druri ATB

[Trinidad]

Drury shows a good raise.

Does it? I agree that's what it should show, but I think many assume it shows at least 10 hcp, or even a good 10.

Well, given that Fluffy posted the problem (and not somebody randomly drawn from "many"), I think it is reasonable to assume that he plays Drury as a good raise. But perhaps he can tell us what 2♣ showed in this particular partnership, Fluffy?

 

Rik

[PhilKing]

Well, given that Fluffy posted the problem (and not somebody randomly drawn from "many"), I think it is reasonable to assume that he plays Drury as a good raise. But perhaps he can tell us what 2♣ showed in this particular partnership, Fluffy?

 

Rik

 

He did - click on the 2♣ bid.

[Trinidad]

I gave West 8-9 HCP and a balanced 4 card raise.

West has already shown 8-9 HCP by rebidding 2♠ (and denied a better hand). He has certainly not denied a three card spade suit. In fact, I would have assumed that he more or less denied four spades with his bidding (but apparently he didn't since he holds four spades).

 

So, your SIM would be realistic if you would give West 8-9 HCP, a balanced hand and 3 spades. Then suddenly 4♠ is not going to look good any more. Even if you allow for all balanced hands with 3 or 4 spades (which overestimates the amount of hands with four spades, since on some West will take a positive action rather than attempt to sign-off in 2♠), 4♠ is going to be poor, since there will still be many more hands with 3 card support than with 4 card support.

 

In addition, a West with four card support will normally accept if East still tries for game after West's 2♠ suggestion to sign-off. (The only reason why West didn't accept the invitation in this case was that he had the worst possible holding opposite the club shortness.) That means that game will normally be reached when East makes a trial bid. There is no need at all for East to blast game. An invitation will get him to the good games and will keep him out of the games opposite the true minimume Drury hands

 

Rik

[awm]

It seems like it'd be good to have a way to bid game opposite a four-card limit raise but not necessarily opposite a three-card raise. You have all the space you could ever need here; maybe:

 

2♦ = please bid 2♠ with a three-card raise, but something else with four

 

And then if still interested opener can make a game try or something over the 2♠ rebid. This hand seems like an obvious game bid opposite a four-card limit raise, but if you move one of responder's spades into diamonds (or clubs, or hearts) then game is not so good.

[Trinidad]

He did - click on the 2♣ bid.

They say that egg is good for your hair. I just wiped it off my face. I will see what it does for my hair.

 

Rik

[rhm]

West has already shown 8-9 HCP by rebidding 2♠ (and denied a better hand). He has certainly not denied a three card spade suit. In fact, I would have assumed that he more or less denied four spades with his bidding (but apparently he didn't since he holds four spades).

 

So, your SIM would be realistic if you would give West 8-9 HCP, a balanced hand and 3 spades. Then suddenly 4♠ is not going to look good any more. Even if you allow for all balanced hands with 3 or 4 spades (which overestimates the amount of hands with four spades, since on some West will take a positive action rather than attempt to sign-off in 2♠), 4♠ is going to be poor, since there will still be many more hands with 3 card support than with 4 card support.

 

In addition, a West with four card support will normally accept if East still tries for game after West's 2♠ suggestion to sign-off. (The only reason why West didn't accept the invitation in this case was that he had the worst possible holding opposite the club shortness.) That means that game will normally be reached when East makes a trial bid. There is no need at all for East to blast game. An invitation will get him to the good games and will keep him out of the games opposite the true minimume Drury hands

 

Rik

You seem to have a confused understanding what the Drury convention is about.

Drury is not a substitute bid for a single raise.

It is a game invitation by a passed based on a fit in opener's major.

With 8-9 HCP, a balanced hand and 3 card support you simply raise 1♠ to 2♠.

There is nothing mystical about 4 card raises.

All else being equal a hand with 4 card support is of course stronger than with 3 card support.

So an equivalent 3 card raise needs some compensating strength to qualify for Drury.

If a balanced 4 card raise with 8-9 HCP is a proper hand for Drury then a balanced 3 card raise with the same HCP is not one.

A balanced 3 card raise with 8-9 HCP is roughly equivalent in strength with a balanced 4 card raise and 6-7 HCP and that I would consider too weak for employing Drury.

I understand that this is hard to comprehend for someone, who never evaluates a hand except by point count.

If you employ Drury with a balanced hand and three card support you should be maximum in HCP and not be 3334.

 

I ran another simulation with these specifications:

Balanced three card raise, 10-11 HCP, not 3♠334 (I would respond 1NT with that, natural and not forcing by a passed hand).

 

Result: With 1000 random deals: 4♠ made on 752 deals out of 1000, almost exactly the same as on my previous simulation.

