[Beowulf] OT: public random numbers?

[David Mathog]

> All nice suggestions, but I wonder if they're truly random.

Random enough in this case - as they are only used to form a seed for a
random number generator, and a seed is only needed "rarely".  So even
though pro basketball scores have definite trends and often look like
(101,95),(103,87),(98,76), these can still create a decent seed value
once sorted and concatenated:  10310198958776
(Lets assume the seed need not be odd.)

> What about using random.org and it being backed-up on archive.org? 
Does that give you the "multiple independent sites" desired?

To some degree, but not as much as the large number of sites that
distribute game scores and stock values.  I originally favored using
stock values until it turned out that those numbers are squishier than
one might have expected, particularly so for indices like the S&P 500
and Dow Jones.  A fellow who works at S&P told me that the opening
prices are prone to timing problems, since at T=0+delta some of the
issues in the index will have traded, and some will not, with the
untraded stock values being filled in with stale values.  I think
similar timing issues affect all the other index values too
(high/low/close).  In these cases, since the index is derived from
formulas, some sites may be independently calculating the values, and
tiny differences in the times the stock values are measured result in
different numbers.  All it takes is one trade difference between the
sample points to change some digits.  When I get some time I still need
to look and see if the high/low/close values for individual stocks are
also variable from web site to web site.  These numbers might be more
reliable for single stocks since they might all trace back to the data
feed from the exchange where the issue trades.

Thanks,

David Mathog
mathog at caltech.edu
Manager, Sequence Analysis Facility, Biology Division, Caltech