[pescetom]

MP

---

Please answer the poll on the basis of your judgement and add any comments here (including which card you prefer/exclude on the basis of your agreements).

[DavidKok]

Spoiler

I don't think the choice of spot card matters much but I would systematically lead the second heart. The choice of suit is primarily based on Bird & Anthias' "Winning Notrump Leads", in particular their tips on page 181:

• Do not lead from broken honour sequences in a 5-card suit on this auction.
  
• Prefer a passive lead over leading from four to an honour, even a passive singleton.
  
• Make the same lead as you would against 1NT-3NT.
  
• Prefer major suits over minor suits strongly.

In addition, since we have such a weak hand I think attempting to hit partner's long suit is preferred over trying to establish our five of clubs/leading away from the spade spots.

For what it's worth, I think this is controversial.

[AL78]

I lead a top (of nothing) heart on the basis of maybe I can get lucky and hit partner's suit, or if not, it hopefully won't give a trick away. Leading a black suit looks dangerous, and a diamond looks very unlikely to gain, less so than a heart, and there has been no Stayman enquiry.

How do you reply in the form of a spoiler?

[pilowsky]

AL78, on 2022-February-16, 17:31, said:

I lead a top (of nothing) heart on the basis of maybe I can get lucky and hit partner's suit, or if not, it hopefully won't give a trick away. Leading a black suit looks dangerous, and a diamond looks very unlikely to gain, less so than a heart, and there has been no Stayman enquiry.

How do you reply in the form of a spoiler?

Spoiler

You put the text inbetween {spoiler} and {/spoiler} but use square '[' ']' brackets and not curly '{' ones.

[mikeh]

I’d lead a heart. For me top of nothing

I did read David’s spoiler after deciding on my lead.

Iav3 mixed feelings about the Bird and Anthias books on leads. The problem, which they acknowledge but (imo) underestimate is that their results come from a double dummy analysis.

Here’s one example, albeit from the book on suit leads. The lead of an unsupported Ace often does remarkably well, but only because the double dummy solver doesn’t need partner’s cooperation. Signalling by either defender is irrelevant and opener can’t fool the defence by falsecarding, etc.

In the real world, all good defenders have agreements about carding, including (as an example) leading the Ace from AK, and what information partner gives on the lead of an Ace.

Thus I think that it’s critical, when trying to learn from their books, to be very alive to double dummy bias, and especially what value it is to study a method where literally partner never signals and yet one always does the right thing.

Having said that, it would imo be just as bad to completely discount everything they suggest.

Here, I’d have led a top of nothing heart both before and after reading their books.

One has to assume, at imps, that one ‘can’ set this contr@ct if one guesses the best lead. That pretty much means playing partner for 8-10 hcp and a suit that can be established

He rates (slightly) to have longer hearts than spades. Our spade holding is dangerous to lead from because the suit could be frozen, in that whichever side first plays it loses a trick. That’s far less likely in hearts.

A minor would be nuts. While opener may have good majors, that’s far from assured but we can be sure that responder doesn’t have a 5 card major and will rarely have a 4 card one..if he does, he should have extras and 4333 shape, and the extras means we’re not beating this anyway.

[smerriman]

I would have expected a double dummy sim to match David's conclusion.

Yet if I sim based on:

West any balanced 20-21
East any 5-11 without a 4 card major (unless 4333)

then a spade lead comes out on top (followed by heart, diamond, club).

This surprises me. But I'm still leading a heart.

[DavidKok]

mikeh, on 2022-February-16, 18:27, said:

Iav3 mixed feelings about the Bird and Anthias books on leads. The problem, which they acknowledge but (imo) underestimate is that their results come from a double dummy analysis.

Here’s one example, albeit from the book on suit leads. The lead of an unsupported Ace often does remarkably well, but only because the double dummy solver doesn’t need partner’s cooperation. Signalling by either defender is irrelevant and opener can’t fool the defence by falsecarding, etc.

In the real world, all good defenders have agreements about carding, including (as an example) leading the Ace from AK, and what information partner gives on the lead of an Ace.

Thus I think that it’s critical, when trying to learn from their books, to be very alive to double dummy bias, and especially what value it is to study a method where literally partner never signals and yet one always does the right thing.

Having said that, it would imo be just as bad to completely discount everything they suggest.

(snipped from the full comment) I fully agree, and whenever they recommend an ace lead it is particularly suspicious. In fact, the only time when I fully support their ace leads is against a slam at matchpoints, where cashing your ace has a realistic chance of preventing the overtrick and they claim this is underrated by most bridge players.
By contrast, my results leading short major suits against 3NT contracts have far outperformed my expectations based on the double dummy analysis - simply because it is often not clear to declarer that I led a short suit. When they later play me for length in the major they perform far worse than double dummy. Of course we disclose that the lead may be from shortness, but it still happens. Of all the lessons in Bird & Anthias I think the twin advice of "against notrump, with a weak hand, try a short major" and "against trumps, don't try the trump suit unless you have to" are seriously underrated chunks of advice, and while generated through double dummy analysis I think these recommendations hold up in actual play.

