UI (16B1b)
You hold....
QJTxx, Kxxx, Tx, Ax, you (West) deal at love all, Teams, and choose to Pass.
LHO bids 3D passed back to you. What do you bid?
Of course that is not the question!
There has been an agreed hesitation by partner over the 3D.
West Dbled and E-W found their way to a making 4H.
Opponents are mildly miffed, and ask for a ruling.
An argument could be made that West has a 'logical alternative' in Pass.
However knowing West well, I know his 'style' is not to make light openings, hence the initial Pass,
but is generally competitive. He really would fall into the 'I was always going to do that' camp,
though I am sure he wouldn't have said that!
My query.
Is 'style' covered by 'class of player' and/or 'methods of the partnership'?
Comments
Actually that is precisely what the question should be! That's how you determine whether or not Pass is a logical alternative. My answer to the question would have been to double, but I would not be surprised to find enough people passing that it be considered an LA.
Yes, it is covered by "class of player" but it can be difficult to evaluate this. I suppose you should try to find people to poll who have similar styles, but failing that tell those you poll of his stylistic preferences.
Yes - Style is covered - and so is demeanour. If you have a very aggressive player then you have to poll very aggressive players, for instance (Hint - look to any who don't have grey hair - if you can find them :) ). Your problem will be to find players who would pass with that hand but are otherwise competitive.
Of course even such players might not bid on the given auction. (The white book suggests one in five, so if you poll 5 people then if one would pass then it is likely that passing is a LA). I am not a mathematician but there is a lot of variance in the results when you only poll 5 people.
Note that there is 16B1b and also 73C1. Did the player "carefully avoid taking advantage of the unauthorized (sic) information"?
This situation is where a poll comes into play. If you can't hold a poll of comparable players then you're left with your own judgement. At worst, you make a ruling and advise players to appeal.
Methods of the partnership are paramount, style of player not so much, since that's open to conjecture. It would not be common, though, for the methods of the partnership to have any written evidence of what they do in this situation.
I'm fascinated by this... and very grateful to Gordon and WeeJonnie for confirming style is included
As Tag suggests not always as easy to tell a player's style as their methods.
I guess at the highest levels you probably know someone's style, and sometimes people do annotate their cards with 'style' type info.
Would you ever go so far as polling those who knew the player better about their style?
Or indeed looking at other results in the session to see if they consistently bid light games and ./ or slams , or
pushed opponents to unmakeable level contracts?
Peter Bushby Suffolk
"Why did you double?"
If I had a penny for every player who tells me "Because I am a very aggressive bidder!", I'd be a rich man!
In the OP case, we have someone who is insufficiently aggressive to open this routine Ro19 hand in 1st at LA at Teams.
"However knowing West well, I know his 'style' is not to make light openings, hence the initial Pass, but is generally competitive."
I would know very few players that well to be aware of such styles. Probably only my most frequent regular partners. I think it is very difficult to be persuaded of a players style. And I don't believe there is anyone who always bids aggressively.
"Why did you double?"
What we are really interested in hearing when we routinely ask this question is something we might not expect to hear. Maybe some systemic requirement we might not guess. Probably not in this situation, but maybe in some other hesitation situations.
"There has been an agreed hesitation by partner over the 3D."
Stating the obvious, but it's always worth checking the use of the Stop Card and that the hesitation was clearly beyond the 10 seconds when there has been a jump bid.
Barrie Partridge - CTD for Bridge Club Live
" ... there is a lot of variance in the results when you only poll 5 people."
Indeed.
Let's say we want to know if passing is a LA to action. I would want at least three to say "pass" before I stop after polling just 5 people. My personal record is about 30 polled when it has taken me that long to get sufficiently away from the EBU 20% figure to be happy enough to give a ruling.
Where I can, though it's not always practical, is to continue a poll sufficiently to give confidence against a possible appeal committee of three. So, before deciding to stop polling, I think to myself "If I poll three more players, could that possibly change my decision whether passing is a LA?". If "yes", then I like to continue polling if I can.
Players usually appreciate knowing the results of polling.
Barrie Partridge - CTD for Bridge Club Live
If you are trying to judge whether Pass is a logical alternative, you need far fewer votes in favour of passing to decide it is, than votes to bid in order to decide it is not.
This started me thinking about a process where you keep polling until it's clear. Statistically speaking, you should decide on your poll size first. If you do (for obvious reasons) want to keep going until the answer is clear then you should define your "acceptable result" parameters in advance. I.e. what is "too close to 20%"? 22%?, 25%?
