[cherdano]

Jinksy, on 2017-October-21, 07:51, said:

If so I would expect it to do better still with 4 human players, since any of the other suits look reasonably likely to save declarer a guess that the DD solver will always get right.

Huh? KQ will also often cost a trick single dummy. Is declarer not allowed to have ♣AJT?

[msjennifer]

Fourth best heart.

[ncohen]

Tramticket, on 2017-October-19, 16:00, said:

8S

2nd choice 8D.

I agree. Go passive against 1N, unless you have a strong sequence. A heart or a club is too likely to give up a trick. Besides, how wrong can GIB be

[lamford]

OK all. A club works very well in that partner has AJxx and you need to start an unblock at some time! A spade and a diamond work badly. I agree with FrancesHinden that a heart is wrong (maybe even at teams too) but partner has ♥ KJTx and you can cash four hearts but then will only beat it one in practice.

[NickRW]

lamford, on 2017-October-25, 06:26, said:

OK all. A club works very well in that partner has AJxx and you need to start an unblock at some time! A spade and a diamond work badly. I agree with FrancesHinden that a heart is wrong (maybe even at teams too) but partner has ♥ KJTx and you can cash four hearts but then will only beat it one in practice.

One specific layout proves nothing. It is clear that everything could work on some layouts and be disastrous on others.

[foobar]

lamford, on 2017-October-19, 14:38, said:

---

matchpoints. rank your choices. 1NT was g11-14

Toss up between ♦ or ♠. Might prefer the former given that the extra length makes leading 8♦ a little safer.

[johnu]

miamijd, on 2017-October-20, 21:32, said:

Who knows what is right? I know I don't, and I'm pretty sure none of you do, either!

Seriously, almost anything could be right.

If there was a good single dummy analyzer, you could get a pretty good idea what lead is best. I don't think there is one, so the best alternative is a double dummy analyzer which has well known discrepancies compared to single dummy play.

Double dummy, clubs turn out to be best, then diamonds, then spades and hearts are worst. You would have to look through the sample hands and make adjustments where real life play doesn't match double dummy play. A couple of people have done deeper analysis with real life hands and found that double dummy is usually pretty close on average. Double dummy analysis is the approach Bird and Anthias used in their 2 books on opening leads.

[nige1]

johnu, on 2017-October-27, 16:59, said:

Double dummy, clubs turn out to be best, then diamonds, then spades and hearts are worst.

Double-dummy simulations often deliver results that surprise me, especially the club lead here.

• What's the relative success for each lead?
• What are the simulation constraints? For example: presumably, LHO has a hand unsuitable for a transfer or for garbage Stayman? presumably, partner's hand is unsuitable for action, assuming that our partnership agree a popular notrump defence?
• What double-dummy simulation software is available for opening-leads? Are there any free programs? Which does JohnU recommend?

[johnu]

nige1, on 2017-October-29, 14:15, said:

Double-dummy simulations often deliver results that surprise me, especially the club lead here.

• What's the relative success for each lead?
• What are the simulation constraints? For example: presumably, LHO has a hand unsuitable for a transfer or for garbage Stayman? presumably, partner's hand is unsuitable for action, assuming that our partnership agree a popular notrump defence?
• What double-dummy simulation software is available for opening-leads? Are there any free programs? Which does JohnU recommend?

I use Dealmaster Pro which isn't free. It's not the most modern interface, but it has an acceptable user interface and is easy to use. You can also compare which contracts make, expected number of tricks per contract which is what I usually look at.

I used some very general assumptions. 11-14 balanced for East, <=11 for West with less than 5 cards in a major, North has less than 6 cards in either major.

For 200 test hands, the best leads were

Clubs - 148
Low diamonds - 140
Low Hearts - 137
♥A - 120
Low Spades - 138

Clubs are the best lead by 4-5% based on the assumptions I used, and the other 3 suits are pretty close and within sampling error. The clear worst lead is ♥A. You could fiddle around with more assumptions about possible holdings, and the relative rankings might move around a little, or you could go through each sample hand and try to adjust for what happens by trying to single dummy analyze the play. You are just trying to get a general idea of the best lead so I think this is good enough.

[nige1]

Thank you, JohnU.

