Join Bridge Winners
Bridge justice on the edge
(Page of 4)

The following refers to several UI posts which were published recently, in particular the suggestion to treat in-tempo actions as UI under certain circumstances. This is no doubt a controversial topic, and you are of course welcome to disagree with my views. But if you choose to disagree, please do not do so blindly.

"UI from an in-tempo action" is a contradiction in terms. The laws are a little vague in this regard, but they use the words "steady tempo" (Law 73, about what is required/recommended), "unmistakable hesitation" and "unwonted speed" (Law 16, as some examples for UI).

I also believe I have seen a document for directors which defined "break in tempo" simply as "deviation from normal tempo". But then, "normal tempo" is not defined anywhere in the laws.

I believe there must be a speed for calls and plays which does not transmit any UI. "In tempo" simply refers to this speed.

An action taken faster than "in tempo" can convey UI, of course - even if, as we are occasionally reminded, it is rarely sanctioned. (Perhaps this is what people mean in this context, namely that "normal tempo" is actually quite slow in some situations. In such a case they would be correct, but the matter of the proper wording is still delicate.)

As an analogy, think about the impact of disclosure provided by partner. If we make a call based on our agreements and our partner explains the agreements correctly, it is highly impractical to treat this as UI.

Basically, it would mean that we could not legally assume we are bidding according to our methods, but we would be forced to act on the assumption that partner has misunderstood us in doubtful situations. This would make it impossible to have a sophisticated auction - clearly this cannot be in the best interest of the game.

Likewise, it would not be practical in the least to treat any tempo as UI. If you think about it, this interpretation would allow directors to penalize virtually any bridge action whatsoever.

UI violations are different from most other infractions because, in order to determine that they have happened at all, it is not enough to look at the events that took place at the table. One must also have a close look at events which did not actually take place. Allow me:

It does not take much to see that there is an insufficient bid, a lead out of turn or a revoke. The occurrence of such an irregularity can objectively be verified. A bid is insufficient if the previous bid was higher. A lead is out of turn if, according to the rules, it was another player's turn to lead. You have a revoke if the player in question did not follow suit although he would have been able to. And so on.

But how do you find out that a UI violation has occurred? You study a player's action and compare it with a hypothetical different scenario, i.e. with something that has not happened at all. No, I am not speaking about LAs; this is about the origin of the UI itself.

In an example hand presented by Kit Woolsey about a week ago, a player won partner's lead against a notrump contract with the ace from A10x and slowly returned the 10. It has been argued that the return would have been much faster from an original holding of A109x. At the same time it has been argued that partner might have been thinking about shifting to another suit, irrespective of his holding in the first suit.

Which one is it? Well, there is no "correct" solution, simply because we do not have the alternative scenario present. We cannot switch to an alternate universe and see how long it takes this player to lead from A109x at trick two. We can only guess.

Michael Rosenberg has repeatedly pointed out - and rightly so - that we would not want the player to play fast from A109x next time and see his partner deduce his holding from his tempo. Therefore, in order to protect the opponents, it has been suggested (in all fairness, he concedes that this is a very difficult situation) to rule against the UI side.

But wait, there is no "next time" yet. If this is about hypotheticals, we might just as well suppose the following: Next time he holds A109x but plays equally slow, and his partner draws the wrong conclusion. Obviously, there will be no complaints by the opponents in the second scenario.

Now, if we rule against this partnership in the first case, we actually do so without cause. The "next time" in a sense proves that this partnership has committed no UI violation. The player has taken his action in the same tempo both times, hence there was no transmission of UI. And still we punish him, not for something he did, but for something he did not do.

Basically, the same applies in a bidding example posted by Ian Casselton a few hours ago. To rule against someone following his partner's in-tempo call means penalizing him based on the purely speculative assumption that his partner would have done something else under different circumstances.

To summarize: Based on what would have happened in the alternative scenario, we run the risk of either penalizing a pair for no good reason or let them get away with sin. Simply put, we might either acquit a guilty pair or convict an innocent one. In statistics, one speaks of Type I and Type II errors. By setting the standards for a score adjustment accordingly, one can reduce the one or the other. But one cannot eliminate both errors, as long as uncertainties and statistical methods are in place.

So the obvious question poses itself: Which is more important, protecting (UI-)innocent pairs from wrongful adjustments, or making sure that potentially guilty partnerships cannot benefit from UI? Depending on how the priorities in this regard are defined, the standards for conviction must be set.

Personally, I prefer the protection of partnerships who have not (yet) been found guilty of UI abuse to be the highest priority. However, this should not just be a matter of personal preferences. The lawmakers must be aware of this dilemma. It is my perception that the requirements for score adjustments in UI cases have been increased on purpose (for instance, the word "demonstrably" was not part of older versions of the laws). I would like to think that the lawmakers inserted this word to emphasize the importance of protecting pairs from arbitrary adjustments in UI situations.

By intentionally lowering the standards in this area, by refusing to actually offer logical demonstrations why one action is suggested over another by UI, by deliberately avoiding logic in favor of lazy guidelines ("huddles always show extras"), we are not just circumventing the law itself, but also the intent - to protect players from random UI rulings - behind the law.

One more thing. People often assert how often the partner of the huddler (or in general the player who has received Unauthorized Information) can read the UI. I have very strong doubts about the accuracy of these observations, at least the way they are presented.

My point is, people tend to remember specifically the occasions where such UI readings were correct. Let me present another very simple example. Suppose we have the uncontested bidding 1NT-2NT, with responder's raise being noticeably slow. Opener holds a hand that would normally pass, but following the above rule he raises to 3NT.

If opener is wrong and his partner was thinking about passing, the opponents accept their good score and move on. Very rarely anyone ever draws attention to these things. As a consequence, the cases which end up in the spotlight are usually the ones where the action taken in possession of UI was successful.

But that means, all the arguments based on empirical evidence are essentially useless. It may well be that the UI is misinterpreted as often as understood correctly, and that just nobody hears of it. This is very similar to cheating investigations: If our judgment is supposed to be on statistical data, we must count the negatives as well.

This is why I view the rants "How comes players always know what the UI means?" unconvincing and, as an attempt to satisfy the "demonstrably suggested" standard, completely unacceptable. It is no secret that people are good at seeing what they want to see. We should stop wanting to see successful actions in the face of UI penalized per se.

Getting Comments... loading...

Bottom Home Top