[pescetom]

Quote

#
# How many clubs does a "natural" 2+ card 1 club opening have ?
#
produce 100000

SC1219 = hcp(south)>=12 and hcp(south)<=19
SC2 = shape(south, 4432)
SC3 = shape(south, 4333) or shape(south, 3433) or shape(south, 4423)
SC4 = (shape(south, 3334) or shape(south, 4414))
SC5plus = clubs(south)>=5 and clubs(south)>diamonds(south) and clubs(south)>hearts(south) and clubs(south)>spades(south)
S1NT = hcp(south)>=15 and hcp(south)<=17 and shape(south, any 4432 + any 4333 + any 5332)
action frequency "of number of clubs" (clubs(south),0,10)
condition SC1219 and (SC2 or SC3 or SC4 or SC5plus) and (not S1NT)

I hacked this script to demonstrate to a partner why it is a mistake to assume that a "natural" 2+ card 1♣ opening will likely be short.
It assumes strong NT, that 1♣ promises 2+ cards (as we will open 1♣ with exactly 4=4=3=2) and that 1♦ promises 4+ cards (and does not deny 4 card clubs), which is the dominant style around here.
If anyone can spot a bug I will be grateful.

[DavidKok]

It looks like a good start to me, but you seem to have missed a number of balanced shapes with 4 clubs (and 4 of a major).
Many people will also include a number of unbalanced 11-counts (or even weaker hands) as well as 20+-counts in their 1♣ opening.

In my Dutch Doubleton scripts I've used similar but slightly different definitions for the hand types in a 1♣ opening, but I include many more hands with exactly 2 clubs. If it's really only 4=4=3=2 hands out of NT range that's for all intents and purposes playing a 3(+) system.

[pescetom]

DavidKok, on 2023-May-16, 11:29, said:

you seem to have missed a number of balanced shapes with 4 clubs (and 4 of a major).

Quite probable, but which? Thanks.

DavidKok, on 2023-May-16, 11:29, said:

Many people will also include a number of unbalanced 11-counts (or even weaker hands) as well as 20+-counts in their 1♣ opening.
In my Dutch Doubleton scripts I've used similar but slightly different definitions for the hand types in a 1♣ opening, but I include many more hands with exactly 2 clubs.

Yes I wasn't bothering too much about shapely 11 counts or freak exclusions like 3NT Gambling. Most 20+ are going to be handled with some other strong opening and again that is very low frequency. I'm more concerned if I missed some balanced shape as above.

DavidKok, on 2023-May-16, 11:29, said:

If it's really only 4=4=3=2 hands out of NT range that's for all intents and purposes playing a 3(+) system.

Yes that is the only real difference compared to a 3(+) system that will always open 1♦ with minors 4-4 and 1♣ with minors 3-3, but I beg to differ that for all intents and purposes the result is the same. There is a significant improvement in 1♦ auctions if responder is certain Opener has 4+ cards. Responder also has clearer inferences in later rounds of a 1♣ auction. I think the only real downside is that it is less nebulous to opponents, particularly if 'Better minor' agreements are not fully disclosed.

[smerriman]

pescetom, on 2023-May-16, 12:47, said:

There is a significant improvement in 1♦ auctions if responder is certain Opener has 4+ cards.

I've never found this to be the case. If you open 1♦ with 4432, your 1♦ opener has 4 cards 96% of the time - responder just acts as if they're certain anyway. And the vast majority of those 4% it doesn't matter, since you're usually in a major fit.

PS - you're missing the (432)4 shapes.

[mycroft]

I have in my notes somewhere (I think it's in "conventions you shouldn't play") that 4=4=3=2s outside of strong NT range are about 3% of all hands that open whatever (might be just 3% of all hands that open). It's "long-established" knowledge, but I have no idea where the proof ever was.

But as a result, I tell people that they should be bidding as if 1♦ promises 4 and 1♣ promises 3, no matter where they put those hands. Panicking over 4=4=3=2 causes more damage than any short fits you end up having to play.

Now, I don't like (but won't complain too much about) 5=5=4=2 systems, as long as they bid as if opener has 3 guaranteed. It's just that (as OP is finding) way too many people don't. I am quite comfortable with "clubs or balanced" 2+ clubs (but the frequency is much higher, and you do have to worry somewhat about it).

