[Apollo81]

IMPs, R/R

Axxx
AKx
Kx
xxxx

(1♦)-X-(1♠)-2♣
(pass)-pass-(2♦)-2♥
(pass)-3♣-(pass)-pass
(3♦)-???

[gwnn]

If I thought my hand was "so good", why didn't I bid 3♣? Pass

[pbleighton]

I would consider double at matchpoints, but it's an easy pass at IMPs.

Peter

[mikeh]

doubling is insane: it is utterly idiotic at imps. And that statement is an underbid.

What do we think is going on here?

Partner bid 2♣... which is hardly a sign of strength. Sure, he has some values, since he need not have bid at all, but he sure doesn't need much. Then he backed in with 2♥. If his (4 card) ♥ suit held any strength, he would/should have bid it the round before. So he is x=4=y=5 or x=4=y=6 with modest values.

We can realistically expect to take 3 or 4 tricks with our hand: would any of us be shocked by the presence of the ♦A behind us?

And we can expect partner to take between 0 and 2 tricks.

A typical EW layout:

QJxxx     Kx
xx         Q10xx
A10xx     QJ9xx
xx         AQ

Ok, that is an unlucky layout from our perspective, but is there anything about the auction that says that it isn't that way?

And note that I've given the opps a 2-2 ♣ fit. Make LHO 5=3=4=1 and RHO 2=3=5=3, and partner can have a stronger hand...and you may have a trump trick and you still can't beat the hand if the opps own the ♣A

Consider the imp odds: assume they make: you lose 11 imps by doubling. Assume they fail by one trick... you win 3 imps. Nice odds? Feeling lucky, p#@k? as dirty harry might say.

Assume they fail by 2 tricks.. you win 7 imps... and does anyone really expect 500 here?

A good rule of thumb, in close imp partscore-into-game doubles is "Say Pass"

A second useful rule, if the first seems too tough, is 'Given the auction, do I have a good chance of down 2?" If the answer is NO, then apply rule 1.

[awm]

I think at MP scoring this is an obvious double. We have 3.5 top tricks, maybe more since the diamond ace is more likely with opener (who has the stronger hand). Hearts are almost guaranteed to cash (partner doesn't have five, neither opponent has five on the bidding, so they are 3334 or 4432 around the table).

So pass wins when partner has zero tricks for making two free bids, or when the diamond ace is behind us and partner has only one trick. Double wins in the most frequent case where partner has at least one trick and the diamond ace is on our right, and could easily net 500 on many lies of the cards.

At IMP scoring I think double is aggressive. It will get +200 most of the time, with an occasional -670 and a somewhat less frequent +500. Assuming the "other table" plays 3♦ undoubled I'd guess the result is something like:

+200 opposite +100 for win 3 about 2/3 of the time
+500 opposite +200 for win 7 about 1/12 of the time
-670 opposite -110 for lose 11 about 1/4 of the time

Overall very close to zero expectation, likely a slight negative. I think it's a reasonable action if looking for a swing, but not something I'd do in a serious match I expected to win.

I think "insane" and "utterly idiotic at IMPs" are extreme statements for a call many of us would make almost automatically at MPs. The forms of scoring are certainly different, but they are not that different.

[mikegill]

Was actually going to post something very similar to awm's response, but I suppose that's already been said. I agree that this seems like an autodouble at MPs - have to protect our 3C and get our 200.
I think this is a marginal IMPs double also (awm's estimate had a -.18 IMP/board expectation) I really don't think +500 is quite so far away here, though. My hand is 4 tricks most of the time here and 3 guaranteed, since opener rates to have about twice as many HCP as responder. And why can't partner have two tricks? He did bid twice on his own. Surely his average number of defensive tricks will be 1 at least on hands where he doesn't pull. There are lots of combinations of outstanding cards he could have that would be worth 2 tricks. If partner is really x4x6, he's going to pull my double to 4C, since he knows I have 4 clubs (unless I doubled on exactly 4333 shape). In fact, arguably he might pull with a hand like xx QJTx xx KQJTx with a really high offense-defense ratio.

[jtfanclub]

Apollo81, on Feb 1 2007, 03:15 PM, said:

IMPs, R/R

Axxx
AKx
Kx
xxxx

(1♦)-X-(1♠)-2♣
(pass)-pass-(2♦)-2♥
(pass)-3♣-(pass)-pass
(3♦)-???

Don't mind me, just musing aloud.

Let's see, opener passed 2♣, then passed 2♥, and finally bid over 3♣.

So...he has no extra strength, no spade support, and was willing to let us play 2♥, but not 3♣. He had a cheap 2♠ bid over 2♥, and declined it, even though his partner (I assume) promised 5♠. Even a doubleton would be enough.

So I guess we're looking at 1-4-6-2 shape.

LHO has 5 spades, RHO has 1, I have 4, partner therefore has 3. But LHO didn't bid 3♦, so I don't think he has 4. Call it 3. That would give partner... 3-4-2-4 shape? Those must be some awful hearts.

