Sinij wrote: "My argument is:

(P1) Some people believe the elections was stolen
(P2) There is no evidence of fraud
----------
(C) There people are delusional "

- There is considerable body of evidence that fraud is possible to pull off. Enough to induce a plausible doubt in the validity of the result.


Sinij wrote: "Additionally, as I disucessed with Derid, such fraud must be very substantial (5 states!) in order to have an effect."

- One voting machine, 1000 voting machines, one tallying server 5 tallying servers - if the same vulnerabilities are present combined with the same lack of security protocol - what difference would it make?

Analogy: The Oracle JVM currently has a large security flaw. If you have a tool to exploit that flaw, is using that tool on 5 different machines that have a JVM installed particularly more difficult than using it on one? The inverse is actually true, when talking about systems security (and by this I include any process not just software systems) security is actually asymmetric. Given a widespread vulnerability, it actually becomes more difficult for the defender because the resources that must be expended by the defender to detect and/or intercept an attack attempt are vastly higher than the resources that must be spend by an attacker to make an attempt.

The 300k number was the smallest known spread with readily available numbers, and reflected just one hypothetical scenario. In reality the vulnerable surface area was much, much larger.



For who could be free when every other man's humour might domineer over him? - John Locke (2nd Treatise, sect 57)