[Cascade]

 nige1, on 2017-March-02, 15:57, said:

We aren't privy to NS methods. Had East passed then 2♠ might well be Pass/Correct. After East's double, South's suit-bids are more likely to be natural. Presumably, redouble would ask for North's 2nd suit. 2N has more than one likely meaning. Anyway, IMO, with ♥ tolerance, South would be likely to pass, in tempo.

2♠ is pass or correct. Not sure if they have a meaning for redouble.

[Cascade]

 barmar, on 2017-March-02, 08:44, said:

They're rewording it in the new laws, and there was quite a bit of discussion in the Yahoo group about it. I'm not sure what the final language is, though.

Rather than change the wording and it is strange to me too as Helene suggested, I think we would be better served with some test that could be applied to determine whether something is suggested. With such a test we perhaps could hope for some better consistency and objectivity in determining whether the something is suggested over another call.

Here we have a discussion that mostly seems to think that 3♣ is suggested over pass if I am interpreting the responses correctly. When this happened the appeal committee that I would otherwise have thought was a very good committee voted unanimously that 3♣ was not suggested by the break in tempo. As I wrote earlier they did not specifically say "3♣ could not be demonstably suggested over pass" which I think was an error and I can't be sure whether the error was in their reasoning or their writing/communicating.

Whatever the interpretation we shouldn't have the situation where one set of fine minds thinks there is no problem and therefore no need for an adjustment whilst another set of equally fine minds thinks that a player should be barred from bidding 3♣ and therefore that an adjustment is necessary.

[nige1]

 Cascade, on 2017-March-02, 20:39, said:

Rather than change the wording and it is strange to me too as Helene suggested, I think we would be better served with some test that could be applied to determine whether something is suggested. With such a test we perhaps could hope for some better consistency and objectivity in determining whether the something is suggested over another call.

Several times, here and elsewhere, I've suggested such a test. Roughly, in summary:

• Poll peers of the player as to what actions they'd seriously consider.
• Add the action chosen by the player to the list of LAs.
• Ask pollees to rank the list.
• Tell pollees about the UI.
• Ask pollees to re-order the list, pretending the UI is AI.
• Check whether the rank of the player's chosen action improves.

When this protocol is impractical, it might still clarify relevant concepts, as a useful thought-experiment for the TD.

As is usual for my suggestions, this attracted ridicule

[mycroft]

My suggestion that we have a series of case studies with the imprimatur of Laws Commission approval for this Law (or, better yet, an Appendix to the Laws) was considered a good idea by the LC member. Having said that, it felt like "that's a good idea, but we both know it won't happen."

Which sort of is my point in that other thread.

[barmar]

 Cascade, on 2017-March-02, 20:39, said:

Rather than change the wording and it is strange to me too as Helene suggested, I think we would be better served with some test that could be applied to determine whether something is suggested. With such a test we perhaps could hope for some better consistency and objectivity in determining whether the something is suggested over another call.

Indeed, that's what much of the discussion was about. Someone recommended something involving determining how the UI changes the expected probability of success of each action. So if it was 60/40 before the UI, and changes to 70/30, you're not allowed to choose the first action, because its probability of success has increased. Furthermore, if it was 80/20 and changes to 60/40, you're not allowed to choose the second action -- it's still less like to succeedthan the first, but its success probability has increased.

[blackshoe]

 barmar, on 2017-March-04, 12:08, said:

Indeed, that's what much of the discussion was about. Someone recommended something involving determining how the UI changes the expected probability of success of each action. So if it was 60/40 before the UI, and changes to 70/30, you're not allowed to choose the first action, because its probability of success has increased. Furthermore, if it was 80/20 and changes to 60/40, you're not allowed to choose the second action -- it's still less like to succeedthan the first, but its success probability has increased.

Hm. Seems to me that second assertion runs afoul of "over another" in the law.

[barmar]

 blackshoe, on 2017-March-04, 14:19, said:

Hm. Seems to me that second assertion runs afoul of "over another" in the law.

I think he was suggesting a wholesale rewrite of that law, to conform with the logic he was proposing. I think the idea is for 16B to express the same thing as 73C (which says you must avoid taking any advantage from the UI), but in a more detailed way regarding logical alternatives.

[mycroft]

We still have to deal with the "I know that 4♠ will be rolled back, but since 6♠ isn't logical here opposite a 'questionable invitiation', let's try that." It's not 80/20, it's 80/18/2; but since the player believes (with the UI) that +200 will be almost the same zero as -100, why not try it?

