GMAT's luck factor

Question

How much luck affects GMAT scores?

Since I learned that almost a quarter of the questions of the real exam are not scored, I have been wondering how much luck can affect one's score.

Let's think about two guys with same skills, same preparation, same performance under pressure, etc. During the exam one is luckier and almost every question that he fails is not scored. However, the other guy is not that lucky and every question he fails is scored. At the end, one might have scored 600 and the other 750.

I'm aware that every question that it is not scored does not affect to the difficulty of the next. Anyway, it is still not 100% fair in my opinion.

I understand that GMAC needs to "test" its future questions before using them, but I think this fact might make GMAT scores unfair as it incorporate a very big "lucky" factor.

What do you guys think?

rid

bigoyal · Answer

yeah. I agree, if you leave something on chance then surely it would be.

For example, if in quant, both the guys are not prepared well for a particular section, say, probability. And the 1st one comes across several probability questions, whereas 2nd guy doesn't see a single probability question.

So, my concept is not to leave anything. Though not always, but one can surely control luck to a great extent.

bb · Answer

Yes and No

Yes: there is a 40 point variance reported by GMAC (that is how much the score will differ if you take it today and next one tomorrow) No: the test is pretty good about figuring out your weaknesses and strengths and managing the luck factor. I can argue either way 8-)

gurpreetsingh · Answer

But still getting experimental question wrong instead of others is a big luck.

nonameee · Answer

Let's solve a simple math problem:

You are studying for an exam which contains 60 questions. You have covered all the questions except for the last 3 (you've covered 95% of all the questions). On the exam you draw three questions and in order to pass it you have to answer all three questions. What is the probability that you will pass the exam?

Answer = 85%

Just think about it: You have learnt 57 questions and yet the probability that you'll pass the exam is around 85%.

What I'm saying is that in order to do well you have to cover a lot of material and still the probability of scoring very high is not as great as you would've thought.

strikethree · Answer

There is a reason that they don't score the experimental questions; it's not like the questions are randomly designated as experimental or not. Even if you take out the experimental questions before the test, you would get the same score since you would be getting the same "real" questions.

Sure, like any test, there might be a deviation from your true "mean". It could be that this score is an outlier. However, it could also mean that your practice scores might have been too high. (or the practice tests were too easy)

You are given a set of questions. It's not anything like a gamble. You, the test taker, are in full control of what you score. If you get it wrong, then you were either tricked or did not know how to complete the problem. Either way, it would be your fault. It's not like you're picking answer choices out of a hat.

Honestly, I don't see what the problem is: It's not like they only let you take the test once. (Which would actually be unfair) Schools don't even average scores -- they take your highest. If that is truly the case (admissions people say it is, but since I'm not an admissions person, I wouldn't know for certain), then the test taker has the upper hand.

Just let it go and study again. Good luck.

shaselai · Answer

From my practice CATs my range was mostly ~670ish with one ever being 700 and i got 720. I know for sure I wouldve done better if I hadnt faced one of my most disliked passage types - dry finance jargon and I got hit with 1 geometry question from the "weird" bin (really cant prepare for) which I am sure cost me on quant especially making me rush on the last 2 questions (made educated guess).So I do think there's a bit of luck involved on whether we get "our worst topic" drawn from the question bank or not.

strikethree · Answer

The GMAT is a 77-question test. No one question is the deciding factor for your score (or sets of 3 for that matter). You can still get a high score even if you receive a question(s) you are not prepared for; it is much more complicated than your formula. If you wanted to score a perfect 800, then yes, every question will likely matter (do you even need to be 100% correct to get an 800?). However, the question is about receiving a high score -- not necessarily a perfect one.

The GMAT is also multiple choice and does not penalize you for guessing. Therefore, your score would likely reflect a better score because you are allowed to guess. In fact, it would probably be much more "fair" if you weren't given choices at all. (But that won't happen)

Without the GMAT scoring algorithm I can't say for certain, but the chances of getting a very low score that deviates from your true mean after 1 or 2 retakes seem very low.

nonameee · Answer

I'm not saying that my example has any real correlation with the GMAT. The only thing I wanted to say is that if you want to get a good score, you'd better study very hard and cover all the topics. The example was just an illustration of my thoughts. Nothing's more.

strikethree · Answer

Yes, I know. But, you also said "still the probability of scoring very high is not as great as you would've thought." Which I thought supported the argument that the GMAT is "unfair".

