8 ms·
Classic agent-based modelling fallacy: "I made an algorithm that replicates an observed distribution, so my description of the algorithm is a potential explanat
by evrydayhustling 9y ago
Classic agent-based modelling fallacy: "I made an algorithm that replicates an observed distribution, so my description of the algorithm is a potential explanation for that distribution.". The fact that you could relabel the relevant variables any which way (the author does, by equivocating status and technology) and product the same outcome should illustrate what a weak form of "evidence" this is.
A better title would be "simple rules to generate asymmetric opportunities", but I guess phrasing this in a way that tells guys the world is unfairly stacked against them gets more internet points.
- Yajirobe 9y ago> "I made an algorithm that replicates an observed distribution, so my description of the algorithm is a potential explanation for that distribution." I mean, technically it is a potential explanation, just not necessarily a correct one.
- niyikiza 9y agoAbsence of evidence actually IS (an albeit weak) evidence of absence. That kind of stuff.
- paultopia 9y agoYeah. This is a classic case of what happens when someone who is good at math but has, how shall I put this gently, zero domain knowledge, gets their paws into something.