 

Rainer Herrmann

[KurtGodel]

You seem to have a confused understanding what the drury convention is about.

Drury is not a substitute bid for a single raise.

It is a game invitation by a passed based on a fit in opener's major.

With 8-9 HCP, a balanced hand and 3 card support you simply raise 1♠ to 2♠.

There is nothing mystical about 4 card raises.

All else being equal a hand with 4 card support is of course stronger than with 3 card support.

So an equivalent 3 card raise needs some compensating strength to qualify for Drury.

If you employ Drury with a balanced hand and three card support you should be maximum in HCP and not be 3334.

 

I ran another simulation with these specifications:

Balanced three card raise, 10-11 HCP, not 3♠334 (I would respond 1NT with that, natural and not forcing by a passed hand).

 

Result: With 1000 random deals: 4♠ made on 752 deals out of 1000, almost exactly the same as on my previous simulation.

 

Rainer Herrmann

 

Fair enough, I'm surprised but can't deny the results. Especially surprised by how good 4S is.

[aguahombre]

Score -100 when dummy comes down:

♠K43

♥Q53

♦T62

♣KJ62

 

Rik

If dummy comes down with that, we will have a discussion about the difference between an invitational raise and a simple 2S raise....and whether we agreed to perpetrate Drury with mere constructive values.

 

ATB is easy on this one. The guy who didn't bid game when partner showed an invite accepts the blame when game makes --- and the credit if it doesn't make.

[Trinidad]

You seem to have a confused understanding what the Drury convention is about.

Drury is not a substitute bid for a single raise.

It is a game invitation by a passed based on a fit in opener's major.

Setting aside what the OP was playing, it is clear that you have an outdated understanding of what the Drury convention is about. In addition, you are misrepresenting what I wrote. I don't know whether it was

• merely inaccurate
  
• silly
  
• poor debating style
  

on your part.

 

But don't worry, I have been silly in this thread myself, so it can happen to the best of us. ;)

 

I never wrote that Drury was a substitute for a single raise. I wrote that Drury was a good raise. To make it 100% clear: That includes the game invitations and the maximum single raises. It does not include the minimum single raises, they just bid 2M.

 

There is no doubt that Drury was originally invented to check whether opener had a true opening or was joking in third (or fourth) seat. But modern Drury has evolved from that. It started long time ago when Drury started to promise support. This was part of "Reverse Drury". Then came the useful space argument (an important argument since Jeff Rubens. Since then: my parents bought their first color TV, Lech Walesa founded Solidarnosz, the wall came down, the cold war ended. I assume you can come up with more recent events, e.g. the Netherlands winning the Bermuda Bowl... twice.):

 

Drury is a 2♣ bid. A single raise is a 2M bid. That means that after Drury, you have room to explore between 2♣ and 2M. This room you don't have between 2M and 2M. (Now, who said that math was complicated?) That, in turn, means that the range for 2♣ can be wider than the range for 2M.

 

When players then realized that:

• 
   
  
• the range for 2♣ should be wider than the range for 2M
  
• they don't want to have any raises in the 1NT response (to make 1NT truely semi-forcing)
  
• modern openings in first and second seat are more aggressive, reducing the need for crazy openers in third and fourth seat.
  

it became clear that Drury should be less used as a checkback for crazy openings but instead should be used to divide all hands with support and 4-11 HCPs over 2M and 2♣. The 2M range takes care of the lower part: 4-7. Drury takes care of the 8-11, and the hands that re-evaluate to more than 11 after the opening.

 

As I said, this is not a development of recent years. Top players have been playing this version of Drury for quite a while now (if they play Drury).

 

To complete the whole Drury scheme and to sketch the mouse trap as it is most often built for you:

Pass-1M

2♣: 8-11, (at least) three card support, or a hand that re-evaluates to something better

    2♦: a decent minimum opening, occasionally a good hand

        2M: 8-9(10)

        All other bids: game tries according to the taste of the partnership*, i.e. decent invitations.

    2M: a (sub)minimum opening. Responder will pass, unless he has a hand that re-evaluated to more than an opening

        All bids: game tries according to the taste of the partnership, a hand that re-evaluated to more than an opening

    3M: according to partnership taste: slam try or game try

    All other bids: game tries according to the taste of the partnership

*short suit, long suit, help suit, Romex, whatever.

 

I hope this has resolved your confusion and catapulted you into the 21st century.

 

You're welcome.

