# Lesson: The GMAT Scoring Algorithm

## Comment on The GMAT Scoring Algorithm

### Doesn't matte the type of

Doesn't matter the type of question, we should go over and answer all of them without considering the level.

### Well, you must answer each

Well, you must answer each question if you want to see the next question (that is, you can't skip questions on the GMAT). That said, I'm not saying that you should never guess the answer and move on. In some cases, that's the best course of action if you want to maximize your score.

### Thank you very much for your

Thank you very much for your information, which is quite different to what I've been told.
I do have 1 small question:
Since, according to your video, my score will be the average difficulty of the correct answers regardless of the number of the correct answers, can I trick the system like this: I'll try and invest most of my time in the 1st 30 questions so that my score will soar up to 700+, and then intentionally answer incorrectly the rest of the test --> I'll still remain at 700+ level? (as it only counts the difficulty of CORRECT answer)
many thanks!

### You're not the first person

You're not the first person to try to unlock the secrets of the GMAT's scoring algorithm :-)
One thing to keep in mind is that you are heavily penalized for having long strings of incorrect answers. So, if you were to devote most of your time answering the first 30 questions, you will undoubtedly have several incorrect responses at the end, which could undermine your strategy.
That said, I encourage you to try this strategy on some official practice tests.

### Ha ha Thank you for your

Ha ha Thank you for your explanation. I came up with this "strategy" right after viewing this video. when I watched other videos on this site, I know that I'll have to pay for 6 wrong answers in a row.
Great work, btw. I don't believe that someone is willing to offer something this good for free!

### This course helping me a lot.

This course helping me a lot. I am really delighted. Thank you.

### Thanks for taking the time to

Thanks for taking the time to say that!

### Again (I have previously

Again (I have previously posted a comment like this on a YouTube video). But with my lack of math skills (from screwing around in high school), this prep program combined with Veritas Prep is finally working for me. Thank you so much, you seem to deeply understand how to teach and you are using that for the greater good. Thanks again and have a great day!

### That's very nice of you to

That's very nice of you to say. Thanks!

### But in the beginning it has

But in the beginning it has higher score gap between the questions, so if i do the first questions right it will increase more than the decrease in the middle or in the end if i do some questions wrong due the waste of time.

### The test-makers say that the

The test-makers say that the first 10 questions are not more weighted that the other questions.

### I have just took a couple of

I have just took a couple of practice tests from Veritas, Manhattan Prep and GMAC. My scores have got a bit of a discrepancy between them and I am starting to freak out.

On Veritas, I'm averaging 650 (4/7 tests) on GMAC 590 and Manhattan Prep 600

There seems to be a huge discrepancy in the kind of questions I have been getting as well. I don't know what to do anymore. Any advice would be appreciated.

### The official GMATPrep tests

The official GMATPrep tests are the most accurate tests. They use official (retired) GMAT questions and the official GMAT scoring algorithm. Also keep in mind that the test-makers (GMAC) collect a lot of data on every question, and every question has been asked in a formal testing environment, so that the well-prepared test-takers are trying their very best to answer every question.

While many test-prep companies (e.g., Veritas & Manhattan Prep) have good practice tests, those companies would never suggest that their tests are as accurate as the official practice tests. The 3rd-party tests don't use retired GMAT questions, they don't use the official scoring algorithm, and the data collected on each question cannot be as accurate, because the students taking the tests aren't doing so in a formal testing environment.

So, while 3rd-party tests are great for practicing your test-taking skills, and for identifying weaknesses, you shouldn't use them to predict your score on test day.

I hope that helps.

Cheers,
Brent

### I am very confused. I scored

I am very confused. I scored a 460 in my diagnostic test of the free official GMAC tests,then after a lot of practice a score of 500 (The Economist free test) even though I felt positive while doing the test, next I attempted the free-kaplan test just after a few days and scored 630 even though I had a feeling while doing the test that I am doing pathetic....was shocked to see a 630 score...... I am confused about where I actually stand...630 is a good score for me... but I am not sure if I would get this score in the real test...any ideas how I should go ahead?

### The third-party practice

The third-party practice tests (from Kaplan etc) are not very accurate. Please read my comments above this post (from February 22).

I suggest that you take another OFFICIAL practice test. This will give you a much better idea of your progress to date.

Cheers,
Brent

### 1. If my target score is 650

1. If my target score is 650, around how many difficult questions must I answer correctly for Quant and Verbal?

2. How does GMAT rate the difficulty level of a question? Would it be possible for me to get a question that GMAT considers difficult for me, and the same question is rated medium for other test-takers?

Thanks.

### 1. Your scores are not based

1. Your scores are not based on how many questions you answer correctly; they're based on the level of difficulty of the questions you answer correctly/incorrectly.

To illustrate this, I'll share an experiment I performed: I took GMATPrep Practice Test #1 four times, and each time I answered every second question correctly (I did this for the quant section only)
Given that I correctly answered exactly half of the questions each time, you'd expect my quant scores to be roughly the same for all 4 tests.
My 4 scaled scores were: 19, 23, 26 and 42
This represents a percentile range from approximately 8th percentile to the 63rd percentile.

In general, students who score between 450 and 650 generally get about 40% of the questions wrong. This is the nature of the GMAT scoring algorithm.
The score YOU receive will depend on the level of difficulty of the questions you answer correctly/incorrectly.

2. The questions you see on test day will have been answered thousands and thousands of times by other test-takers. Given this, it's easy for the test-makers to assign a precise level of difficulty for each question. Since all test-takers have different strengths and weaknesses, it's very likely that each person will perceive difficulty levels differently.

Cheers,
Brent