[pescetom]

smerriman, on 2023-May-16, 13:41, said:

I've never found this to be the case. If you open 1♦ with 4432, your 1♦ opener has 4 cards 96% of the time - responder just acts as if they're certain anyway. And the vast majority of those 4% it doesn't matter, since you're usually in a major fit.

I remember it as closer to 6%. All the same I agree that it would make little difference if you do actually agree to act as if it promised 4 cards, including fixing trumps with a 4 card raise, but that doesn't seem to be the way most play.

Be that as it may, to each his own poison. My point to partner was that with 2+ clubs natural it is foolish to be paralyzed by fear of 2 cards as responder, or gleefully make a natural overcall as opponent. Run the script to see the probability of holding 5 or 6 cards.

[smerriman]

I ran a quick sim and the chance of a 1♦ opener being 3 cards was 4.6%.

Adjusting your sim for the missing hands, the chance of a 1♣ opener being 2 cards was also 4.6% So the argument to get them to switch systems is almost as convincing as the argument to not switch.. the point in both cases is that you can just bid as if it won't happen.

[pescetom]

smerriman, on 2023-May-16, 14:18, said:

I ran a quick sim and the chance of a 1♦ opener being 3 cards was 4.6%.

Adjusting your sim for the missing hands, the chance of a 1♣ opener being 2 cards was also 4.6% So the argument to get them to switch systems is almost as convincing as the argument to not switch.. the point in both cases is that you can just bid as if it won't happen.

Thanks. There is also of course the factor that the more you yourself have, the more likely it is that they are indeed short. Which is an issue for the 3+ school whereas for 2+ it almost becomes an opportunity, assuming you are aware of the odds.

What bemuses me is that opposite a natural 2+ 1♣ opening, many opponents playing 3+ and even some playing 2+ themselves are happy to give up a 2♣ overcall as Michaels and even to play it as natural.

[DavidKok]

I play something that might be best described as 5552, 1♦ only being 4 on 1=4=4=4 or 4=4=4=1 shape. I think it is sensible to play a 2♣ overcall of my 1♣ as natural - it is very similar to a standard 1♦-2♣ auction, with all the usual problems.

[pilowsky]

One reason people (well me...) are anxious about 2+ club systems is that partner will pass and leave you in the s**t for a big number.
So an additional factor is whether or not partner is likely to have a hand that is unable to make any kind of bid.

[smerriman]

You're probably going to be equally unhappy about a 3 card 1♦ opening being passed out - though that's very rare - but undoubled at the 1 level when the opponents probably have the majority of the points isn't even often a bad result.

But if anything, I think the opposite logic would apply there - over 1♦ it's exceedingly rare that nobody is going to bid a major; but over a 2 card 1♣ opener it's even more exceedingly rare than nobody can bid 1♦ either, when you only have 3 of them.. and if partner can't, they probably have support.. so it seems a shorter 1♣ is actually safer if that were the concern.

[Cascade]

pescetom, on 2023-May-16, 11:11, said:

I hacked this script to demonstrate to a partner why it is a mistake to assume that a "natural" 2+ card 1♣ opening will likely be short.
It assumes strong NT, that 1♣ promises 2+ cards (as we will open 1♣ with exactly 4=4=3=2) and that 1♦ promises 4+ cards (and does not deny 4 card clubs), which is the dominant style around here.
If anyone can spot a bug I will be grateful.

The scripts I use include:

1. For balanced hands shape(south, any 4333 + any 4432 + any 5332) and spades(south)<5 and hearts(south)<4 and diamonds(south)<4 to get all balanced hands with 2-5 clubs and not 4-4 in the minors.

2. I also usually include all 11 hcp hands that are not balanced (or perhaps not a 4441) and 10 hcp hands with six clubs. A hack to do that is to put hcp(south)>=10 and hcp(south)+clubs(south)>=16 and clubs(south)>=5. That will get all 11 hcp with five clubs and 10 hcp only if there are six clubs. (We need another condition for the 4441s.)

