strong diamond system questions
#1
Posted 2011-April-20, 10:34
infolab.stanford.edu/~qsun/bridge/mellon.pdf
In general, what is the supposed advantage of such strong diamond systems as opposed to strong club systems, or sttandard? It seems like a strong diamond system is nothing more than a standard precision 1C with a nebulous 1D but with 1D and 1C reversed and slightly different point ranges for the various opening bids. am I missing something here?
#2
Posted 2011-April-20, 11:00
newchemist, on 2011-April-20, 10:34, said:
infolab.stanford.edu/~qsun/bridge/mellon.pdf
In general, what is the supposed advantage of such strong diamond systems as opposed to strong club systems, or sttandard? It seems like a strong diamond system is nothing more than a standard precision 1C with a nebulous 1D but with 1D and 1C reversed and slightly different point ranges for the various opening bids. am I missing something here?
Many strong diamond systems such as Magic Diamond are evolutionary offshoots from Strong Pass systems
Strong pass systems got banned. Folks decided that they could play some games with the Fert, 1♣, and 1♦ openings...
Voila... Carrotti --> Magic Diamond
#4
Posted 2011-April-20, 17:42
Anyway, the point is that the precision 1♦ opening is something of an ugly beast. Opener can hold a pretty wide range of hands, from a diamond one-suiter to a minor two suiter to a variety of three-suited patterns to a balanced hand (which might have as few as two diamonds). Finding the best contract opposite this sort of an opening can be difficult, and it creates a number of problems especially in contested auctions. In particular, deciding whether you have a thin game or not over 1♦ (which often depends on how well your hands fit together) can be pretty painful. It's just not clear which cards are likely to be "working" and which are not.
Since you will often want to stop at the part-score level, having an extra step to sort out the "limited catch-all" seems pretty useful. We find that opening these hands 1♣ (instead of 1♦) allows a lot of flexibility; for example we have something close to a relay structure that even functions pretty well in competition. We can also get out in 1M when it's right, whereas playing precision you have to accept that opener will often raise or rebid 1NT (if you responded on garbage). So in principle we get substantially better bidding after the 1♣ "catch-all" than we would after a precision 1♦.
In exchange, we lose a step by opening 1♦ strong instead of 1♣ strong. But people bid over strong 1♣/1♦ a lot (some people play that a 1♠ or greater overcall is almost mandatory). So we don't really lose a step after all. Also, a great percentage of the strong sequences will end in game or beyond, so we potentially have more space to sort out where we belong.
The overall system also has the property that we reach 1NT by a balanced opener quite quickly without disclosing anything significant about responder's hand (in comparison to a traditional precision structure where responder's longest suit is often indicated prior to the 1NT rebid, or where notrump might be wrong-sided). Assuming that we have really good methods over 1NT openings, we can then re-use those methods after 1♦-1♥-1NT, 1♣-1♦-1NT, and so forth. This makes the overall system a lot simpler, and (again assuming our methods over our 1NT are really good) should also be quite effective.
It may also be worth mentioning that Recursive Diamond (and its later version Mellon Diamond) are designed around accurate bidding of light game contracts. Most strong club systems are designed around accurate slam bidding, but we believe there are substantial pickups available from bidding light games with a good fit and (maybe more often given the aggressiveness level of modern IMP players) from avoiding lousy light games when the hands do not mesh well.
a.k.a. Appeal Without Merit
#5
Posted 2011-April-20, 18:15
I also for a few years had a partnership with a strong 1♦ opening, in a canape context. The comments regarding the opponents usually bidding over the strong opening are extremely valid truths. In fact, as a sort of recognition of this, ourn system notes began with contested 1♦ auctions and then later, as an afterthought, discussed the rare occurrence of a pass by LHO.
In that canape structure, 1♣ as the "catch-all" was a slight gainer if the auction was uncontested. Without describing all of the reasons, assume as an example that canape is in use, where 1♣ or 1♦ must be the call to show a longer major with shorter or equal in one (or either) minor. If 1♦-P-1♥(cheapest call)-P-?, Opener cannot bid both majors easily. If 1♣ is the catch-all:
1♣-P-1♦-P-?
1♥ = 5 hearts, 4+♦ (or ♣, or ♣/♦)
1♠ = 5 spades, 4+♦ (or ♣, or ♣/♦)
If both openings face intervention, then nothing matters. If neither faces intervention, the strive for better handling of the intermediates (the real focus of canape structures) over stronger hands (especially in a Roman-ish style of heavy 1♦ openings -- stronger than Precision (15)16+ and more like (17)18+) suggests saving that extra step for the intermediates.
