to the Q. You can't hold up, so you take K immediately. Now what? You can play on the double finesse, but the odds are against us because of the opener on the right. The only real alternative is to play on first and hope that suit behaves. If that's the case, you have 8 tricks and perhaps some possiblities for a 9th.
You continue a small to the Ace and let the J run. RHO is smart enough not to cover immediatly, but you're still playing the odds. When the J holds, you have only 1 chance of making 5 tricks and that's if they split 3-2. You're lucky and continue to play off all your s.
On those s, RHO gets into "trouble": he has to keep QJ or he gives away 3 tricks, and has to keep A or I get 2 tricks for free. The only thing left to do is discard s, which are high (but he doesn't know that for sure). Because of these discards I can count out the hand, and play to the K without fear, to develop my 9th trick.

Saturday, October 14, 2006

Money Bridge on BBO

I've been playing Money Bridge for more than a week now, and it's quite an interesting setup imo.

For those who don't know what Money Bridge is:
- You play against 1 human player.
- No kibitzers are allowed at the table.
- You both get coupled with a GIB, a very respectable computer player. You have to play it's system, which is some sort of 2/1GF without too many gadgets.
- It uses total point scoring.
- The stakes are chosen between 0.1 cent and 2 cents per point.
- You need to be able to pay for a loss of 2500 points (which is also the maximum per hand).

It's definetly cheat proof, and skill should be rewarded in the long run. If you're only playing a few games however, you'll need some luck, since it's total points. In other words, if you don't get good hands, you'll probably lose money unless your opponent bids too high every time. You have to put trust in the dealer that it will give you as much points as your opponents get.

What tactics should one use? Well, basicly you should bid and play like you're playing imps:
- V games need 35% chance.
- NV games need 43% chance.
- Small slam needs 50% chance (any Vul).
- Grand slam needs 67% chance (any Vul).
- Try to make your contract, overtricks are worth only a few more cents so don't risk your contract for these.

At the moment there are still some minor problems with GIB's actions. Sometimes it bids crazy and there's nothing you can do about it. You won't get refunds or anything similar from BBO, so I guess you just have to hope it happens to your opponents as well so you can get your money back that way. I've seen him bid a vulnerable 6 with 0hcp after a 2 opening with interference (which would go -1). But guess what: my GIB partner sacrificed in 6 and my human opponent went to 7, promply doubled by my GIB partner. Who's fault was that really? 5 made, so apparently he made a pre-sacrifice, and his human partner ruïned it.
I've seen even worse things at my side: after a 1 opening he responded with a 5332 with 5 s from QJxxx and nothing else. After my reverse he showed 8+HCP by rebidding ! 3NT-3 was the final result, losing 300 points without good reason imo...

Perhaps some of you get a negative feeling about GIB, but not everything about GIB is bad, in contrary. I really like to see him play! He seldom makes mistakes and usually makes a trick more than you'd bet your money on. He also finds killing leads and defenses, which all make up for some poor bidding decisions imo. The poor actions are more exception than rule, so it's definetly a worthy partner to have!

One note: don't play high stakes if you don't have enough money. I've already heard someone complain that his first experience was a turnoff. His opponents played 2 games: a vulnerable game and a vulnerable slam, both made. Result: lost some money (not the end of the world), but the most horrible part was that the table got closed because he didn't have enough funds to pay for a possible disaster (2500 points). So he didn't get a chance to win his money back unless he put more money on his bankrole...

Monday, October 09, 2006

Don't push me

Playing a team match, you hold following hand:
-
AT863
QT42
AQ85

The bidding starts with partner showing 11-14HCP with a pretty balanced hand (can be 4441, 5422 and sometimes 6m322):
1NT! - 3 - Dbl - pass
4 - pass - ?
Do you still try something?

I decided not to because of my poor holding, and huge chance that partner holds some high s. Now however, my LHO came in with 4 and partner bid 5! This clearly means he hasn't got any lost values in , so I bid 6 because he had to have 'something' in .

