GF or limit raise?

[benlessard]

If partner has 4H then its surely GF worthy, if he has 3H then its probably a limit if he has 2 then its surely a limit.

 

KQxxx

KQxx

??

?? + a little extra

 

KQxxxx

Qxx

Axx

x

 

KQxxx

Kx

Axxx

xx

 

KJxxx

Kx

KQxx

xx

 

Qxxxx

QJx

AKx

xx

 

My feeling is that if your opening requirement a are a bit under these hands (clear but minimum openings) then a limit raise is probably better.

This is 1 advantage of non-Gf relays, having opener make a 2nd bids before GF.

[Cascade]

Here is a double dummy simulation with this hand opposite an ordinary hand with five spades and nothing more distributional than 5-4-3-1.

 

hcp Prob more than ten tricks

 

9 0.241525424

10 0.45631068

11 0.652694611

12 0.765822785

13 0.857142857

14 0.946666667

15 0.966666667

 

Most players don't open all eleven counts so this seems to suggest that Game Forcing opposite a sound opening is reasonable.

[655321]

But what are the chance of 10 tricks opposite 5332 11 and 12 - these are the only hands that would pass a 4 card limit raise. All 5431 will bid 4♠, so shouldn't be included in the simulation.

[Codo]

And with how many 5431 hands do you pass a limit raise?

I think any 12 HCP hand with 5431 is a simple 4 Spade over 3 Spade, so these hands do not count.

Please make the simulation for 5332s too, I would really like to see the results.

 

If you open all/most 7 loser hands with 11 HCPs, this is a limit raise to me.

If you play sound openers, this is GF.

 

The poster claims to play both, which seems to be impossible to me. :)

[matmat]

Cascade: "Here is a double dummy simulation with this hand opposite an ordinary hand with five spades and nothing more distributional than 5-4-3-1."

 

nothing MORE distributional than 5-4-3-1. i assume this includes the more balanced hands

[skjaeran]

This is a close to max limit raise for me.

 

We don't need much of an excuse to accept a limit raise though.

[CSGibson]

And with how many 5431 hands do you pass a limit raise?

I think any 12 HCP hand with 5431 is a simple 4 Spade over 3 Spade, so these hands do not count.

Please make the simulation for 5332s too, I would really like to see the results.

 

If you open all/most 7 loser hands with 11 HCPs, this is a limit raise to me.

If you play sound openers, this is GF.

 

The poster claims to play both, which seems to be impossible to me. :)

Let's say sound within context of what I've seen posted here.

[Cascade]

Cascade: "Here is a double dummy simulation with this hand opposite an ordinary hand with five spades and nothing more distributional than 5-4-3-1."

 

nothing MORE distributional than 5-4-3-1. i assume this includes the more balanced hands

Yes precisely I said:

 

hcp between 9 and 15 inclusive;

 

exactly five spades;

 

four or fewer hearts;

 

four or fewer diamonds;

 

four or fewer clubs;

 

one or more hearts;

 

one or more diamonds;

 

one or more clubs.

 

Here is the relevant code:

 

             (spades(south)==5) and 
            (hearts(south)<=4) and
            (diamonds(south)<=4) and
            (clubs(south)<=4) and

            (hearts(south)>=1) and
            (diamonds(south)>=1) and
            (clubs(south)>=1) and

            (hcp(south)>=9) and
            (hcp(south)<=15)

[Cascade]

Here are the numbers for 5-3-3-2 and 5-4-2-2 hands

 

hcp Prob ten or more tricks

9 0.178571429

10 0.32195122

11 0.556701031

12 0.736842105

13 0.792792793

14 0.93258427

15 0.962264151

[655321]

Here are the numbers for 5-3-3-2 and 5-4-2-2 hands

Trust me, I will bid game as opener with 5422.

Can someone do a simulation for 5332 only?

 

These numbers are interesting - the numbers for 10 and 11 HCP dropped substantially with the restriction, but there was not much difference for 12 or more.

Also, do these numbers tend to be similar to real life - ie does the Deep Finesse finding all the missing queens get balanced by Deep Finesse making the best opening lead?

