[pran]

shyams, on Feb 25 2010, 09:49 PM, said:

StevenG, on Feb 25 2010, 12:54 PM, said:

Where do the hands come from? from the ECats website may be of interest.

I remembered this; and had hunted for the information on the ECats website earlier once. I'm quite curious to know whether this is really legal.

Also, though the stated aim is to have nearly the same # of HCPs etc, I think it is meant to work only across the entire set of 32 or 36 boards. Most clubs would only have the first 24 or 26 boards in play -- rendering the HCP balancing exercise a bit futile.

Of course it isn't legal: It violates the fundamental criterion for deals being random.

[pran]

Vampyr, on Feb 25 2010, 09:25 PM, said:

pran, on Feb 25 2010, 07:02 PM, said:

StevenG, on Feb 25 2010, 01:54 PM, said:

Where do the hands come from? from the ECats website may be of interest.

Indeed it is. According to this the deals have never been played before so it is not a replay. However the deals have been manipulated in that some of them have been filtered out (not to be played) invalidating the dealing process.

And there is the problem that came up in another thread -- each player is given approximately the same number of HCP (or the same number of good hands) so that predictions may be made at the end of a session. This is just awful.

That is precisely the point.

Take a fundamental example: You throw a (fair) coin ten times and get "head" all ten times. What is the probability for getting "tail" on the next throw?

Answer: Exactly 50%

Slogan: "A coin has no memory".

If you can tell the probability for a particular outcome of a process based on knowledge of previous outcomes of this process then the process is not random.

If a set of deals has been filtered to produce a particular statistic, for instance giving each direction approcimately the same total HCP then this set of deals is no longer random.

[mycroft]

Or, as I said when asked that question a week ago, "slightly more than 50%". That "slightly more" being the chance that the coin is weighted (such that p(10H) > 2^-10), or that the guy is flipping a double-headed coin (non-zero)...

But Sven put the "fair" in quotes to avoid that. I'm not always fair.

[pran]

mycroft, on Feb 26 2010, 01:18 AM, said:

Or, as I said when asked that question a week ago, "slightly more than 50%".  That "slightly more" being the chance that the coin is weighted (such that p(10H) > 2^-10), or that the guy is flipping a double-headed coin (non-zero)...

But Sven put the "fair" in quotes to avoid that.  I'm not always fair.

I suspect that you misread the question? If the coin is not "fair" and you throw ten heads in a row the probability of obtaining tail on the next throw is far less than 50%, not slightly more.

[mycroft]

pran, on Feb 26 2010, 01:59 AM, said:

I suspect that you misread the question? If the coin is not "fair" and you throw ten heads in a row the probability of obtaining tail on the next throw is far less than 50%, not slightly more. :rolleyes:  :unsure:

of course I did :-)