Here is my full script

oneclub = hcp(north)>=10 and 
not (hcp(north)>=20 and shape(north,any 4333 + any 4432 + any 5332 + any 5422)) and 

(
(
((hcp(north)>=12 and hcp(north)<=14) or
(hcp(north)>=18 and hcp(north)<=19)) and
shape(north, any 4333 + any 4432 + any 5332) and 
diamonds(north)<4 and spades(north)<5 and hearts(north)<5
) or

(
hcp(north)>=11 and shape(north, 4414)
) or

(
hcp(north)>=10 and 
hcp(north)+clubs(north)>=16 and
clubs(north)>=5 and
clubs(north)>diamonds(north) and 
clubs(north)>hearts(north) and 
clubs(north)>spades(north) and 
not (hcp(north)>=15 and hcp(north)<=17 and shape(north, any 5332))
)
)

[thepossum]

I have an observation followed some possibly interesting preliminary results

Nobody has discussed the impact of seat position which I find strange

On initial analysis (and missing lots of other conditions) it you seated 2nd after a pass the chance of 2 clubs goes up

Of course I haven't incorporated the obvious condition discussed above about what you have in your hand and whether it matters how many clubs your partner has
- which is obviously a more systemic and game theoretic consideration which I will leave to the experts
EDIT I am trying to incorporate preempts and cover all different seat positions

As a nobody with half a brain isn't the point that you have some way to express your hand strength and shape vaguely accurately

I am sitting on approximately 5% replicating the above script - does that sound right - and does it matter and is it worth my time?
Lets simplify and assume opening bid is seat one. Then you get 1.25%. Am I right? - plus if it is seat two plus if it seat 3 plus if it seat 4 and I am looking at my own hand minus the chance that everyone has passed and I have all the cards too etc

[DavidKok]

I don't think the effect is that large. In my opening simulations I include second, third and fourth hand assumptions by conditioning the relevant other hands to not have an opening bid (constructive or preemptive). The effects of this on shape are very minor - most of the hands that are worth an opening in hearts are worth an opening in clubs if we swap the suits around. People upgrade a bit more with both majors (but not a lot more than people upgrade with a major+minor, or both minors), giving a slight distortion. People also have weak two openings in the majors (usually) but not in the minors (usually). Lastly conditioning on certain HCP ranges has an impact on the ratio of balanced versus unbalanced hands, but not on (say) primary clubs unbalanced hands versus primary spades unbalanced hands. The effect should not be as big as you're finding.

[thepossum]

ok thanks


I am only here trying to replicate

Maybe the earlier seats depend on who your opps are

[helene_t]

pilowsky, on 2023-May-16, 18:48, said:

One reason people (well me...) are anxious about 2+ club systems is that partner will pass and leave you in the s**t for a big number.
So an additional factor is whether or not partner is likely to have a hand that is unable to make any kind of bid.

On the plus side, if you play weak jump shifts, partner won't have a 6-card suit when they pass your 1♣ opening. And opponents apparently didn't have a suit which they found worthwhile to bid at the 1-level (although, to be fair, opps can have a biddable club suit, especially if they play that 1♣-(2♣) is michaels). Also, if they double 1♣, you can still scramble to any suit at the 1-level.

It is also relatively safe for partner to bid 1♦ on a subminimal hand although I am not really sure if it's a good idea to do so if you play vanilla so that opener will rebid 2NT with any 18-19 balanced hand.

So while I don't have any data tp back it up, my feeling is that 1♦ on a 3-card suit is more risky that 1♣ on a doubleton. And 1M on a 4-card suit would make me more nervous, also.

Safest option is, presumably, to play something like Dutch Doubleton (or some variants of T-Walsh) that allow you to play 1NT when opener has 18-20 points.

[mikeh]

I play in two partnerships in which we open 1C on 2+. In one, we have 2+ only with 4432. In the other, we are even more extreme than David. 1D promises an unbalanced hand, so even 3352 is 1C if out of our notrump range. Thus, for us, 1C is frequently 2.

As for being passed in 1C….as is increasingly common we very rarely pass 1C. It’s not forcing. The new ACBL CC asks for the minimum point count shown by a response to 1C and I have been writing ‘4’ but probably shouldn’t….it’d be more accurate to say ‘3’…..however, we will pass some 3 counts and, indeed, some 4 counts.

Transfer Walsh helps make this work, especially a version that has 1C 1R 1N show a big hand…for us, 17-19 (good 19’s open 2N), so we don’t have that dreaded 1C 1M 2N situation when responder has 3-4 hcp. For us, the 2N rebid shows a four card raise with 17+ hcp, and responder can transfer to get out at the 3-level.

