Search This Blog

Saturday, June 16, 2018

A Guide to JEE Counseling 2018

I wish I had written this blog a couple of weeks ago. Actually, I did start it, but then suddenly, without any prior planning, I had to go to US for 2 weeks. The reason for urgency was simple. There is something that I want to suggest that I have not suggested in the prior years of writing such a blog post. So let us start with that first. (By the way, most of the general principles would be applicable to anyone deciding which engineering college to study in. So feel free to forward it to those who are looking for colleges other than IITs too.)

Visit a few IITs within the next 10 days.

If you can afford it, that is. Today itself, you should shortlist your options to a few IITs. This shouldn't really be difficult, at least for men, since they know their rank and last year's closing ranks. While the closing ranks will change from year to year, they will not change drastically in most cases (but be aware that they can change and fill up choices accordingly). So start with programs which closed just ahead of your rank and take the next 20 programs in the order of closing ranks. Despite whatever I or anyone else is going to tell you, the bottomline is that you are likely to consider these 20 programs more seriously than anything else. If there are some programs in this list, which you will certainly not join, remove that. If you indeed have a strong preference for any other program, since you have read about it somewhere, or someone has encouraged you to consider this, add that too to your list of serious programs. In this list, all the programs may belong to 5-10 IITs. Some IITs may be too far to travel or there may be logistics issues (no reservation in trains, flights too expensive). Don't travel to them (but read about them as much as possible, connect to them and ask if there is anyone in your city who can answer your questions). But whichever IIT you can travel, do that immediately.

Almost all IITs have people willing to talk to potential students and parents and even show them around if they come to campus. Many IITs are doing open houses, specific event to answer queries of students/parents together. All this information should be there on their website. If an IIT does not have any link for potential student, seriously consider removing all their programs from your list. If they don't care to attract students, they are not likely to treat you well for four years. (Treat you well, not in terms of physical assault, I can assure you, but in terms of flexibility in curriculum, rules of the hostels, and so on.) If they do have contact persons mentioned on the website, write to them your program, which department you would like to visit, and what you would like to do. When you go to the IIT, visit the hostel where you may be staying in the first year, talk to students, preferably more than one from different years and preferably at least one from the programs you are considering, talk to at least one faculty in each program that you are considering. This can give you huge insights into the IIT that you would not get through second hand accounts on quora.

I am emphasizing it this year because last year my daughter was in the same situation. We decided that we will go to all those colleges which were the top choices of my daughter. About one week of India tour, and this gave us so much greater insight and comparative information from our perspective that we could not have got by reading all about them. What is the importance of flexibility in curriculum. What kind of freedom you would enjoy in the hostels. What is the quality of infrastructure on campus, particularly your likely hostel. You would assume that these are minor questions compared to what kind of placement a program has. But if you visit a few IITs, you would know why they are very important questions. And even as far as placement is concerned, you would get a better perspective on that by personally visiting the place.

Previous blogs:

The following older blogs will help as well.

A Guide to JEE Counseling 2015

A Non-Guide to JEE Counseling 2016

What would I do if I had JEE Rank 1 (2017)

Conflicts of Interest Disclosures:

I am an alum of IIT Kanpur and a Professor of IIT Kanpur and hence would have a natural interest in promoting IIT Kanpur.

I have been a Guest Professor of IIT Gandhinagar (similar to adjunct faculty elsewhere) and I visit them very frequently (more than 50 visits so far, many of them paid for by them). And I am mighty impressed by them.

There is very little ongoing relationship with any other IIT. But in the past, I have taught at IIT Bombay, I have had research collaboration and joint papers with IIT Delhi faculty. I have co-guided students with Directors of IIT Roorkee, IIT Bhilai and IIT Jammu, with many joint publications. And in general, I have friends in most IITs. I doubt if any of this has affected my views, but it is for the reader to take any bias into account. In any case, take all my views as just another input to your decision making process, and ensure that the final decision is yours alone. Don't blame me a few years later when things don't work out. (Though I won't mind an email of appreciation if things do work out as projected.)

Now back to questions.

Should we prefer a program or an IIT?

