BBO Discussion Forums: Do you come in? - BBO Discussion Forums

Jump to content

  • 3 Pages +
  • 1
  • 2
  • 3
  • You cannot start a new topic
  • You cannot reply to this topic

Do you come in?

#21 User is offline   msjennifer 

  • PipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 1,366
  • Joined: 2013-August-03
  • Gender:Female
  • Location:Variable private
  • Interests:Cricket,Photography,Paediatrics and Community Medicine.

Posted 2020-September-22, 04:45

Sir,I shall double for three reasons
1)THERE WILL NOT BE another chance.
2)We play LEBENSOHL so no chance of going overboard.
3)I am an aggressive bidder.
Thanks.
0

#22 User is offline   nudnikbp 

  • PipPipPip
  • Group: Full Members
  • Posts: 84
  • Joined: 2019-January-09

Posted 2020-September-22, 05:22

Pass. Mikeh pinpointed the reason: if partner has a hand in the 10 high card point range, you're likely to get too high.
0

#23 User is offline   Stephen Tu 

  • PipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 4,076
  • Joined: 2003-May-14

Posted 2020-September-22, 10:12

So I did some simulations, and I now believe double is best on a frequency basis. Our main fear is that partner puts in 3d in a 4-2 fit or 5-2 fit and that this is bad. But it looks like this happens only ~20% of the time by my estimate.

Almost half the time (~ 48%), partner has spades, we are presumably happy; partner won't be able to overcall/double on a decently large fraction of them. (of course we may subsequently overbid/underbid these, but some of these will presumably be cancelled out by opps competing in hearts when we were about to play an unfortunate diamond fit, or 5-2 diamonds getting out for down 1 when they were making hearts, etc.)

About 10% or so of the time partner has longer clubs then diamonds without spades.

About 40% of the time partner has diamonds w/o spades, and diamonds>=clubs. But somewhat over half these rate to work out, because either partner is strong enough to FG with 3H (or try 3nt himself) and we manage to bring in 3nt, or partner has 6+ diamonds and we are presumably not too unhappy.
1

#24 User is offline   DavidKok 

  • PipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 2,246
  • Joined: 2020-March-30
  • Gender:Male
  • Location:Netherlands

Posted 2020-September-22, 11:49

And what fraction of the time do we land on our feet after a pass?
0

#25 User is offline   FelicityR 

  • PipPipPipPipPip
  • Group: Full Members
  • Posts: 980
  • Joined: 2012-October-26
  • Gender:Female

Posted 2020-September-22, 11:55

View PostStephen Tu, on 2020-September-22, 10:12, said:

So I did some simulations, and I now believe double is best on a frequency basis.


Thank you for doing a simulation, Stephen. I've come in late on this discussion.

My view, for what its worth, is 5422 is not the best shapes generally at bridge - I believe Benito Garozzo discussed this shape somewhere - but with a minimum hand I would Double 2 with my regular partner because I would be showing s but not necessarily both minors, or a stronger hand with a different shape but not necessarily a 4 card suit.

However, if the opening bid had been 2 and I had 2425 shape I would consider pass even though we use Lebensohl, because the opponents are non-vulnerable and have the higher ranking suit and can probably outbid us.
0

#26 User is offline   Stephen Tu 

  • PipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 4,076
  • Joined: 2003-May-14

Posted 2020-September-22, 12:31

View PostDavidKok, on 2020-September-22, 11:49, said:

And what fraction of the time do we land on our feet after a pass?


I didn't calculate this but I would assume a great deal worse. The problem is West cannot come in with a balanced hand and 3cd hearts without like a 15 count for a 2nt balance. And if distributional can't double/overcall without very close to opening hand either IMO. Over anything East should drive to game IMO, so basically anything where partial is the limit you cannot reach. Or hands where you get out for -1 undoubled when they were making hearts.

