[PhilKing]

lamford, on 2013-April-03, 07:40, said:

Interesting. Stefanie thought 3NT might win at pairs too. Given that the expectancy is 8.0 tricks, it might be better to double on even less, as there is no redouble. Always doubling it was hugely better than never doubling it, despite the 11 950s. The contracts that were three or more off hurt declarer a lot. One problem with my sim is that it is Deep Finesse for both sides, so there will be no 3NT making on the lead! Declarer makes some very thin 3NT to balance those out.

If ever there was an auction where a trick is going on the lead this is it - total tricks on my (micro) sim were 268 - an average of 8.375. It was only the aggressive doubling strategy that gave 1NT the win.

[Vampyr]

gordontd, on 2013-April-03, 01:26, said:

Nobody suggested you should do anything to be considered a nice guy.

I thought it was strongly implied by ggwhiz, and I think that maybe some others share this view. Possibly not members of this forum though!

[lamford]

PhilKing, on 2013-April-03, 09:50, said:

If ever there was an auction where a trick is going on the lead - total tricks on my sim were 268 - an average of 8.375. It was only the near-perfect doubling strategy that gave 1NT the win.

My average was 8.02 over 1000 deals. You will have a high standard deviation for 24 boards. Doubling every 3NT gave 1NT a clear win, so I don't think your sim is large enough. I shall do 10,000 when I have time.

[PhilKing]

lamford, on 2013-April-03, 10:07, said:

My average was 8.02 over 1000 deals. You will have a high standard deviation for 24 boards. Doubling every 3NT gave 1NT a clear win, so I don't think your sim is large enough. I shall do 10,000 when I have time.

But it will still be double dummy sim, so doing a million or whatever makes no difference. There's no point opening 3NT if they are guaranteed to make the best lead.

I understand that double dummy and single trick averages tend to be close, but not in this auction!

On:
A J 8 7 3
9 8 7
Q 10
10 7 3

Partner held seven hearts and a stiff spade, so your sim will find the heart lead, for instance, for down 1 instead of up 1.

And a diamond on:

J 8 3
7 6 5 4
Q 7 2
10 4 3

For down one instead of up two.

And a diamond on:

J 10 7 3
A 9 7
Q 10 3
7 5 4

for down one instead of making 9.

These three boards account for half the difference, with an overtrick or undertrick frequently appearing elsewhere.

[mycroft]

How are these sims (both in terms of DD tricks and projected score) taking into account that second-hand passed? So you have the information that, *given lefty has a 1NT opener and he has a barred partner*, your partner has decided to pass (tell me he wouldn't make a lead-directing 1H bid maybe a little light? After all, partner also heard the NT opening, and shouldn't be hanging him, no?) and you know where 26 points are, and you're going to double on a blank lead and a flat 10?

Your goal in doubling is the 800+; of course, difference between 400 vs a partscore and 550 isn't that much, but 650, 750...? And, of course, 550 vs 400 or 420 isn't great, and 650 is worse...

For instance, I'd throw out the first of Phil's counterexamples; 7 hearts and a stiff spade, in this auction, isn't going to pass under the NT bidder unless it was the J-seventh 3 count. On the other hand, I'm sure I'd be throwing out a bunch of other hands that weren't passing, even without 12 high.

Not sure it would affect anything, just asking.

Note: I don't see anything in the OP about MPs vs IMPs. It really makes a difference, especially to the speculative double; *if* it's a partscore deal, you're playing against only the one or two pairs that bid game at MPs, but it's the extra -2, -3 IMPS vs +lots more; however, it's a hand everyone will get to game on, double is throwing away your average in favour of a top/bottom gamble at MPs, whereas at IMPs it's basically an even gamble - if it makes, you're -4 (with no overtricks), if it's down, you're getting +2 (with only one downtrick), and you get the "what if we're the only ones in NT" extra.

I'm still not sure it's a gamble I'd be willing to take with teammates, on a total crapshoot. We've already got the benefit of them having to guess the contract.

[PhilKing]

mycroft, on 2013-April-03, 12:42, said:

For instance, I'd throw out the first of Phil's counterexamples; 7 hearts and a stiff spade, in this auction, isn't going to pass under the NT bidder unless it was the J-seventh 3 count. On the other hand, I'm sure I'd be throwing out a bunch of other hands that weren't passing, even without 12 high.

You may have a point. 2♥ on all heart preempts is probably correct to avoid a uinilateral penalty double.

