Your understanding is incorrect. But I read How to Lie With Statistics too, so I'm happy to help you with more sources for your perusal. Keep in mind that merely invoking How to Lie With Statistics does not refute a statistical analysis, you actually have to identify the problem.

1. This document goes into much greater depth about the statistics you're criticizing: http://maamodt.asp.radford.edu/Serial%20Killer%20Information..., including a coverage of serial killers with 2 or more versus 3 or more victims.

2. The FBI's profile of serial killers, from their behavioral science unit, in which they explicitly call out the popular conception of high IQ serial killers as being a myth: https://www.fbi.gov/stats-services/publications/serial-murde...

3. This study conducts a meta analysis of published literature that finds no correlation between verbal intelligence and traits traditionally associated with psychopathy: https://www.ncbi.nlm.nih.gov/m/pubmed/28836093/ (also on bioRxiv if you'd like to read the non-paywalled version).

4. This study finds that, for a set of intelligence metrics and psychopathy checklist traits, psychopathy and intelligence are orthogonal: http://journals.sagepub.com/doi/abs/10.1177/0093854807309806

5. For good measure, since you like popular media like true crime TV, I'll throw in a Scientific American article for you: https://www.scientificamerican.com/article/5-myths-about-ser...

> But I'm really not finding your source compelling. So my mind has not been changed.

I suppose I should be transparent with you: I don't debate with people online to persuade them, I debate with people online to persuade the audience. I don't particularly mind if you're convinced, but it is important to me that misinformation not be spread unchallenged. So now we're at a crossroads. I've provided empirical evidence from a variety of cited sources, and you've provided an anecdote from an undergraduate course you took years ago. My sources might be incorrect! The beauty of science is that you can challenge its findings. I welcome you to challenge the sources I've provided as well (though I'd remind you that you've continually abstained from providing any of your own). But if you don't, that's also totally fine, because I think I've sufficiently demonstrated my point either way. Considering your lukewarm dismissal of an academic dataset based on a disagreement with one particular data point out of 206, it appears true crime television and your recollection of an old psych class are better authorities than published research.