What's declarer's 'advantage'?

[Hanoi5]

There was this little article in this month's Bridge World about declarer's advantage in comparison to double dummy analysis. Jeff Miller examined 48,704 results from the World Championship in Shanghai and determined something. What he compared were the contracts played and the results at the table with the double dummy analysis from certain program. His results are interesting yet foreseeable, I think.

 

If you were not spoilt by reading the magazine, what do you think he concluded? If you've read the article or if the percentages are posted here, do you agree with those results? Do you think it can change from one tournament, level, scoring, etc?

[helene_t]

My guess would be that declarer is advantaged in games and partscores, but not in slams. In a slam, declarer is on lead most of the time so he has the most opportunities to make a "mistake".

 

I think it depends somewhat on the level. Intermediates like myself generally declare OK but defend badly, so at most clubs and low-level tournaments declarer has a bigger advantage. I don't think it matters much whether it's IMPs or matchpoints, or whether it's a suit contract or NT.

[hrothgar]

Here's my thinking:

 

Lets start by asusming that folks are relatively sane in their bidding. In theory, the higher the contract, the more information that is known about declarer's hand.

 

Next, let's ignore opening leads for the moment and only consider situations in which the defenders can see the dummy.

 

For example, consider the following two auctions

 

(1M) - P - (2M) - Float

 

(1M) - P - (4M) - Float

 

I suspect that the defender's are in a better position to place cards when they are defending against higher level contracts.

[mikeh]

My Bridge World always arrives late, so i haven't seen the article yet.

 

I am not a fan of double dummy analysis, so I am either going to feel vindicated or I am going to have to rethink my position B)

 

I expect that the result will be that declarer will generally do better than double dummy.

 

I expect that declarer's edge will diminish as the level increases, because, as a rule, the higher the level, the less flexibility/ambiguity/uncertainty there will be.

 

Reasons why, in my expectation, declarer has an edge include:

 

1. Control of his hand and dummy... defenders can end up on different wavelengths, but declarer knows all of the offensive assets

 

2. The opening lead can blow a trick (or more) and a tempo... more tricks swing on the opening lead than on any other play, and in most cases the opening lead is made by a player who has less information than either he or anyone else will ever have during the play. Thus declarer, before following to trick one, knows more than did leader choosing his lead. Knowledge is power.

 

As to the extent of the edge? I would guess it is about .3 to .6 of a trick, and falls below that at the 5 level or higher.

[whereagles]

I expect that the result will be that declarer will generally do better than double dummy.

Agree with this. Other advantages seem to me too intangible to quantify.

[Trinidad]

Declarer's advantage is that he can control his partner. There cannot be any miscommunication between the dummy and the declarer, whereas defenders are often working on two different plans.

 

This means that declarer's advantage is maximal when the room for misunderstandings by the defense is largest. That occurs when little is known about declarer's hand and when the defense needs to make a lot of decisions, i.e. when the defenders have a lot of 'points' and declarer has made few and vague bids. Normally that happens in partscore contracts.

 

Another example of this situation is after a preemptive raise (e.g. 1♠-Pass-3♠ (weak)- All pass). The defenders may well have the majority of the HCPs, and very little is known about declarer's hand. It is easy for the defense to blow two or more tricks.

 

The declarer's advantage is minimal if the dummy has been the captain in a relay auction. When dummy has asked for declarer's hand and the distribution and placement of all honors are known, the defenders (should) know pretty well what the plan should be.

 

Unfortunately, after 12 rounds of highly accurate bidding, declarer can usually claim the slam in trick 1...:)

 

Rik

[barryallen]

Does the double dummy analysis include the lead being given, I would assume so?

[kenrexford]

Hence "grabbit."

 

Years ago, in Washington DC, I learned a "new convention" which is one of my favorite.

 

1X-P-1Y-P-

1Z-1NT!

 

!="1NT has not been bid yet." Or, "grabbit."

 

The idea was that a sufficiently large amount of the time, Declarer is +1 on the play. So, whoever declares 1NT wins.

 

If N-S make 1NT, E-W make 1NT.

If N-S is -50 in 1NT, or -100 VUL, or -100 NV but doubled, E-W make 120.

If N-S is -100, or -200 VUL, or -300 NV but double, E-W make 400/600.

 

Stuff like that.

[awm]

I'll buck the trend and say that declarer's advantage is more significant in game and slam contracts.

 

If defenders were always double-dummy, then declarer would have a disadvantage. So the advantage comes from defenders being more likely to make mistakes than declarer. While there are certainly players who are "bad defenders" (or for that matter "bad declarers") the only fundamental reason defending is harder than declaring is the opening lead, where the defense can see only one of the four hands and therefore has a lot less to go on. A high percentage of the "defensive errors" made in top flight bridge are on the opening lead.

