ThatDudeKnows wrote:
I think it's a common misconception that ballparking isn't effective on official questions or that it isn't effective on more difficult official questions. I'll do a thorough breakdown on non-geometry questions at some point, but since this happens to be a geometry question, check THIS out!!
If you reread what I wrote, I think you'll recognize that you're mischaracterizing what I said. I've studied the effectiveness of estimation (and of every other 'strategy' recommended in prep books) using large pools of official questions. And it turns out estimation (combined with a bit of logic) is, by far, the most successful guessing strategy on the GMAT. In fact, for PS, it's really the only effective guessing strategy. I've taken repeated official practice tests where
I rely only on logic and estimation (and a random number generator when I need to guess) for PS, and on a guessing strategy for DS I won't describe here, and without solving anything I can score in the Q40-Q44 range consistently. I'm defining 'estimation' a bit loosely, so I'm including basic conceptual reasoning (e.g. in a weighted average situation, I'm using the fact that we know the overall average is closer to the larger group's average). That's applying those conceptual principles and estimates perfectly, so it's not the score someone should expect applying those principles imperfectly, but it's still very effective.
The problem is, I can't get beyond that level. And there's a reason for that: if an estimation strategy works on, say, this hexagon question in this thread (and the hexagon is approximately the circle, so estimation clearly will get you very close), every high level test taker will see that, even the high level test takers who can't see how to 'do the math'. So every high-level test taker will get the question right, one way or another. And that means the question cannot be a high-level question, because that's how they determine question difficulty: they administer questions as experimental questions, and see how well test takers at each ability level answer them. If every high-level test taker gets a certain question right, it is essentially by definition not a high-level question. So these strategies do start to fail the harder questions get.
To me the best practice is this: if solving seems annoying or time-consuming, glance at answer choices. If they're far apart, estimate liberally. If they're close together, you can't be sure, after estimating, that you'll be able to identify the right answer, so use it as a fallback only, if you can't find a better way to answer. And get good at math (by which I mostly mean: get good at conceptual reasoning), because that's what the hardest GMAT questions are really testing.
Anyway, I just wanted anyone reading this discussion to have two points of view, from two people who have thought about these questions a lot.
Not to be pedantic, but I'm certainly not suggesting relying "only on logic and estimation." In fact, I noted that logic, estimation, and using the answer choices work on 10 out of 17 medium official questions and 7 out of 9 hard official questions. The implication is that you'll need to do real math on the others, but 16 out of 26 is
, so it's a heck of a lot better than "never," it certainly applies to more than just "lower level geometry questions," and it is without a doubt the single most useful thing you can learn in geometry (more valuable than any content-based studying). When you CAN use those tools, they allow you to answer questions more quickly (so you have more time for questions that do require real math), cause far fewer silly mistakes (which directly improves your score), and avoid the stress and brain damage that happens when you're crunching numbers and find yourself stuck (which helps you stay on track mentally through the rest of exam, indirectly improving your score).
close
08 Jun 2020, 02:38
15%
(02:21)
wrong 