[Cascade]

Here are the numbers based on specific hand type:

 

   
   5431         5422         5332
9  0.582568807  0.299559471  0.100961538
10 0.71563981   0.491712707  0.225490196
11 0.875776398  0.644859813  0.478021978
12 0.880239521  0.686567164  0.617021277
13 0.98         0.863636364  0.776119403
14 1            0.972972973  0.867647059
15 1            0.933333333  0.968253968

 

Each simulation was 1000 hands in total over all hcp. There were more hands with 9, 10, 11 hcp than 14 or 15 hcp since they are more frequent.

[655321]

Thanks Wayne.

[Cascade]

Here are the numbers for 5-3-3-2 and 5-4-2-2 hands

Trust me, I will bid game as opener with 5422.

Can someone do a simulation for 5332 only?

 

These numbers are interesting - the numbers for 10 and 11 HCP dropped substantially with the restriction, but there was not much difference for 12 or more.

Also, do these numbers tend to be similar to real life - ie does the Deep Finesse finding all the missing queens get balanced by Deep Finesse making the best opening lead?

The analysis that I have heard about suggests that declarer has an advantage over double dummy at low-levels (e.g. 1NT) and a disadvantage over double dummy at high-levels (Grand slams). Somewhere between they are roughly balanced.

[whereagles]

Totally GF for me. Not even a min... lol.

[Codo]

Here are the numbers based on specific hand type:

 

   
   5431         5422         5332
9  0.582568807  0.299559471  0.100961538
10 0.71563981   0.491712707  0.225490196
11 0.875776398  0.644859813  0.478021978
12 0.880239521  0.686567164  0.617021277
13 0.98         0.863636364  0.776119403
14 1            0.972972973  0.867647059
15 1            0.933333333  0.968253968

 

Each simulation was 1000 hands in total over all hcp. There were more hands with 9, 10, 11 hcp than 14 or 15 hcp since they are more frequent.

Thx a lot.

 

So these days where you bid 4 M with 26 points are long gone.

 

With prime 11 HCPs and 4 card support in a 4432 hand to a "nearly" random 11 HCP 5 card major hand you better be in game, because it makes double dummy more then 50 % of all times.

 

Maybe I must improve my declarer game, I often fail to make ten tricks in this case. But I will give it a try again.

[kenrexford]

This hand might be GF for many people but limit for others.

 

I and my partner open "shapely nine-counts" too often to force game here.  In fact, 1♠-P-3♣!-P-3♠ is a sequence that is usually followed by "good luck partner" and "can't you take a joke?"

OP stated what the requirements are for an opening bid in that partnership, what your own partner would open with is not really relevant.

Oh. Missed that.

 

I'd still bid 3♣, and then bid 3NT even if partner suggested a signoff, then. That is, if I understand the meaning of "sound" openings.

 

The examples given, like three kings and a queen all separated in a 5332 11-count, are not consistent with my idea of what "sound" openings means.

[P_Marlowe]

Hi,

 

limit raise, espesially if you show 4 card support.

I agree, it is a max. for the limit raise, and that you

could certainly make a case for counting this as 7

loosers.

 

But than partner will usually accept the limit raise anyway

unless he is dead min. and bal.

And if this happens, it will certainly not be a bad idea to

play 3S.

 

With kind regards

Marlowe

[jdonn]

If going by Wayne's results I think that makes bidding game totally clear. I have to admit they suggest it would be a bit more successful than I expected though. BTW, very many 5332 13s would pass a limit raise I assume, as well as 5422 11.

[Cascade]

So these days where you bid 4 M with 26 points are long gone.

I recall from doing this sort of analysis previously that with a 9-card trump fit (5-4) only around 23 hcp are needed for game.

[Cascade]

If going by Wayne's results I think that makes bidding game totally clear. I have to admit they suggest it would be a bit more successful than I expected though. BTW, very many 5332 13s would pass a limit raise I assume, as well as 5422 11.

