2C overcall results in -800

[HighLow21]

No, there are still the same number of unaccounted-for non-club cards in the mix, it makes no difference. Think about it...

Yes, but that is only one part of the calculation: the number of non-club cards in the mix.

 

The other part of the equation is the expected number of non-diamond cards in partner's hand, and as that number goes up, so does the expected number of clubs. (And so does the expected number of spades and hearts!)

[HighLow21]

These writers try to explain their judgement in layman's terms, and sometimes they get it wrong. Besides, some of them are trying to make a buck or two, because professional bridge doesn't exist.

Mike Lawrence is a 3-time world champion, a member of the Bridge Hall of Fame, and one of the all-time leading masterpoints winners. His book on overcalls is considered one of the top two or three texts in history on the subject, and when it was written in 1980, is was THE authoritative book on the subject.

 

No, it shows he is wrong in one aspect.

I can show you that he is correct with a mathematical proof if you prefer.

[Statto]

Demonstrably false.

 

... length in RHO's suit decreases partner's likely length in that suit and, therefore, on an expectation basis, the number of cards he holds in any other suit.

Please demonstrate then. I would like to see how this flawed analysis holds up. :rolleyes:

[655321]

What am I to search on, and what am I going to find that negates the award-winning writings of a world champion? Besides, simple mathematical logic dictates that he is correct.

 

Well I don't know, I tried searching on +Lawrence + overcall +length, you might be able to think of other search terms.

 

This post gives results of a simulation:

http://www.bridgebase.com/forums/topic/20253-overcalling/page__st__40__p__207460#entry207460

 

The thread itself has posts ranging from intelligent discussion right down to jtfanclub posts.

[hrothgar]

I can show you that he is correct with a mathematical proof if you prefer.

 

I'll see your proof and raise you a Monte Carlo simulation...

 

Changing the number of Diamonds in West's hand from three to four has negligible impact on East's club hold

I didn't bother with a formal significance test, but I doubt that there is a statistically significant relationship

 

 

one_diamond =

 

spades(south) < 5 and

hearts(south) < 5 and

 

(

(diamonds(south) >=4 and diamonds(south) >= clubs(south))

 

or

 

(diamonds(south) == 3 and clubs(south) == 2)

)

 

 

two_clubs =

 

clubs(west) == 5 and

spades(west) <= 4 and

hearts(west) <= 4

 

condition

 

one_diamond and

two_clubs and

diamonds(west) == 3

 

action

 

average(clubs(east))

[Statto]

I'll see your proof and raise you a Monte Carlo simulation...

 

Changing the number of Diamonds in West's hand from three to four has negligible impact on East's club hold

Close, but try no effect. No simulation needed.

[HighLow21]

Please demonstrate then. I would like to see how this flawed analysis holds up. :rolleyes:

Very well, just to keep it simple:

 

Assume opener has 4+ diamonds. I have 4 diamonds and 5 clubs, so ignoring my exact major suit layouts, this accounts for 13 of the cards. 39 are unknown, and partner has 13. Therefore on an expectation basis, partner has a 13/39 (1/3) claim to every unknown card.

 

The unknown cards are:

13 spades

13 hearts

5 diamonds

8 clubs

 

Partner has, on average:

13 x (1/3) = 13/3, or 4.3 spades.

13 x (1/3) = 13/3, or 4.3 hearts.

5 x (1/3) = 5/3, or 1.7 diamonds.

8 x (1/3) = 8/3, or 2.7 clubs.

 

On average, you usually have a club fit (if not at least club tolerance). It gets even better if you add a 5th diamond to your hand or opener's hand. Even if he only has 2 clubs, he may have a diamond ruffing value.

 

Now change your minor shape to 1-5. Now, there are 42 unknown cards, but they are 13 spades, 13 hearts, 8 diamonds, and 8 clubs.

 

Partner has, on average:

13 x (13/42) = 169/42, or 4.0 spades.

13 x (13/42) = 169/42, or 4.0 hearts.

8 x (13/42) = 104/42, or 2.5 diamonds.

8 x (13/42) = 104/42, or 2.5 clubs.

 

Now, partner only rates to have a club fit about half the time. In addition, just as often as not, dummy will be of no value for ruffing anything (in a club contract) and you may get tapped at trick 2. Granted a major suit fit is now more likely, but that offsets the decrease in the chance of a club fit. (Plus we can overcall in a major or takeout double with any hands having a 1-5 minor suit shape.)