[dokoko]

Assuming advanced players using unusual methods not thoroughly discussed, i bet their "agreement" is as follows:

1. If a player judges action is called for but doesn't know how to act, he will hesitate (naturally as there is no simple solution to his problem). Partner will get the message and choose the best action from his point of view. This will be a guess and can be argued as such (and accepted by most TDs) if successful. Obviously advancer didn't suggest the specific action taken.

2. First Player will normally not hesitate if he is unsure whether to take action at all, because he will fear that partner will act (on the hesitation) and will run when it might be better to stay.

This kind of agreement obviously is never explicit but any experienced player will "feel" it is wrong to hesitate in situation 2 (because of previous accidents) but it's ok to hesitate in situation 1 (because of some good results). This even applies to casual partnerships as the claimed experience isn't partnership-specific.

In fact I can't remember ever being asked "What would you bid if you were unsure of your methods?" by a TD resolving an UI situation. And even then, I guess, away from the table most ppl would first decide what methods would apply if undiscussed and then select their call based upon the method chosen. This is not the pragmatic approach a reasonable player would use at the table.

So my reasoning as TD would be as follows:

1. By choosing your unusual methods you are responsible for any problem that may cause.
2. A hesitation in unclear situations tends to require partner to act.
3. So any action is deemed influenced by UI unless there is evidence to the contrary.

You may put it the other way round: A pass without hesitation suggests partner is happy in 2h, the hesitation suggests partner isn't that happy. Hence a non-pass is suggested by the UI.

[blackshoe]

 dokoko, on 2017-August-06, 04:18, said:

2. A hesitation in unclear situations tends to require partner to act.

I don't buy this. It is contrary to law.

[dokoko]

 blackshoe, on 2017-August-06, 07:51, said:

I don't buy this. It is contrary to law.

Obviously it's contrary to the law, but that's how humans act.

[blackshoe]

I think you misunderstood. It is contrary to law to rule that a hesitation tends to require partner of the hesitater to act without investigating the LAs and which are demonstrably suggested over which others.

[barmar]

Figuring out what a hesitation suggests is hard. I think what he's saying is that directors are biased in a particular direction, and it's difficult to overcome that. Directors are, after all, only human.

[blackshoe]

Well, maybe. Is Cthonic a director?

[barmar]

 blackshoe, on 2017-September-11, 19:17, said:

Well, maybe. Is Cthonic a director?

Gratefully, no. And if it were, it could probably calculate a way to issue a penalty for you misspelling its name.

Now you've got me imagining Chthonic visiting a "certain North London club" -- there may be a explosion of insufferability.

[blackshoe]

ROFL!

I knew it was wrong, but I didn't have time to figure out why.

[barmar]

 blackshoe, on 2017-September-12, 20:01, said:

ROFL!

I knew it was wrong, but I didn't have time to figure out why.

Because you're merely a human, of course!

[Zelandakh]

I would tend to agree with Helene that a BIT tends to suggest running and certainly "could demonstrably have been suggested". Sure it is not the slam dunk thatr a XX would be but 3♣ after a BIT is surely more appealing than without so it has to be penalised if it works out.

[blackshoe]

 Zelandakh, on 2020-February-29, 05:19, said:

I would tend to agree with Helene that a BIT tends to suggest running and certainly "could demonstrably have been suggested". Sure it is not the slam dunk thatr a XX would be but 3♣ after a BIT is surely more appealing than without so it has to be penalised if it works out.

"could demonstrably have been suggested" is old news. The current law says "is demonstrably suggested".

[Zelandakh]

 blackshoe, on 2020-February-29, 10:50, said:

"could demonstrably have been suggested" is old news. The current law says "is demonstrably suggested".

Well I think I already demonstrated the logic as to why 3♣ is mathematically more likely to be right after a BIT than without, which would seem to disqualify it. This is one of the issues here and crosses over with the BIT appeal thread. Say we have a position in which the estimated probabilities of success are P 80%, XX 5%, 3♣ 15%. After UI the probabilities change to P 30%, XX 50%, 3♣ 20%. If asked by a TD, any player understanding the situation would say that XX is suggested by the UI. But the odds of 3♣ working have also increased and it is therefore also demonstrably suggested. So far, so good but let me give another example now that is more difficult. Say the probabilities after UI are actually P 30%, XX 60%, 3♣ 10%. Now the odds of 3♣ working have gone down, so no problem, right? Wrong! Think of it from the point of view of an experienced player. They understand that XX is going to be ruled against and so have to decide between P (the ethical call) and 3♣. Between these, the odds of success are P 75%, 3♣ 25%. So, for a class of player that understands that XX is not biddable, the relative odds of 3♣ working have increased even here. My impression, perhaps false, is that top players understand the maths of UI situations much better than TDs and have mostly learned how to get around them without recording the disaster that the ethical choice might lead to.