In your equation, as the number of reviews (c) increases, the ranking (r) would go down. (And not in the good way.) so that's why that's not working if you'd want more reviews to lead to a higher score.
I doubt TER is accounting for number of reviews, but possibly. The world may never know.
As xyz said, the new chemistry and incall scores are effecting the ratings, but we don't know how exactly that's being computed.
Both chemistry (call that H) and incall (I) are 1 to five stars, for a total possible of ten. I looked at one profile with both H and I scores and tried to get to the overall rating listed. With just a slight bit of rounding error, it looked like they may be combining H and I as a new X out of ten parameter and then averaging across the three.
That would be [a+b+(H+I)]/3=r
The ()'s around H and I aren't mathematically necessary, but there for clarity.
I can't say this is what TER is doing, but on the one score I checked, it's possible. WAY too small a sample to try to claim it's been reverse engineered.
The incall score is not required, so a provider could have an H but no I, which would mean they're getting screwed over if this theory is correct. (And not in the good way.) In a case like that, maybe they double the H, to try to make up for the missing I. So in the equation above, the (H+I) would become (HH).
Or maybe this is all crap. Who knows?
Also, if you're checking the math and just looking at who's on the list or not, remember what someone here said about minimum number of reviews and at least one review in the last 90 days to even be eligible. I think you'd do better to just stick to the names on the list and try to get them in the right order with your math.
Actually, TBH, I think you'd do better to drop it. Don't get obsessed, just be you.
I'm not sure why you want to know. I think you may need to find a way to be okay with never knowing. Good luck.