>>44
That's (almost) equally unlikely.
If we just take the first 8 characters of the tripcode and feed that back into the hash function:
aaaabbbb -> zzzzcccc(xxx)
zzzzcccc -> aaaabbbb(xxx)
Equivalent to attempting to find
x = f(f(x)) for some
x. If
x = f(x) doesn't exist for any
x, then it's possible that neither does
x = f(f(x)), because the set of all 8 char inputs
X is equivalent to the map
f(X) -> Y, just in different order.
Taking a simple ``hash'' function with no collisions:
f(x) = x+1%100 bounded on 0 <=
x < 100, then there is no
x that satisfies
f(x) = x. Similarly, there is no
f(f(x)) = x, due to the property of the function, you have to do 100 cycles before the input can equal the output. Which is equivalent, in this case, to feeding the entire set of its inputs to the function, until you get back to where you started.
So even in the DES hash, it's possible you'd have to perform 64
8 = 281,474,976,710,656 cycles before you can satisfy the equation.
The chance is still absurdly low. It gets higher and higher the more cycles you use, but the rate at which the chance goes up is
dependent on how many collisions the function has.