If you start saying West should start balancing super-light in case East felt compelled to pass on hands like this, and East should not drive game, you hurt your bidding considerably when West holds stronger hands and East has to hold back, your ranges are too wide with too little space to sort things out, plus you get in trouble when South has the strong misfit hand instead of East.
0

#27 User is offline   mycroft 

  • Secretary Bird
  • PipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 7,114
  • Joined: 2003-July-12
  • Gender:Male
  • Location:Calgary, D18; Chapala, D16

Posted 2020-September-22, 16:22

And what fraction of the time do we get to 4 with 22 high? And what fraction of the time would we end up in 3 with that same 10-count when I have another card? All of your arguments about "balancing super-light" (which I think is frankly "about this hand", but I absolutely know that I overcall preempts very conservatively) apply in direct, if you will always come in with frankly, a pretty crappy weak 1NT opener.

Preempts Work, and they work because those kinds of decisions can't be made with great accuracy when you take away the 1 level.
When I go to sea, don't fear for me, Fear For The Storm -- Birdie and the Swansong (tSCoSI)
0

#28 User is online   pilowsky 

  • PipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 3,628
  • Joined: 2019-October-04
  • Gender:Male
  • Location:Israel

Posted 2020-September-22, 16:48

View PostDavidKok, on 2020-September-22, 04:33, said:

This idea of 'shouldering the burden yourself' is noble, but probably not winning bridge. As they say "preempts work", and Stephen already gave an excellent example of why the treatments over 1NT and 2 should be different despite the gap between these bids not being massive. The fact is that after a preempt by opponents you are short on bidding space, and frequently will have to bid based on informed guesses and odds instead of scientific investigation. In other words, you cannot afford to wait for the perfect hand that will make partner happy no matter what, often you'll have to risk a poor outcome for a chance at a good one. Preempts work.

As for partner passing, this would typically show either a five-card heart suit (or a solid four-card suit), or a decent number of points (let's say 9-12) with a notrump distribution, 4 hearts and values in the heart suit, gambling that 3NT is one off but 2X can be set. At this vulnerability the latter hand type might well bid anyway, red against white is not the time for speculative penalties. Regardless your A9 are wonderful news for partner, and while you are on the light side your hand has OK values for a takeout double (but not the ideal shape). Slam down the ace of hearts and collect a good penalty without trouble. Instead, the auction I'm worried about is partner pulling to 3 (possibly with a convention).


Well, just to put our minds at ease, I gave the hand to the robots on a teaching table. I have a funny feeling that I read somewhere that Mike Lawrence had a hand in designing GIB - I might be wrong. In his complete book on doubles, he states "a five-card suit is gold".
The robots overcalled 3 and after I bid 3 they made their way to 4. I let them play it out.
Here's what happened.
Non legit hoc
0

#29 User is offline   Stephen Tu 

  • PipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 4,076
  • Joined: 2003-May-14

Posted 2020-September-22, 17:04

View Postmycroft, on 2020-September-22, 16:22, said:

And what fraction of the time do we get to 4 with 22 high? And what fraction of the time would we end up in 3 with that same 10-count when I have another card?

I would imagine pretty low, unless the 10 count opposite had a lot of shape to justify a GF. You have invitational sequences available; 3S and for some people also 2nt followed by 3S. This hand of course would decline, and if you have another king you would presumably accept.

Now, 3S will go down on some of the worse 10 counts, and that would be sub-optimal, but you may find that they were making 2H and partner wasn't going to balance if you didn't act so you were never really getting +110 and -100 might be better than -110. It'll probably be pretty difficult for the opener's partner to find the axe.

Quote

All of your arguments about "balancing super-light" (which I think is frankly "about this hand", but I absolutely know that I overcall preempts very conservatively) apply in direct, if you will always come in with frankly, a pretty crappy weak 1NT opener.
This is obviously a dead minimum, but it's definitely easier for someone to come in with doubleton heart than tripleton heart. IMO it works better if the short hand is slightly aggressive by a point maybe rather than balancing seat being very aggressive by a king or so just because it's balancing.

