[EL] Put very succinctly: (re same birthdates)

[Adam Morse]

No, because you don't care whether there are two people named John Jones
with the same birthdate.  You care whether (and how many) pairs of people
share both the same birthdate and the same name.  The "does there exist?"
question we care about doesn't have a specific name in it; it is:
"Is there any set of two people in this set for which (a) they have the same
birthdate (day and year) and (b) they have the same name?"

The "How many ...?" question we care about is:
"How many sets of people exist in this set for which all members of the set
(a) have the same birthdate and (b) have the same name?"

Part of the point is that you can't do a simplistic combination of these
probabilities.  Your 1:10,000 ratio tells us that each birthdate in the
overall pool will have roughly 1/10,000 times the total number of
birthdates.  So, out of a pool of 600,000 voters, every individual birthdate
has 60 people born on it (on average--there will be peaks and troughs, of
course.)  The question then is how many of those pools of 60 people have
multiple people with the same name in them?  The answer will be surprisingly
large.

Or going the other direction, you could divide the pool of voters up by
names.  If Robert Smith is 1 in a thousand, then a pool of 600 people with
the name Robert Smith will exist.  Because 600>180, you would then predict
that there would be some pair of Robert Smiths with the same birthday.
 Iterate over all names, and you predict numerous duplicates. (1 in a
thousand is probably high, although I'm not sure; Smith as a last name is
about 1 in a 100, and Robert is about 3 in 100, which suggests that Robert
Smith is 3 out of 10,000; but that's simplistic, because there will be some
correlation between first names and last names based on ethnicity, etc.  But
even so, there are then numerous different combinations that are each 3 in
10,000 or 2 in 10,000, and we don't care which name has the duplications).

--Adam

On Sun, Sep 11, 2011 at 2:57 PM, Bev Harris <bev at blackboxvoting.org> wrote:

> The question of whether two specific persons are likely to have same
> birthdate
> is calculated from how many possible birthdates exist, not the number of
> people
> in the room.
>
> The only relevance number of people in the room has is AFTER you have
> calculated
> the probability that two SPECIFIC people would have the same birthdate.
>
> For purposes of identifying duplicate voters the right question to ask is
> "Are
> there two people named John Jones that have the same birthdate?" not "Does
> John
> Jones have the same birthdate as any other voter on the list?"
>
> Right?
>
> Bev Harris
> Founder - Black Box Voting
> http://www.blackboxvoting.org
>
> * * * * *
>
> Government is the servant of the people, and not the master of them. The
> people, in delegating authority, do not give their public servants the
> right
> to decide what is good for the people to know and what is not good for them
> to
> know. We insist on remaining informed so that we may retain control over
> the
> instruments of government we have created.
>
>
>
> ----------------------------------------------------------------
> This message was sent using IMP, the Internet Messaging Program.
>
> _______________________________________________
> Law-election mailing list
> Law-election at department-lists.uci.edu
> http://department-lists.uci.edu/mailman/listinfo/law-election
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://webshare.law.ucla.edu/Listservs/law-election/attachments/20110911/26b37fbe/attachment.html>