In other words, brute force using the sort of mnemonics and tricks that people are advised to use to help them remember a "strong" password. Password strength is a spectrum, and there's a fair difference between "strong enough to make a random criminal go find an easier target" and "strong enough to keep out a determined, funded attacker targeting you personally".
Or it's possible that a lot of these suspects just used really crappy passwords. People are silly that way.
I can't find the source either, but this slashdot post has more details http://it.slashdot.org/comments.pl?sid=1449178&cid=30152...
Most passwords out there are very insecure because they use words that look, sound or are composed of words from dictionaries, that increases the likelihood of cracking the password quite a bit. They can essentially split the load, let some machine brute force sequentially and others perform a dictionary attack.
Even better they should generate a custom dictionary by dumping the strings from any non-encrypted data from the same user. That is what I would do at least.
Pedophiles don't tend to be the sharpest tools in the shed so I'm guessing dictionary words do the trick 90% of the time.
When in doubt make your password something like j3vm#sdq-oj3ew7!d. I call it an "8 by 8" and have committed myself to memorizing a bunch of them for everyday use.
You know, users in general don't tend to do very well at picking passwords. The great success story of educating users on password strength is that, after years of telling people to use both numbers and letters, the most popular password went from being "password" to "password1".
Not to mention that it's probably unwise to assume that criminals of any sort "don't tend to be the sharpest tools in the shed". Maybe it's true, maybe not, but underestimating an adversary's competence is asking for trouble.
It feels good to say mean things about our enemies, but I find it implausible that someone's credibly found a link between pedophilia and low intelligence. (Presumably people don't admit to being pedophiles, so you can't find a random sample of them to study. You can study pedophiles who get caught, but presumably those are the least intelligent pedophiles, so you can't draw conclusions about pedophile intelligence from them. I think it's safer to assume pedophiles have a normal distribution of intelligence.)