[han]

You are on opening lead against 3NT with

Q10
10xx
J1087xx
xx

The auction was:

p - p - 1NT - p
2C - p - 2D - p
3NT - all pass

What would you lead and why?

[CSGibson]

 han, on 2012-August-23, 11:47, said:

You are on opening lead against 3NT with

Q10
10xx
J1087xx
xx

The auction was:

p - p - 1NT - p
2C - p - 2D - p
3NT - all pass

What would you lead and why?

Either way, matchpoints or imps, I'm leading the J of diamonds. They don't have a huge overpowering HCP NT (passed hand), this is the least likely to cost a trick, and if declarer is scared of my estabished diamonds, he might take finesses into partner.

[TimG]

Diamond Jack, because I don't want to do the non-obvious thing and find that the obvious thing would have easily beaten 3N.

[JLOGIC]

Maybe I will be convinced, but I still think diamonds is the most likely way to beat the hand. If partner has 11 points and 5 spades they will often open the bidding, 6 spades they will often preempt, etc.

If you get the threat of diamonds set up, you also have pressure on declarer and maybe they can misplay even if they could make. They don't know the diamond count sometimes or that we have no entry or w/e. And of course sometimes you can just run diamonds or the opps are 2-2. Maybe the opps will duck one too many diamonds and we can switch, whatever.

On the other hand, the SQ lead might mislead partner, and if declarer knows that you have QT doubleton earlier on he will probably peg your hand type quite well and play it very well. The argument against this is that usually if the SQ lead is wrong the play/defense won't matter since we have 3 HCP with their suits splitting, so maybe this is a non factor. And if it's right, partner will know your holding usually.

I think double dummy simulations downplay the importance of these things, or maybe I'm just being too unwilling to change my ways from conventional wisdom! I could still be convinced that just going all in on your good spades because the situation is so desperate could be right, maybe I just need more time.

[bluecalm]

Double dummy sim barely prefers Qs to Jd (I excluded 11+ with 5 cd major and any hand with 6+ major) but it solves some problems for declarer. I am still in Jd camp.

[JLOGIC]

FWIW I think low diamond is better than the DJ.

[SteveMoe]

♠Q trying to hit partner's length - s/he has all the entries. Pard doesn't rate to have more tha 1 or 2 ♦s.

[yin970902]

I like ♠Q
Partner has the other honor, I will give an assisting lead.

[rduran1216]

Imps I'm leading the spade Q all day. Isn't it likely partner has a few entries and if we hit in with J9xxx or KJ9x or better we've done something very useful with our opening lead. A diamond can work when declarer is 3325 and we catch partner with something like Hxx, but that's a really narrow window. We know partner has values, and spade length, sure diamond can be right, but if I'm trying to beat this, I would never lead a diamond, we just have too much info that a spade is better.

[akhare]

Q♠ -- David Bird's book made me do it ...

[han]

This hand came from the finals of the transnationals and the world mind games. In the transnationals both Meike Wortel and a Canadian player whose name I forgot had to lead after this auction. Wortel choose the diamond jack, the only suit to beat the contract, and the Canadian player led the heart ten.

Wortel, who thought that it was close between the spade queen and a diamond, argued that the auction showed that partner was maximal for his pass, and therefore was not so likely to have a 5-card major. Also, she thought that not leading a diamond was a swingy action, which she thought she wouldn't need in order to win.

Can we test this lead problem with a double dummy simulation? Finding the right specifications (what exactly does it mean that LHO and partner did not open the bidding?) is not an exact science, but with some effort reasonable specifications can be found. What is more problematic is the double dummy nature. If the contract depends upon finding the spade queen, the double dummy declarer will find it. For this reason I'd expect the spade queen lead to cost less often than in practice.

My double dummy results for 2000 hands:

Spade queen beats it 104 times.
Heart 10 beats it 41 times.
Diamond jack beats it 86 times.
Small diamond beats it 70 times.

It seems to me that the difference between the spade queen and the diamond jack is too small to be confident that the spade queen is best in practice. On the other hand, I think it is fair to say that the spade queen is a serious alternative, while the heart 10 was not.

[TimG]

You might be able to do some simple things to get an idea for how often the ♠Q gives away the contract (or an overtrick) against a single dummy declarer. You could count how often declarer has exactly 9 double dummy tricks with Kxx/AJx or Axx/KJx or similar.

I don't think overtrick IMPs should be ignored here. Intuitively, I would expect the ♠Q to give an extra trick far more often than the ♦J and those overtick IMPs add up when we're talking about beating the hand only 5% of the time.

[JLOGIC]

Again, I question why we are leading the jack and not low. Yes, of course a double dummy sim will like the jack since if declarer has A9 opp Qx or Ax opp Q9 he will play low double dummy but the queen in real life.

[bluecalm]

Quote

Wortel, who thought that it was close between the spade queen and a diamond, argued that the auction showed that partner was maximal for his pass, and therefore was not so likely to have a 5-card major. Also, she thought that not leading a diamond was a swingy action, which she thought she wouldn't need in order to win.

This Pszczola is a lucky guy to partner such a player.

[CSGibson]

I think DD sims will overrate the spade Q because declarer will always find it doubleton anyway in a DD sim. Similarly, I think that Justin's suggestion of a low diamond will work more frequently than DD sim suggests.

I'd be more interested in knowing when a spade Q lead gives up the 9th trick without a fight, as opposed to when it it necessary to beat the hand.

As to Justin's comment, it would not surprise me to find dummy with Hx and declarer with HH9x in diamonds, where leading the J is necessary not to blow a trick. I think the point of leading diamonds is not necessarily to set the suit up, though it is nice that's a possibility, but instead to avoid blowing a trick when they rate to be in a closish 3N.

[rduran1216]

This is not a close 3NT. hearts are splitting, partner's clubs are poorly placed, etc. The argument for the spade Q is that we need to get partner's hand going if we have any shot.

[rogerclee]

 bluecalm, on 2012-August-24, 11:42, said:

This Pszczola hanp is a lucky guy to partner such a player.

FYP

[han]

That I already knew Roger.