Receiving feedback is invaluable, but it loses its worth when reviews are merely spam for credits.

I understand how important it is to ensure the authenticity of the process, so I'd like to be transparent in how I detect and deal with fakes without being so specific it makes it easier for people to game the system.

No system is perfect and you may be able to get around it if you want to spam reviews or bot. Please don't. It hurts everyone and I'd rather not have to ban you.

Detection

1. Time spent per review. I track how long each review takes and how long a reviewer spends looking at each picture in a profile (or even if they check all the images). Reviews under a certain time threshold won't earn credit or affect the reviewed profile.
2. Profile Score Deviation. When a user gives a score for a profile, I want to measure how different it is from the average score that profile has received. To do this, I calculate the squared difference between the user's score and the average score, and then take the average of these squared differences. I further add weights to so that profiles with fewer scores have less impact on the overall measure.
3. Profile Score Distribution. If someone spams "5" over and over it's pretty obvious it's fake. They may try to mix it up and spam different random numbers, but people are actually quite bad at faking randomness. I don't want to go into too much detail in this section since that could help people evade detection.