[kereru67]

Mendelson advocates leading the Q from AQTxx against NT contracts. Has this been analyzed mathematically? Seems to me any lead could be "right" or "wrong" depending on the distribution in the other three hands. Which lead has the best chance of success?

[Cyberyeti]

Probably depends on the auction, 2N-3N the Q may well be wrong as the J is more likely to be on table than the K (although could be doubleton). 1any-1N-3N the reverse is true.

[sfi]

One of the top Australian players was written up for the queen lead a couple of years ago. Unfortunately the layout on this particular hand was:

---

It really depends on where you expect the honours to be and whether partner is likely to have an entry and another card in the suit.

[whereagles]

As mentioned, leads from AQTxx and similar holdings depend more on the auction than in any lead convention.

The Q works well is LHO has the strong hand (Kxx/Jxx more likely).
The A works well when you have an entry (hoping for dummy to come up with Kx or Jx).
Small may also work well.
And sometimes the best lead is lead another suit!

[Cthulhu D]

Isn't it almost always going to be best to lead another suit? I mean it's not as bad as KJxxx, but it has the same attributes.

[whereagles]

A lead from a suit like this comes up on top more often than we'd think. At least if I recall Bird/Anthias correctly.

[Cyberyeti]

Cthulhu D, on 2014-October-19, 05:33, said:

Isn't it almost always going to be best to lead another suit? I mean it's not as bad as KJxxx, but it has the same attributes.

Well it might be that partner's trick and 4 tricks from your suit are essential before declarer cashes his 9. Kxx over, Jx under being a prime example where you want to lead the Q.

[Cthulhu D]

whereagles, on 2014-October-19, 05:41, said:

A lead from a suit like this comes up on top more often than we'd think. At least if I recall Bird/Anthias correctly.

This exact suit combination is actually used as an example 5 times, and they recommend leading from it (selecting the A from AQTxx) against 1NT-3NT only, otherwise lead another suit. In the one example it is correct, they note that a significant part of the effect is to enable you to switch to hearts when it is right. Otherwise they think it is basically terrible:

1NT-3NT lead the ace from a major
1NT -Failed Stayman - 3NT - don't lead this suit when it's a minor
1NT - All Pass - Don't lead this suit when it's a major (make a passive lead from Kxxx)
1S-1NT-All Pass - Don't lead this suit when it's the oM
2NT - All pass - Don't lead this suit when it's a major

Overall we can see it does well when you need to go active - 1NT-3NT when you have a hand that is strong (I think the example has 10 HCP) and thus partner is broke.

Similar suits also consistently do awfully. Leading away from tenances against 3NT seems super bad unless you know they have a long cashing suit in dummy or otherwise that it is a race to get 5 before they get 9.

Quote

Well it might be that partner's trick and 4 tricks from your suit are essential before declarer cashes his 9. Kxx over, Jx under being a prime example where you want to lead the Q.

Well.. yeah, obviously. But it could also give declarer a 9th trick. And the question isn't 'against 1NT-3NT' where this might be true, it's against NT contracts generally - and it's generally terrible. Any situation where a passive lead is called for it's bad.

[Cyberyeti]

Cthulhu D, on 2014-October-21, 05:05, said:

Well.. yeah, obviously. But it could also give declarer a 9th trick. And the question isn't 'against 1NT-3NT' where this might be true, it's against NT contracts generally - and it's generally terrible. Any situation where a passive lead is called for it's bad.

Against 1N-3N I'd never lead the Q, but would normally lead the suit.

I'd rarely lead the Q any time I expect the K to be on my right, depends on the auction whether I'd lead the suit.

In a club pairs, I'd lead small from the suit every time unless the suit had been bid by opps or I was sure K was in dummy, that's what will happen at every other table and I hope to outdefend the rest of the field once I see dummy.

[whereagles]

Chtu, I was thinking more of 3NT contracts.

In other NT contracts it's more dubious yes.

[Cthulhu D]

Cyberyeti, on 2014-October-21, 05:29, said:

In a club pairs, I'd lead small from the suit every time unless the suit had been bid by opps or I was sure K was in dummy, that's what will happen at every other table and I hope to outdefend the rest of the field once I see dummy.

Bird Anthias have that as a anti percentage action. Against 1NT passed in, with a construction with AQTxx Kxxx Jx Tx they have the 4th from the king 12% more likely to beat the contract, and on average taking an extra .3 of a trick at matchpoints. If the K is a smaller honour (say Jxxx), the heart lead is gonna look even better.

Against 1S-1NT-All pass holding 5xx AQTxx 3 10xxx the spade and club lead both rank ahead of any lead from AQTxx.