As you say, speculative doubles needn't be outlandish, since we are gaining anyway when they are too high.

[lamford]

mycroft, on 2013-April-03, 12:42, said:

How are these sims (both in terms of DD tricks and projected score) taking into account that second-hand passed? So you have the information that, *given lefty has a 1NT opener and he has a barred partner*, your partner has decided to pass (tell me he wouldn't make a lead-directing 1H bid maybe a little light? After all, partner also heard the NT opening, and shouldn't be hanging him, no?) and you know where 26 points are, and you're going to double on a blank lead and a flat 10?

That would be fine but second-in-hand is not allowed to have the bid cancelled and the auction revert to him. Either the 1NT is accepted, or any replacement call is made and the 1NT bidder's partner is barred.

[lamford]

PhilKing, on 2013-April-03, 10:11, said:

But it will still be double dummy sim, so doing a million or whatever makes no difference. There's no point opening 3NT if they are guaranteed to make the best lead.

I understand that double dummy and single trick averages tend to be close, but not in this auction!

I agree that the opening lead could be crucial, but looking at the sample of 100 I produced, passive leads were a huge winner. You know that they will have an average of 24 points, and dummy need not be balanced, so giving a trick on the lead is to be avoided.

And there were lots of hands where I would not expect to make 3NT but DF did. Some would suggest that I would not expect to make 3NT on the majority of hands, but the examples were declarer avoiding finesses and endplaying both opponents in turn, or playing some exotic squeeze. And I now think the opponents should double freely on balanced 8-counts. Failing to double when it goes 3 off is much more expensive than doubling when it makes 3 overtricks.

[PhilKing]

lamford, on 2013-April-04, 08:09, said:

I agree that the opening lead could be crucial, but looking at the sample of 100 I produced, passive leads were a huge winner. You know that they will have an average of 24 points, and dummy need not be balanced, so giving a trick on the lead is to be avoided.

As discussed elsewhere, passive leads are overrated by DD programs, because whenever they pick up partner's suit, the double dummy declarer was going to get it right anyway, and partner always finds the right switch where mere humans will sometimes erroneously return partner's suit.

It is entirely possible that the received wisdom of leading fourth highest on occasion is not completely wrong.

[lamford]

PhilKing, on 2013-April-04, 08:33, said:

As discussed elsewhere, passive leads are overrated by DD programs, because whenever they pick up partner's suit, the double dummy declarer was going to get it right anyway, and partner always finds the right switch where mere humans will sometimes erroneously return partner's suit.

It is entirely possible that the received wisdom of leading fourth highest on occasion is not completely wrong.

The passive leads were chosen by inspection, not by the program; I don't think there is a SIM that allows one that level of choice. If there is a SIM that plays "normal" bridge, it will be interesting to compare the results.

[mycroft]

lamford, on 2013-April-04, 08:03, said:

That would be fine but second-in-hand is not allowed to have the bid cancelled and the auction revert to him. Either the 1NT is accepted, or any replacement call is made and the 1NT bidder's partner is barred.

No, either the 1NT is accepted, or the auction reverts to the correct dealer, who passes by Law, and partner passes, and North bids 3NT.

So, by the time the *double* has to be made on the flat 10, East knows that partner, knowing that North knows that partner MUST pass, and that North has a 1NT opener, and that partner knows that North has a 1NT opener, chose to pass rather than make a call which could be a little off (as partner isn't going to hang you for it unless he thinks the points are all under him, right?)

[Edit: I didn't realize I was answering the wrong statement the first time.]

[lamford]

mycroft, on 2013-April-04, 09:53, said:

No, either the 1NT is accepted, or the auction reverts to the correct dealer, who passes by Law, and partner passes, and North bids 3NT.

So, by the time the *double* has to be made on the flat 10, East knows that partner, knowing that North knows that partner MUST pass, and that North has a 1NT opener, and that partner knows that North has a 1NT opener, chose to pass rather than make a call which could be a little off (as partner isn't going to hang you for it unless he thinks the points are all under him, right?)

[Edit: I didn't realize I was answering the wrong statement the first time.]

You are right, and I think second in hand should just bid any five-card suit on all hands at the one level, preparing the lead against 3NT. Pass should be 0-7 without a 5-card suit. All opening bids just pass and double 3NT. The only time it then goes 3NT all pass is when both defenders are 0-7. Trying this strategy on my 100 hands gave 3NT a negative expectancy and made it worse than Pass and way behind 1NT.

