[awm]

pclayton, on May 25 2007, 07:18 PM, said:

I play OS. Just because something is unfamiliar doesn't mean that its needs advance discussion.

As a matter of policy, we pre-alert in team games, even though we don't have to. We don't pre-alert in pair games.

All of OS, except for the 1N response to the garbage overcalls (which we don't play) is GCC. Much of OS is essentially natural.

It isn't too difficult to devise that a double of a 1NTO operates like a redouble. In my experience, where people get confused is doubling the 2 suited jump overcall. It 'sounds' negative, but many play it as penalty.

The 2N overcall comes up so infrequently, and its frequently 'our hand', so devising a defense isn't worth the bother usually.

Yes, it's GCC. Then again, canape-strong diamond systems are GCC and require pre-alert. Opening eight-counts is GCC and requires pre-alert. Frequent preempts on poor five-card suits is GCC and requires pre-alert.

It's quite easy to devise a defense to overcall structure. But if you've never seen it before, it's easy to have difficulties when you and partner aren't on the same page. Several different people who play this method have described "accidents" that opponents had against them (basically not knowing their defense) that lead to top boards. If this experience is as prevalent as it seems, it suggests a pre-alert might be a good idea.

The fact is, opponents opening light at the one-level is nice to know, but almost nobody changes their defensive methods because of this. A lot of people look only at the "front of card" before starting a round, so won't notice overcall structure. A lot of people ask the opponents "general approach" rather than look at a card, and won't hear about overcall structure. It just seems strange that defenses to opening bids which are very non-standard, very unfamiliar, and frequently result in misunderstandings by the opening side are "not a pre-alert" whereas opening methods as vanilla as opening a few points light require a pre-alert.

Obviously you can argue that a top-flight pair will look at the opponents' card in detail, or that they'll have seen overcall structure before and have a prepared defense, or have meta-defenses... but by that argument nothing would require a pre-alert (and in fact you could argue that openings/responses don't require an alert because they're on the CC). It just seems to me that if any method that's on the general chart requires a pre-alert (and some do) then overcall structure should require one.

[pclayton]

Adam - I might be confused on this. Why would any GCC treatment require a pre-alert? I thought only mid-chart required a pre-alert.

[pbleighton]

Quote

Adam - I might be confused on this. Why would any GCC treatment require a pre-alert? I thought only mid-chart required a pre-alert.

Pre-alerts are required for any system which may be fundamentally unfamiliar to the opponents, such as canape, those where hands with < 11hcp are routinely opened at the 1 level, and those where 2 bids may be Qxxxx or 3 bids may be Qxxxxx. 4 card majors and strong club systems are excluded, unless they qualify for a pre alert for another reason.

EHAA, for example, is pre alertable.

There is no specific provision for GCC legal bids being pre alertable. The ACBL does encourage alerting wherever you think it's appropriate, even though it may not be specified in the regs. This could be construed (or not) to apply to pre alerts.

On a practical note, I've played a lot of pre alertable systems, and most of the opponents absolutely hate the pre alerts. They just want the bids alerted. I will therefore not pre alert something the ACBL doesn't say I need to pre alert.

Peter

[Fluffy]

They sent a letter to all members of AEB (spannish NBO).

It as 4 or 5 pages to say you must alert everything but stayman, an even prealert many other things.

Nobody has followed the rule, and people do exactly the same as before.

[helene_t]

It's impossible to devise an alert procedure that satisfies everybody. I used to hate the Dutch alert procedures but since I realized that most other countries have even more ridiculous procedures and that many people hate the alert rules that I like the most (and vice versa) I've become more sympathetic to it. For example, I very much like anouncements, but Jeff Rubin wrote in an editorial in Bridge World that they are abused so often that they should be abandoned, and probably he knows more about how they work in practice than I do.

In particular, it's difficult to devise alert procedures that work at all levels from coffee-house to Bermuda Bowl, yet a jungle of different alert procedures is not desirable either. We already have three dimensions (SO, yearly changes, and IRL/online) of different procedures.

The general rule in the Netherlands is that natural calls are not alertable. This conflicts with the general rule that you should alert everything that may be misunderstood by opps if not alerted. Surprisingly, not everybody thinks that the "alert misunderstandable calls" precedes over "don't alert natural calls" although, as I read the laws, the precedence is quite explicit. I can imagine two reasons for that:
- "Misunderstandable" is too subjective to be enforceable. You might alert everything on the basis of the fact that Gozilla might not understand that a 1♦ opening promises diamonds, that a 1♥ response to 1♦ is forcing etc. On the contrary, "natural" can be defined in a way that makes it almost culture-independent.
- The main reason for alerting is that opps may play different defenses against different methods. Often, standard methods apply against natural methods while lead-directing doubles are used against artificial methods. While not bullet-proof, the information that a call is "artificial" gives more clues to the choice of defense than the information that "it might be different from what you are used to".