[LBengtsson]

3♠. It would be good to know if the opponents have Stayman only, Puppet Stayman, or Muppet Stayman on their convention card. I like the ♥ top of nothing lead but it might lose time. If you had ♠J109x you would have lead the ♠J and I see no reason not to lead ♠3 here. It will not deceive partner and might put declarer on a guess at trick one.

I know at MP the idea is not to give away overtricks to the ops but you should still play as if you can break the contract also. It is a guess whether partner has useful cards in ♠ or ♥ 50/50 I guess, but you have extra card in ♠ and possible useful ♠J9. It might be losing lead also, but you do know that until board is over

[P_Marlowe]

Major, ... I am going with a heart.

To beat it, we need partner to have more strength, than we have,
in which case, trying to set up his suit seems to be a better shot,
than trying to set up our suit. Heart has the advantage, that we may
hit a 5 carder, spade will usually at best hit a 4 carder.

If partner has nothing in hearts, it wont cost anything.

[P_Marlowe]

smerriman, on 2022-February-16, 18:40, said:

I would have expected a double dummy sim to match David's conclusion.

Yet if I sim based on:

West any balanced 20-21
East any 5-11 without a 4 card major (unless 4333)

then a spade lead comes out on top (followed by heart, diamond, club).

This surprises me. But I'm still leading a heart.

How large is the diff between hearts and spades?
It may have to do with the 9 of spades.

[eagles123]

Seems pretty much like a guess between spades and hearts, I go with a spade.

[pescetom]

LBengtsson, on 2022-February-16, 22:55, said:

It would be good to know if the opponents have Stayman only, Puppet Stayman, or Muppet Stayman on their convention card.

Good point.
EW are typical intermediates, not great at bidding but they can't go far wrong here.
2NT is 20-21 balanced with traditional shapes only (5M332 is possible but no other 5cM).
Stayman over 2NT is vanilla 4-card with no discussion about developments.
Major transfers over 2NT are vanilla 5+card with no discussion about developments.


Thanks to those who voted so far.
I'll leave the actual hand for tomorrow to get a few more unbiased votes in.

[Disclaimer: I voted myself, but only because I said I would have led spades long before opening this thread]

[pescetom]

mikeh, on 2022-February-16, 18:27, said:

I’d lead a heart. For me top of nothing

I did read David’s spoiler after deciding on my lead.

Iav3 mixed feelings about the Bird and Anthias books on leads. The problem, which they acknowledge but (imo) underestimate is that their results come from a double dummy analysis.

Here’s one example, albeit from the book on suit leads. The lead of an unsupported Ace often does remarkably well, but only because the double dummy solver doesn’t need partner’s cooperation. Signalling by either defender is irrelevant and opener can’t fool the defence by falsecarding, etc.

In the real world, all good defenders have agreements about carding, including (as an example) leading the Ace from AK, and what information partner gives on the lead of an Ace.

Thus I think that it’s critical, when trying to learn from their books, to be very alive to double dummy bias, and especially what value it is to study a method where literally partner never signals and yet one always does the right thing.

Having said that, it would imo be just as bad to completely discount everything they suggest.

Here, I’d have led a top of nothing heart both before and after reading their books.

One has to assume, at imps, that one ‘can’ set this contr@ct if one guesses the best lead. That pretty much means playing partner for 8-10 hcp and a suit that can be established

He rates (slightly) to have longer hearts than spades. Our spade holding is dangerous to lead from because the suit could be frozen, in that whichever side first plays it loses a trick. That’s far less likely in hearts.

A minor would be nuts. While opener may have good majors, that’s far from assured but we can be sure that responder doesn’t have a 5 card major and will rarely have a 4 card one..if he does, he should have extras and 4333 shape, and the extras means we’re not beating this anyway.

I agree with most of what is said here (and with the OT of DavidKok about my partner's frequent trump leads for that matter). I'm a bit skeptical about the hearts longer than spades argument: yes it is quite likely, but my spades are 4 cards and that is a certainty. I confess I would have led ♠3 here, but I'm willing to keep an open mind that hearts may be safer (and get round to reading the book).
Just beware that this is MP as declared in OP, so holding them to 9 tricks by doing the obvious may well be a good score.

[DavidKok]

pescetom, on 2022-February-17, 10:24, said:

I'm a bit skeptical about the hearts longer than spades argument: yes it is quite likely, but my spades are 4 cards and that is a certainty.
[...]
Just beware that this is MP as declared in OP, so holding them to 9 tricks by doing the obvious may well be a good score.