I'd suggest one of the thesis of the book is that leading away from an honour is really sub optimal, particularly if it's a king. Leading away from Tenaces is also bad. Leading away from a KJxxx type tenace is diabolical.

It's sure possible not to agree with the results, but it is an illuminating look at things.

whereagles, on 2014-October-21, 05:37, said:

Chtu, I was thinking more of 3NT contracts.

In other NT contracts it's more dubious yes.

It's worth considering that the example they use there it is also benefiting strongly from the major suit bias. One of the Quiz answers suggests that when the AQTxx suit is a minor in an auction that suggest a major suit bias (failed stayman -> 3NT), the lead is terrible.

If a major suit bias is in effect, and you have length in the minors, that makes it even more probable that partner has major cards and not much in the minors. As the idea of the AQTxx lead is to catch partner with a fitting honor, the less fitting cards he has the worse it is.

So I guess I'd suggest that we only want to lead from AQTxx when

A) The auction has been non invitational directly to game, and thus it is more likely we need to race for our 5 winners
B) There is an effect at work that suggests partner is likely to have fitting cards - e.g. It's a major and they bid an auction that creates major suit bias e.g. 1NT-3NT and thus partner is more likely to have fitting cards? Otherwise pick another suit.
C) And in these circumstances, we are going to lead the A, not the queen.

Therefor to answer the OP's question, don't lead the Q ever, mathematical analysis suggests it is bad.

[Cyberyeti]

Bird/Anthias's analysis is very dubious in that IIRC it's DD, ie that you will always find the perfect switch.

Leading the A is terrible if you have no side entry.

And you miss my point about the club game, even if not leading the suit is a marginal favourite, getting 0 when it was essential, when I reckon to score 60% if I find the same lead against the normal contract as will be made at all the other tables is just not worth it.

Also if you don't ever lead from xxx in these circumstances, partner can sometimes draw some very powerful inferences about the rest of your hand, and is much more likely to get the subsequent defence right. We've had a series of unmitigated disasters from partner leading from xxx against NT because I simply cannot work out what he's got.

[WellSpyder]

Cyberyeti, on 2014-October-21, 06:34, said:

Bird/Anthias's analysis is very dubious in that IIRC it's DD, ie that you will always find the perfect switch.

Yes, it's DD. The forums have debated this before, but I can't help agreeing with your feeling that analysis based on DD analysis may not be a very strong guide to what works best in practice.

Quote

Leading the A is terrible if you have no side entry.

One of the big advantages of leading A rather than Q or a lower one is that you may be able to work out what to continue with, if for example dummy has Kx or Jx. Double dummy this isn't an issue anyway, so A won't appear to gain.

Quote

We've had a series of unmitigated disasters from partner leading from xxx against NT because I simply cannot work out what he's got.

I also think DD analysis underplays the benefit of partner having a clearer idea of what you have and therefore what to do next.

[Jinksy]

Cyberyeti, on 2014-October-21, 06:34, said:

We've had a series of unmitigated disasters from partner leading from xxx against NT because I simply cannot work out what he's got.

Which card do you lead? My P and I play top of nothing specifically three small, which has helped a bit with this problem - usually it's not going to matter too much to him whether you've led from two or three small.

[Jinksy]

Cthulhu D, on 2014-October-21, 05:43, said:

I'd suggest one of the thesis of the book is that leading away from an honour is really sub optimal, particularly if it's a king. Leading away from Tenaces is also bad. Leading away from a KJxxx type tenace is diabolical.

That's not how I interpreted it. In a major, leading from such a suit vs 3N with no majors shown by the opps seems like the best lead most of the time if you have has much as one likely outside entry. Perhaps against a 1N 2N / 3N type auction you might prefer a passive lead in the other major if available, but with eg xxx KJxxx Kxx Qx and the auction 1N 3N, I'd be pretty surprised if a ♥ wasn't the top lead.

[Cyberyeti]

Jinksy, on 2014-October-21, 11:26, said:

Which card do you lead? My P and I play top of nothing specifically three small, which has helped a bit with this problem - usually it's not going to matter too much to him whether you've led from two or three small.

We lead middle, but that's not usually the problem, it's that I have to find his side Qxxx from one of my two KJxx holdings and fail, where the rest of the room is leading the suit.

Look at the hand I posted as "Declarer plays slowly" or similar. Partner should have been able to work out exactly what I had (or needed to have) given the knowledge that I don't lead from xxx/10xx unless I have to.

[Jinksy]

Cyberyeti, on 2014-October-21, 12:25, said:

We lead middle, but that's not usually the problem, it's that I have to find his side Qxxx from one of my two KJxx holdings and fail, where the rest of the room is leading the suit.

