It's weird.
By law (? statute? regulation? rule? whatever), Texas has defined 70% as the score required to pass. But Texas doesn't get to determine the passing score. NCEES determines the passing score (called the cut score), and it's different for every exam and different every time the exam is administered.
NCEES determines the cut score by having a committee of existing PEs weigh the difficulty of each question and assess what percentage of PEs should reasonably be able to get that question right. It's a whole process. But the takeaway for this discussion is that the passing score is different for every exam (Civil: WRE, vs Civil: Construction, vs Mech: HVAC, etc). And it changes every time the exam is administered (Civil: WRE April 2019 is different than Civil: WRE Oct 2019, etc). Also we don't get to know what the cut score is.
So NCEES says the (unknown) cut score is passing, and Texas says you need 70% to pass. That doesn't match up. So what Texas does, is they arbitrarily define whatever the cut score is for each exam as 70%. That part is understood. But how the percentages change as you move away from 70% is not understood. From what I've heard people say, it doesn't appear to be linear. I assume it's some statistical distribution, but who knows.
Here's an example: In the "Cut Score" thread someone who didn't pass posted that their diagnostics from NCEES said they got 46/80, which is 57.5%. Their Texas provided result said they got a score of 69%. We know 70% is passing, so they were probably pretty close, but because it's weird and nonlinear, we still don't know exactly what the cut score is. (That being said, people like to speculate and to try and figure it out.)
Thank you. This somehow makes some sense. Basically, our exam answer sheet---->Black box 1 (by NCEES)---->Black box 2 (by state boards)----->result! The better you prepare the higher the chance of passing!