[mycroft]

Well, second hand probably shouldn't pass with a real opener, after it does go AP or p-p-1NT-...you'd be way behind over having shown your 14-with-5-hearts. Playing a weak NT, I probably would pass 1NT openers, though :-)

[lamford]

mycroft, on 2013-April-04, 13:29, said:

Well, second hand probably shouldn't pass with a real opener, after it does go AP or p-p-1NT-...you'd be way behind over having shown your 14-with-5-hearts. Playing a weak NT, I probably would pass 1NT openers, though :-)

But we are told by a reputable source that North bids 3NT 100% of the time, and never adopts a mixed strategy. So there is no risk in passing.

[PhilKing]

lamford, on 2013-April-05, 06:11, said:

But we are told by a reputable source that North bids 3NT 100% of the time, and never adopts a mixed strategy. So there is no risk in passing.

We had both overlooked the (very appealing) second seat one-of-a-suit strategy to show a weakish hand with a decent suit plus an entry. If they declarer that strategy, I will counter with some passes over their pass(non mixed obv - just pass bad 15s 1NT on good 15, 3NT decent 16 or 17) and judgment over their bids.

Anyway, I am now persuaded most 15 counts are not worth 3NT. And this hand looks like a downgrade today.

[lamford]

PhilKing, on 2013-April-05, 07:07, said:

We had both overlooked the (very appealing) second seat one-of-a-suit strategy to show a weakish hand with a decent suit plus an entry. If they declarer that strategy, I will counter with some passes over their pass(non mixed obv - just pass bad 15s 1NT on good 15, 3NT decent 16 or 17) and judgment over their bids.

Anyway, I am now persuaded most 15 counts are not worth 3NT. And this hand looks like a downgrade today.

Do they need to have that strategy on the CC against a third in hand 1NT out of turn? And I presume the one of a suit is alertable. Interestingly, they may not be allowed the agreement to open on less than some rule of 18 or other in EBU land. If they are not, the 3NT strategy becomes more appealing.

[PhilKing]

lamford, on 2013-April-05, 07:27, said:

Do they need to have that strategy on the CC against a third in hand 1NT out of turn? And I presume the one of a suit is alertable. Interestingly, they may not be allowed the agreement to open on less than some rule of 18 or other in EBU land. If they are not, the 3NT strategy becomes more appealing.

I don't think they are allowed a discussed non-standard strategy.

[Zelandakh]

That seems impossible Phil. That means that discussing the situation on BBF and reaching firm conclusions would put you in an impossible situation at the board owing to an opponent's infraction: "I am sorry, I cannot bid this board since my partner and I have discussed non-standard meanings for bids in this position." Yet the same agreement is legal by "bridge logic" deductions? Not buying it. If you are right about the regulations saying this then it is silliness. Are you sure it does not say that such an agreement is illegal after an infraction by our side, but legal if we are the NOS? If there is one section of the Laws/regulations that needs a good suggestion on "Changing Laws" it is surely this one!

[PhilKing]

Zelandakh, on 2013-April-05, 07:51, said:

That seems impossible Phil. That means that discussing the situation on BBF and reaching firm conclusions would put you in an impossible situation at the board owing to an opponent's infraction: "I am sorry, I cannot bid this board since my partner and I have discussed non-standard meanings for bids in this position." Yet the same agreement is legal by "bridge logic" deductions? Not buying it. If you are right about the regulations saying this then it is silliness. Are you sure it does not say that such an agreement is illegal after an infraction by our side, but legal if we are the NOS? If there is one section of the Laws/regulations that needs a good suggestion on "Changing Laws" it is surely this one!

I'm not up with the modern EBU rules, but when I was a junior I know we were not allowed a defence to insufficient bids. I know not whether this is analagous or still in force.

[lamford]

PhilKing, on 2013-April-05, 08:03, said:

I'm not up with the modern EBU rules, but when I was a junior I know we were not allowed a defence to insufficient bids. I know not whether this is analagous or still in force.

gordontd quoted a relevant regulation:
Under Law 40B3 (d) a pair is allowed to vary, by prior agreement, its understandings during the auction and play consequent on an irregularity by either side, except that following its own insufficient bid a partnership may not change by prior agreement the meaning of a replacement call so that it is brought within the criteria of Law 27B1 (b).

I don't think this means that the standards for an opening bid can be changed, and the rule will apply at whatever level one is playing. I think that an agreement to pass with a hand that will double 3NT will fall foul of:

"It is not permitted to play an opening pass to show values."

That really is nonsense.