 

Now obviously it's easier to lead against an informative auction than an uninformative one. But it's not clear what this means: for example 1nt-6nt is not a very informative auction and 1♣-1♥-1♠-2♥-Pass gives a lot of information.

 

I think the bigger effect from the level of the contract is the critical nature of the lead. Defending a partial it is likely that the defense will get in on many occasions. It is also likely that opening leader's partner has a fair number of cards. This means that a tempo lost from a passive lead is unlikely to be fatal (you're getting in lots of times) and also that an active lead is unlikely to blow a trick (partner has a decent hand, so will often have some help in the suit lead). In a game or slam contract there is often quite a lot riding on the lead. The tempo lost from a passive lead could easily cost the contract (who says you're ever getting in again?) and an active lead is fairly likely to hit partner with nothing in the suit (partner won't have much) and blow a trick.

[paulg]

Hence "grabbit."

 

Years ago, in Washington DC, I learned a "new convention" which is one of my favorite.

 

1X-P-1Y-P-

1Z-1NT!

 

!="1NT has not been bid yet."  Or, "grabbit."

 

The idea was that a sufficiently large amount of the time, Declarer is +1 on the play.  So, whoever declares 1NT wins.

 

If N-S make 1NT, E-W make 1NT.

If N-S is -50 in 1NT, or -100 VUL, or -100 NV but doubled, E-W make 120.

If N-S is -100, or -200 VUL, or -300 NV but double, E-W make 400/600.

 

Stuff like that.

The reason for one of my answers in the current Australian Bridge bidding forum.

 

But I was not so explicit and if only I had used the term 'grabbit' I'm sure they'd publish my comment :)

[barryallen]

I'll buck the trend and say that declarer's advantage is more significant in game and slam contracts.

 

If defenders were always double-dummy, then declarer would have a disadvantage. So the advantage comes from defenders being more likely to make mistakes than declarer. While there are certainly players who are "bad defenders" (or for that matter "bad declarers") the only fundamental reason defending is harder than declaring is the opening lead, where the defense can see only one of the four hands and therefore has a lot less to go on. A high percentage of the "defensive errors" made in top flight bridge are on the opening lead.

 

Now obviously it's easier to lead against an informative auction than an uninformative one. But it's not clear what this means: for example 1nt-6nt is not a very informative auction and 1♣-1♥-1♠-2♥-Pass gives a lot of information.

 

I think the bigger effect from the level of the contract is the critical nature of the lead. Defending a partial it is likely that the defense will get in on many occasions. It is also likely that opening leader's partner has a fair number of cards. This means that a tempo lost from a passive lead is unlikely to be fatal (you're getting in lots of times) and also that an active lead is unlikely to blow a trick (partner has a decent hand, so will often have some help in the suit lead). In a game or slam contract there is often quite a lot riding on the lead. The tempo lost from a passive lead could easily cost the contract (who says you're ever getting in again?) and an active lead is fairly likely to hit partner with nothing in the suit (partner won't have much) and blow a trick.

That makes sense to me assuming the lead is a variable in the analysis. You may only require one error in game / slam which most of the time comes from the lead.

Again the assumption here is that this is points based, not tricks based.

[Phil]

At IMPs, I think the higher the contract, the higher declarer's advantage is eroded. There are less cards each defender can hold, and a good defender can usually work out what is required.

 

MPs is different. The optimal line for declarer may change depending on the the number of tricks declarer is trying to take. There's a big difference defending against 6♥ and defending against a declarer who is quietly trying to take 12 tricks in 4.

[jdonn]

Obviously the answer is declarer has a greater advantage at lower level contracts, in terms of percentage chance of making more tricks than are available double dummy. However it's not clear where declarer has a greater advantage in terms of something like total points.

[awm]

Another issue is that declarers make mistakes. This needs to be deducted from "declarer's advantage."

 

In a slam contract, declarer can see the great majority of the high cards. There are often only a few reasonable lines, and a good declarer will almost always find the best line.

 

In contrast, for a partial there are lots of important cards declarer can't see. There are a huge number of possible lines, and working out the "percentage" for each is often untenable. So declarers make a lot more mistakes in partials.

[mycroft]

I think the difference is between "making the contract" and "beating double-dummy".

 

There are lots of slams I've made because they guessed wrong on the OL. But it's an either-or proposition; the hand's over once dummy comes down. In 1NT, there's more of a chance for the defence to go wrong - especially against 1NT float - later in the play. Which might make the difference between -50 and +90, true; but might also make the difference between +90 and +120, or between -100 and -50. In the higher contracts, it's much more likely that the defender's mistake will be a contract-making mistake. It's an interesting thought, though.