If I get lesser rank but higher GATE SCORE IN 2019 GATE compared to this year 2020 then what should i put in MTECH admission while filling up the form . Because in maximum, IIT form , they are having only one row to fill up the valid gate score and rank

1) if 1 lakh students took the GATE in 2015 and 2 lakh students take the GATE in 2016 then are the ranks across both years comparable? or is the percentile comparable?
Think about it for a second…. if we assume that the average distribution of intelligence/skills of participants is constant across both years then the percentile score is comparable whereas both rank and total score are not
2) if we assume that the test is absolute/perfect and always of constant difficulty then the absolute score is comparable. (faulty assumption, anyone who knows anything about standardized testing will understand the problem here)
3) if we assume that the top 2000 people who take are from one planet and the others are from another planet (different level of intelligence) then the rank is comparable. THIS IS A VERY UNREASONABLE ASSUMPTION…….

So in practical terms:
GATE result is valid for 3 years now
lets suppose in 2014 cs gate was taken by 1.5 lakhs
in 2015 , taken by 1.2 lakhs
in 2016 lets say its taken by 2.5 lakhs.
lets say someone scored rank 1000 in 2014 vs another guy rank 2014 in 2015 and another guy rank 1000 in 2016, who is more deserving???
you might say well it depends on his absolute score? nope because absolute scores can only be compared if the degree of difficulty across all years was exactly the same.
you might say rank 1000 across all three years is the same? well nope because its not a reasonable assumption from a statistical point of view. In fact in years in which the number of GATE takers was low it would most likely be easier to get a better rank.
Let me convert it into simpler terms.
You go to a local tournament in any sport with 100 people and place 1st, is that the same as placing 1st in a tournament of 1000 people? in general its not as impressive but if the distribution of skills is different then it could be (example placing 1st at a national level competition among 100 people vs placing 1st in a district level competition with 1000 people). However in GATE the level is already fixed. It is a national level competition and the distribution of intelligence of students is assumed to be similar.

The only way rank is useful is if you make the assumption that every year the number of ridiculously talented people who take the GATE is constant rather than making the more reasonable assumption that the % of ridiculously skilled/talented people is constant.
With all due respect to the person who wrote this article, please improve your understanding of statistics/probability distributions and percentiles.

and is it possible that someone having lesser MARKS than me will have a higher SCORE than me????

gurukul kangri haridwar