[pran]

If A is a subset of B then every occurrence that satisfies A will also satisfy B while some occurrences that satisfy B will not satisfy A.

Now If A promises two denominations, at least one of which is specified, while B promises one specified denomination (with no reference to any other denomination) then A is a subset of B.

(Unless of course, if B denies the existence of a second denomination)

[gordontd]

Tramticket, on 2019-February-04, 04:18, said:

It seems to me that a strict interpretation of the subset test would render the test almost pointless since it will be rarely, if ever, fulfilled. But a better interpretation is to not consider the test as separate and use the concepts of "subset" and "similar" as informing each other.

I think you are right, both as to what the words actually mean and as to how we are being told to interpret them in practice.

[Tramticket]

gordontd, on 2019-February-04, 07:26, said:

I think you are right, both as to what the words actually mean and as to how we are being told to interpret them in practice.

Thanks

[blackshoe]

Tramticket, on 2019-February-04, 04:18, said:

It would seem to me that these three tests are not to be considered as stand-alone independent tests, but the concept of "subset" is used to inform our interpretation of "same or a similar meaning".

I don't know why you would think that. It seems clear to me that they are independent tests.

[barmar]

blackshoe, on 2019-February-04, 09:24, said:

I don't know why you would think that. It seems clear to me that they are independent tests.

I agree. The word "or" between the tests seems pretty clear.

[Vampyr]

pran, on 2019-January-23, 18:14, said:

They have changed their course in recognition that TD skills today are significantly improved from what you could expect some 40 years ago.

This just demonstrates that they haven’t visited a club in the past 40 years.

[Tramticket]

blackshoe, on 2019-February-04, 09:24, said:

I don't know why you would think that. It seems clear to me that they are independent tests.

Ok, look at the two examples (6) and (7) on page 8 of the Tom Kooijman commentary. Neither example meets the subset test (the point counts will generally be different), but I would argue that the symmetry in the two examples means that they must have the same level of "similarity" if we ignore the subset test. So why is example (6) considered a Comparable Call? As Gordon notes, "Hands with spades and a minor are a subset of hands with spades, not the other way around." This makes sense intuitively: we don't want players to be able to provide additional information to partner by making an insufficient bid or call out of rotation first.

So I would conclude that the reason (6) is a Comparable Call is because the 2♠ over-call is an "imperfect subset" of the 2♠ opening, which means that the over-call can be considered "similar" because little significant additional information is given by the call out of rotation. By contrast, (7) is not close to a subset and we should not consider this over-call as "similar" because there is significant extra information provided by the call out of rotation (the presence of a second suit).