Quote

Preempts Work, and they work because those kinds of decisions can't be made with great accuracy when you take away the 1 level.

Obviously you are going to get some wrong, but the main question is whether you get more frequent good results than bad ones by acting rather than passing. I know I'd personally much rather double on this hand than balance overcall 2S on my 8/9 count or balance double with 4342 on 9 count. Or you could get conservative and pass both sides and avoid 3d disasters but you'll lose a lot when 2s makes or you drive them into 3H down 1.
0

#30 User is offline   mikeh 

  • PipPipPipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 12,849
  • Joined: 2005-June-15
  • Gender:Male
  • Location:Canada
  • Interests:Bridge, golf, wine (red), cooking, reading eclectically but insatiably, travelling, making bad posts.

Posted 2020-September-22, 19:15

View PostStephen Tu, on 2020-September-22, 10:12, said:

So I did some simulations, and I now believe double is best on a frequency basis. Our main fear is that partner puts in 3d in a 4-2 fit or 5-2 fit and that this is bad. But it looks like this happens only ~20% of the time by my estimate.

Almost half the time (~ 48%), partner has spades, we are presumably happy; partner won't be able to overcall/double on a decently large fraction of them. (of course we may subsequently overbid/underbid these, but some of these will presumably be cancelled out by opps competing in hearts when we were about to play an unfortunate diamond fit, or 5-2 diamonds getting out for down 1 when they were making hearts, etc.)

About 10% or so of the time partner has longer clubs then diamonds without spades.

About 40% of the time partner has diamonds w/o spades, and diamonds>=clubs. But somewhat over half these rate to work out, because either partner is strong enough to FG with 3H (or try 3nt himself) and we manage to bring in 3nt, or partner has 6+ diamonds and we are presumably not too unhappy.

I don’t understand this post. A proper simulation should not conclude that ‘we are presumably happy’ to have doubled when partner has spades.

Surely whether we are happy depends on whether the double realistically gets us to a better outcome than would a pass.

Playing 4S doubled, down 2, for example, would likely leave us extremely unhappy. Reaching 4S making would be good, but maybe we’d reach it after a pass.

In other words, analyzing a simulation requires a huge amount of largely subjective analysis. If the point was that double reaches spades often, that’s of no use at all.
'one of the great markers of the advance of human kindness is the howls you will hear from the Men of God' Johann Hari
0

#31 User is offline   Stephen Tu 

  • PipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 4,076
  • Joined: 2003-May-14

Posted 2020-September-22, 19:56

View Postmikeh, on 2020-September-22, 19:15, said:

I don't understand this post. A proper simulation should not conclude that 'we are presumably happy' to have doubled when partner has spades.

Surely whether we are happy depends on whether the double realistically gets us to a better outcome than would a pass.

Playing 4S doubled, down 2, for example, would likely leave us extremely unhappy. Reaching 4S making would be good, but maybe we'd reach it after a pass.

In other words, analyzing a simulation requires a huge amount of largely subjective analysis. If the point was that double reaches spades often, that's of no use at all.


It's a first step ballpark estimation, not intended to be a down to the last percentage estimate of how much better one action is than another. I was interested in the relative frequencies partner has spades, to the nightmare scenario where partner bids 3D.
I think if anyone told you that partner was guaranteed to bid spades and wasn't going to bid diamonds you would rather double then pass, no? Of course it's not going to work out 100%, but I think it's clear bidding will work out a ton more often then passing.

Now if you want to bore down into greater detail, I can. For example, I looked at hands where partner has 6-9 hcp and 4+ spades, where he is unlikely to act if we don't. Where South doesn't have a huge fit or say a strong NT hand that would 2nt inquiry or whatever. Here we are making 2S+ some 86+% of the time and down 1 only another 12 or so. While they are making 2H+ 90+%. Surely we want to bid, no?