The entire auction:
1NT! - 3 - Dbl - pass
4 - pass - pass - 4
5 - pass - 6 -pass
pass - pass

Partner's hand was:
A85
QJ975
A76
K6

Slam was laydown after the working finesse. Now my LHO complained that I first passed on 4, and later bid 6. Imo he has no reason to complain, because it's because of him I bid this way. He pushed us by giving us more information than we were entitled to after his partner's great action: he passed first, and later he supported. This gave my partner the opportunity to show his good holding without wasted values in .

A good rule with few exceptions: either support immediatly and to the right level, or never support partner for the rest of the deal.

Wednesday, October 04, 2006

LEHO & HELO carding

There are many carding methods around, of which most focus either on showing count or on suit preference. Lots of experts like some sort of count, and lots of them hate it and swear by suit preference, because count gives away too much information to declarer. I'll leave that discussion for another time.

There are at least 2 methods which allow you to show both count AND suit preference with a single card! They're called LEHO and HELO, acronims for:
LEHO = Low Even High Odd
HELO = High Even Low Odd
They work best when playing UDCA or something similar, and when partner leads a suit where you have some length.

Suppose for the rest of the post that you're playing UDCA with LEHO.

Say you're defending 4, you hold T92 and partner leads Ace (apparently he wants to give you a ruff). You will discourage and/or show an odd number of cards, and you have 2 cards to do that: T and 9. Here comes LEHO in play:
- If you want the LOWEST remaining suit, you play an EVEN card (LEho)
- If you want the HIGHEST remaining suit, you play an ODD card (leHO)

So in our example:
- 2 would encourage partner to continue .
- 9 asks for a switch.
- T asks for a switch.

Important note: this is most useful in a long suit, so you might want to apply this only in a suit you bid.

Saturday, September 30, 2006

A missed opportunity

Suppose you're playing with me and you hold the following hand:
AKJT752
A9
7
932

All vulnerable, your partner (= me) as dealer, imps scoring, the bidding goes:
pass - pass - 4 - 5
all pass

You may disagree with the 4 bid, but I like it.

I lead a and you take a look at following dummy:
83
J8
KQJ
KJT765

Declarer wins with J in dummy, and leads J which you cover. Now the question arises: how can we defeat this contract? Imo there's a lot known about the entire hand.
- Your partner didn't lead which suggests he holds 2 or more (or 0, which is unlikely because declarer would hold 4 from the Q and wouldn't bid 5).
- He lead , your short suit, which suggests that he's long in , doesn't hold the Ace, but still wanted to give us a ruff.
- Declarer played J, which suggests he holds either KQ or KT. Looking at the vulnerability, I'd rather give him KQ or even KQT, so there won't be another trump trick unless you can ruff.
So there are 2 suits left to try and reach partner for a ruff: and . Looking at the auction and HCP alone, I think you shouldn't give partner A. There's a lot more to say for Q, since we think partner has 2+. Declarer can easily be singleton or void to explain his bidding. There's one more chance for success: partner having a void in either black suit.

So which suit do you play, and which card?

The key is in the lead: if partner played a high , he says you can reach him with , but if he lead a low , he's asking for a return.

At the table, I held Q96, declarer had a singleton. I lead a small instead of a high one, partner took A immediatly but continued K. After analysing this hand, I saw my error. Partner said he thought about a low at the table, but didn't do it because he hoped for another trick. I was glad he realised his argument for playing a high was wrong. We missed an opportunity to defeat 5, but I guess we learned from that mistake...

Tuesday, September 26, 2006

Defensive blunder

Lets talk about the following hand, might be a good lesson:

Dealer:South
Vul:EW
Scoring:imps
-
Q984
AT9754
AKQ
KQ965
T3
6
86432
AT72
A72
QJ83
J7
J843
KJ65
K2
T95

Suppose you are sitting in East. Your opponents had a bit of an agressive auction:
pass - pass - 1 - pass
1 - pass - 3 - pass
4 - pass - 4 - pass
5 - pass - 6 - Dbl
all pass

Partner leads K, dummy ruffs with 4 and plays Q.

First question: do you take your Ace or not? There's no real reason to take it. Playing low gives away a trick (double dummy), but you make it very hard on declarer to get to 11 tricks, since he can't draw another trump! I think most good players would hold it up once.

