Urgent: Strategy solicited; GMAT on 14th March

March 14? That's pi day (3/14)! It's your lucky day, I can feel it.

First of all, your #1 job on quant is to get rid of the careless errors. Four or five or seven dumb mistakes on easy questions will absolutely torpedo your score. You can study all of the advanced math questions you want, but if you make silly mistakes, the algorithm will kick you in the nuts over and over again until you're writhing on the floor in agony. It's the nature of adaptive tests.

On every question you do--whether in practice or on a test--focus on reading the question twice before you start scrawling numbers on your page. Sure, it takes an extra few seconds, but you absolutely can't afford to miss the little modifying phrases that change the question ("x is positive" or "x is a two-digit integer", etc.), and you definitely can't afford to make small transcription errors. As you're completing the question, double-check any simple algebra and arithmetic if you have a tendency to make dumb errors in those areas.

And before you click "next", re-read the question again, and ask yourself: "How are they trying to screw me?" By now, you've done enough studying that you've seen nearly every trap that the GMAT throws at you. You're making bad errors because you're letting the clock get in your head, and you're not taking the extra few seconds that you desperately need to spend to avoid those errors.

As you practice, focus on developing a good rhythm: read the question twice, check your work twice, read the question again, and look for traps and dumb errors before you move on. Be absolutely religious about this. On every question. Without exception. Yes, it takes a few extra seconds, but you can't afford NOT to spend those seconds.

Also, look carefully through your practice tests, and identify questions that took you more than three minutes. Are you being stubborn on the harder questions? Maybe you need to focus on letting go of those questions quickly. If you're overmatched by a question, don't waste your time on it. You can miss tons of questions and do extremely well on the quant section, but you can't afford to put yourself in a huge timing hole, which then causes you to rush through questions and make more dumb mistakes.

So seriously: focus on developing a good rhythm to reinforce your accuracy, and then focus on consciously walking away from questions when they give you trouble. If I could wave a magic wand and make you 1) never make errors on questions that you understand, and 2) never waste time on questions you don't understand, then you would absolutely score at least a 47 on your next test, and possibly a 48 or 49. I would bet my life on that if I had to, Sachin.

On verbal, the unglamorous reality is that you're probably a slow reader, and there isn't a whole lot you can do to improve your reading speed in the short run. Some native speakers also struggle with reading speed, but non-native speakers are more likely to have a hard time completing 41 questions in 75 minutes. And no, it's not fair.

When your test is a week away, all you can really do is focus as much as you can on efficiency (especially on SC, since you can save at least some time by quickly eliminating certain answer choices if you're quick to recognize the GMAT's favorite SC rules), and try not to let the timing issues rush you into bad errors. The verbal algorithm fundamentally works the same way as its quant counterpart: dumb errors on easy questions can really kill your score, and if you do well enough on the first, say, 35 questions, you can afford to screw up on the last six without completely wrecking your score.

Whatever you do, concentrate on maintaining accuracy early in the test. You don't want to make a dumb error because you're rushing through question #5, just for the privilege of taking a look at question #36 or #41. That's always a terrible tradeoff on an adaptive test. My usual advice is to do the verbal questions accurately and methodically until you start to run out of time toward the end of the section, and then guess at the very end if you have to. All else being equal, I just don't think it makes sense to skip question #21, either--in theory, a mistake earlier in the test will hurt you more than a guess later in the test. You might get lucky and guess on experimental questions in the middle of the test, but that seems a little bit dangerous, since we don't know exactly how many experimental questions are on the test, and you'll need some luck if you want to miss the experimental questions only.

Then again, if you're substantially worse (in terms of both speed and accuracy) at, say, CR questions, then maybe you could start skipping them a little bit earlier. That would make some sense, if you're confident that you're stronger on the other question types. But in all honestly, I would pour as much energy as possible into becoming more efficient on SC at this stage--I think that's probably more likely to help your score than any particular guessing strategy. Lucky guesses definitely help, though.

Good luck, Sachin!