This then got me onto your key comment: "If I poll three more players, could that possibly change my decision whether passing is a LA?", which leads to "what is a good first sample size"?
Let's go with your "three more people", and assume that your first sample is not representative of the population as a whole (otherwise why are you going on?).
If you poll 5 people, and get 2 or more (40%) going for the LA, then that outcome will not change with three extra people (at best 2 out of 8 - 25%).
If you get 1 (20%) going for the LA, then only if the next three don't go for the LA does the outcome change.
If you get no-one going for the LA, you're probably not going to go on, even though you could get a change of outcome.
Extending these numbers, you would have to poll at least 13 people initially to potentially get a "no LA" outcome that would not change. How many of us do that?
Thanks for all the responses.
I think the main point is that 'style' is covered.
Though I agree it is difficult to judge, in this case West was one of my regular partners.
I am also a bit worried that this opens the door to the 'I was always going to bid that!' brigade.
Yes, my initial question would be the one to poll.
But this was Swiss and I needed an answer, if only for the purposes of assignments.
I thought the Dbl was justified, not least because at all the other eleven tables, E-W played a H contract.
(But I have no idea if the auctions were similar.)
As I said N-S were only mildly miffed and didn't appeal.
Subsequently the consensus of the great and the good (aka those in the pub!) was for some action,
with one esteemed exception.
That was about eight people, a little short of Jeremy's thirteen!
"But this was Swiss and I needed an answer, if only for the purposes of assignments."
While it is ideal, it isn't at all essential. You usually have to wait till players have finished their match before you can poll anyone, and you may well hold up the event if you want to give a ruling before assigning for the next round. Assignments can be made for the next round or rounds using the table result, and if the table result subsequently gets adjusted, that's no problem for EBU Swiss Pairs Scorer or even Scorebridge.
Barrie Partridge - CTD for Bridge Club Live
It's worth pointing out that a ruling leading to an adjustment never creates a mismatch for the purpose of Swiss assignments.
Interesting post - Thanks, Jeremy. As Gordon also suggested, polls tend to be much shorter the more obvious it is that passing is a LA (sticking to the simple situation of wanting to know if Pass is a LA!) and a good proportion of polls are like that, where we can stop very quickly. We do need to go a bit longer if no one is considering passing, but I'd probably stop after about 7 or 8 polled. It's when just one or two start saying they'll pass that makes a poll longer.
Barrie Partridge - CTD for Bridge Club Live
Speaking as the dissenter referred to I've little doubt that quite a few, even a majority, would protect.I would have asked some other players in the top half of the field as the match took place near the top of the field. The doubt I expressed was whether 80%+ would have bid or doubled. I think pass is an LA. The results at other tables are, IMO, irrelevant because, as noted, the auction may have been different. I don't think it was realistic to start consulting and seeking to find people with a similar style.
I think a lot of nonsense is spoken about sample sizes for polls, and practical limitations should also be considered. I'm sure statisticians could come up with calculations for sample sizes to give suitable statistical power to our poll, to reduce the chances of making a type I error (deciding an action is a logical alternative when really it isn't) or a type II error (deciding an action isn't a logical alternative when really it is) with 95% confidence, or something like that, but what we are doing is rather different than that. To start with, the guidelines in the White Book are quite vague about how to define a logical alternative, with phrases such as "fewer than about one player in five" and "serious consideration" and "more than just an isolated exception". When we ask players what they would do, we rarely get a "yes-or-no", "bid-or-pass" answer. The doubt that a player shows when they think hard and finally decide to bid indicates that pass (or some other action) is being given serious consideration, and that can be taken into account.
If I poll three players and they all give the same answer confidently, without hesitation, that's pretty good evidence that there's no alternative being considered, and that I can discount anything else as a logical alternative. If one or more of them think for a long time, I can ask what else they are considering, and then perhaps poll another player or two to see if anyone would actually choose the action, but I rarely poll more than three players, and only in cases of doubt would I go up to as many as five. The danger of polling as many as eight or even thirteen players, apart from it taking up too much of the TD's time, is that you end up asking players who are ever less suited to be considered peers of the original player, or those you trust less to give an honest, considered opinion. And what you're trying to pin down is quite a vague and slippery concept in the first place.
Large polls may boost a TD's confidence, but I think that confidence is based on shaky foundations. My advice is to conduct a small poll, but do it well, choose your subjects carefully, those who are as near to peers of the player as possible, and those who are willing to take time to give a considered answer, and take into account the nuances of their answer. I never feel ill at ease taking the results of such a poll into an appeals hearing.