0

#32 User is offline   mikeh 

  • PipPipPipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 12,849
  • Joined: 2005-June-15
  • Gender:Male
  • Location:Canada
  • Interests:Bridge, golf, wine (red), cooking, reading eclectically but insatiably, travelling, making bad posts.

Posted 2020-September-22, 20:25

View PostStephen Tu, on 2020-September-22, 19:56, said:

It's a first step ballpark estimation, not intended to be a down to the last percentage estimate of how much better one action is than another. I was interested in the relative frequencies partner has spades, to the nightmare scenario where partner bids 3D.
I think if anyone told you that partner was guaranteed to bid spades and wasn't going to bid diamonds you would rather double then pass, no? Of course it's not going to work out 100%, but I think it's clear bidding will work out a ton more often then passing.

Now if you want to bore down into greater detail, I can. For example, I looked at hands where partner has 6-9 hcp and 4+ spades, where he is unlikely to act if we don't. Where South doesn't have a huge fit or say a strong NT hand that would 2nt inquiry or whatever. Here we are making 2S+ some 86+% of the time and down 1 only another 12 or so. While they are making 2H+ 90+%. Surely we want to bid, no?

You and I seem to have very different ideas about simulations. Of course, I ‘knew’ that partner has 6-9 hcp and 4+spades, I’d want to bid. In my wirld, catering to a narrow set of hands partner might hold does not lead to a useful simulation.

As I mentioned in my first post, doing it properly requires an enormous amount of work, and is just about impossible to do objectively.

One could try, but it would likely require analyzing, individually, at least 100 deals, and deciding what each player would do with each hand. So that if two good analysts examined the same hands, they’d likely draw different conclusions. I’d expect one who began feeling double was good would find that confirmed and one who felt it was a poor action would find the same hands to confirm that view as well😃 There’s just too much subjectivity.
'one of the great markers of the advance of human kindness is the howls you will hear from the Men of God' Johann Hari
1

#33 User is offline   Stephen Tu 

  • PipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 4,076
  • Joined: 2003-May-14

Posted 2020-September-22, 21:13

Well, you are welcome to deal out a bunch of hands and take a look at them yourself in excruciating detail and get a feel for what is happening. I know that if partner has spades I am in good shape. And the diamond scenario is relatively low. The frequency difference is enough that I personally don't feel I have to go through 100 hands in double dummy detail. After looking at several printouts of sample hands I'm pretty sure I want to double without carefully counting exactly how many tricks are available on each board.



0

#34 User is online   pilowsky 

  • PipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 3,628
  • Joined: 2019-October-04
  • Gender:Male
  • Location:Israel

Posted 2020-September-22, 21:28

View PostStephen Tu, on 2020-September-22, 21:13, said:

Well, you are welcome to deal out a bunch of hands and take a look at them yourself in excruciating detail and get a feel for what is happening. I know that if partner has spades I am in good shape. And the diamond scenario is relatively low. The frequency difference is enough that I personally don't feel I have to go through 100 hands in double dummy detail. After looking at several printouts of sample hands I'm pretty sure I want to double without carefully counting exactly how many tricks are available on each board.




Like all matters of Judgement, it varies inversely with the (hours of wakefulness multiplied by your blood alcohol level) divided by masterpoints. This is a well-known fact.
Non legit hoc
0

#35 User is offline   nullve 

  • PipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 2,234
  • Joined: 2014-April-08
  • Gender:Male
  • Location:Norway
  • Interests:partscores

Posted 2020-September-23, 01:01

Here are 100 deals

Spoiler

dealt randomly (using the Dealer script

Spoiler

) on the condition that

  • North has something resembling a classic Weak 2 in terms of hcp and shape (so a hand with 5-S6H and 5-10 hcp that doesn't meet the rule of 20);
  • East has 12 hcp and 4225 shape (because we want to know to do on "similar" hands, not just one that we will almost certainly never be dealt again).