At the table East cashed A, so we cash and have another decision to make.

The key now is to figure out what declarer needs to make his contract, and what dangers he'll have to cope with. It's clear that he'll need his length (which you control), and he'll also need to draw trumps at some time without us getting another chance to play some s. To set up his s, declarer will need entries (you only have 2) and 2 trumps to ruff. So it might be possible to lock up the hands.

There's another question that is probably very interesting: how many s does declarer have? We can't know for sure, but it's probably 2-3 since most players don't cuebid partner's suit with a singleton. If he has 3, then you can probably give partner a ruff, but in that case you'll always have a trick so there's no need to play . If he does have a singleton or void, he'll be totally ruined anyway (with stiff he needs to ruff 3 s, so he'll need a 5 card trump).

So what do you continue?

It's necessary that you mess up declarer's plan. You suppose declarer has Kx (which is the only dangerous one for us). He will be able to set up his s only if he can draw your last trump when he's in dummy (since he'll need his other trumps to ruff s), ready to cash his remaining s. So there are 2 possibilities to mess that up:
- let declarer ruff a : declarer will have to hope for a 3-3 split in now, which isn't the case (he'll need to draw trumps, and will only be able to ruff 1 more , not 2)
- mess with the entries. You only have 2, you declarer will need 2 s to get to dummy twice. Now that you'll ruff the 3rd he'll be down.

At the table, East played a , which is imo the least useful continuation possible. Now my partner (I was sitting North) found a very nice line of play: K, A, ruff a , to dummy, ruff another , to dummy, draw the last and claim for contract.

If you visualize declarer's plan, you'll have 'double dummy vision' playing this hand.

Sunday, September 24, 2006

Too greedy

Here's a fun one, too bad opponents made lots of mistakes or this would've been a candidate for hand of the year:

Dealer:West
Vul:EW
Scoring:imps
432
J54
K9
AT432
JT8
AQT8
T5432
5
Q65
32
AQJ86
J76
AK97
K976
7
KQ98

The bidding was quick and simple:
pass - pass - pass - 1NT
pass - 3NT - all pass

Some might not agree with the 1NT opening, but ok... So you're declarer in 3NT by South (nicely wrongsided): do you see any chance of making this? I guess not, but at least you get a little chance after the lead of J. The dangers are quite clear: , but also for communication between the hands or an entry. There's still not much chance of making this, except if opponents mess it up (or if both Aces are onside, but still risky). The lead told me the location of Q, so there might be some endplay possibilities. The hands aren't played double dummy, so mistakes can still be made. Lets give opponents some opportunities ;-)

I took A and started immediately with cashing 5 s. West discarded 4 small s and East 2 and 3. They played standard carding, so both Aces would be wrongsided! Chance for success now seemed very low, but opponents made huge errors during their careless discarding! They were both too greedy, they both wanted to make all their length tricks (which is impossible anyway) instead of just trying to defeat the contract. I still wasn't home, but if RHO would make a mistake in I could endplay him. So I played small and luckily I saw East play small (which gives me the contract). K was easy, followed by a small to the Q. Poor East only had AQJ86 left, and decided to play low to my King (it doesn't matter what he plays, because he'll get it back later). 8 tricks were there already, and I had to find another one. is the only chance, and to my big surprise RHO didn't follow. So LHO had kept AQT8 and would be endplayed as well! The King was covered by the Ace, but West couldn't do anything else than playing giving me J as my 9th trick.

3NT= obviously was a miracle: nobody bid it, and nobody made 9 tricks in their partscores.

I hope my opponents learned something from this hand, which is never to be too greedy in an undoubled NV contract with imps scoring.

Monday, September 18, 2006

The long run...

You hold:
AQT3
JT2
K973
98

You were dealer, all green, imp scoring, we get a quick auction:
pass - 3 - pass - 3NT
all pass

Since you don't know how solid opponents are when they preempt, but can't take much risks, you decide to go for a semi agressive lead: a small .

Dummy looks great for declarer:
65
K
654
AQJT432

Partner takes the Ace and returns 2 to your K. You're already glad you took these tricks, but the hand isn't over yet. You might've just given declarer some overtricks. The question is: what do you continue, and why?