[Tramticket]

nige1, on 2017-October-29, 14:15, said:

Double-dummy simulations often deliver results that surprise me, especially the club lead here.

• What's the relative success for each lead?
• What are the simulation constraints? For example: presumably, LHO has a hand unsuitable for a transfer or for garbage Stayman? presumably, partner's hand is unsuitable for action, assuming that our partnership agree a popular notrump defence?
• What double-dummy simulation software is available for opening-leads? Are there any free programs? Which does JohnU recommend?

I use Lead Captain. It's not free, but it seems to work reasonably well. I'm on computer at work at the moment, so can't run the deal through the program.

[lamford]

Tramticket, on 2017-October-30, 02:07, said:

I use Lead Captain. It's not free, but it seems to work reasonably well. I'm on computer at work at the moment, so can't run the deal through the program.

Thanks, TT. I have ordered Deal Captain. I was surprised and wiser from learning that the club lead was best double-dummy. I feel sure it would be best single-dummy as well, as I have found that both declarer and defenders benefit almost equally from DD play.

[NickRW]

lamford, on 2017-October-30, 08:23, said:

...I have found that both declarer and defenders benefit almost equally from DD play.

One thing to be prepared for is 'misleading' partner. For example Bird and Athias reccomend doubleton leads quite highly, whereas some human players often seem to have an aversion to them and do not enter the possibility you might have made such a lead into their thinking. So you can end up in the spectacularly cr@p position of improving your leads and worsening your results as a consequence.

[johnu]

NickRW, on 2017-October-30, 11:07, said:

One thing to be prepared for is 'misleading' partner. For example Bird and Athias reccomend doubleton leads quite highly, whereas some human players often seem to have an aversion to them and do not enter the possibility you might have made such a lead into their thinking. So you can end up in the spectacularly cr@p position of improving your leads and worsening your results as a consequence.

Bird and Anthias also recommend singleton leads as well as doubletons. I was playing with somebody who didn't know Bird from Anthias, and in a 3NT contract, got an inspired singleton ♠Q lead through a suit bid by dummy. In a very misguided attempt to block the suit, partner went up ace in dummy, sure that opening leader had some length in the suit led. That was a spectacular failure as RHO had ♠KJ10xx with only 1 entry. Maybe not the best example because declarer misplayed, but without a singleton lead, there would have been no opportunity to misplay.

But I agree that the defense can be unbelievably bad if one of the defenders misguesses whether the lead was from shortness or not.

[Lorne50]

nullve, on 2017-October-21, 07:41, said:

I think the ♣K (or the ♣Q, if the K would ask for an unblock) is the Bird-Anthias lead on this hand.

FWIW a 1000 deal DD analysis with south opening a weak NT, West fixed, North with less than 11 points and no 5 card major suggests nothing between 3 suits (ave tricks made 6.8) but clubs a clear winner (ave tricks made 6.5).

No attempt made to weed out hands that transfer to a minor or make a 4th seat protective call, but Bird-Anthias is based on similar computer based analysis and may have taken time to weed out those cases.

[steve2005]

PhilG007, on 2017-October-20, 13:38, said:

♥2 Unlike a suit contract,it's perfectly safe to lead away from aces
at No Trumps This suit could also be declarer's weak point.

People don't like a heart lead as it might give away a trick. You can give a trick away in any suit (club obvious don't lead)
Remember opps playing weak 1N,if field is in 1N they will know a 4-card suit held and opener's better minor so you are potentially at a disadvantage.
If opps have missed a 4-4 major fit, which is possible it probably doesn't matter what you lead-you will get good board.

I see nothing wrong with a heart lead, what the heck partner can even have 5.

[WesleyC]

I've got a lot of experience with double dummy simulation, and this hand is almost the textbook example of a situation where I would not trust the double dummy result - a club lead will work out *significantly* worse in practice.

In the likely case that declarer holds the club ace and club length, there's a decent chance that you will win both of your clubs if you avoid leading one.

The double dummy effectiveness of the club lead is also artificially boosted because when partner does hold the Club Ace, you will *always* find the killing switch.

More than any other contract, defender start their defense vs 1NT completely in the dark, with almost no clue about declarers distribution and high card location. Leading a long suit at tricks one sends an invaluable message to partner that will often be helpful in guiding the defense.

At MPs I'm happy to lead a diamond.