It depends. Are you interested in a particular discipline or may be two disciplines. If yes, go for those disciplines. If there is no specific discipline interest, then choose an IIT. For example, people around a rank of 100 ask me whether they should prefer CS at IITD, or EE at IITB. I don't want to answer this directly (since I don't want to be held responsible for any of your decisions). So I ask the student to tell me what s/he would have done if the rank was 1 and not 100. Of course, they would have chosen IITB/CSE. I ask them to reflect on why they would have chosen IITB/CSE without any doubt. Is it because they have heard IITB is a great place, IITB has Mood Indigo, IITB has maximum number of companies coming up for placement and other stuff like that, then they should choose IITB/EE over IITD/CSE. On the other hand, if they want to join that since it is the most prestigious program (based on last year's closing ranks), most toppers choose this, etc., then look at which program has the second best closing rank. If they say, they want to study CSE and within CSE departments, they feel IITB is the best, then I say, choose IITD/CSE. Most students have a clear idea of what should be the order of choice for the first few programs, but when they come to those programs which they are more likely to get admitted to, they are confused. The only thing you need to do is to think about why you were sure of the ordering of top choices. That will tell you what to do next, if you are honest with yourself.

IITs versus other institutions (both abroad and at home)!

Abroad: In the past I have been reluctant to recommend that after a good JEE rank, you consider a university abroad, mainly on account of cost-benefit ratio and culture shock to an 18-year old. While the issue of affordability and cultural differences remain, I am increasingly getting convinced that cost-benefit ratio is in favor of top universities of the world. The flexibility that you have in good universities of changing your major, of combining two very diverse set of majors, broad based education, number of projects and the quality of projects, emphasis on other attributes like team work, a greater emphasis on ethics and professionalism, there is a big difference between IITs and the top universities in the world. So, if you can afford it, and you are confident of handling cultural differences, consider those options seriously.

Science (IISc/IISERs): In the past, I have suggested that if you are interested in science, consider IISc seriously, and then IISERs. But I am going to change that order. IISERs are more tuned to working with under-graduate students and that makes a big difference.

Non-STEM universities: I did mention Ashoka university in one of the blogs earlier. Computer Science without a broader education in areas like Maths, Design, and Social Sciences is incomplete. IITs will certainly give you enough and more Maths, but only a sprinkling of social science (and that too sometimes courses only meant to complete the requirements on a transcript). CS at a place like Ashoka is an option worth considering, even though they haven't been able to build a strong CS department yet. And I think it is definitely worth considering a non-STEM education in entirety, if engineering (or IIT) is only a way to success in career and not a passion. I had visited other places like FLAME (Pune), and Premji University last year and was suitably impressed by what they are offering. Also the five year integrated program at IIM Indore is a great option to explore, particularly for those who are keen on management as a career.

NITs/IIITs: In the past, my stand has been that if you are passionate about a discipline which is not available in IITs, take it up in the next best institute. I am going to change that stand a little bit. Take it up in the next best institute, only if that place is really good. Otherwise, take up something closer in an IIT. For Computer Science for example, I will strongly recommend IIITs at Hyderabad and Delhi. They have disadvantage of small campus with less diversity, but CS is great. So, after the old 5 IITs, may be you want to consider them seriously. And particularly at IIIT Delhi, there are so many interesting combinations which are likely to be so important in future. Think about them seriously. For most engineering disciplines, BITS Pilani is a good option. But I would be reluctant to recommend NITs even for lower ranked students who are not getting their top choices in IITs. What has changed in recent times is the ability to study on one's own through online courses. So if you can't get your favorite discipline in IITs or any really good university, take up something else (hopefully with some overlap) and study some basic subjects online, give GATE/GRE and go for MS/MTech in your favorite discipline after a quality education at the under-graduate level.

Comparison within IITs: 

In earlier years, I had suggested a small difference between the old IITs with IITB as my favorite, then IITD and IITK, and finally IITKGP and IITM. But the difference was always small. Last year, I had suggested that perhaps IIT Gandhinagar would be my top favorite. This year, I would like to suggest that there is not much difference between these six IITs. IITB has had a good fortune of having a series of good leadership. But that leadership deficit in other IITs has reduced with Directors of all these institutes doing quite well in the last few years (and we have a new Director at IITK, who happens to be from IITB).