I haven't had that problem yet. Surely much of the time after your opening lead is discouraged, you can give suit preference either by continuing the suit (admittedly risky when it might mislead P about your holding, but must be worth it sometimes when you're choosing between pip cards) or by a smith peter later in the play?

Quote

Look at the hand I posted as "Declarer plays slowly" or similar. Partner should have been able to work out exactly what I had (or needed to have) given the knowledge that I don't lead from xxx/10xx unless I have to.

Can you be more specific? I don't want to google around for a thread whose name is 'similar' to a key phrase

[Cyberyeti]

Jinksy, on 2014-October-21, 15:05, said:

Can you be more specific? I don't want to google around for a thread whose name is 'similar' to a key phrase

I thought you'd read it, was posted within the last few days:

http://www.bridgebas...lowly-what-now/

If you look at #7, you get the kind of logic I'm talking about.

[Cthulhu D]

Jinksy, on 2014-October-21, 11:29, said:

That's not how I interpreted it. In a major, leading from such a suit vs 3N with no majors shown by the opps seems like the best lead most of the time if you have has much as one likely outside entry. Perhaps against a 1N 2N / 3N type auction you might prefer a passive lead in the other major if available, but with eg xxx KJxxx Kxx Qx and the auction 1N 3N, I'd be pretty surprised if a ♥ wasn't the top lead.

To the research! Against 1NT - 3NT

With two outside aces and KJxxx in diamonds against 1NT-3NT they suggest a diamond

xx x KQxxx KJxxx they suggest a major

xx KJxxx Ax KJxx major

KJTxx KJTx xxx x - spade

Q xx KJxx KJxxxx - Q Spades

So that's 3/5 for the lead, plus against both 1NT-3NT and 2NT-3NT Tx KJxxx xxxx Jx the heart lead is better at imps, but the spade lead is better at matchpoints.

Some other auctions!

xx Jxx AKJxx xxx - against 1D-1S-1NT-3NT the spade or club are winning leads

Aginast 1NT all pass, KJxxx Jxxx Ax xx - Heart is best

Against 1H-1NT Qxxxx xx x KJxxx Spade or Heart

Against 1NT-2NT Jxxx Jx xx KJxxx Heart-Spade (club is ~25% worse than the preferred heart)

Against a failed stayman invite xx Qxx KJxxx Axx Spade or heart

Failed stayman again KJxxx xx Axxx Tx Club or heart

2NT- All Pass AQx x KJxxx Txxx Club is strongly best.

I guess adding that all up, I am left with the conclusion that on any invitational or part score auction, leading away from the tenace is probably bad. Against game auctions it is a calculated risk to find a fitting honour with partner and establish the 5th card and may be worth doing - but it's not at all clear cut!

[Jinksy]

Cyberyeti, on 2014-October-21, 17:31, said:

I thought you'd read it, was posted within the last few days:

Maybe, but I have a hard time keeping track of everything I've seen on here

Quote

http://www.bridgebas...lowly-what-now/

If you look at #7, you get the kind of logic I'm talking about.

Yeah - but this seems to corroborate what I was saying. E must have had a chance to signal suit preference on the ♥s. On the assumption that he could read you for junk in ♦s (easier here if you lead the 10 as top of nothing, though I'm not sure if that's good long-term strategy), he only has two suits to signal preference between.

It's still risky to trust him enough to lead away from your actual ♠ holding, but in a long-term partnership, by the time the third ♥ comes to you, you must have been able to signal more comprehensively. For eg, if the first and second cards offer count and the second suit preference, what would be the difference between him playing 4-9-8 and 8-9-4?

As you say, when you switch to the ♣, P could still play the J to discourage (or if you trust this declarer to be exactly 5233, he could play a pip).

I agree with lamford that you don't want to mindlessly give SP all the time, but assuming this is IMPs, IMO E can see fairly early that there's little other hope of beating the contract than for you to have good Ss (which maybe S's 3N bid hints at anyway, given that he presumably has a few points in the minors for it, not to mention S's failure to touch the suit, despite that seemingly excellent fragment in dummy).

I'm in too much of a hurry to try to analyse the defence in depth, and couldn't add much to what others have said even if I wasn't, but I'm just arguing for the general principle that precise signalling agreements with your P seem like they could solve most instances of this sort of problem. Sure, followed slavishly they'll give info away to sharp opps, but theoretically most of the time you can probably read the necessary switch, so signalling is partially a crutch. If your opps are better enough than you on the day that they'll gain more from your signalling than you will, then you were probably never going to beat them anyway