[awm]

benlessard, on 2013-September-30, 22:08, said:

This either mean that we are pretty poor for leads or that a DD lead isnt real bridge.

I don't think either of these conclusions follows. Certainly we are not as good on opening lead as double dummy. However, humans never play as well as double dummy; even a computer with unlimited processing power won't play as well as double dummy because of incomplete information. Arguably the opening lead is the toughest part of the game because we have least information (can't see dummy, no signal from partner yet, etc).

The fact that overall human results are close to double dummy (especially at the slam level) is interesting but not really indicative. There are hands where declarer is likely to do worse than double dummy (for example, where the only way to make is some anti-percentage play like dropping a stiff king in an eight-card fit). There are other hands where the defense is likely to do worse than double dummy. All the overall human results suggest is that these hands are close to equally common... not what will happen on any particular hand.

With that said, I think there is room for improvement on opening leads, and the findings of this book are certainly interesting. It's well-known though that both ace leads and passive leads fare better double dummy than single dummy. The main reason is two assumptions: that giving declarer a guess he could make himself will never cost a trick (true double dummy, but not single dummy) and that the defense will never err after the opening lead (again true double dummy, but not necessarily single dummy).

[benlessard]

I agree with 99% of your post except that your argument seems to strengten my conclusion not attack it (i was a sarcastic when we said we are all poor on leads)

If you lose one fourth ! of tricks on every 3Nt contract so at the end declarer has 8% extras chance of making the contract how come you cannot come to the conclusion that leading DD and real bridge is totally different ? A fourth is really huge IMO, most of the time I blow a trick on lead im still pretty much convinced ive made the correct lead, so overall im pretty sure we don't do that many mistakes on leads.

What seems to work in DD isnt dont seems to based on any real bridge principle, its simply that real bridge lead have a big inherent luck factor that cannot be avoided.

Im not saying there is any lesson to be taken from DD, just that the numbers will never really mean a lot because on lot of hands its simply a matter of seing dummy vs not or a matter of luck.

Of course I should really read the book before continuing.

[helene_t]

I think it would be very difficult to reach any firm conclusion about effectiveness of different lead styles based on empirical research.

If the lead of an unsupported ace tends to work well then it could be offset by a decrease in efficacy of leads of supported aces because of the lack of inference that partner has the King. So we would probably have to compare the success of pairs that adopt one lead style with pairs that adopt another lead style. But obviously it could be that players that lead in a particular way just tend to be better defenders.

Maybe if we could identify a large number of pairs in which the two partners have very different lead styles and then see if the results are different when A is on lead than when B is on lead. There are still possibly confounding issues. But if several analyses pointed in the same direction we would have a reasonable indication.

[rhm]

awm, on 2013-September-30, 23:12, said:

The fact that overall human results are close to double dummy (especially at the slam level) is interesting but not really indicative.

Not indicative of what?

Quote

There are hands where declarer is likely to do worse than double dummy (for example, where the only way to make is some anti-percentage play like dropping a stiff king in an eight-card fit). There are other hands where the defense is likely to do worse than double dummy. All the overall human results suggest is that these hands are close to equally common... not what will happen on any particular hand.

True. I still do not see what that refutes.
I would never claim that a double dummy line on any single deal has to be the right one. Nor an opening lead

But if double dummy would claim for a specific hand over a large number of samples card x wins by a large margin over any other card against some contract, you would have to provide very strong arguments why that is a fluke and another card is superior single dummy in the long run.
We know that different leads win on different possible layouts of the remaining 39 cards, so what is surprising that there is an 8% difference in success between double dummy leads and human leads at the top level?
That partner will not find the right continuation every time is true for active leads as for passive leads. No convincing arguments have been brought forward why passive leads are inherently worse in that respect.
Of course if someone will always assume the lead has to be from longest and strongest against all indications, this somebody will not defend well. Give him this book to read. Maybe it is time to make less simplistic rules past the beginner stage what to lead against 3NT.
If I lead an ace I will not find the right continuation every time.
True.
Do I find always the right continuation when I lead anything else the next time I have to take a decision? You must be joking and I am not known for suffering from an inferiority complex.

Rainer Herrmann

[whereagles]

well, I see there is no current trend towards switching gears w.r.t. opening leads default cards

(damn, that was convoluted)

[uday]

What am I missing?

Let's say partner is on opening lead, and I'm just minding my own business over there.

All else being equal, I would prefer him to lead a card that, double dummy, will beat the contract rather than a card that, double dummy, allows the contract to make.

Wouldn't/Shouldn't we all?

