[P_Marlowe]

Hi,

my partner suggested to switch from Lavianthal
to direct discards.
Could anybody comment on the plusses and minuses
of using direct discards.
If possible also comment on differences between
suit and nt contracts.

Thanks in advance for your ccmmens.

With kind regards
Marlowe

[Apollo81]

I assume you mean something like "high encourages" or "low encourages" by "direct discard." I will refer to "high encourages" as "standard" in the rest of the post.

Let's say that you can only afford to pitch from one suit (happens more often that inexperienced players realize). Even with no signals partner can nearly always narrow down his choice to 2 different suits using bridge logic. If the suit you're discarding from is one of them, high encourages that suit and low encourages the other logical suit. If the suit pitched is not a logical choice of switch, then your discard is suit preference anyway since "encouraging" isn't logical. Either way you can always communicate what you want.

Methods like Lavinthal and Odd Even seem to me like a crutch for inexperienced players that haven't yet developed the bridge logic to eliminate suits as possible switches. Odd Even may gain once in a blue moon when you have the right card available to communicate what you want AND partner wouldnt read your card playing standard AND partner would make a mistake without your signal, but a lot more often you are just forced to communicate something different than what you want or just part with a card that you need to keep. Lavinthal (suit preference discards) seems strictly inferior to standard since the method determines what suit (the one discarded) isn't a logical choice as opposed to your situational judgment.

[bid_em_up]

I have found that, for the most part, I really prefer Lavinthal Discards vs. NT, first discard only.

Why? You don't have to discard a possible winner from your best suit in order to encourage that suit.

However, there are also some disadvantages to them as well. Such as when you can't really afford to pitch from either of the remaining two suits (you need to retain your holdings to protect those suits), but you really don't want either of them led. Unfortunately, if you now pitch from your best suit....partner will refuse to lead it, and lead one of the remaining suits according to your pitch from your best suit. This can be quite aggravating at times.

Another advantage to Lav., imo, is that it only takes one discard to convey the message of what you want.

By direct discard, I assume you mean, you encourage what you want led or discourage what you don't want led. This takes either one encouraging discard or two discouraging (one in each of the remaining two suits, unless already known). If you can't afford to give one encouraging, then partner only has a 50/50 shot at getting it right if he wins a trick before you have a chance to make your second discouraging discard.

Versus suit, I prefer direct discards where high encourages and low discourages, and prefer to give negative inferences first unless absolutely necessary to do otherwise.

[awm]

I've found that there is a substantial advantage to lavinthal discards against notrump contracts. The issue is that fairly frequently there are two suits from which I cannot afford to discard -- one being my long/strong suit which I want partner to switch to (where I'm hoping to establish length tricks for our side) and the other being a side suit where I'm trying to keep length to prevent declarer from running the suit (say I hold JTxx or the like). When the auction is something like 1NT-3NT, it's often unclear early in the hand just what suit(s) declarer holds, and it becomes harder for partner to infer which suits not to lead.

Against suit contracts, if I really want a suit lead I can usually pitch an encouraging card from that suit, plus a lot of auctions to suit contracts are more informative than 1NT-3NT, and eliminating the trump suit (and/or playing suit preference when declarer draws trumps) can usually help remove any ambiguity in standard signals.

[david_c]

Apollo81, on Mar 13 2007, 05:31 PM, said:

Let's say that you can only afford to pitch from one suit (happens more often that inexperienced players realize).  Even with no signals partner can nearly always narrow down his choice to 2 different suits using bridge logic.

I don't think this is true. Indeed, I've tried playing a method which relied on there only ever being two suits you need to worry about, and it didn't work! Basically there are two different situations you might find yourself in. Assuming a NT contract:

(i) More than one suit has been led. In this situation there is nearly always at least one suit which can be ruled out. Methods which focus on a straight encouraging/discouraging message seem to work well.

(ii) Only one suit has been led. For example, the opening leader cashes a few rounds of a suit and partner shows out. Or opening leader finds a bad lead and declarer leads the suit straight back. In this case it very often happens that you might need to show interest in any one of the three remaining suits. Mckenney or other suit-preference methods seem to work better here.

My preferred methods now distinguish between these two situations. Against a suit contract, (ii) only applies if the suit led is trumps.

[P_Marlowe]

"Direct discards": The suit discarded is the suit requested.

My partner says, that he quite often needs to protect
honors in short suits, i.e. cannot discard from those suits,
instead he thinks discarding from a long suit (he is interested in)
is often more an option.

With kind regards
Marlowe

[Apollo81]

P_Marlowe, on Mar 13 2007, 01:12 PM, said:

"Direct discards": The suit discarded is the suit requested.

My partner says, that he quite often needs to protect
honors in short suits, i.e. cannot discard from those suits,
instead he thinks discarding from a long suit (he is interested in)
is often more an option.

With kind regards
Marlowe

What's wrong with high or low encouraging?

Usually in the situations your partner mentioned he has only one choice of suit, but more than one choice of card within that suit.

It's "standard" for a reason

[mikeh]

Let me commend to you the Grannovetter book: A Switch in Time. This is a brilliant book, imho. It describes the Obvious Switch principle.

This is a fundamentally different concept in carding.

I have played most signalling methods, including Lavinthal, udca, odd/even and standard, and firmly believe that the Obvious Switch principle is vastly superior.... altho I do caution against its overuse... as with all signalling methods, there is too much of a good thing in that observant declarers can pick you apart if you give away too much information.