 

Rik

[D4rw1n]

My first instinct was to look at the hand like this:

 

♠KJxx

♥xx

♦Txx

♣xxxx

 

opposite

 

♠A97xx

♥ATx

♦AJ98

♣x

 

and being unlucky I don't play weak major raises today, I open 1♠ and partner raises to 2♠ (again, a minimum hand)

Is opener supposed to bid game now, thinking he's going to make it? (responder doesn't raise invite, after all, only 4 points)

This lead me to think that what responder should value is the fourth trump and three ruffable clubs (assuming trumps break)

 

Second instinct was that I hope you didn't run your simulations double dummy, since then a single finesse in diamonds is enough unless they are 3-3 with honours split. And the spades can be played for no losers unless QTx behind or 4-0, which gives us 3♦+1♥+6♠

 

And before I get any responses about "but the simulation wasn't just for defenders hands", my point is that trump suits like QTx vs A97xx are, double dummy, one losers at most (mere mortals can lose 2 tricks) and diamonds like Kxx vs AJ98 or Qxxx vs AJ98 have only one loser (if the QTx or king is offside)

 

- J

[cherdano]

I am not sure what Rik and Rainer are arguing about. There are many partnerships who play drury as promising approximately a limit raise - 10+ hcp with typical 3-card support shape. There are many who play it as a good raise - a good 8 hcp with 3-card support would be enough.

My preference is for the latter agreement - you are at the level of 2C with a known major suit fit, certainly we can handle a range of 8-11 hcp (and we can live with the fact that a hand with 4 trumps and 11 hcp has to force to the 3-level). But obviously playing it as a limit raise also has advantages - the best drury auction is obviously 1M-2C-4M (no information leakage, no additional chances for opponents to double), and it will happen more often if drury promises a limit raise.

[D4rw1n]

Their disagreement would seem to come from the fact that Rainer doesn't want to play at the three level, (if you need to go that high you might as well take a shot at 4 because that you can do without giving information) but Rik finds it acceptable.

 

If we start worrying about information leakage, we should get the advantages to all our 1M - 2"M" sequences, because it's not worth it to transfer the problem away from drury just so that it exists somewhere else. One way to do that is

 

Pass - 1M - ?

2♣ = 8-9 points, 3+SUPP

2♦ = 10-11 points, 3+SUPP

2M = 6-7 points, 3+SUPP

1NT = can have 4-5 points, 3SUPP

3M = 4-5 points, 4SUPP

 

Quick simulation showed that if you have 10+ points in passed hand and 4cSUPP, for 4M not to be worth the shot partner needs to have worse hand than you. (or just 4 cards for opening) Transfering this information to other sequences we have the rebids

 

Pass - 1M - 2♦ - 2M - 4M = I had 4+SUPP

Pass - 1M - 2♣ - 2♦ = Do you have 4+SUPP

Pass - 1M - 2M - 4M = here I'm gonna need a little extra strength anyway because dummy has less entries, so no need to ask how long the SUPP is

 

- J

[rhm]

I am not sure what Rik and Rainer are arguing about. There are many partnerships who play drury as promising approximately a limit raise - 10+ hcp with typical 3-card support shape. There are many who play it as a good raise - a good 8 hcp with 3-card support would be enough.

My preference is for the latter agreement - you are at the level of 2C with a known major suit fit, certainly we can handle a range of 8-11 hcp (and we can live with the fact that a hand with 4 trumps and 11 hcp has to force to the 3-level). But obviously playing it as a limit raise also has advantages - the best drury auction is obviously 1M-2C-4M (no information leakage, no additional chances for opponents to double), and it will happen more often if drury promises a limit raise.

As I said before I agree that the lower limit of a Drury raise is slightly less than a limit raise, but it should be really a maximum single raise.

To explain the difference lets do this by example:

 

Take the West hand

 

[hv=pc=n&w=skj43hq5dt62ckj62]133|100[/hv]

 

For me this is a limit raise, but only just. Change a black jack for the ten and I would not consider it a limit raise. I would still employ Drury

 

[hv=pc=n&w=skj4hq5dt632ckj62]133|100[/hv]

 

Not worth a limit raise, but I would still employ Drury

 

[hv=pc=n&w=skj4h53dt632ckj62]133|100[/hv]

 

Now the hand is a raise to 2♠, but too weak for Drury.

 

At least 95% of all balanced hands with 8 HCP with three card support are not strong enough for Drury.

 

Rainer Herrmann

[wanoff]

As I said before I agree that the lower limit of a Drury raise is slightly less than a limit raise, but it should be really a maximum single raise.