I have some stats on DD-lead-efficiency of human beings. It is difficult to interpret. My original thinking was along the lines of measuring skill by assuming that - all else being equal -

A stronger player (leader, i suppose) is more likely to find a DD winning lead than a weaker player on any given hand

Think that's true?

What was interesting also was that since each hand has 1-12 "best" leads (discarding hands where all leads are best) , and that this number varies from hand to hand, that I had to normalize the aggregated data somehow. That is, my statistic for each player looks something like this

"On any hand from the set of N hands where this player was on lead, assume there was exactly 1 best lead. How often did he find it?" and also
"What % of the time over N boards with varying numbers of DD-perfect leads did this player find one of the winning leads?"

I don't know if this is a rating or a partial rating, but I'll say this - cheaters seem to cluster at the high end of this statistic. I'll also say that this is quite computationally intensive, and slow as heck.
I ran it over one month's hands in the MBC or something, and it took forever, or at least about a week. I suppose I could find a way to speed it up if I cared enough.

I'll present my data in more detail if anyone cares, at least in aggregate with names removed.

[GreenMan]

uday, on 2013-October-02, 15:11, said:

What was interesting also was that since each hand has 0-13 "best" leads

Seems as if this should be either 1-13 or 0-12 depending on whether you classify the case where all leads are DD equal as "all best" or "none best". Or am I misreading?

[uday]

My typo - it is 1-12 ( i discarded hands where all leads were equal from a DD POV)

[GreenMan]

I see, thanks. The study looks interesting.

[gnasher]

uday, on 2013-October-02, 15:11, said:

Let's say partner is on opening lead, and I'm just minding my own business over there.

All else being equal, I would prefer him to lead a card that, double dummy, will beat the contract rather than a card that, double dummy, allows the contract to make.

Wouldn't/Shouldn't we all?

I would prefer him to lead a card that, single dummy, will beat the contract rather than a card that, single dummy, allows the contract to make. Putting it another way, all else isn't always equal.

[rhm]

deleted

[rhm]

gnasher, on 2013-October-03, 03:39, said:

I would prefer him to lead a card that, single dummy, will beat the contract rather than a card that, single dummy, allows the contract to make. Putting it another way, all else isn't always equal.

This is a banality, irrefutable true and entirely worthless. Uday's comment is different, subject to arguments maybe, but not worthless.
DD acts as a proxy for single dummy and of course when in doubt you prefer the real thing to the proxy.
Nevertheless there are reasons why sometimes a proxy is used, because some uses can only or better be done with proxies.
Single dummy lead and play more often than not means a comedy of errors, bad guesses or failures without errors, which can hardly be simulated nor predicted.
After all there is one declarer and two defenders and everyone of them is not confined to a single mistake.

I am a long term subscriber to the Bridgeworld and for decades every month there is in the MSC an opening lead problem.
Quite frequently all suits find backers and sometimes even the vote is quite evenly split.
The arguments for each and every card is interesting, sometimes illuminating, but in the end leaves you none the wiser. Quite often some comment everything could be right
You simply do not know how often, which lead would work. How to decide? Does majority vote convince you?

Some scenarios can be simulated double dummy. Bird/Anthias uses 5000 deals per hand.

I still have to see the single dummy simulation over a meaningful sample, which would be workable and where the "single dummy analysis", whatever that is, is robust and you can trust.
I do not know how to predict errors, even though I understand the art of making opponents guess or uncertain.

Note, I am not claiming or arguing DD simulation should be your only source of inspiration for opening leads.

Rainer Herrmann

[PhilKing]

I assume Andy's point is that the best double dummy lead will not necessarily work because it may require subsequent DD play.

In broad terms, a more passive style may be DD best, but make the subsequent defence harder.

[rhm]

PhilKing, on 2013-October-04, 04:27, said:

I assume Andy's point is that the best double dummy lead will not necessarily work because it may require subsequent DD play.

In broad terms, a more passive style may be DD best, but make the subsequent defence harder.

I understand that but where is the evidence for this claim?
It can not be the lead in itself, since many lead high when leading passive.
The view that an attacking lead is often a hit and miss affair is a bit simplistic. If true, play and defense would be boring after an attacking lead.
Even if there would be evidence I would tend to argue finding the killing defense is often hard.
What is even tougher is changing lifelong habits.
As you say DD does not argue for a passive style, just for a more passive tendency in some scenarios and it is my impression that at least some successful pairs from Italy have this approach.
I would need a lot of persuasion to use a sub-optimal strategy just because the defense will be simpler to execute if it worked.
It is a bit like claiming you should never lead an unsupported honor including an ace, because the defense for partner is tougher if he can not assume you have led from a honor sequence.
(Arguing not to do so because it is a general loosing strategy would be a very different matter.)

