BBO Discussion Forums: 5 card spades and weak responder - BBO Discussion Forums

Jump to content

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

5 card spades and weak responder Advantage (or not) of opening 1NT with a 5-card major

#1 User is online   pescetom 

  • PipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 7,294
  • Joined: 2014-February-18
  • Gender:Male
  • Location:Italy

Posted 2020-October-27, 06:54

I hacked this rough script to see whether it was vantageous in terms of score to open 1NT rather than 1♠ with 5332 spades opposite a weak responder.
It assumes that opener is vulnerable, responder will transfer in any suit but can not Stayman at this strength, opponents will not let us play 1♠ but otherwise keep quiet. 

#
# NT vs Spades comparison
#
nt1517= hcp(north)>=15 and hcp(north)<=17
weak= hcp(south)>=4 and hcp(south)<=7
invite= hcp(south)>=8 and hcp(south)<=9
thingame= hcp(south)>=10 and hcp(south)<=10
safegame= hcp(south)>=11 and hcp(south)<=12
resp5332= shape(south, any 5332) and hearts(south)<=3 and spades(south)<=3
xfer= spades(south)>=5 or hearts(south)>=5 or clubs(south)>=6 or diamonds(south)>=5
s5 = spades(north)==5 and shape(north, any 5332)
produce 1000
condition nt1517 and s5 and weak and (not xfer)
x1N_score= score(vul,x1N,tricks(north,notrumps))
x2S_score= score(vul,x2S,tricks(north,spades))
outcome = x1N_score == x2S_score ? 0 : (x1N_score > x2S_score ? 1 : -1)
action frequency(outcome,-1,1)


Run:
Frequency :
   -1        567
    0        114
    1        319
Generated 1476991 hands
Produced 1000 hands
Initial random seed 1603879365
Time needed  381.751 sec


The result is that 5771% of the time opening 1NT leads to a worse score, 3229% a better score, 11%never the same.
[NOTE: percentages edited after elimination of a bug spotted by @nullve]
Not good news if true, particularly at MP.
I expected this, both from experience and from previous discussions, but not so severe.
Maybe one does not notice it as happening frequently because this precise situation is rare (1 in 1477 of all boards of this simulation).
Any thoughts/comments, faulty premises or errors spotted?
1

#2 User is offline   Cyberyeti 

  • PipPipPipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 13,919
  • Joined: 2009-July-13
  • Location:England

Posted 2020-October-27, 07:21

I'd be interested in seeing the equivalent for a weak NT although you'd need to remove the "no stayman" clause. On the odd occasion I played a strong NT we still had garbage Stayman there, but I know many don't.
0

#3 User is offline   nullve 

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

Posted 2020-October-27, 08:09

View Postpescetom, on 2020-October-27, 06:54, said:

Any thoughts/comments, faulty premises or errors spotted?

Not an error but

invite= hcp(south)>=8 and hcp(south)<=9
thingame= hcp(south)>=10 and hcp(south)<=10
safegame= hcp(south)>=11 and hcp(south)<=12
resp5332= shape(south, any 5332) and hearts(south)<=3 and spades(south)<=3

does nothing right now because of

condition nt1517 and s5 and weak and (not xfer)

0

#4 User is offline   nullve 

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

Posted 2020-October-27, 08:33

Here

Spoiler

are 100 deals that satisfy your condition.