0

#36 User is online   pilowsky 

  • PipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 3,628
  • Joined: 2019-October-04
  • Gender:Male
  • Location:Israel

Posted 2020-September-23, 01:26

Perhaps we could resolve this with a 12 board challenge format? Step outside boys. Posted Image
Non legit hoc
0

#37 User is offline   AL78 

  • PipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 1,967
  • Joined: 2019-October-13
  • Gender:Male
  • Location:SE England
  • Interests:Bridge, hiking, cycling, gardening, weight training

Posted 2020-September-23, 02:27

View PostFelicityR, on 2020-September-22, 11:55, said:

However, if the opening bid had been 2 and I had 2425 shape I would consider pass even though we use Lebensohl, because the opponents are non-vulnerable and have the higher ranking suit and can probably outbid us.


I don't see why it should make any difference whether the opening is in hearts or spades. They can outbid you in spades, but so what? If they outbid you, you have forced them a level higher which might mean they go down. If you don't bid, they don't have to outbid you and can play at the two level. This is assuming they will bid over you.
1

#38 User is offline   DavidKok 

  • PipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 2,246
  • Joined: 2020-March-30
  • Gender:Male
  • Location:Netherlands

Posted 2020-September-23, 02:53

View Postpilowsky, on 2020-September-22, 16:48, said:

Well, just to put our minds at ease, I gave the hand to the robots on a teaching table. I have a funny feeling that I read somewhere that Mike Lawrence had a hand in designing GIB - I might be wrong. In his complete book on doubles, he states "a five-card suit is gold".
The robots overcalled 3 and after I bid 3 they made their way to 4. I let them play it out.
Here's what happened.

This does not put my mind at ease at all. The cold 4 is missed, and NS are doing something rather fancy (not to say 'ridiculous') on trick 5. Also South should grow a pair and bid 4 instead of 3. And I just noticed the vulnerability doesn't match the poll.
0

#39 User is offline   bluenikki 

  • PipPipPipPipPip
  • Group: Full Members
  • Posts: 561
  • Joined: 2019-October-14

Posted 2020-September-23, 08:45

View Postmsjennifer, on 2020-September-22, 04:45, said:

Sir,I shall double for three reasons
1)THERE WILL NOT BE another chance.
2)We play LEBENSOHL so no chance of going overboard.
3)I am an aggressive bidder.
Thanks.


In your partnership, what is the minimum strength for a *jump* advance here?

If you were playing with a stranger who had agreed to lebensohl, what would you expect that minimum to be?
0

#40 User is offline   mikeh 

  • PipPipPipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 12,849
  • Joined: 2005-June-15
  • Gender:Male
  • Location:Canada
  • Interests:Bridge, golf, wine (red), cooking, reading eclectically but insatiably, travelling, making bad posts.

Posted 2020-September-23, 09:38

Hi nullve

I appreciate the effort, but think that generating hands for east invalidates the procedure. Not all 4=2=2=5 12 counts are the same. Bear in mind that I think that most players see the east hand as borderline (imo, pass is the better call but obviously there will be many hands where it is not, and I think the bidders see it similarly but in reverse)

The actual hand has poor texture and a horrible diamond holding. A lot of your examples have all their hcp in their long suits, improving the hand significantly. Also, Kx in hearts is almost a good as AX, in terms of the heart suit, and leaves an extra point for the long suits.

Ironically, I think a more appropriate simulation is a lot easier to do (one hand is known) and definitely a lot easier to analyze. Finally, double dummy results are simply wrong, unles you know people who can both bid and play double dummy. No double dummy analysis wil, help you decide how the auction would go, let alone hoe the hand would be played in real life.
'one of the great markers of the advance of human kindness is the howls you will hear from the Men of God' Johann Hari
0

  • 3 Pages +
  • 1
  • 2
  • 3
  • You cannot start a new topic
  • You cannot reply to this topic

1 User(s) are reading this topic
0 members, 1 guests, 0 anonymous users