People have asked me about IITH as that seems to be the favorite (along with Gandhinagar) among the new IITs. Unfortunately, I have never visited them. I am aware of some very interesting curriculum innovations that are taking place there, and I have a lot of respect for its Director. But without seeing things on the ground, I wouldn't put them in the same group. (But knowing the reason for not putting them in the same group, you could appropriately handle information about them.)

I would put the other older IITs next. These include IIT Roorkee, IIT BHU, and IIT Guwahaty. Then the remaining so-called 2nd generation IITs. And finally, the so-called 3rd generation IITs. I have no idea where to place IIT (ISM). I think it is a very unique institute with some very strong departments, but at the same time some rather ordinary departments. So, difficult to put it in a group.

Of course, any ordering is simplifying things a lot (and that has been my problem with NIRF or any other ranking). What may be right for me, may not be right for you. So do your own research. Decide factors that you would like to base your decision on, and then decide what should be the ranking of different IITs. The above grouping should be used only if you are totally clueless about what you want and like.

Metro versus non-Metro IIT: This is a non-question, only asked by students who have already decided to stay in IITD or IITB, and are looking for confirmation of their bias. If the answer is "Metro" would you prefer IIT Madras over IIT Kanpur. Would you prefer IIT Hyderabad over IIT Kanpur. If the answer  to these questions is a No, then Metro is hardly a factor in your mind and you are only using this to convince your mind that you should be studying at IITB or IITD. I, of course, run a lonely battle over social media trying to convince everyone that location is immaterial in today's world, even less so for an undergraduate student.

Which discipline to prefer?

Assuming that you have no keen interest in any discipline since if you do have some interest, follow that interest.

First of all, all disciplines can lead to a great career, or could lead to a poor career. That in the previous years, students from certain disciplines had received a higher average "package" in the campus placement, is no proof or even an indication that people choosing that discipline to study today will be earning the maximum 50 years from now. In fact, people earning the maximum 50 years from now will be those who learn just one skill, how to keep learning throughout your life. The best jobs today couldn't be predicted 20 years ago, and we can't predict today the best jobs of 20 years from now, let alone 50 years. So, don't even attempt to find out placement statistics, since they will strongly influence your choice, whether you want it or not.

Second, if you are sure that you don't want to follow a science or engineering career, take up a discipline which is least competitive in the IIT of your choice. Of course, avoid disciplines that you would hate. And perhaps, choose an IIT which is less competitive. (By less competitive, I mean programs where lower ranked students are more likely to join.) Take part in extra-curricular activities. Enjoy your stay. Take advantage of huge learning opportunities outside the classroom. Make friends. This network will help you a lot in future.

Third, if you are not sure of what discipline to prefer, join an IIT with a lot more flexibility in curriculum in terms of ability to do second major, ability to do minor programs, etc. So, if after a semester or two, you want to study different things, there is a greater likelihood of your being able to do. Do not depend on branch change. Very few students will be able to do that in any IIT. I would not be able to tell you which IIT is more flexible than the other. But the Counselling brochure will have some information about them - the rules for branch change and existence of second major/minor, etc. Talk to people in those IITs to get more details. A bit of research now will help you a great deal. And as I suggested above, visiting the IIT before freezing the choice list will be the best way to get a feel for such issues.

Fourth, if nothing else works for you, but you do want to choose a discipline and not an IIT first, choose Computer Science or a related discipline (like Maths and Computing). May be, it is my bias since I am in Computer Science department, but it seems to me that the skills we help you build are useful in all other disciplines today. But this is also the most sought after discipline. So this may mean that you end up with an IIT which does not have its own campus right now.

Programs (4/5 year, BS/BTech, etc.):

Single degree versus Dual degree: I used to be a strong supporter of dual-degree programs. I am no longer sure about that stance. I have started to think that to force a student to make a choice of discipline of an under-graduate program is bad enough but to force him/her to make a choice of discipline for a graduate program when s/he hardly knows about anything is not good. Even earlier, I discouraged students from joining programs which are very narrow (so you decide not just a broad discipline for MTech, but a narrow sub-discipline for MTech). But now, I think, unless you are very sure about your interests and passion, avoid dual-degree programs.