As you can see, there are many deals where EW would have competed after a 1N opening, giving NS another chance to find a superior 2♠ contract.
1

#5 User is offline   Cyberyeti 

  • PipPipPipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 13,919
  • Joined: 2009-July-13
  • Location:England

Posted 2020-October-27, 08:55

Also are you assuming that if partner responds 1N to 1♠ you're passing with 17, I think you're playing 2N from the wrong side rather than 1N.
0

#6 User is online   pescetom 

  • PipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 7,294
  • Joined: 2014-February-18
  • Gender:Male
  • Location:Italy

Posted 2020-October-27, 09:44

View Postnullve, on 2020-October-27, 08:09, said:

Not an error but

invite= hcp(south)>=8 and hcp(south)<=9
thingame= hcp(south)>=10 and hcp(south)<=10
safegame= hcp(south)>=11 and hcp(south)<=12
resp5332= shape(south, any 5332) and hearts(south)<=3 and spades(south)<=3

does nothing right now because of

condition nt1517 and s5 and weak and (not xfer)


That's intentional, they are for the next instalments :)
0

#7 User is online   pescetom 

  • PipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 7,294
  • Joined: 2014-February-18
  • Gender:Male
  • Location:Italy

Posted 2020-October-27, 09:47

View Postnullve, on 2020-October-27, 08:33, said:

As you can see, there are many deals where EW would have competed after a 1N opening, giving NS another chance to find a superior 2♠ contract.


Good point, thanks. I had decided to ignore this (and also the possibility of opponents winning the auction) but thinking about it, it is worth quantifying.
0

#8 User is online   pescetom 

  • PipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 7,294
  • Joined: 2014-February-18
  • Gender:Male
  • Location:Italy

Posted 2020-October-27, 09:52

View PostCyberyeti, on 2020-October-27, 08:55, said:

Also are you assuming that if partner responds 1N to 1♠ you're passing with 17, I think you're playing 2N from the wrong side rather than 1N.

1N is semi-forcing: no question of pass, but 2N is also right at the top end for 1N equivalence; 1S opener will usually rebid a 3card minor and I assumed (out of laziness/simplicity) that leads to 2S.
0

#9 User is offline   Vampyr 

  • PipPipPipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 10,611
  • Joined: 2009-September-15
  • Gender:Female
  • Location:London

Posted 2020-October-27, 10:17

View Postpescetom, on 2020-October-27, 09:52, said:

1N is semi-forcing: no question of pass, but 2N is also right at the top end for 1N equivalence; 1S opener will usually rebid a 3card minor and I assumed (out of laziness/simplicity) that leads to 2S.


Your preferred system was not taken into account in the simulation or the comment you quoted.
I know not with what weapons World War III will be fought, but World War IV will be fought with sticks and stones -- Albert Einstein
0

#10 User is online   smerriman 

  • PipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 3,754
  • Joined: 2014-March-15
  • Gender:Male

Posted 2020-October-27, 13:27

I think it's pretty clearcut that if you knew in advance the choice was between playing 1NT and 2M, you'd always rather play 2M.

But that seems to have no relation to the reasons people decide to open 1NT with a 5cM, which is to help with all of the other cases (including competition).
Reverse engineering GIB, part 1: A treatise on the insanity of bidding simulations
0

#11 User is online   pescetom 

  • PipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 7,294
  • Joined: 2014-February-18
  • Gender:Male
  • Location:Italy

Posted 2020-October-27, 15:28

View PostVampyr, on 2020-October-27, 10:17, said:

Your preferred system was not taken into account in the simulation or the comment you quoted.


Not sure what you mean by that.
The simulation (as explained) assumed that such bidding would more often than not converge on 2♠.
The comment I quoted (if you mean that of cyberyeti, as I assume) was not easy to parse, but I took it as asking whether North might pass 1NT (no) and whether we would end up in 2NT if not (almost certainly no).
0

#12 User is offline   nige1 

  • 5-level belongs to me
  • PipPipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 9,128
  • Joined: 2004-August-30
  • Gender:Male
  • Location:Glasgow Scotland
  • Interests:Poems Computers

Posted 2020-October-27, 17:07

View Postpescetom, on 2020-October-27, 06:54, said:

I hacked this rough script to see whether it was vantageous in terms of score to open 1NT rather than 1♠ with 5332 spades opposite a weak responder. It assumes that opener is vulnerable, responder will transfer in any suit but can not Stayman at this strength, opponents will not let us play 1♠ but otherwise keep quiet.
Spoiler

The result is that 71% of the time opening 1NT leads to a worse score, 29% a better score, never the same. Not good news if true, particularly at MP. I expected this, both from experience and from previous discussions, but not so severe.
Maybe one does not notice it as happening frequently because this precise situation is rare (1 in 1448 of all boards of this simulation).
Any thoughts/comments, faulty premises or errors spotted?