 

I realize the difference is small, but the expectation of finding a fit a slightly more often, alongside the increased likelihood that the trumps can be used to ruff diamonds, is an important consideration.

[Statto]

Now change your minor shape to 1-5. Now, there are 42 unknown cards, but they are 13 spades, 13 hearts, 8 diamonds, and 8 clubs.

Um, how can there be 42 unknown cards? Are there 3 cards in my hand that I'm not allowed to look at?

 

Besides the 13 cards in my own hand which I know about, I also know RHO has some diamonds, maybe only 3, but that still limits it to 36 unknown cards in the first case, if we're just interested in distribution.

[jillybean]

LOL

 

I'm still sticking to my requirement for 6 decent clubs before I make a vulnerable 2 level overcall.

[the hog]

You do not need a solid suit to overcall. You just need a suit; this is not even a non vulnerable vs vulnerable overcall.

[neilkaz]

I am a big fan of aggressively overcalling 1♦ with 2♣ but here it goes too far, esp. when Red. The suit is weak enough that you often won't want it lead if LHO declares and while on BBO, weak opps never seem to catch someone for making a gargage bid like this, good opps will catch you and 2♣X making isn't game.

 

I commonly see people with this hand here but no Q♠..ie 10 hcp and the same rotten suit passing and then overcalling into a live unlimited auction. I passed PD's supX a month ago vs one of those opps who likes to pass and then o/c junk and racked up 1400.

 

I need a better suit than this to overcall at the 2 level vulnerable and certainly more tricks if things go badly.

 

.. neilkaz ..

[Statto]

The other part of the equation is the expected number of non-diamond cards in partner's hand, and as that number goes up, so does the expected number of clubs. (And so does the expected number of spades and hearts!)

It's simpler if you just stick to considering the non-♣ cards. Opener has some number of clubs. The remainder will be randomly distributed amongst the other 2 hands yet to bid. Whether your other cards are ♦ or ♥ or ♠ will not affect that.

[Cthulhu D]

This is actually a pretty interesting hand - the traveller would be a masterpiece played at the local club: Anyone else for a first in ekrens opening with east's hand? Or 5 hearts + another (which could be spades). I think I'd pre-empt with south's hand too, but he's lucky in that I think 2S played by south makes. This is going to be a swingy board.

 

Whoever sticks their neck out here is in big trouble.

[HighLow21]

It's simpler if you just stick to considering the non-♣ cards. Opener has some number of clubs. The remainder will be randomly distributed amongst the other 2 hands yet to bid. Whether your other cards are ♦ or ♥ or ♠ will not affect that.

Hmmmmm... you might be right about this and I apologize for being so forceful in my earlier posts. One thing, though, is that whatever club support (if 2+) likely to be more useful if partner is short diamonds.

 

I need to go think about this. :huh:

[MickyB]

My having diamond length reduces partner's expected diamond length.

 

Partner having short diamonds increases his expected club length.

 

However, just because

 

A implies B, and

B implies C

 

does not mean

 

A implies C

[ArtK78]

I read Lawrence's book many years ago. I know he liked to overcall good 4 card majors when he had length in RHO's suit on the theory that it was more likely that partner had a fit if he rated to have less cards in RHO's suit. I never heard this theory extended to overcalling a lower ranking suit at the 2 level.

 

Having said that, the 2♣ overcall on the hand in the OP was an atrocity and deserved the result that it got. It would have received a similar result if it were a six card suit. It should receive a similar result even if it were nonvul or even at favorable vulnerability. But overcalling at the two level vul on QT9xx? To be charitable, it is not my style.

 

There is nothing wrong with overcalling at the two level with a five card suit. But QT9xx is not a suit. It serves virtually no purpose to overcall on this hand. The bid is not lead directional. The bid wasn't made because the overcaller had a good hand and had to take some action. The bid is barely obstructive as it doesn't take up enough of the opponent's bidding space to make the risk worthwhile.

 

So, if there is a lesson to be drawn from the result of this atrocity, it is don't overcall bad suits at the two level.

[Statto]

However, just because A implies B, and B implies C ... does not mean A implies C

Actually it does in formal logic. But "implies" is the wrong word here, as there are no guarantees B-)

[HighLow21]

My having diamond length reduces partner's expected diamond length.

 

Partner having short diamonds increases his expected club length.