Does Nomenclature matter? BS/MSc/BTech for Maths and Computing, for example: There are similar programs in different IITs with vastly different nomenclature for the degree given at the end. Maths and Computing combination is perhaps the most significant example of this. The nomenclature makes no difference to private sector. (Don't sue me if tomorrow if you find one company that prefers some nomenclature. I would advise you to avoid that company anyway.) Similarly, it is very unlikely that any quality educational institute will worry about the nomenclature (they may bother about whether you have a 4-year bachelors or a 5-year masters). But government entities and those regulated by government entities will worry about the nomenclature. If there is a job opening for BTech in Computer Science, they would be very reluctant to accept someone who has a BS in Maths and Computing even though the latter may satisfy all requirements what the skills and knowledge that one should have for the job. Similarly, if AICTE allows an MTech to teach, they may not allow an MSc in the same discipline with same number of years to teach. But most IIT graduates do not consider technical jobs in the government anyway. So in general, you can ignore the nomenclature issues.

Should you drop a year and try again?

The answer is an unequivocal no, unless you can identify a specific reason for your poor performance that is unlikely to repeat next year, something like an illness during the exam. Just saying that I couldn't prepare along with 12th, and now that I will be preparing exclusively for IIT, will be able to do better, does not work.

I am sure you have asked people on social media this question, and you have heard from lots of people that dropping a year helps. But remember, that answer is there because anyone who has improved his/her performance wants to tell the world how smart their decision was. And anyone who has not improved his/her performance would not want to tell the world how unhappy they feel at having wasted a year for no gains. So you are hearing from a small set of students, and you are not hearing from a representative set of students.

Studying in the drop year is not easy. All your friends are gone. You don't feel like socializing with anyone, and your motivation levels are generally low because you thinking about your poor performance of the last year. Of course, some overcome all this, and perform well, and they are the ones who are telling everyone how great their decision was to drop a year. At the beginning of the year, it is very attractive to think that I too will be like these super-motivated folks, and if you indeed can work like them, you too will be rewarded. But most people lose their motivation during the year and even perform poorly in the second attempt.

Impact of Women Reservation

Difficult to predict as this is the first year of reservation. But I would guess that the closing ranks of all programs except the top few most popular programs will slide as the women candidates take up the reserved seats in the more popular programs. I would also expect the closing ranks to slide a bit because every year a few more students are taking up options outside the IIT system (including universities abroad, or places like IIIT Hyderabad/Delhi, BITS, Ashoka, etc.) But let it not be a factor in your choice. Fill up as many options as you would not mind studying. Don't save on typing. But do not fill up any option that you will regret studying later on.

This is the end for now. I may add more information, if there are questions that I have not addressed. Please feel free to send me an email at sanghi on

Friday, June 15, 2018

JEE Advanced 2018 Results: IITs Blunder

To understand what went wrong with the ranking list of JEE Advanced 2018, we need to understand the difference between eligibility and ranking cutoff (and admission cutoff) and various legal and other requirements governing admissions. In this article, I first explain the process of admission, which could be quite boring to most readers of this blog, but I do this for completeness sake. Sorry for a very long article. Those of you, who understand the admission process, please go directly to the next section "What happened to JEE Advanced this year."