I often open notrump with a 5-card major. For game purposes, if he chooses to explore for a 3-5 fit, partner can use some variation of 5-card Stayman (e.g. 3♣ = Muppet).

For partscores, opponents rarely defend double-dummy. And are more likely to lead a major than a minor to 1N.

You are well-placed If you play Gazzilli, and open 1♠ with 16+ HCP.

With 15 (or a good 14), however, you can avoid rebid problems by opening 1N.
Guthrie.tech
0

#13 User is offline   Vampyr 

  • PipPipPipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 10,611
  • Joined: 2009-September-15
  • Gender:Female
  • Location:London

Posted 2020-October-27, 17:56

View Postpescetom, on 2020-October-27, 15:28, said:

Not sure what you mean by that.
The simulation (as explained) assumed that such bidding would more often than not converge on 2♠.
The comment I quoted (if you mean that of cyberyeti, as I assume) was not easy to parse, but I took it as asking whether North might pass 1NT (no) and whether we would end up in 2NT if not (almost certainly no).

What I meant is that a semi-forcing or forcing NT was not a condition of the simulation. What Cyberyeti meant was that opener will usually raise the 1NT response with 17.
I know not with what weapons World War III will be fought, but World War IV will be fought with sticks and stones -- Albert Einstein
0

#14 User is offline   mycroft 

  • Secretary Bird
  • PipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 7,107
  • Joined: 2003-July-12
  • Gender:Male
  • Location:Calgary, D18; Chapala, D16

Posted 2020-October-27, 17:56

Any pair who lets 2-of-a-fit play more often than 1NT-AP deserves what they get. This is why.

The assumption that they will never bid over strong NT died thirty years ago; the assumption that after 1♠-2♠, they will never bid is dying; frankly, it's more likely that they will pass out 1♠ than 2.

What you've shown is that a trump fit tends to play a trick better than in NT, if extra strength isn't there. And by taking away all the transfer hands (and not taking away the Garbage Stayman hands, but that's not really a big deal), you're increasing the odds of a spade fit fairly significantly:

[spades]   [sim] [-"not transfer"] [difference]
    0	       2          28          -2.5%
    1	      78         129          -5%
    2	     271         297          -3%
    3	     369         302          +7%
    4	     280         181          +10(!)%
    5+         0          63          -6% (I was surprised it was that high, but it seems to be consistent through runs)


The assumption that you would be allowed to play in the (almost 30% likely) 5-4 fit at the 2 level with 19-23 high seems even less reasonable than the 37% 5-3 fit.

But whether in isolation NT or spades will score better with 5♠332s isn't really the point. The reason to open 1NT with 5332, especially 5♠332 is so you don't have to worry about a flat medium hand in your responses to 1♠-1NT or 1♠-2m. Any losses from 1NT-AP auctions have to be compared against the gains from the hands where we do open 1♠ and there's that nice gap between 14 and 18, unless partner has a second suit; whether it's a competitive auction or just a "work out the level" one.
When I go to sea, don't fear for me, Fear For The Storm -- Birdie and the Swansong (tSCoSI)
0

#15 User is offline   nullve 

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

Posted 2020-October-27, 18:26

View Postpescetom, on 2020-October-27, 06:54, said:

Run:
Frequency :
   -1	     710
    0	       0
    1	     290
Generated 1447933 hands
Produced 1000 hands
Initial random seed 1603798540
Time needed  417.215 sec


The result is that 71% of the time opening 1NT leads to a worse score, 29% a better score, never the same.

