eton123
Disclaimer: I don't know if someone has already posted something like this but if so feel free to delete this.
So I took the GMATs for the 3rd time yesterday and ended up with the same score of 700 so I'm going to take a break for a while but because I got the same score I figured I would try and figure out how much each subsection of verbal is worth in case that is beneficial to anyone. My official GMAT scores are the following (and for each score I bought the ESR which is where my subsection scores are coming from).
GMAT 1: 700 (45Q 40V)
GMAT 2: 680 (47Q 36V)
GMAT 3: 700 (46Q 40V)
On my 1st GMAT (based on the ESR) I received CR 37, RC 51, SC 38
On my 3nd GMAT (based on the ESR) I received CR 51, RC 39, SC 35
(If only I got could a CR 51 and RC 51 on the same exam

)
I set them equal to each other with CR = X, RC = Y, SC = Z = 1-X-Y
So:
37X + 51Y + 38(1-X-Y) = 51X + 39Y + 35(1-X-Y)
solved for Y and got Y = (17/9)X -(1/3)
plugged everything back in and I got:
CR is worth (57/212) or 26.9% of Verbal
RC is worth (37/212) or 17.5% of Verbal
SC is worth (118/212) or 55.6% of VerbalI was definitely surprised how much SC was worth.
Anyway, hope this helps for anyone who was curious.
Firstly, 700 is a
really good score, and I wish you all the best for the admissions process.
A couple of points for this post:
1. I see that you've set up 2 equations to solve for X, Y, and Z. Did you add any other information to this?
2. If you didn't, you'll notice that (CR 0.200, RC 0.044, and SC 0.756), (CR 0.400, RC 0.422, and SC 0.178), (CR 0.340, RC 0.309, and SC 0.351)... all satisfy the equations you've set up. In fact, some (of the infinitely many possible) combinations will lead to negative weights!
The math is not that important though.
The most important point here is that the basic assumption that the final score is derived from the subscores is not correct. The GMAT does not even look at question type when it scores a test. Making this approach even more unreliable is the fact that the subscores are included just to help test takers with basic analysis and are not very reliable at all. And to really complicate things, scaled scores usually indicate a range of ability estimates. That is, the first V40 and the second V40 might not indicate the same level of performance

Take a look at
this post. It addresses a very similar analysis on the weighting of subscores, though in that case the recommendation (to prioritize RC) actually came from a test prep company.