I tried the first 84 questions only two days before my Aug 17 GRE, intending to "oil" my mind for the Quant sections. The unofficial score is 170. But I would be lying if I say these questions are not tough. Instead, they are so darn tough and deserve to be categorized as math olympiad for would-be GRE takers.

First of all, an idea for solving majority of those problems I tried are not straightforward at all, i.e., it is so difficult to come up with a potential idea solving the problem within 20 sec (executing the idea and finding the solution is the next step). To be sure, one of the reasons is that I saw them only for the first time and thus more practice can improve it. But still, I think this non-straightforwardness makes them unlikely to occur in real GRE tests (wondering whether they may occur in GMAT though since I never took one).

Maybe I made a somewhat unsubstantiated claim here as I only took 2 GRE tests. So the next is my second thought. Having developed an idea in my mind when solving these questions, the next step is to execute the idea. The more serious issue with these questions is that executing the idea takes rather long especially taking into account the average time of 105 sec for each question in the Quant section. Of course, not every question will be this hard. But the chance of having a question that takes several minutes to solve while the rest take only 1/3 of time or even less is very slim, I think. Not to say, there are full of traps in majority of these problems, which further reduces the likelihood that you will see one of them in a real GRE test.

The above reasons will be irrelevant if there are some super smart tricks that lead you to a right answer within 100 sec and circumvent the long process required by ordinary ideas, which is also unlikely as GRE is not an IQ test. GRE quant tests have two dimensions: 1) accurate understanding of basic concepts and 2) speed of basic calculation. Whether these "100 Hardest Problems" have their place in GRE or not, a reasonable strategy of getting a high score of GRE Quant should create some doubt of including these problems in the training. Instead of cracking one "Hardest" problem using 5 min and losing 3 easy questions due to carelessness, a better strategy would be to practice more with those GRE-likely and unsurprisingly easy questions so that you are faster without sacrificing the accuracy. Well, if you have enough training time and also aim for 170, then it does no serious harm trying these out, even just for fun.