Also, nv our 1N is 10-13 so 1C 1R 2M (responder’s) shows 14-16, thus we don’t need to jump to 3M, nv, on 15-16. Since we’re nv half the time, this also reduces the risk of getting too high when responder is weak.

It seems to work well, in that we’ve not gone for -200 or greater due to languishing in 1C or getting too high after a light response (which is not claiming we’ve never gone down, just that we rarely get to a silly level).

The unbalanced diamond seems to work well, primarily because we play a lot of gadgets which are based on the known unbalanced nature of the hand. We don’t, for example, need 1D 1M 2N for the balanced 17-19/18-19, so we use it for 6+ diamonds 17+ hcp, fewer than 3 cards in the major unless very strong….2N is, in essence if not theory, forcing to 3D, and we have more gadgets over it. 1D 1M 3D shows 6+ with 3M, non-forcing but strong.

And so on.

As for how we bid over a 2+ 1C….by the opps….any club overcall is natural and we use 2D as Michaels.

Yes, for many players the 2+ is low frequency but our experience has been that it’s very difficult to catch us speeding in clubs even when opener has real length. It’s difficult to pass a negative double for penalties when sitting under declarer and having no idea how many clubs responder has.

[pescetom]

smerriman, on 2023-May-16, 13:41, said:

PS - you're missing the (432)4 shapes.

Oops Although not (32)44 shapes of course.
So much for making code readable and efficient before publishing... my first definition did not have this omission so I've reverted to that one.
Of course that might have problems too.

thepossum, on 2023-May-17, 04:19, said:

I have an observation followed some possibly interesting preliminary results

Nobody has discussed the impact of seat position which I find strange

On initial analysis (and missing lots of other conditions) it you seated 2nd after a pass the chance of 2 clubs goes up

Of course I haven't incorporated the obvious condition discussed above about what you have in your hand and whether it matters how many clubs your partner has
- which is obviously a more systemic and game theoretic consideration which I will leave to the experts
EDIT I am trying to incorporate preempts and cover all different seat positions

As a nobody with half a brain isn't the point that you have some way to express your hand strength and shape vaguely accurately

I am sitting on approximately 5% replicating the above script - does that sound right - and does it matter and is it worth my time?
Lets simplify and assume opening bid is seat one. Then you get 1.25%. Am I right? - plus if it is seat two plus if it seat 3 plus if it seat 4 and I am looking at my own hand minus the chance that everyone has passed and I have all the cards too etc

I added a crude possibility to switch first/second seat and it looks as it makes almost no difference, as Davidkok suggested.
Second seat gives a tiny decrease of short holdings and a tiny increase of longer holdings.

#
# How many clubs does a natural 2+ card 1 club opening have ?
#
produce 100000

SC1219 = hcp(south)>=12 and hcp(south)<=19
SC4minus = (shape(south, any 4414) or shape(south, any 4432) or shape(south, any 4333)) and diamonds(south)<4
SC5plus = clubs(south)>=5 and clubs(south)>diamonds(south) and clubs(south)>hearts(south) and clubs(south)>spades(south)
S1NT = hcp(south)>=15 and hcp(south)<=17 and shape(south, any 4432 + any 4333 + any 5332)
Epass = hcp(east)<6 or (hcp(east)<11 and (hearts(east)<6 and spades(east)<6))

action frequency "of number of clubs" (clubs(south),0,10)
#------------ Uncomment one of the two conditions to set first/second seat
condition SC1219 and (SC4minus or SC5plus) and (not S1NT)
#condition Epass and SC1219 and (SC4minus or SC5plus) and (not S1NT)

[thepossum]

My attempts at making any sense of this issue are still at the earlier stages but I have worked out the chances of opening in each seat and trying to rule out a balancing NT - which strangely has a big impact - but since the chance of being in balancing seat is fairly small (my fudge factors came up with around 5% chance of balancing) who knows - probably almost non-existent with the way many people bid these days - like ruling out a multi (I just wanted to bid anything)

By my reckoning the average number of clubs in your hand drops from seat 1 to 3 then increases in seat 4 - with similar patterns in chance of 2 clubs (not much diff in 1 to 3 but definitely in seat 4)

EDIT QUick clarification - I am talking the chance of opening 1 club in each seat