Most of it was supposed to be directed to the OP anyway.

nonameee · Answer

No, no, that's not what I meant. GMAT is fair in a sense that it measures your ability to think quickly under limited time constraints. And also, in order to do very well you actually have to study a lot, even more than you think.

Posted from my mobile device

Posted from my mobile device

strikethree · Answer

Yep, must have remembered incorrectly. Thanks for the catch.

Although, minus the experimental questions, there are fewer "real" test questions that count. (I believe the number is about 1/4 to 1/3 experimental)

gmattester123 · Answer

IMO, 40 points is too large a margin and in most cases, will make the difference between an accept or reject - especially for the B-schools that "hunt" for higher GMAT score applicants to boost their overall class GMAT average.

A 10 point variance is not much luck, but 40 points is, especially when the overwhelming majority of GMAT testers are within the 500 to 600 score range.

trunksli · Answer

man, 40 pt variance sounds really high, especially when you get in the uber competitive ranges. One point in section score is a 5% difference in people you score higher than. I don't like how one three digit number is consuming my life right now, and there's a 40pt variance.

gmattester123 · Answer

Perhaps bb can confirm this with official statistics from GMAC, but I would hazard to guess that over 60% of students that write the GMAT obtain a score between 500 to 700 (60% is a conservative guess too). If this is true, that means the majority of students are "squished" within a 200 point range and within that 200 point range, there is a 40 point variance!

mission2009 · Answer

Nice evaluation by one n all, but the luck factor will be present in each and every test !

Warlock007 · Answer

Dear BB i would like to know how does it manager the LUCK factor.......

deadlycat · Answer

Luck is definitely a factor but because the GMAT is a CAT, it's much less of a factor than you think. Think of it this way, one guys's true ability level is 600, if he get question higher level than that, he usually get it wrong, if he get question lower than that, he usually get right.

Now assuming our guy got lucky and correctly guessed one or two questions that he's supposed to get wrong. Well the next question will be hard, and the next one after that and possibly the one after that. For as long as his lucky score is higher than his true score, he'll constantly be bombarded with questions that are out of his league. In the long run, his score will drift back to the true value. Of course the GMAT has finite number of questions so luck remains a factor, but it's much less than the usual paper based tests such as the LSAT.

Also, the scenario you described is not a proper comparison. You'll can't compare random experimental questions to level corrected real questions, that's like comparing apples and oranges. You'll need to compare the real questions on both of them against each other, but again scores will tend to drift back to their true mean even with some good or bad luck during the test.

GMATPill · Answer

Luck is indeed very important. Some people perform well on the actual GMAT test. Some tighten up and choke.

One way to control luck in your favor is to practice GMAT questions with the right psychological mind set. A lot of students study the GMAT with lots of distractions and don't focus properly. Maybe a question here and there--and then a dinner break.

What you may not realize is that the GMAT is a mental marathon---you need to stay focused for a long period of time. If your study environment does not mimick the conditions of the actual test---you may end up feeling "unlucky" and choke on the exam.

Do yourself a favor and pretend your study sessions are like the actual GMAT questions/conditions. Visualize yourself for success and you'll be in a better position to capitalize on the "LUCK" factor.

You can read more about  GMAT Visualization and Psychology Strategy  here.

Warlock007 · Answer

Hey GMATPill 
thats a nice post indeed
but i have some queries
1 probably this varied result may be attributed to the un-scored questions
2 how to get rid of the worst effect of these 10 un-scored questions (which is greater than 40 points acc to GMAC)

	Manhattan Prep: $99 Bonus Score Higher on Your GMAT Exam. Guaranteed.
	Target Test Prep - Save up to $599 $400 Discount + $99 Bonus GMAT Club Tests
	e-GMAT: 60% Off on GMAT Focus Get to 735+ from any starting point
	Magoosh increases your score, guaranteed - 15% off ALL Magoosh GMAT plans with code GMATCLUB15 + Free GMAT Club Tests.
	Experts' Global - ~~$1720~~ $960 End-to-End GMAT Prep + App Support: 150+ Videos, 2000+ Questions, 15 FLTs +
	GMATWhiz - $299 ~~$499~~ Experience the course through a free trial.
	GMAT Club Tests - Free Included with every course purchase of $149+ View all deals

GMAT's luck factor

Prep Toolkit

Top 5 GMAT Debrief Videos

615 to 715 in 15 Days (Ria's Strategies)

More than Quant/Verbal Abilities: Speed vs. Accuracy

745 with Only Self-Prep and GMAT Club

From 555 to 765: 210-point Improvement

Perfect 805 Score Debrief by Julia

My Rewards