We also bounced back and forth, as the loss of that step for strong hands is somewhat painful and because practice seemed to show that typically unbalanced imntermediate catch-all sequences screamed for intervention more often than the unknown power openings.
-P.J. Painter.
#6
Posted 2011-April-20, 18:30
I managed to come up with a pretty nice transfer-walsh scheme over 1♣, which completely solved all the mess I used to have after the standard precision nebulous 1♦ opener and was quite resilient to overcalls from 4th seat (less against overcalls in 2nd, but still very playable).
I also invented a transfer response scheme to the strong 1♦, which I tweaked to be symmetric with rebids after a negative response, i.e.
1♦ 1♠/1NT/.../2♥ = bal/clubs/.../spades, 8+ hcp
1♦ 1♥ (0-7 hcp)
1♠/1NT/.../2♥ = bal/clubs/.../spades
The 1M openers were also rather funny. I added a permanent Drury-like 2♣ relay so that one could open 1M on just about any 10-count without going over the 2M level when responder has an invite.
The system used a 14-16 NT and weak 2s in all suits.
All this worked rather well, ON PAPER. In practice I never got to play it because pard couldn't be bothered LOL. It's still on version 1.0 ha!
#7
Posted 2011-April-20, 19:21
awm, on 2011-April-20, 17:42, said:
Anyway, the point is that the precision 1♦ opening is something of an ugly beast. Opener can hold a pretty wide range of hands, from a diamond one-suiter to a minor two suiter to a variety of three-suited patterns to a balanced hand (which might have as few as two diamonds). Finding the best contract opposite this sort of an opening can be difficult, and it creates a number of problems especially in contested auctions. In particular, deciding whether you have a thin game or not over 1♦ (which often depends on how well your hands fit together) can be pretty painful. It's just not clear which cards are likely to be "working" and which are not.
Since you will often want to stop at the part-score level, having an extra step to sort out the "limited catch-all" seems pretty useful. We find that opening these hands 1♣ (instead of 1♦) allows a lot of flexibility; for example we have something close to a relay structure that even functions pretty well in competition. We can also get out in 1M when it's right, whereas playing precision you have to accept that opener will often raise or rebid 1NT (if you responded on garbage). So in principle we get substantially better bidding after the 1♣ "catch-all" than we would after a precision 1♦.
In exchange, we lose a step by opening 1♦ strong instead of 1♣ strong. But people bid over strong 1♣/1♦ a lot (some people play that a 1♠ or greater overcall is almost mandatory). So we don't really lose a step after all. Also, a great percentage of the strong sequences will end in game or beyond, so we potentially have more space to sort out where we belong.
The overall system also has the property that we reach 1NT by a balanced opener quite quickly without disclosing anything significant about responder's hand (in comparison to a traditional precision structure where responder's longest suit is often indicated prior to the 1NT rebid, or where notrump might be wrong-sided). Assuming that we have really good methods over 1NT openings, we can then re-use those methods after 1♦-1♥-1NT, 1♣-1♦-1NT, and so forth. This makes the overall system a lot simpler, and (again assuming our methods over our 1NT are really good) should also be quite effective.
It may also be worth mentioning that Recursive Diamond (and its later version Mellon Diamond) are designed around accurate bidding of light game contracts. Most strong club systems are designed around accurate slam bidding, but we believe there are substantial pickups available from bidding light games with a good fit and (maybe more often given the aggressiveness level of modern IMP players) from avoiding lousy light games when the hands do not mesh well.
You make a very good case for the strong diamond (and I've been wondering about it) but you seem to favor the strong club. Why is that? Also, do you think there is a trend toward one or the other approaches in the expert community?
#8
Posted 2011-April-21, 01:08
#9
Posted 2011-April-21, 02:40
Quote
to be much less common than the "let's clobber their club" people like to think. For one
thing, the opponents are vulnerable half the time, and when they are, they cannot really
afford to bid very much unless they actually have something.
The biggest disadvantage for strong 1♦ is that you have less bidding room for strong hands.
#10
Posted 2011-April-21, 03:33
1♣=strong hand, 0-3 spades
1♦=strong hand, 4+ spades
George Carlin
#11
Posted 2011-April-21, 05:14
gwnn, on 2011-April-21, 03:33, said:
1♣=strong hand, 0-3 spades
1♦=strong hand, 4+ spades
Kind of funny that you mention it. I did think about that concept for some time, except for the extreme damage that would do to intermediate hand auctions. The next step in the thought process was of course to somehow combine that concept with a Polish Club concept.
-P.J. Painter.
