[Dshield] Undersea cable cuts

Pete Cap peteoutside at yahoo.com
Thu Apr 17 19:53:01 GMT 2008

I wasn't trying to downplay John's post.
I just would really like to see some actual statistical analysis applied to this sort of thing.

I think that, globally, we have a cable break once every three days, so the odds of getting 5 breaks in 12 days is near unity.  They would probably be different if we accounted for the geographical clustering, but I doubt it would be very much different.
This means, for the conspiratorially minded, that it's really hard to tell an attack on the cables from a random event.  But, we could establish some kind of threshhold by asking how many cuts in how many hours in how small an area we'd need to observe to begin to think it was suspicious--a metric like "4 cuts among cables servicing the UK in 4 hours might be suspicious."
We'd need to survey the cables from public data.  The routes are pretty well known.  This is why I asked if anyone wanted to help out.  Any takers?

Best regards,

Be a better friend, newshound, and 
know-it-all with Yahoo! Mobile.  Try it now.  http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ

More information about the list mailing list