I think the BF should resolve this issue by stating either:
- Natural calls are never alertable (no exceptions), or
- You must alert only on the basis of deviance from common standards. Naturalness per se is irrelevant.

Of course no matter which of the two they choose a lot of people will get angry.

In conclusion, I think the Dutch procedures are not that bad. Some minor justifications would make them better, though:
- Why alert PS? The not-alertable 2♣ response to 1NT could mean anything depending on Stayman version, and PS falls well within that range. As Apollo says, it would make more sense to alert opener's rebids after PS.
- Why alert Walsh? Makes no sense. In particularly in NL where many advanced pairs play some kind of soft-Walsh which may or may not be alertable depending on the mood of the TD. If you play major inversion, you really need to alert but opps won't notice because the Walsh-alerters cry wolf.

[whereagles]

We have a nice one in Portugal. Two of our alert rules are:

#1 "Vulnerability-dependant bids must be alerted"
#2 "NT openers showing balanced hands need not be alerted"

So what happens if you play a variable 1NT? Nobody knows!!!

[han]

pclayton, on May 27 2007, 11:38 AM, said:

Adam - I might be confused on this. Why would any GCC treatment require a pre-alert? I thought only mid-chart required a pre-alert.

See my own post in this thread:

What I like least is that alert procedures is that many duplicate players think that the rules are more important than the main idea: warning the opponents about your unusual methods. But given how poorly the rules are understood by the average bridge player, I guess that cannot be avoided.

[KiwiBridge]

From the NZCBA manual:

During the auction:

Do alert all conventional calls other than those listed below. A conventional bid is not only a bid whose meaning is not related to the bid (e.g. a Multi 2D has nothing to do with diamonds) but also a bid that conveys extra information (e.g. 2H showing hearts and another is conventional because there is additional information that the hand contains a heart suit – making it conventional).
Do alert bids that are outside reasonable expectation. For example: a non-forcing response by partner in a new suit would require an alert.
The only exceptions are:

Do not alert any natural No Trump opening bid and simple 2C Stayman response (you should pre-alert the HCP range - if it is different from what is generally accepted as the basic system).
Do not alert any 3 card+ minor or 4+ major opening bid at the one level unless there is additional information (e.g. canape).
Do not alert a strong 2C opener and the negative 2D response.
Do not alert any double or redouble irrespective of meaning
Do not alert any cue-bid -- a cue-bid is defined as a bid in a known opponents suit, or a bid of the suit they have bid irrespective of the meaning of that bid.
Do not alert any bid at the 4 level or higher.
The last three categories of calls referred to above are considered to be self-alerting and therefore are exempted from alerts. Your opponents are expected to ask about their meaning if it is important to them to know before they make their own call.

[foo]

cardsharp, on May 25 2007, 08:12 AM, said:

There is often much debate about alerting on BBO, but in f2f bridge we all play under different alerting regulations and I wondered what you especially liked, and disliked, about the ones you regularly use.

<snip>

Any other ideas I should be pushing the SBU to consider?

One class of alerts that I think all Sponsoring Organization's should be considering more carefully is those involving unusual negative inferences.

Simple example.
Playing Support X's, Opener's raise showing 4+ and their X or XX showing exactly 3 are both alertable. OTOH, Opener's other calls have the strong inference that Opener has 2-. =and it's not altertable= (at least in any jurisdiction I've played).

That's a simple example. There other examples, particularly in rarer systems, that have far more wide-reaching implications.

Since the guiding principle is that the opponents should know everything Our hand that we know, this lack of "negative inference" alerts IMHO can be a serious problem in terms of maintaining equity.

[jdonn]

foo, on May 29 2007, 12:09 AM, said:

One class of alerts that I think all Sponsoring Organization's should be considering more carefully is those involving unusual negative inferences.

Simple example.
Playing Support X's, Opener's raise showing 4+ and their X or XX showing exactly 3 are both alertable.  OTOH, Opener's other calls have the strong inference that Opener has 2-.  =and it's not altertable= (at least in any jurisdiction I've played).