Eligibility requirement implies that those who do not meet this requirement are not likely to perform satisfactorily in the program under consideration.
If eligibility means the minimum standard to follow a program, why is eligibility different for different categories. One reason is that people from disadvantaged backgrounds often work harder and are able to make up for some deficiencies that may be there at the time of admission. And hence a small difference may be allowed in the eligibility requirements of General candidates and others. Indeed, the courts have decided that the eligibility requirements between General and OBC-NCL (Non Creamy Layer) can be at most 10 percent. However, when it comes to admission of SC/ST students, courts have pronounced that the educational institutions must provide additional support to them to succeed, and hence in their case the difference in the eligibility requirements can be larger. (I am not aware of any legal case regarding Persons with Disability or PwD, but IITs have decided that their eligibility will be similar to those of SC/ST students, essentially admitting that they will be provided additional support to succeed.)
The admission of a student in a particular category can be allowed only if they meet eligibility requirement of his/her category.
IITs have a rather complex set of eligibility conditions (which wasn't the case when I sought admission to IITs a few decades back). These include performance in 12th class (75 percent marks or 80 percentile performance, which ever is lower), when one has passed 12th class (in the year of admission, or previous year, and in case of an improvement exam, the year of first passing the 12th class is considered), minimum number of marks in JEE Advanced in each subject (10 percent for General, 9 percent for OBC-NCL, and 5 percent for SC/ST/PwD), and minimum total marks in JEE Advanced (35 percent for General, 31.5 percent for OBC-NCL, and 17.5 percent for SC/ST/PwD). In addition, one needs to be in top 1 lakh approximately in JEE Mains for General and similarly in top few for other categories.
Finally, IITs look at all those candidates who meet all eligible requirements and rank them. The ranking order is the total marks in JEE Advanced with some tie breaking rules. Everyone who meets all eligibility requirements of General class gets a Common Rank. All reserved class students who meet eligibility requirements of their respective categories get a Category Rank. It means that students belonging to reserved classes may have a General rank too, in addition to their Category ranks.
The Ranking Cutoff is the performance of the student who has the last announced rank. Since IITs announce a rank for all those students who are eligible for admission, the ranking cutoff is same as eligibility requirement in their case. Please note that it is not necessary for the two to be same. If the number of eligible candidates is very large, IITs could decide to rank only as many as they think will be able to fill all seats. But it is convenient to announce all ranks and then ask only some of them to participate in the counseling and in case seats do remain vacant, ask more students to fill up their choices.
The Admission Cutoff is the performance of the student who is the last person to get admission in a particular program in a particular category. Note that admission of reserved class students depend solely on their rank among eligible students and does not depend on the admission cutoff of general students.
To consider the implications of all that is written above, let us take a hypothetical university which has only one program, and it has announced that the eligibility is 60% marks in 12th class for General candidates, 55% marks for OBC-NCL, and 50% marks for SC/ST/PwD. A large number of applications are received, much more than the number of seats, and they announce a ranking for all eligible applications. Then they announce admission cutoffs for different categories. For example, they may say that it is 98% for General, 85% for OBC-NCL, and 60% for SC/ST/PwD. This means that all students with 98% marks or more will get admission to the program. Among OBC-NCL students, all those with 85% marks or more will get admission, and so on.
Note that the difference in the cutoff for OBC-NCL and General can be more than 10 percent. Similarly, the difference between SC/ST and General can be very large as well. One cannot have a rule which says that OBC-NCL students' admission cutoff can be at most 10 percent less than General, or that SC/ST admissions cutoff can be at most 20 percent less than General, etc.
What happens if the number of eligible candidates in a particular category or in all categories is less than the number of seats (or some people who had earlier shown interest by applying now do not want to seek admission). 
There are only two options in this case. One, you have a smaller class than what you planned for. That is, leave the seats vacant. Two, you change the eligibility conditions. Changing eligibility conditions means that we are either willing to admit academically unprepared students, or our earlier eligibility conditions were arbitrary and ad-hoc and therefore, there is no harm in changing them.

What happened to JEE Advanced this year?

When the result was prepared, it was found that less number of students were eligible for admission to IITs than any time in the past, compared to the number of seats on offer. The Joint Admission Board (JAB) which consists of all Directors and all JEE Chairpersons decided that they will announce the results by sticking to the eligibility conditions that had been announced long ago. In public statements, they said that the number of successful candidates were 1.6 times the number of seats, and they were confident that all seats will be filled by these eligible students.

This only means that none of the leadership in our greatest institutes had done their homework. Let us look at the claim of 1.6 times, for example. They said that there are 11279 seats and 18138 eligible candidates. This was patently false, since there are about 800 extra seats to accommodate women reservation as well. And therefore, the ratio of eligible candidates to seats is only 1.5 and not 1.6. That JEE Chairperson keeps repeating 1.6 as mantra only shows his ignorance and perhaps a secret desire to not have 14% reservation for women.

