[lamford]

---

IMPs. An everyday auction. If you ask, dummy will answer that South may have a five-card major and often upgrades good 14s and can sometimes be (5 4 2 2). Your lead. (You have no agreement with partner regarding tray placement, but you know that he is not 0-4, both from the auction and that he has not coughed.)

[mgoetze]

Small heart seems to stand out as the obvious lead.

[lamford]

mgoetze, on 2015-September-10, 04:02, said:

Small heart seems to stand out as the obvious lead.

And I would expect it to come out ahead in a poll of experts ... But is it the right lead?

[BillPatch]

I am a bridge player and not a resulted. Woolsey says that the right play is the one that works on hat particular deal, but even he would say that a better bridge player tries to find the correct bridge play, the one that produces the highest expected value according to the scoring method in use. At Matchpoints the safe low heart is a stand out, Playing IMPs I would expect two other options to score almost as highly. The 8 of hearts lead also attacks the long suit, but invites partner to try to return into my other suit with strength. The Q of spades may allow us to run the suit when partner has Kxxx(x)(x). It may give up a trick, but sets up partner's long suit.

My expected ratings at IMPs are fourth best heart 10, 8 hearts 8, Q spades 7, third best heart 6. Of course if the SOTM requires a home run the Q of spades sticks out as clearly best.

[lamford]

BillPatch, on 2015-September-10, 08:29, said:

I am a bridge player and not a resulted. Woolsey says that the right play is the one that works on hat particular deal, but even he would say that a better bridge player tries to find the correct bridge play, the one that produces the highest expected value according to the scoring method in use. At Matchpoints the safe low heart is a stand out, Playing IMPs I would expect two other options to score almost as highly. The 8 of hearts lead also attacks the long suit, but invites partner to try to return into my other suit with strength. The Q of spades may allow us to run the suit when partner has Kxxx(x)(x). It may give up a trick, but sets up partner's long suit.

My expected ratings at IMPs are fourth best heart 10, 8 hearts 8, Q spades 7, third best heart 6. Of course if the SOTM requires a home run the Q of spades sticks out as clearly best.

I can live with the eight of hearts lead, but I think the queen of spades is likely to come to a sad end, a bit like Herman in the opera of the same name. It puts all your eggs in one basket. There was a better lead in my simulation.

[cherdano]

I am sure ♠A is the best double dummy lead. It may well also be the best single dummy lead.

[lamford]

cherdano, on 2015-September-10, 08:54, said:

I am sure ♠A is the best double dummy lead. It may well also be the best single dummy lead.

I agree completely, and that is what my simulations showed. (% beaten: 12% AS, 9% heart, 7% minor; of course they often have extra values). A friend in a North London club led the ace of spades, and found his partner with ♠Kxxxx. He was accused of having seen the hand records by the only declarer to go off! Such is the paranoia in the bridge world at the moment.

David Bird says he did a DD simulation with not dissimilar results.

[WesleyC]

In spite of the DD analysis, I don't think the ♠A is the best lead here.

a) Double Dummy simulation over-rates unsupported ace leads. As Cherdano points out, single dummy you won't always find the killing switch.

b) Overtricks are costly, especially when you're only slightly improving your chance to beat the contract. On this hand I'd guess that the ♠A lead blows an overtrick about 30% of the time - roughly 0.3 IMPs/board. This is about the same as the 0.3 IMPs/board you gain by setting the contract 3% of the time.

c) Most pairs play double in this auction as lead directing for one of the majors. Your style here would have a big influence on the right lead.

[lamford]

WesleyC, on 2015-September-10, 09:53, said:

In spite of the DD analysis, I don't think the ♠A is the best lead here.

a) Double Dummy simulation over-rates unsupported ace leads. As Cherdano points out, single dummy you won't always find the killing switch.

b) Overtricks are costly, especially when you're only slightly improving your chance to beat the contract. On this hand I'd guess that the ♠A lead blows an overtrick about 30% of the time - roughly 0.3 IMPs/board. This is about the same as the 0.3 IMPs/board you gain by setting the contract 3% of the time.

c) Most pairs play double in this auction as lead directing for one of the majors. Your style here would have a big influence on the right lead.

The simulation was done by looking individually at the hands that could be beaten. If there was a winning switch, it was obvious.

Yes, it did cost the overtrick sometimes, but not as often as you might think. If the king of spades was with declarer, we would usually make our second spade trick. If in dummy, it cost just a tempo.

Oh, I forget to mention, this pair was not in a hospital for the criminally insane, and would not have doubled with ♠Kxxxx. And they had no method of signalling this holding either. Out of interest, some pretty good players, Phil King, Richard Probst and Boye Brogeland on a very similar hand, think the ace of spades is right.

[aguahombre]

WesleyC, on 2015-September-10, 09:53, said:

Most pairs play double in this auction as lead directing for one of the majors. Your style here would have a big influence on the right lead.

I don't think Kxxxx of spades is likely to produce a double. But, the failure to double might decrease the expectation of success from a heart lead. At IMPs I would think the Spade bullet to be an excellent shot --- I might not have thought of it at the table, but if an opponent did that to me at IMPs I would (after seeing his whole hand) admire it rather than report it.

[WesleyC]

I just ran a really quick and dirty sim myself, and found the ♠A lead was significantly better than your 12% vs 9% suggestion.

The other thing I hadn't considered, was that although the spade lead does often cost a trick (by establishing declarer's ♠K), Cashing out 3 or 4 rounds of spades often saves multiple tricks (even if it doesn't beat the contract) because the layout is very friendly for declarer. In fact over the 10,000 hands, the over-trick IMPs ended up about even.

So in summary, I stand corrected - ♠A lead for the win!

[lamford]

WesleyC, on 2015-September-10, 11:07, said:

I just ran a really quick and dirty sim myself, and found the ♠A lead was significantly better than your 12% vs 9% suggestion.

The other thing I hadn't considered, was that although the spade lead does often cost a trick (by establishing declarer's ♠K), Cashing out 3 or 4 rounds of spades often saves multiple tricks (even if it doesn't beat the contract) because the layout is very friendly for declarer. In fact over the 10,000 hands, the over-trick IMPs ended up about even.

So in summary, I stand corrected - ♠A lead for the win!

Good of you to post again and thanks. I set 15-17 for the opener and 10-14 with no 4 card M, balancedish, for the responder, which might have led to lower values. Some of those could have been slam tries! Also I only looked at the hands where a set was possible, and checked them individually. Good point about cashing three or four rounds of spades.

[Jinksy]

At IMPs I would consider the AS lead standout, and I'd counterclaim something about best behaviour about any pair who accused someone who'd found of cheating.