Actually the acbl has gone the other way (and I agree with them). The raise promising 4 when you play support doubles used to be alertable but hasn't been for a couple years now.

[paulg]

foo, on May 29 2007, 06:09 AM, said:

One class of alerts that I think all Sponsoring Organization's should be considering more carefully is those involving unusual negative inferences.

Simple example.
Playing Support X's, Opener's raise showing 4+ and their X or XX showing exactly 3 are both alertable.  OTOH, Opener's other calls have the strong inference that Opener has 2-.  =and it's not altertable= (at least in any jurisdiction I've played).

That's a simple example.  There other examples, particularly in rarer systems, that have far more wide-reaching implications.

Since the guiding principle is that the opponents should know everything Our hand that we know, this lack of "negative inference" alerts IMHO can be a serious problem in terms of maintaining equity.

Negative inferences are often a topic of debate, but the view that I've seen from the ACBL and EBU is that they are not generally alertable (BOCTAOE).

The problem that the SOs face with negative inferences is a proliferation of alerts. Alerting every call is just as bad and confusing as not alerting so they have to draw a line somewhere. Negative inferences tend to fall just on the non-alert side of that line.

[helene_t]

From a philosofical point of view, there's no difference between negative inference and positive inference. For example, one could argue that the strong pass in a weak opening system is not alertable since it just shows a hand that doesn't meet the criteria for a (weak) opening. This may sound pedantic but seriously I don't think the term "negative inference" is clear enough to be used in law text.

I think it's important to teach players the purpose of the alert procedure, namely to warn opponents about calls that have a different meaning than what they might assume without the alert, especially when it might influence the defensive bidding of opps. Of course this will always conflict with the letter of the law because the letter of the law must be enforcable and therefore less subjective. But even if the SO chooses a very mechanistic approach (alert all artificial calls such as Stayman, alert no natural calls such as strong 3-openings), it is important that players understand the underlying idea, because:
- They may alert some calls that are not alert-requirering but still can be alerted, especially if playing against beginners who have no idea about the differences between "beginner standard" and "advanced standard". Even if the law does not require this, there is still something called good sportsmanship.
- They should pre-alert any surprising agreements that must not be alerted (such as artificial doubles against 1NT).

[Gerben42]

The big problem is this: Information value. The alert should have information value so if you alert anything that is different from SAYC, for example, players playing a different system would have to alert almost everything, thus giving their opponents the disadvantage of having to suffer meaningless alerts, causing the really unusual things to be swamped.

For example playing 2/1 with gadgets and weak NT, consider this auction:

1♦ alert, cannot be a weak NT
- 1♠ alert, can have longer minor
2♣ alert, can have 4♦ 5♣
- 2♠ alert, invitational
Pass

On the other hand this information may be useful when defending.

[mink]

Some said that players need to understand the purpose of alerting - the general principle that alerts are there to warn opps that a bid has an unusual meaning. I strongly disagree. Unusual, if not exactly defined by the sponsoring organization, is a subjective term - what is unusual for a little old lady may be standard for a top player and vice versa. If you allowed the alerts to be based on opps understanding of what is standard and what is unusual, it would require us to alert a transfer if you know that opps do not play transfer and not alert it if they play transfer. This is ridiculous.

Rather, when designing alerting regulations, the sponsoring organization must implicitly define what should be regarded as standard and define the regulations such that for any call it is clear by the regulations if it is alertable or not. The standard has to be a simple standard of course, as most players will not remember a complex set of alerting rules.

My favorite rules would be:

never alert:
- a pass after that partner might pass, too
- a double that is for penalty or strength-showing or info or optional or negative
- a redouble that is strength-showing or sos
- a cue bid in opp's naturally bid suit
- a bid of 4 club or higher, except for an opener
- a 1 club opening showing at least 2 club cards and denying a 5 card major

always alert:
- a bid in opp's naturally bid suit showing you have this suit yourself

in uncontested auctions do not alert
- stayman 2c and 2d by opener
- transfer to major and and the next bid by opener

in uncontested auctions do alert
- bids usually used for stayman and transfer if they mean something else

If a bid is not covered by the above rules, do not alert if
- a suit bid shows at least 4 cards in this major or 3 cards in this minor or
- an nt bid shows a balanced or semi-balanced hand or is non-forcing

Alert all other bids.

These rules do not refer to strength, therefore an unusual strength is never a reason to alert. This is because the rules are much simpler this way. They also do not contain the word unusual, which is necessarily interpreted differently by different people.