There were other issues with this. As we have explained above, admissions are done in different categories. So while overall, the number of eligible candidates may be 1.5 times the number of seats, one had to look up whether this is true for each category of students. If they only had looked at numbers category wise, they would have noticed that this time the number of eligible candidates in OBC-NCL category were actually less than the number of seats reserved for them. On the other hand, for SC category the number was 2.3 times the number of seats. How could this be missed by everyone in JAB. It shows how completely unprepared they come for the meetings where such important decisions are to be taken. That says a lot about the leadership of our IITs.

Media reported some extremely stupid statements. I find it difficult to believe that even this clueless leadership could have made those statements, and perhaps there is a communication gap between IITs and Media. For example, it was said that in case the number of seats remaining vacant is large, we can always announce a second list then. That special admission round could be held at that time. This was stupid because you would have converted OBC-NCL seats to General in the usual rounds. Then when you announce the second list in which there would be more OBC-NCL students, would you tell these students that they can't get any seat since all vacant OBC seats have been converted to General (as per the law), or would you tell those General candidates, sorry, we are canceling your admission now that we have more OBC-NCL candidates. Obviously, the solution of extended list had to be implemented now, and not later.

Also, the last General candidate to get any seat in IIT system last year had a rank of about 15,000 (Architecture courses at Kharagpur and Roorkee). This time, the total number of students who had a Common Rank was less than 15,000. While the number of seats had gone up by 1000. How were they sure that this year, the number of candidates were enough to fill all seats.

It would have been at least courageous to announce that IITs recognized that the number of eligible students is less this year and that they will leave seats vacant, since they didn't find enough good students. Not admitting this, and yet announcing only a small list of successful students was stupid to say the least.

Of course, if they had indeed admitted that they don't intend to fill large number of seats in IIT BHU and IIT ISM and may be some other IITs as well, it would have been based on a completely arbitrary eligibility condition.

In general, the eligibility condition should be based on some research. Is there any evidence that statistically speaking students getting 40% marks in JEE perform significantly better than those getting 35% marks. Also, have IITs not compromised with their pre-announced eligibility conditions in the past. What does giving 18 bonus marks do other than compromising eligibility conditions. Indeed, every single year ever since the system of JEE Mains and JEE Advanced has been created, bonus marks have been given to students on some pretext or the other. Most boards in the country give bonus marks. So why suddenly we are saying that this year despite the less number of eligible students we will not give any bonus marks to anyone and artificially increase the pool of eligible students.

I am glad that finally, JAB has decreased the eligibility condition. But I am sad that ministry had to intervene to get it done. The ministry intervention is nothing but an interference with IIT autonomy, but if the entire leadership of all 23 IITs combined will act stupidly, the political leadership is not going to remain silent. I wish IIT leadership does not give occasions for political leadership to intervene.

Why IITs have 35% as eligibility requirement?

Let us go back in history and see how IITs admitted students then. When I was seeking admission a few decades back, the eligibility condition was simple. Pass 12th class with 33% marks with any subjects (not even PCM was required). Same eligibility for all classes.

But there was something hidden, and these were pre-RTI days and hence remained hidden. After JEE results were tabulated, they will decide individual subject eligibility. Those who did not get that many marks in that subject were not eligible for admission. This eligibility condition was never announced but was implemented, and was different for General and SC/ST. (There was no OBC or PwD reservation at that time.) Every year, the eligibility could be different, but it would usually be a number which would make about 15-25% students to be eligible. If we take the set of students who were eligible in all three subjects, it would typically be 15-20% of those who have taken the exam. Out of these about top 3% would be assigned a rank, and top 2% would actually be offered admission. The rank was purely based on total score.

The idea at that time was that IITs wanted to admit only those students who were good in all three subjects, and did not want to admit students who were excellent in two subjects and poor in third, and hence the subject wise eligibility, which was never publicly announced.

For SC/ST, there was another hidden condition. They had to score a total of at least 75% of the total scored by the last general student on the rank list. The idea was that if the gap between the General and SC/ST students in the same class was very large in terms of their academic preparation, that would lead to very poor academic performance by SC/ST students and may lead to termination of their programs and worse. However, the seats reserved for SC/ST were filled up by admitting students against the vacant seats for a one-year preparatory program. This was an excellent program, and most SC/ST students who finally took admission to IITs after one year were able to compete well with the rest of the class at times even better than the students who had directly got admission that year.