To explain the difference lets do this by example:

 

Take the West hand

 

[hv=pc=n&w=skj43hq5dt62ckj62]133|100[/hv]

 

For me this is a limit raise, but only just. Change a black jack for the ten and I would not consider it a limit raise. I would still employ Drury

 

[hv=pc=n&w=skj4hq5dt632ckj62]133|100[/hv]

 

Not worth a limit raise, but I would still employ Drury

 

[hv=pc=n&w=skj4h53dt632ckj62]133|100[/hv]

 

Now the hand is a raise to 2♠, but too weak for Drury.

 

At least 95% of all balanced hands with 8 HCP with three card support are not strong enough for Drury.

 

Rainer Herrmann

 

That's a slight change since you originally said you'd bid game over Drury.

Isn't this simply a partnership agreement queastion where East bids game over a 4cd limit raise or a fitting 10HCP 3cd raise.

[rhm]

That's a slight change since you originally said you'd bid game over Drury.

Isn't this simply a partnership agreement queastion where East bids game over a 4cd limit raise or a fitting 10HCP 3cd raise.

Why?

I would bid game with the East hand if West employs Drury and I could not care less, which hand West held, and my simulations seem to indicate that this is the correct action.

I believe missing game here is not caused by a partnership agreement question, but by bad evaluation, probably caused by over-reliance on point count.

By the way opener explained the 2♣ bid as "drury fit 10+ with support"

East did not appreciate the strength of his hand after receiving a Drury raise.

 

Rainer Herrmann

[gordontd]

Score -100 when dummy comes down:

♠K43

♥Q53

♦T62

♣KJ62

 

Rik

That's not a 2C bid.

[gordontd]

2♦ is not a natural game try. It is a generic game try, or, you could say that it confirms a "real" opening as opposed to a 3rd seat subminimal opening.

I don't think that's how it's commonly played any more, and the lack of alert in the diagram suggests to me that it's not how this pair play it.

[wanoff]

Why?

I would bid game with the East hand if West employs Drury and I could not care less, which hand West held, and my simulations seem to indicate that this is the correct action.

I believe missing game here is not caused by a partnership agreement question, but by bad evaluation, probably caused by over-reliance on point count.

By the way opener explained the 2♣ bid as "drury fit 10+ with support"

East did not appreciate the strength of his hand after receiving a Drury raise.

 

Rainer Herrmann

 

The middle hand does not even look like 50% let alone 75%.

If West thought Drury showed 4 and East thought it showed 3+, then indeed it would be an agreement problem. If not, I agree it's bad evaluation.

[rhm]

The middle hand does not even look like 50% let alone 75%.

If West thought Drury showed 4 and East thought it showed 3+, then indeed it would be an agreement problem. If not, I agree it's bad evaluation.

Agreed, but this is a misunderstanding.

Bidding games is not done because it is a sure thing but because it is the percentage action, maximizing your expected score over a distribution of possible hands.

When a simulation shows that out 1000 random deals 750 make game and 250 not, some of those 250 partner hands may have no chance making game whatsoever and some may go down due to bad splits.

The middle one is a dead minimum Drury and it has a lots of wasted values in clubs in addition.

Still game is not without chances, in particular single dummy, if the East hand is not revealed in the bidding.

 

Rainer Herrmann

[gnasher]

But obviously playing it as a limit raise also has advantages - the best drury auction is obviously 1M-2C-4M (no information leakage, no additional chances for opponents to double), and it will happen more often if drury promises a limit raise.

Is that significantly better than 1M-2♣;2♦-4M, which is the auction you should have when playing wide-range Drury?

[cherdano]

Is that significantly better than 1M-2♣;2♦-4M, which is the auction you should have when playing wide-range Drury?

It's not a big difference of course, but it still tells opponents that declarer wasn't strong enough to bid 4M directly.

As another example, surely 1M-2M is a better auction than 1M-2C-2D-2M?

[Finch]

I am not sure what Rik and Rainer are arguing about. There are many partnerships who play drury as promising approximately a limit raise - 10+ hcp with typical 3-card support shape. There are many who play it as a good raise - a good 8 hcp with 3-card support would be enough.

My preference is for the latter agreement - you are at the level of 2C with a known major suit fit, certainly we can handle a range of 8-11 hcp (and we can live with the fact that a hand with 4 trumps and 11 hcp has to force to the 3-level). But obviously playing it as a limit raise also has advantages - the best drury auction is obviously 1M-2C-4M (no information leakage, no additional chances for opponents to double), and it will happen more often if drury promises a limit raise.

 

The reason not to play it as 10+ hcp is if you open virtually all 11+ HCP hands. Then it will rarely come up.

[gwnn]

As another example, surely 1M-2M is a better auction than 1M-2C-2D-2M?

How about

1M-2M; pass (~4-7 hcp)

 

versus

1M-2M; 3x-3M; pass?

 

Now we're a level higher with the defence knowing something more about opener's hand.