And I thought this was obvious.

[han]

Q

KQ10x

Q10xx

10xxx

 

1D - (4S) - ??

 

All vulnerable at IMPs. Standard American openings.

[Fluffy]

With full values it would be close, but with a stiff queen I think it is an easy pass

[gnasher]

I think it's obvious.

[CSGibson]

Pass

[pooltuna]

Your choices should be pass or double not 5♦!

[mtvesuvius]

Definitely pass.

[Phil]

Obvious pass.

 

And 5♦ over partner's double.

[BunnyGo]

Obvious pass.

 

And 5♦ over partner's double.

 

Out of curiosity, what would 4NT be over partners double?

 

FWIW, I think the pass is far and away my first choice too.

[gnasher]

Interesting. I thought double was obvious.

[lamford]

Interesting. I thought double was obvious.

I agree with you, and did a simulation, where partner has 12+ and 3+ diamonds, without five hearts, and North has 7-8 spades with the ace, king of the suit (he is vul) and another card. I looked at spade, heart and diamond contracts. 4S made 28% of the time, 5D 24% of the time, and 5H 10% of the time. I suspect partner will pass most of the time, including a lot of the time 4S is making, but the big gains are when he bids and there is a double game swing. I don't mind losing 790 some of the time. Partner might have Jxx of spades as well, and they might be misguessed or not pickable up because declarer has 8.

 

But maybe double is more penalty oriented in US, where I think most of the posters here are from.

[655321]

I think I like pass.

 

I could easily see myself doubling at the table though, partner has opened, we have 9 HCP, shortage in their suit (and for me double is takeout, not cards) but somewhere there has to be a limit, and an Aceless working 7 count probably falls below it.

 

Maybe it seems an obvious double to lamford because he plays a weak NT, the biggest difference I find between playing a weak NT and a strong one is that playing a weak NT when partner opens 1m it is 'always' right to compete, playing a strong NT it is usually better not to stretch in competitive auctions, partner always seems to have a balanced minimum.

[Phil]

Out of curiosity, what would 4NT be over partners double?

 

FWIW, I think the pass is far and away my first choice too.

 

4N is normally two places to play. The combinations would be:

 

Hearts + Clubs: equal length

Hearts + Diamonds: 2 card difference

Clubs + Diamonds: 2 card difference

[han]

I also did a simulation but used different specifications from lamford's. For example, I disallowed 7222 shape for the overcaller, and allowed 4S on AJ-8th (but not on AJ-7th). I also didn't allow partner to hold a 1NT or 2NT opening.

 

I found that 4S made 53% of the time and 5D 20% of the time. The 53% is a large difference from lamford's number.

 

Of course, these numbers don't prove that pass or double is better. In order to draw such conclusions one would need to look at what happens after pass or double, and this is both difficult, time intensive and somewhat subjective.

 

I dealt 200 hands and checked manually how many of those would reopen after a pass. I selected 71 hands, much more than I expected. Most of those would reopen with a double. For example, one of the hands with which I thought that I would double was x Jxxx AJxxx AKx. If you disagree then you would get a lower number. I should say that most of the reopening hands were clearer than that one.

 

I ran the double dummy simulation on the hands that were left, and found that 4S still makes on 53% of the left-over hands, but 5D makes on only 10% of these hands. To recap, you could originally make 5D on 40 out of the 200 hands, but on 27 of those partner would not pass out 4S.

 

I'm tempted to say pass is better, but I already thought so before I started the simulation :).

 

At the table the call was 5D, which lost some imps when opener had Jxxx Ax KJxx Kxx and both games were down.

[gnasher]

I ran the double dummy simulation on the hands that were left, and found that 4S still makes on 53% of the left-over hands, but 5D makes on only 10% of these hands. To recap, you could originally make 5D on 40 out of the 200 hands, but on 27 of those partner would not pass out 4S.

 

You seem to have missed out a relevant statistic (insofar as any of these statistics are relevant): on the 53% of hands where 4♠ makes, how often does 5♦ go only one or two down?

[whereagles]

Dbl. They're not making overtricks (I hope!) and can easily go -500. If pard pulls, my hand is adequate. Sort of... lol.

[han]

You seem to have missed out a relevant statistic (insofar as any of these statistics are relevant): on the 53% of hands where 4♠ makes, how often does 5♦ go only one or two down?

 

I don't think that this is a particularly relevant statistic. With some of those 53% hands, we'll defend 4S no matter whether we pass or double. Then it does not matter how well 5D does.

 

To do a proper investigation, we'd have to figure out what happens after a pass or double with each of the hands. And even if you and I both did this, we would get entirely different outcomes. (by which I don't mean that I think you would do a bad job.)

[mfa1010]

I think that this is a rather clear pass. A bunch of random queens are not worth so much at this level - we can't double with so little. I'm sure that if partner would always do the right thing (sacrifice in 5♦, almost never bid slam, etc.) then double might be OK (but not more than that), but in real life I think it is a loser.

[gnasher]

I don't think that this is a particularly relevant statistic. With some of those 53% hands, we'll defend 4S no matter whether we pass or double. Then it does not matter how well 5D does.

 

Bidding a making 5♦ instead of taking 100 against 4♠ undoubled would be worth 11 IMPs. Saving for 200 against a making 4♠ would be worth 9 IMPs. You seemed to think the first category relevant, in that you told us how often we could make 5♦. Why isn't the second category almost as significant?

 

Similarly, defending 4♠x = when we could have defended it undoubled costs 5 IMPs. Saving for 500 against a making 4♠ gains 3 IMPs. I don't see what you hope to learn from measuring one without measuring the other.

 

To do a proper investigation, we'd have to figure out what happens after a pass or double with each of the hands. And even if you and I both did this, we would get entirely different outcomes. (by which I don't mean that I think you would do a bad job.)

I agree - it's difficult to get meaningful results from simulating a high-level competitive auction.