Rainer Herrmann

[helene_t]

If my lead of the king against notrumps promises KQTx(x) then partner knows it is usually safe to encourage with the jack from Jx(x). If I might have lead from KQxx(x) it could easily cost a trick to do so. So when I actually have KQTxx and partner plays low I won't know whether continuing will help declarer complete his Bath Coup.

DD the king from KQxx(x) is superior while standard is x against notrump.

[rhm]

helene_t, on 2013-October-05, 03:54, said:

If my lead of the king against notrumps promises KQTx(x) then partner knows it is usually safe to encourage with the jack from Jx(x). If I might have lead from KQxx(x) it could easily cost a trick to do so. So when I actually have KQTxx and partner plays low I won't know whether continuing will help declarer complete his Bath Coup.

DD the king from KQxx(x) is superior while standard is x against notrump.

This is running in circles.

The typical argument against DD goes time and again:

The defense has a guess single dummy DD would not have or declarer has no guess DD he would have single dummy. It comes in different guises like here or declarer has a two way finesse for the queen etc.

Even the DD proponents do not doubt that. DD is not single dummy play, but the argument of the proponents is that these effects cancel each other.
Otherwise DD results would not be a close proxy for single dummy results.

With regard to your example I am pretty sure that if you encourage with the jack, unless you can see a switch is urgently required, will usually be best whether partner has led from KQTx(x) or KQxx(x).
(By the way some play that the king asks for an unblock and the queen would be the right card if you do not have the ten. I like this method)

The main DD argument was that it seems more often better to secure at least one trick in the suit and the tempo and avoiding giving declarer a cheap trick rather than what you gain by taking your slightly better all out chance establishing this suit by starting low.
We humans are impressed by deals when we find leading low would have been the only chance to beat the contract and jump to conclusions.
Evolution has not made judging statistical trade-offs a top priority.
Human experience is known to be biased, far more obvious and also easier to show than that DD simulation is biased.

Rainer Herrmann

[NickRW]

I would like to know exactly what "DD" means in this context.

It could mean (and I suspect it does mean) the computer chose the best lead for this exact deal (i.e. it knows the layout of all 4 hands). Or, it could mean that the computer chose the average best lead with a range of possible layouts consistent with some sort of bidding sequence and what that was supposed to show (i.e. a single dummy solution is found by repeated DD analyses - which is, I think, what robots actually do when playing for real).

The former would, for example, favour an ace lead from Ax where partner has Kxxxx and the declarer and dummy are distributed 3-3. But, obviously, that layout won't happen on all deals that are reasonable for a particular contract/bidding sequence.

Nick

[rhm]

NickRW, on 2013-October-07, 06:31, said:

I would like to know exactly what "DD" means in this context.

It could mean (and I suspect it does mean) the computer chose the best lead for this exact deal (i.e. it knows the layout of all 4 hands). Or, it could mean that the computer chose the average best lead with a range of possible layouts consistent with some sort of bidding sequence and what that was supposed to show (i.e. a single dummy solution is found by repeated DD analyses - which is, I think, what robots actually do when playing for real).

The former would, for example, favour an ace lead from Ax where partner has Kxxxx and the declarer and dummy are distributed 3-3. But, obviously, that layout won't happen on all deals that are reasonable for a particular contract/bidding sequence.

Nick

The book takes hands as lead problems and for each hand samples the remaining 39 cards with constraints consistent with the bidding. Sample size: 5000 deals.
On the 5000 deals it is then checked how often which cards as opening leads will defeat the contract double dummy or for matchpoints will generate the most tricks for the defense.

Rainer Herrmann

[whereagles]

I've seen a claim somewhere on DD that went something like this: (sorry Rainer, can't give you the exact reference )

1. DD is more helpful for declarer at slam level since you always take the right finesse and/or get the endplay right.

2. At partscore level DD is more helpful for the defence, as they always get the shifts right. Declarer doesn't have much of a choice of plays at this level.

3. At game level, stuff sort of evens out.

[NickRW]

rhm, on 2013-October-07, 07:20, said:

The book takes hands as lead problems and for each hand samples the remaining 39 cards with constraints consistent with the bidding. Sample size: 5000 deals.
On the 5000 deals it is then checked how often which cards as opening leads will defeat the contract double dummy or for matchpoints will generate the most tricks for the defense.

Rainer Herrmann

Thanks. That lends more credibility to it then.