Karl

[david_c]

helene_t, on May 29 2007, 10:24 AM, said:

- They should pre-alert any surprising agreements that must not be alerted (such as artificial doubles against 1NT).

I really hate this one. In my opinion, pre-alerts should not be used as a substitute for alerts. This is because:
(i) You can't expect people to remember their opponents' systems.
(ii) It takes a fair amount of effort to explain and understand a pre-alert correctly. If the agreement in question doesn't actually come up, then all that effort is wasted.
(iii) If it does come up, it's much easier to explain at the time than it would have been beforehand, because you don't have to explain what situation it applies to.

In any case, there's absolutely no need for this sort of regulation. If you're prepared to write "non-penalty doubles of 1NT openings must be pre-alerted" into your regulations, how is it any more difficult to write "non-penalty doubles of 1NT openings must be alerted"? I can understand a reluctance to introduce exceptions into the regulations, but it seems to me that you're adding an exception whatever you do, just in a different place.

Pre-alerting does have a useful function, but that is in warning opponents about any particularly unusual conventions that they may need to discuss their defence to in advance. It shouldn't a be a substitute for alerting.

[helene_t]

david_c, on May 29 2007, 03:27 PM, said:

helene_t, on May 29 2007, 10:24 AM, said:

- They should pre-alert any surprising agreements that must not be alerted (such as artificial doubles against 1NT).

I really hate this one. In my opinion, pre-alerts should not be used as a substitute for alerts. [......]
In any case, there's absolutely no need for this sort of regulation. If you're prepared to write "non-penalty doubles of 1NT openings must be pre-alerted" into your regulations, how is it any more difficult to write "non-penalty doubles of 1NT openings must be alerted"? I can understand a reluctance to introduce exceptions into the regulations, but it seems to me that you're adding an exception whatever you do, just in a different place.

Point taken. I was wrong.

What we usually do is to pre-alert stuff which we can't alert because of regulations (such as artificial doubles of 1NT) as well as stuff that may or may not be alertable depending on the mood of opps and/or TD. Especially weak jump shifts. Some people get upset if you don't alert WJS, claiming that the law says they must be. Other people get upset when you do alert WJS, claiming that the law prohibits alerting of WJS.

Of course this is not an ideal situation. Either the laws should be clearer and/or better understood, or at least we should trust opps to acknowledge that since everybody interprets the law differently, they can't expect us to alert as we should according to their interpretation of the law.

[pbleighton]

Quote

I really hate this one. In my opinion, pre-alerts should not be used as a substitute for alerts.

Ditto.

In fact, pre-alerts should be eliminated.

Peter

[paulg]

England and Scotland do not have pre-alerts. However our methods, specifically transfer responses to a short 1♣ opener and leading low from small doubleton, are unusual here and come up very frequently, so we do pre-alert everyone to them.

It is a little more tedious in the ACBL where we must pre-alert five or so other (Mid Chart) methods that are much rarer and, imo, easier to defend. You can see the opposition's eyes glazing over after the first two.

But I agree with David that the idea of pre-alerts is reasonable if it can be done quickly, concisely and is something that the defence would need to discuss.

Paul

[jtfanclub]

pbleighton, on May 29 2007, 08:39 AM, said:

Quote

I really hate this one. In my opinion, pre-alerts should not be used as a substitute for alerts.

Ditto.

In fact, pre-alerts should be eliminated.

Peter

I disagree. I think that the number of pre-alerts should be expanded, but kept short on each. I mean, if on the computer I can say

"Hi, we play Precision with upside-down carding, Journalist leads, and odd/even discards"

How much harm can it do to pre-alert this on face-to-face, even in a pairs game?

The little stuff that doesn't come up very much, that I don't like those in pre-alerts, although I'd be willing to (for example) pre-alert the Multi 2♦ if that means we can play it in GCC.

[paulg]

I was wrong, Scotland does actually require pre-alerts. Everyone does it but I thought this was a matter of courtesy rather than regulation, but our directives state:

SBU Directives said:

§2 At the start of each round, you should exchange convention cards with your opponents and inform them of:

• Your basic system
  Examples: Acol, strong club, five-card majors with forcing no-trump
  
• Any unusual aspects of your system
  List these items in a clearly identified area of the convention card. Examples: canape (bidding a shorter suit before a longer one), strong two over-one responses, artificial opening bids such as the multi-2♦, weak jump shifts, weak jump overcalls.

England requires you to exchange convention cards but there is no requirement to pre-alert (although I've yet to find anyone object to it).

Paul