#12
Posted 2011-April-21, 09:24
straube, on 2011-April-20, 19:21, said:
Trying to say which is better out of strong ♣ or ♦ is a bit like saying which is better out of strong or weak NT. Or 5cM versus 4cM versus 5443 openings. People obviously have strong opinions about these things and mount some good arguments, but convincing sceptics with arguments that are truly encompassing and quantative (i.e. actually proving your point) is another matter altogether.
One thing I would say in reference to someone else's comments in this thread is that it is often said that the 1M openings in strong ♣ are better because they are more limited (the 1♣ opener being a point or two less usually than 1♦ in a strong ♦ system). Well, I would say that this has merit in the uncontested auction. I would also say that, in the contested auction, having the 1M opener as relatively weak at its top end benefits the opponents in that it is a little easier to balance. So I think this argument is relatively weak - given that there are a lot of contested auctions.
Nick
A flippant PS. I'm sure you could devise a decent system around a strong 1♥ opener - with I suppose 1m being a transfer to the majors and 1S being minors. Nobody is going to play it - but I'm sure it would be basically playable.
#13
Posted 2011-April-21, 09:56
George Carlin
#14
Posted 2011-April-21, 10:46
#15
Posted 2011-April-21, 10:53
George Carlin
#16
Posted 2011-April-22, 19:20
As for why we switched to 1♣ strong, there are four main things I can say about that:
(1) Relays are really good in uncontested strong club auctions. Even though some opponents do like to bid, the occasional uncontested relay sequence can be a big winner. We didn't have these after the 1♦ openings (losing a step really costs there); in fact our 1♦ auctions were never really that effective.
(2) A lot of the advantage from the 1♣ catch-all was not actually due to the step saved, but rather due to our allocation of hand types (with minor one suiters opening 2m intermediate). We still do this in our strong club system (we just switched 1♣/1♦ openings and changed the follow-ups, but our other openings didn't change much).
(3) There is an issue with notrump ranges. We really like playing a strong-ish notrump range (like 14-16 or 15-17) and Mellon diamond works better with a different set of ranges (basically our response structure to 1♣ catch-all misses too many major-suit fits if the balanced option is too weak, so we use 1♣=13-15 if balanced and 1NT is either weaker (NV) or stronger (V) than this... but we didn't really like that 1NT range).
(4) Mellon diamond seems to randomly wrong-side an awful lot of contracts (for example after 1♦-1♥/1♠). While this occasionally produces a win too, it seemed to increase variance a lot in situations where that was disadvantageous, and while we can deal with increased variance in situations that we feel are a net win (else we'd be playing 2/1) playing contracts from the weaker hand for no reason is probably a net loss.
a.k.a. Appeal Without Merit
#17
Posted 2011-April-22, 21:41
NickRW, on 2011-April-21, 09:24, said:
A flippant PS. I'm sure you could devise a decent system around a strong 1♥ opener - with I suppose 1m being a transfer to the majors and 1S being minors. Nobody is going to play it - but I'm sure it would be basically playable.
It's been done--Reese's Little Major from the 1960's was structured precisely this way.
#18
Posted 2011-April-23, 06:49
mikestar13, on 2011-April-22, 21:41, said:
Yes. I vaguely recalled after writing what I did, "wasn't that the basis of the little major" - but wasn't sure.
Nick
#19
Posted 2011-April-24, 09:30
1♣: any medium (say 14-16) non-NT hand or minimum (11-13) 63/54 minors, 31 majors
1♦: any strong (17+) or weak NT, F1, all others nf
1♥: min 4-5 no 6c suit
1♠: min 4-5 deny 4♥, no 6c suit
1N: "14-16" no 4cM, 5422, 6322 allowed
2♣: min 6♥ or 7+♥, void
2♦: min 6♠ or 7+♠, void
2♥: min 6♣ 4c suit or 7+♣, void
2♠: min 6♦ 4c suit or 7+♦, void
2N: min 55 minors
3X: min 7+, no void
There's some symmetry when responder relays 1♦ (waiting) over 1♣ and opener simply bids normally, having gotten extra strength off his chest.
As much as I'd like to have nat 2-level openings (read: GCC), I need the extra space to get a weak misfit/inv+ relay for second suit. I've stuck the flat minor 1/2-suiters into 1♣ (rebid NT 'offshape' nf) so we should be able to scramble some fit from the 2M openings. If nat (and probably flat) responder is under much more pressure to shut up and leave in the M misfit. "Minimum" can be stretched out with shape, so the 2+lv openings is basically constructive preempt/intermediate. System emphasis is heavily on bidding exact shape.
This is completely untested and unfinished, I have no idea how and if it works.