So not even once did 1N and 2♠ go off by the same number of tricks??? Something must be wrong...

Maybe

outcome = x1N_score == x2S_score ? 0 : x1N_score > x2S_score ? 1 : -1

doesn't mean what I thought it does, which (in pseudo-code) is:

if (x1N_score == x2S_score) 
	outcome = 0 ;
else 
	if (x1N_score  > x2s_score)
		outcome = 1;
	else
		outcome = -1;


EDIT:

Running the script with

outcome = x1N_score == x2S_score ? 0 : (x1N_score > x2S_score ? 1 : -1)

instead of

outcome = x1N_score == x2S_score ? 0 : x1N_score > x2S_score ? 1 : -1

:

Frequency :
   -1	     567
    0	     114
    1	     319
Generated 1476991 hands
Produced 1000 hands
Initial random seed 1603879365
Time needed  381.751 sec

This post has been edited by nullve: 2020-October-28, 04:16

0

#16 User is offline   nullve 

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

Posted 2020-October-27, 19:30

[Deleted pointless experiment]
0

#17 User is offline   dsLawsd 

  • PipPipPipPip
  • Group: Full Members
  • Posts: 299
  • Joined: 2017-September-15

Posted 2020-October-28, 01:09

Interesting, but computers are not good at parsing suit texture OR
honor location in general. If the hand looks no-trumpish then
1NT gives some bailouts whereas strong suits give protection vulnerable at spades.
On a good day things go well. On a bad day whatever I choose turns out rather poor.
And course there are some 17 point hands and some 17! point ones.
At least one of my partnerships past, this issue changed our
1 NT to 14-16.
I suspect spades being the boss suit tends to be chosen whenever
doubt exists...

That is why I keep a daily record with remarks for regular
partnerships to analyze + or - results. That means, of course,
that you needs quite a lot of them to be statistically
meaningful.

Thanks for this interesting view.
0

#18 User is offline   bluenikki 

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

Posted 2020-October-28, 07:22

View PostCyberyeti, on 2020-October-27, 07:21, said:

I'd be interested in seeing the equivalent for a weak NT although you'd need to remove the "no stayman" clause. On the odd occasion I played a strong NT we still had garbage Stayman there, but I know many don't.

The term garbage stayman is an abomination.

Regardless, even if you play 1NT - 2♣ - 2♦ - 2♠ as forcing or invitational, surely you and everyone else, when dealt

xxx
xxxx
xxxxx
x

will reply 2♣.
1

#19 User is online   pescetom 

  • PipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 7,294
  • Joined: 2014-February-18
  • Gender:Male
  • Location:Italy

Posted 2020-October-28, 07:48

View Postnullve, on 2020-October-27, 18:26, said:

So not even once did 1N and 2♠ go off by the same number of tricks??? Something must be wrong...


I thought so too, but couldn't spot why. Thanks :)

I will correct code and results in the OP to avoid confusing anyone new to the thread.
The results now look more like what I expected.
0

#20 User is online   pescetom 

  • PipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 7,294
  • Joined: 2014-February-18
  • Gender:Male
  • Location:Italy

Posted 2020-October-28, 10:07

View Postnige1, on 2020-October-27, 17:07, said:

I often open notrump with a 5-card major. For game purposes, if he chooses to explore for a 3-5 fit, partner can use some variation of 5-card Stayman (e.g. 3♣ = Muppet).

For partscores, opponents rarely defend double-dummy. And are more likely to lead a major than a minor to 1N.

You are well-placed If you play Gazzilli, and open 1♠ with 16+ HCP.

With 15 (or a good 14), however, you can avoid rebid problems by opening 1N.


That's the basis of our system too, with a native 5-card Stayman. But as it is under end of year scrutiny I wanted to look at where it was costing us or might not be working optimally (for instance, when responder with 5m332 should Stayman at various strengths, which is what I will be looking at next).
0
  • 2 Pages +
  • 1
  • 2
  • 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