Scoring: IMP

---

So looking at that...if partner bid 2♥ the first time, and that was the end of the auction, we'd have 2♥, 2♠, 1♦, and 2♣ losers. Down 2. Which would mean if they go down we have a good board, and if they make we have a good board...provided we don't double. Not tough to move the cards around so that they make and we're down 3. Almost impossible to move them around so that we can make 2♥.

So I don't X. But then, my partner would never have bid the hand this way with 3-4-2-4 shape, the opponents must be extra conservative or lying. Right?

[Apollo81]

I would not double because I think partner will have 0 defensive tricks at least as often as he has 2. Under these assumptions, double has a worse expected value than pass. I agree that it is a percentage MP double.

Pard had 3424 shape and 1 trick, as is probably very normal for this auction. The previous poster's suggestion for the full deal was pretty close to right.

[mikeh]

awm, on Feb 1 2007, 04:50 PM, said:

I think at MP scoring this is an obvious double. We have 3.5 top tricks, maybe more since the diamond ace is more likely with opener (who has the stronger hand). Hearts are almost guaranteed to cash (partner doesn't have five, neither opponent has five on the bidding, so they are 3334 or 4432 around the table).

So pass wins when partner has zero tricks for making two free bids, or when the diamond ace is behind us and partner has only one trick. Double wins in the most frequent case where partner has at least one trick and the diamond ace is on our right, and could easily net 500 on many lies of the cards.

At IMP scoring I think double is aggressive. It will get +200 most of the time, with an occasional -670 and a somewhat less frequent +500. Assuming the "other table" plays 3♦ undoubled I'd guess the result is something like:

+200 opposite +100 for win 3 about 2/3 of the time
+500 opposite +200 for win 7 about 1/12 of the time
-670 opposite -110 for lose 11 about 1/4 of the time

Overall very close to zero expectation, likely a slight negative. I think it's a reasonable action if looking for a swing, but not something I'd do in a serious match I expected to win.

I think "insane" and "utterly idiotic at IMPs" are extreme statements for a call many of us would make almost automatically at MPs. The forms of scoring are certainly different, but they are not that different.

Where do you get these percentages from???

You must play with more conservative partners and against more aggressive opps than I do.

I would expect no more than xx Jxxx xx KQxxx from partner: with that hand, how would you bid in the given sequence? Surely exactly the way the post showed?

If you accept that as a reality, please, please explain to me your expectation that you will usually break even by doubling?

Give LHO QJxxx xxx AJxx x and opener Kx Q10x Q109xx AJx... please point out where these hands are inconsistent with the bidding.

Forget +500. Forget +200... heck, even -670 is a steal... you are -870!

And when you compare, your explanation is? 'I misread the auction?... I thought partner promised more defence? I didn't think the opps had their bids?'

Now, am I saying that double will never work? Of course not... but you must play in a different game than I do when you can sprinkle chances of -870 around, fighting for a minor suit partscore.

While I doubt that my 'percentage' guessing is any better than yours, my effort looks like:

+500 5%
+200 50%
-670 40%
-870 5%

Which leads to an imp expectancy of about - 3 imps for the double. I don't know many teams I can spot 3 imps a board and expect to win.

Note that this analysis makes double a good action at mps and insane (idiotic) at imps... and, yes, the games ARE that different. Why do you think that Barry Crane won so few big team games?

[pclayton]

I think double is silly (there's that word again) at all forms, including MP's.

RHO probably is 2=3=5=3 (no support double, no rebid of 2♦); LHO sounds 5=3=4=1, which makes pard 2=4=2=5. Sure we are cashing 2♥, 1♠ and the likely ♦K, and...., and....? Pard isn't scoring a spade ruff, and I can't lead trump from my side.

Pard couldn't jump over 1♠, so assets are rather limited.

[dcvetkov]

I dont think that dbl even at MP is so clear cut. I expect this to make more often then not, so I may pass, and doubling in IMP is well, maybe insane is pretty close.
You are probably outgunned in MP, partner is close to broke, and diamonds are breaking well for them ( 2-2) And there may not be any club tricks coming and one is a premium. ( Playing partner for 2 tricks is well just simply unrealistic).
So if you are looking at 3.5 tricks, they may be the end of it. It looks this contract will depend on where is the ace of D, or whether you can cash 3 hearts, and thats too much of a gamble

My 2 cents

[awm]

I think that this game is too hard for me if I cannot count partner, who made not one but two free bids in competition for a single trick.

Perhaps one should simply never double partscores. After all, LHO (who has shown very few values) could have all the aces behind my kings and kings behind my aces, and all the suits I'm hoping to cash tricks in could break 3-1 or 4-1, and partner might've bid a bunch of times on no values just to push the opponents up?

[dcvetkov]

I am not saying it wont work. Probably its about 50% whether you gonna get your hoped for 200 in MP. But you have to be willing to accept occasional -670. +100 should be decent or below average, and -670 will be a zero.
And yes, partners occasionally will stretch to bid to steal the hand, even without values for free bid.
I estimate there are 18 trumps, and one of those deals where both sides can make 9 tricks.
This hand has nice defensive values though. Anything less, and I would pass at MP with not much thought. and BTW, parther heard you dbl and choose not to dbl