I am more and more convinced that the best approach is to minimize signalling altogether: playing a count or attitude or suit preference card only when you have reason to think that partner needs the information. Thus, if as sometimes happens, you can tell that you hold the important defensive cards, your discards should be uninformative.

Of course, when you are trying a new method, or if you are playing mainly against un-observant opps, it can be useful to signal everything for a while, until you are comfortable.

[kenrexford]

The "Lavinthal Discard" is defined as a special agreement, and yet it really is nothing amazing other than a logical extrapolation of logical card play, coupled with agreed defaults.

To explain. I'll bet that everyone "uses" Lavinthal Discards, at least on rare occasions. Suppose you want to send a message to partner about clubs, as opposed to diamonds. For example, hearts may be trumps, and you want a club switch if partner gets in, or you are planning to save clubs in a squeeze. Whatever.

With "Lavinthal Discards," you would pitch a small spade to indicate a club message.

Without Lavinthal, suppose you had to save all of your minor cards, are out of hearts, but are known to have a plethora of useless spades (maybe you opened 3♠). Furthermore, just to avoid wiggling, suppose that you spade length and holding is 100% established. (Maybe Opener ruffed the first spade in his hand.) Normal logical card play would suggest that a small spade pip, as opposed to a high spade pip, suggested a club signal, right? This is a Lavinthal Discards, and it would be readable by a good partner even without discussion.

It seems that Lavinthal would be most useful as a default when three suits offer legitimate contextual options, or where two suits only are in focus but it is "known" that the person cannot pitch in either suit, regardless of his "preference" or desired signal. Regular carding makes sense when only two suits are in focus and when the person can be expected to have space to make a pitch in the disfavored suit, or when a person might want to send positive messages in two of three suits simultaneously (negative for one carrying positive for both remaining).

Ideally, pure logic might suggest that a pip is either Lav or ct/att, deciphered by reference to known circumstances, but that idea is very difficult to use in practice.

All that Lavinthal Discards as an agreement do, then, is to define that the discarded suit is the one with useless pips and establish that suit as usable for suit-preference signals without the necessity for logic and circumstances defining when a suit is immaterial and the pips, accordingly, used for suit-preference.

[pclayton]

mikeh, on Mar 13 2007, 09:24 AM, said:

Let me commend to you the Grannovetter book: A Switch in Time. This is a brilliant book, imho. It describes the Obvious Switch principle.

This is a fundamentally different concept in carding.

I have played most signalling methods, including Lavinthal, udca, odd/even and standard, and firmly believe that the Obvious Switch principle is vastly superior.... altho I do caution against its overuse... as with all signalling methods, there is too much of a good thing in that observant declarers can pick you apart if you give away too much information.

I am more and more convinced that the best approach is to minimize signalling altogether: playing a count or attitude or suit preference card only when you have reason to think that partner needs the information. Thus, if as sometimes happens, you can tell that you hold the important defensive cards, your discards should be uninformative.

Of course, when you are trying a new method, or if you are playing mainly against un-observant opps, it can be useful to signal everything for a while, until you are comfortable.

Mike, you might be the 1st high level player I know of that plays Obvious Shift. if you've mentioned it before, I hadn't noticed.

I place OS in my favorite partnership, and I really like it. Aside from keeping the rules straight which becomes 2nd nature after awhile, and the propensity to over-signal, it is a superior method I think.

What I'm amazed at is the lack of other high-level pairs that play this, although I thought I heard Meckwell plays something similar.

Do you know of other pairs that play this? Have others looked at it, but discounted its merits? Or, is this just one of those great methods that hasn't been picked up by the mainstream expert community?

[mikeh]

Thank you... I do not consider myself to be a 'high level player', merely a high level poster, but thank you

I remember reading, years ago, that Zia-Rosenberg played some variation on OS.

I don't recall reading of any others who play OS as such, but I suspect that many pairs incorporate some elements of the ideas into their approach to signalling.

[mike777]

"Mike, you might be the 1st high level player I know of that plays Obvious Shift"


LOL ROFL, well the Granovetters are not doing that poorly themselves, they may be close to being high level players

[paulg]

Link to Obvious Shift Principle courtesy of Dan.

[ArcLight]

>I am more and more convinced that the best approach is to minimize signalling altogether: playing a count or attitude or suit preference card only when you have reason to think that partner needs the information. Thus, if as sometimes happens, you can tell that you hold the important defensive cards, your discards should be uninformative.


Mike, how can your pard know when you are signaling "for real" as opposed to just playing random cards or following suit.
Keep in mind taht you are better than most of us.
How can a good intermediate (or advanced) player "know" when their pard is signaling (becaus ethey need to know) or not.

I've seen Justin and other experts make similar statements, but I'm not sure on the criterial.

- Obviously if dummy has a long suit in NT with no other entries - signaling Count makes sense

When else is it "obvious" to pard you are signaling "for real - when they need to know"

[mike777]

If you are asking about "OS"

The "obvious" shift part in the carding means, if you discourge at trick one, what suit are you telling partner to shift to.
1) You can never TELL partner to shift to trumps so that is out. She needs to figure that out on her own.
2) In general vs, NT the the OS suit is dummies weakest suit.
3) In general vs. suit contract it is dummies weakest side 3 card suit.

4) Trick two and all others tricks are almost always suit preference.
5) Count is very seldom given except in "obvious" hold up situations.
6) If you wish you can add o/e at your very first discard without a problem with "os"

I should add something very important here. Using your bridge knowledge and experience and counting and visualizing is still very important.


http://www.bridgetoday.com/index.php

[Halo]

Generally I think it is hard to do better than count, so long as you have the energy to pay attention and think. I often play attitude.