And no doubt that is right. Our judgement should be better than some double dummy analysis so that when we get to game with a 5-3-3-2 it should be better than when the computer suggests it is right with a random 5-3-3-2.

[bhall]

Wayne,

 

In your simulations, was each deal run through Deep Finesse or a similar double-dummy routine? I am not sure, but that may tilt the odds in favor of declarer. Does anyone know? In particular, the odds of developing an extra ♣ trick get better when the ♣ holdings in the defenders' hands are known.

 

Also, a longer simulation with exactly 12 HCP would be of some interest. I would guesstimate that only about 154 (+ or - 12) hands fell into the 5332 12-pt category, out of your 1000-hand run.

[Cascade]

So these days where you bid 4 M with 26 points are long gone.

I recall from doing this sort of analysis previously that with a 9-card trump fit (5-4) only around 23 hcp are needed for game.

Here is a rough and ready (1000 hand simulation double dummy) result for a 5-4 for without either partner having a singleton:

 

hcp odds 10+tricks
21   0.14973262
22   0.268867925
23   0.358208955
24   0.562130178
25   0.694915254
26   0.840707965

 

It looks like my recollection maybe based on vulnerable at IMPs or maybe I allowed singletons in the past.

 

23 hcp is marginal but as I said earlier we would hope to get to some of these using decent judgement and of course avoid the bad ones.

 

24 hcp is plenty at any colour or form of scoring.

 

Here is the code:

 

(spades(north)==5) and
(spades(south)==4) and
(hearts(north)<=4) and
(diamonds(north)<=4) and
(clubs(north)<=4) and
(hearts(north)>=2) and
(diamonds(north)>=2) and
(clubs(north)>=2) and

(hearts(south)<=4) and
(diamonds(south)<=4) and
(clubs(south)<=4) and
(hearts(south)>=2) and
(diamonds(south)>=2) and
(clubs(south)>=2) and

(hcp(north)+hcp(south)>=21) and
(hcp(north)+hcp(south)<=26)

[Cascade]

In your simulations, was each deal run through Deep Finesse or a similar double-dummy routine? I am not sure, but that may tilt the odds in favor of declarer. Does anyone know? In particular, the odds of developing an extra ♣ trick get better when the ♣ holdings in the defenders' hands are known.

 

Also, a longer simulation with exactly 12 HCP would be of some interest. I would guesstimate that only about 154 (+ or - 12) hands fell into the 5332 12-pt category, out of your 1000-hand run.

Yes I use GIB's double dummy engine.

 

Declarer has an advantage over double dummy at lower levels. Double dummy has an advantage over declarer at high-levels. I don't know where the break even point is. There is a thread over on rec.games.bridge on this subject.

 

Actually 12 hcp came up only 141 times out of 1000.

 

Here are 1000 12 hcp hands opposite the hand in the opening post in this thread:

 

tricks frequency
7 [space] [space] [space]7	
8 [space] [space] [space]56	
9 [space] [space] [space]348	
10 [space] [space] 472	
11 [space] [space] 111	
12 [space] [space] 6

[PrecisionL]

I choose 2NT (compressed 'Bergen') = Artificial G.I. or better with 4 trumps. Opener rebids according to losers: 3♣ = 5 or less, 3♦ = 6, 3♥ = 7-8 (if♥ trumps], 7 if ♠ are trumps, 3♠ = 8.

 

If partner bids 3♠, I pass otherwise raise to 4♠.

[655321]

If going by Wayne's results I think that makes bidding game totally clear. I have to admit they suggest it would be a bit more successful than I expected though.

Agree, also agree that the analysis shows game making much more often than I would have thought. So my judgement could be off (always very possible :D ), or the double dummy analysis of 4M might overestimate how often game makes, or both.

 

BTW, very many 5332 13s would pass a limit raise I assume, as well as 5422 11.

Do many people play limit raises like this?

To me, a limit raise asks you to bid game unless your hand is really bad. So while it is probably possible to construct a 5332 13 count where rejecting the invite looks right, to say that very many 13s would pass a limit is not how I play limit raises.