The secrecy and the absence of RTI helped IITs continue with this system till 2005. In 2006, it was alleged that the subject wise eligibility conditions were decided in a malafide fashion to keep a specific JEE candidate out of the final list even though his performance was better than any cutoff decided in the past. RTI, which had just become the law, helped the cause. The newspaper headlines forced IITs to stop these hidden conditions and announce all eligibility requirements in advance.

Bitten by what happened in 2006, IITs pretty much decided to not have subject wise eligibility conditions. The subject wise eligibility conditions were kept so low that there was no danger of disqualification of any student in future. It was 20 percentile score in each subject in each category. Now, 20% of the students giving JEE get close to 0 marks in a subject. So, the marks based eligibility was so low that it did not bar anyone from selection on this count. However, when media asked for the exact cutoffs under RTI and put that out as a news item, it was embarrassing to see single digit marks being touted as eligibility condition. The right thing to do at that time would have been to completely do away with subject wise eligibility (or enforce a decent number as the eligibility), but IITs decided to have a 2-digit number as eligibility. They announced that the minimum eligibility in each subject would be 10 percent for General and 5 percent for SC/ST

What was remarkable about reducing subject-wise eligibility was that it was taken without any application of mind as to what will be the implication of this on the quality of admissions. There was a long standing principle in the admission process which I just stated a little while ago. That IITs will admit only those students who were good in all three subjects and not admit students who were excellent in two and poor in one. This principle was being thrown out of the window without any discussion in Senate of any IIT. It was said that in the era of RTIs, we had to be transparent and hence this. This was a stupid logic, since transparency did not mean giving away your principle of admission that had stood the test of time for 45 years. If they had decided the eligibility to be 20 percent and 15 percent for General and SC/ST respectively, that too would have been transparent. But, IIT leadership wasn't much better at that time.

What is interesting is that in 2007, the batch that was admitted based on the new eligibility had lots of students who were exactly the type we were avoiding for the past 45 years. While JEE and IIT system would refuse to do any research on the implications of their decision, I decided to enquire from my contacts in different IITs if they felt that the student cohort was any different that year, and in multiple IITs, the impression was that a much larger number of students were weak in Maths in that batch. At least two IITs reported that a significantly larger number of students had received F grades in Maths courses in the first year. Of course, not a scientific study, and impressions should not be used to change policies, but impressions such as these should have encouraged IITs to do a proper scientific investigation. But that was never done. And we continue to offer admission to students who are excellent in two subjects and poor in the third.

May be I should clarify that I am not against the new admission philosophy. What I am against is that rules for admission have been changed without even understanding that the admission philosophy has been changed as a side effect. If someone were to propose with good arguments that we change our admission philosophy, it was discussed in faculty at large, and then changed, it would be fine with me. But the changes in IIT JEE are either forced by MHRD or by fear of an adverse ruling by the courts. In my 25 years at IIT Kanpur, I rarely have had a discussion on how to change JEE so that we get better students.

Anyway, coming back to subject wise eligibility conditions, removing subject wise eligibility (or requiring such low marks) would have had a serious repercussion on how we admitted SC/ST students. You could not deny admission to any SC/ST student who had received such low marks in JEE. The previous policy of only admitting SC/ST students who had received 75% marks of the last ranked general candidate would have been struck down by the courts, if people knew about it. And in the RTI era, IITs could not continue doing so. But the system of admitting only good students directly and rest through a one-year preparatory program had worked extremely well and really trained these students to compete with everyone else on an equal footing. IITs did not want to give up on that. IITs knew that if we were to fill up all SC/ST seats directly, they will not be able to compete with the other students and lots of them will face their programs being terminated, or will take more than 5 years to graduate (which is happening now).

Indeed, in 2008, the landmark judgment regarding OBC reservation (and a later case seeking clarification) resulted in a very clear principle. You cannot restrict admission in reserved classes based on the difference between the marks obtained by general category and reserved category. So what was felt as illegal was actually declared as illegal, though in the context of OBC reservation.

So another eligibility condition was created, which was to simulate the condition of only good SC/ST students getting direct admission and remaining through one-year preparatory course. This was the total marks in JEE.