If you intend to find a passive lead, hearts are significantly safer than spades.
As for the length: the primary strategy of setting their 3NT on an auction like this is "find the 5-card suit our side owns, establish it, run it". The odds that partner has 5 hearts on this auction should be quite a bit higher than the odds that partner has 5 spades. Furthermore, even if partner does have spade length (and/or values) our hand is weak. We likely do not have the entries to run the fourth spade if we successfully establish it, and even then it is possible that only gets us up to 4 tricks. If partner has 4 hearts and at most 4 spades I suspect attempting to set up their suit is better than attempting to set up ours, because partner presumably holds a stronger hand.
Choosing a lead is always a gamble, but I think the odds favour leading a heart here.

[Cyberyeti]

DavidKok, on 2022-February-17, 12:00, said:

If you intend to find a passive lead, hearts are significantly safer than spades.
As for the length: the primary strategy of setting their 3NT on an auction like this is "find the 5-card suit our side owns, establish it, run it". The odds that partner has 5 hearts on this auction should be quite a bit higher than the odds that partner has 5 spades. Furthermore, even if partner does have spade length (and/or values) our hand is weak. We likely do not have the entries to run the fourth spade if we successfully establish it, and even then it is possible that only gets us up to 4 tricks. If partner has 4 hearts and at most 4 spades I suspect attempting to set up their suit is better than attempting to set up ours, because partner presumably holds a stronger hand.
Choosing a lead is always a gamble, but I think the odds favour leading a heart here.

My partner always has 5 spades if I lead anything else, the ♠9 adds some safety.

[pescetom]

DavidKok, on 2022-February-17, 12:00, said:

If you intend to find a passive lead, hearts are significantly safer than spades.
As for the length: the primary strategy of setting their 3NT on an auction like this is "find the 5-card suit our side owns, establish it, run it". The odds that partner has 5 hearts on this auction should be quite a bit higher than the odds that partner has 5 spades. Furthermore, even if partner does have spade length (and/or values) our hand is weak. We likely do not have the entries to run the fourth spade if we successfully establish it, and even then it is possible that only gets us up to 4 tricks. If partner has 4 hearts and at most 4 spades I suspect attempting to set up their suit is better than attempting to set up ours, because partner presumably holds a stronger hand.
Choosing a lead is always a gamble, but I think the odds favour leading a heart here.

Fair enough, but as reminded to mikeh, this is MP and we may not need or realistically be able to set 3NT.
The odds of finding partner with 4 spades are reasonable: if he doesn't have values in our chosen suit we are going nowhere anyway, if he does then our J9 plus a gold standard side entry are about what he should expect.
Or so I would have reasoned at the table, at any rate.
It's always a gamble, I agree.

[smerriman]

P_Marlowe, on 2022-February-17, 02:36, said:

How large is the diff between hearts and spades?
It may have to do with the 9 of spades.

Out of 500 deals, 343 were ties, 88 favored the spade, and 69 favored the heart. While small, that's still pretty significant.

I tried using the same 500 deals, but swapping the 9 of spades with the 4 of spades. The spade was still favored.

A lot of people have been mentioning they're leading the three card heart suit trying to hit partner's length. That makes sense at IMPs; setting up that 5 card suit is how you beat the contract. In this case, with the points heavily weighted to declarer and dummy entries likely limited (where leading into declarer's suit will give him an extra tempo vs having to cross to dummy to do it), MP will be more about not giving away a trick.

Perhaps the fact your hearts are shorter increases the chance of *declarer* having length there, which makes it more dangerous. For example, 23 of the wins for spades were when declarer had a 5 card heart suit. But only 11 of the wins for hearts were due to declarer having a 5 card spade suit.

[sfi]

smerriman, on 2022-February-17, 18:44, said:

Out of 500 deals, 343 were ties, 88 favored the spade, and 69 favored the heart. While small, that's still pretty significant.

I tried using the same 500 deals, but swapping the 9 of spades with the 4 of spades. The spade was still favored.

My thought (and reason for leading a spade) is that I need less from partner in spades to avoid blowing a trick than I do in hearts. I know Antoff/Bird suggests a heart here, but I've never fully bought into this. It does seem pretty close though, and I'm gratified that smerriman's data supports my impression.

[pescetom]

So here is the full board.

MP

---

The auction matches the hands, but on this layout only a clubs lead (chosen by 8% in this poll) will hold Declarer to 9 tricks rather than 12.
Nothing really odd about that, every dog has his day.

But what surprised me (and sent me looking for a reality test) was that at national level only 65% of declarers in NT made 12 tricks and almost 25% (28 out of 122) were held to 9 tricks. I don't have lead statistics except for my own club (spades 5, hearts 3, clubs 1) but as far as I can see, 9 tricks must be the result of a clubs lead - even blocking diamonds does not explain it.

[P_Marlowe]

sfi, on 2022-February-17, 19:40, said:

My thought (and reason for leading a spade) is that I need less from partner in spades to avoid blowing a trick than I do in hearts. I know Antoff/Bird suggests a heart here, but I've never fully bought into this. It does seem pretty close though, and I'm gratified that smerriman's data supports my impression.

If a DD simulation favors heart, Anthias / Bird would favor the heart, their suggestions are based on DD simulations.