Analysing this hand should get you on the right track to defeating this contract:
- is useless: K is quite a high card, and you might as well set up declarer's suit.
- has no future left at all. Partner returned a small out of respect for your lead (you might have a great source of tricks). If he had anything in , he would've returned a bigger card, so partner had A2.
- is one of the more interesting suits: the K might be declarer's only entry to a bunch of s, but you still need to find K in partner's hand for this to be successful. So what if partner has the K? He'll probably return a to your AQ to defeat the contract with ease. So the only question which remains is: what if declarer has K? Then you really NEED to take some tricks very fast, or block declarer's suits (which is impossible). There are 3 ways of making tricks 'quickly': either find A in partner's hand so he can return , or just cash s from the top (if partner has K) or hope partner has J and K. Which brings us to the next possibility...
- A is probably most interesting of all: you'll get a signal from partner. The only time this won't work is when partner has K or A.

So what's best? Imo the auction can help us here: the 3NT bidder will probably have a honour. Otherwise bidding 3NT is just suicide: he has to count on 6-7 tricks. This already makes vs 50-50.
What about A or K? One of the suits is definetly problematic for declarer because of our holding and dummy's K. But I don't think there's a way to tell for sure, except declarer's sanity. He only has QJT8, and he's quite poor in the Majors as well. Since opponents clearly preempt quite solid, you may expect quite a poor hand on your right.

If percentages don't give us an answer, all that's left is 'the long run':
- Lead a when partner has K gives declarer 7+2+2 = 11 tricks.
- Lead A when partner has A gives declarer 7+2 and maybe K = 9-10 tricks.

And if you're right:
- gives you 4 tricks, so 3NT-2
- gives you 2-4 tricks (depends if partner has the J or a 5 card), so 3NT-1/-2/-3. However, if partner has A chance of J (or a 5 card) has decreased significantly, so it's probably only 3NT-1.

In both situations A is more successful in the long run.
You may wonder what happens if partner doesn't have either. Then the maximum you'll make is A, so you might as well take it before opponents can discard dummy's s on A and QJ.

On the actual deal you would've been lucky: partner had K94 and could easily encourage , whatever method you play.

Thursday, September 14, 2006

One for the defense

Here's a nice hand from a few days ago:

Dealer:North
Vul:NS
Scoring:MP
6
T4
AT964
AT982
K875
QJ86
J72
Q4
AJ4
A753
KQ3
KJ3
QT932
K92
85
765

The bidding went:
1* - Dbl - pass - 2
pass - 2NT - pass - 3NT
all pass
(1 shows around 9-14HCP with 4+, and can contain longer )
My RHO probably bid 2NT just to try to rightside the contract.

I was sitting South and had to lead. I generaly don't like agressive leads, and there was every sign not to on this particular hand.
- Partner will probably be short in and chances are big that I won't have an entry if I can setup my suit anyway.
- Leading from under Kxxx gives the entire show away.
- Declarer will probably have a double stop.

So I decided to lead a , hoping to reach my partner's longest suit. We lead 1/3/5 and I thought my count would be important: I lead the 5. Partner nicely held up his Ace and declarer won with the J. Now K was played and partner took it immediatly, gave the suit a second of thought and correctly played a low to keep communication. This only fails when I had 53, but you might still get your trick back later. So the communication was still open. Declarer won in dummy and finessed to my Q. I had an easy continuation, and partner did what we worked for: cash the tricks. Declarer discarded in his hand and in dummy, so partner returned a small to his Ace. Now declarer tried a - squeeze (or splitting 3-3) by cashing his s. On the last I hold T92 K, and dummy still had K87. I had to keep my s since partner was all out (I already knew declarer had A as well), so I discarded my K. Lucky my partner had the T, otherwise declarer's squeeze would've worked. -2 was a great result.

GIB's Double Dummy Solver says we can only make 5 tricks. Declarer messed up by keeping his communication. If he kept K87 QJ J in dummy, he can easily take A, play to the J, a back to his Ace, and the last . NOW I'm squeezed in and ...