 

However, just because

 

A implies B, and

B implies C

 

does not mean

 

A implies C

Ummm, actually, yes it always does, by the transitive property of logic. :-)

[HighLow21]

So, if there is a lesson to be drawn from the result of this atrocity, it is don't overcall bad suits at the two level.

Agreed, unless the rest of the hand is great, the suit is 6 cards, and/or the vulnerability is favorable. (And even then, it's usually marginal.)

[Antrax]

This thread was pretty interesting (thanks, JB!) until it turned into a cesspit of appeals to authority. A pity.

What I don't really understand is how the sixth club matters. Let's say I give you your partner's ♣7 and in return he gets one of your diamonds. It seems that makes things much worse, not better. No?

[BunnyGo]

This thread was pretty interesting (thanks, JB!) until it turned into a cesspit of appeals to authority. A pity.

What I don't really understand is how the sixth club matters. Let's say I give you your partner's ♣7 and in return he gets one of your diamonds. It seems that makes things much worse, not better. No?

 

The original hand was just not going to be a good overcall, even with a small 6th club (move some spade honors to clubs and then you're talking). And hopefully, when we do have 6 cards instead of 5, we didn't take it from our partner ;)

[jillybean]

This thread was pretty interesting (thanks, JB!) until it turned into a cesspit of appeals to authority. A pity.

What I don't really understand is how the sixth club matters. Let's say I give you your partner's ♣7 and in return he gets one of your diamonds. It seems that makes things much worse, not better. No?

I didn't mean to imply a 6th club would make this hand ok, it wouldn't. I have experienced several -800's, and worse by overcalling on crappy 5 card suits. The advice I was given is that I should have a decent 6 card suit or a solid 5 cards and outside values. So far this has worked well for me.

[gwnn]

Ummm, actually, yes it always does, by the transitive property of logic. :-)

Yes I'm afraid MickyB used the wrong word there. It might be better to use "suggest". So, for example, (suppose we live in the 60's):

 

This hair product will make your hair grow.

Those who have long hair are more likely to be addicted to narcotics and oppose the war.

Therefore, my grandpa who is using the hair product is a hippie.

 

Another one:

Those who sit in the sun doing nothing as a kid will likely need glasses after a while.

Those who wear glasses are statistically more intelligent than those who don't.

Therefore, if you sit in the sun doing nothing long enough, you will get smarter and smarter!

[phil_20686]

A pretty good criteria for a 2 level over call is that it should be "better than a wk nt". So if you are 5332 you need extra values to tilt cost benefit in your favour.

 

Hands like xx Axxx x Axxxxx you have to overcall over 1d though as you potentially have a huge hand if partner has a fit and he may well have too many diamonds to get in himself later.

 

Lentgh in RHO's suit is not generally a good thing in an overcall. Despite the intuitive proposition that partner should be more likely to have support it turns out not to be true. I am not going to attempt to "prove it" but we did do simulations on the forum that basically conclusively proved that it made no difference. In fact we showed slightly more than that: It is also (much) more likely the hand is a misfit. To see this recall that with you having 4 cards in RHO's suit, it is considerably more unlikely that the opponents have a large fit. Since we tend to have fits when the opponents have fits and vice verse, the odds of us having a nine card fit are considerably reduced if you have four cards in RHO's suit.

 

Worse, if you have honours in rho's suit, often the total tricks are considerably reduced. Our overcalls have the biggest gain when both us and them have relatively pure fits, with some decent shape, and there are a large number of total tricks. Thus a holding like KQTx in opponents suit is a strong warning sign not to overcall.

[mycroft]

On the other side of the coin (and I certainly wouldn't recommend 2♣ on that hand, so I'm not commenting about that!) - if you don't go for 800 sometimes, you're not bidding enough.

 

One of the things I tell newer players is that "almost certainly, you're too cautious. *Also*, you're bidding on the wrong kinds of hands. So, agree with your partner that for the next month, whenever you think about bidding, you'll do it. Play in good games. You *will* get doubled, you *will* go for big numbers. Smile, laugh, go on to the next hand, and keep doing it. At the end of the month, look at the hands where you went for numbers, and try to work out why they were wrong (and remember, sometimes it's just "unlucky"). Look also at the hands where you *didn't* go for numbers, and try to work out why they worked. *Now*, back off to "saner"; you'll have a much better idea what saner means. And you'll likely feel comfortable with two other things: a) how to play impossible contracts with equanimity, and as best you can (which is a very good skill to learn), and b) you'll have a better idea as defender when the opponents have stepped out and a big number is available."