Initially, based on the previous distribution of marks, it was felt that 35 percent as eligibility requirement was a safe requirement that would always result in sufficient number of general candidates to fill up all seats. And SC/ST eligibility was proposed as 75% of that.

But that is when the issue became political. The two tier admission policy of IITs, which I believe has helped SC/ST students tremendously, was under attack by activists, and it was said that IITs are deliberating causing most SC/ST students to graduate in 5 years. The reality was that many of these students would not graduate in even 5 years if they were given direct admission without the solid support that the one year preparatory course provided. But IITs succumbed under pressure and decided to have SC/ST eligibility as 50% of General students. So it was announced as 17.5 percent, which means that most seats are filled now directly, and there is hardly any student in the preparatory program. The number of students in the preparatory program is so few that only a few IITs run that program on behalf of all other IITs.

What should be done going forward?

If we are only looking at how to avoid the embarrassment of this year's JEE fiasco, then the solution is simple. Reduce the General cutoff to 30 percent, and in future, if there is still an issue of less number of eligible students, just add bonus marks to all participants (like what CBSE does). That will take care of it.

But in my opinion, each such fiasco is really an opportunity for the IIT system to think of admission process as a whole. Who is the right student for us. How do we test for that. How do we decide our eligibility requirements to ensure that we attract such students.
In 2011, when the then Minister wanted to make significant changes to IITs' admission process, the IIT system had promised that we would have a standing committee that would look at the admission process on an ongoing basis and keep making suggestions to JAB. A committee was indeed made, but with a very limited mandate and a limited timeframe. It is time that we make a standing committee to look into these issues from a long term perspective. The JEE system currently is geared to somehow finish the task each year. It is not geared to think of any innovation. And IITs must encourage research on different aspects of admission.

Added on 15th June:

A few friends have told me that I am being harsh towards JEE, and if there were a handful of seats (they put the number in single digits) vacant, it wouldn't make a difference, and the pre-announced eligibility conditions are legally binding. So let me make a prediction here. At the end of counselling session in July, there would not be more than 10,000 students from the first set of 18138 selected candidates who will accept admission in IITs. So we are talking about 2,000 seats being filled in by new students (or remaining vacant). I think a few women seats would still remain vacant. So I am criticizing the result for having a few vacant seats. It is enormous number of seats and many of the less popular programs would have got 0 students.
I agree that there is a legal issue in changing eligibility conditions. But as I mentioned above, IITs have compromised eligibility conditions every single year either by directly reducing the eligibility or by adding bonus marks. And they have not been challenged on this count. To challenge this would require the existence of a person who stands to lose by this. I don't think there is any such person in this universe.

Wednesday, May 30, 2018

Life Lessons from a course on Computer Networks

I have been teaching for 25 years. The course that I have taught the maximum number of times in this period is "Computer Network." I try to ensure that at the end of the course, students have understood how data moves from one place to the other in Internet, be it an email, or a webpage, or a file download. They should also be able to design a network for their home/office/building/campus, etc. They should be able to write a network application, or design a protocol, and so on.

But I hope they learn more than that.I like to teach this course because this has so many important lessons on how to lead our own lives. On the other hand, we some times design protocols based on what works in our own life, in human interactions.

We start with a lesson that one hopes everyone had learnt as a kid. In any group, if someone else is speaking, you should wait for him to stop before speaking. This is nothing but Carrier Sense Multiple Access (CSMA) protocol. We then improve upon it by saying that as you begin speaking, listen if someone else is starting too. If two persons happen to start around the same time, then they should both stop and one of them then restart speaking. Well that is how your good old Ethernet works. We call it CSMA/CD. CD for Collision Detection.

When you have to communicate with someone, would you call them up and start speaking non-stop even before you hear a "Hello" from the other side. Or would you first say, "Hello," and ask if they are free to talk (Connection Establishment). If they give the go ahead, speak only a little and wait for some response, which could just be that they have understood what you have just said. If they ask you to wait for a few seconds, you stop speaking for that much time (Sliding Window Protocol). Well, that is the difference between UDP (User Datagram Protocol) and TCP (Transmission Control Protocol). No wonder, as in real life, even in computer communication, most applications use TCP. No one wants to communicate with people who don't listen.

When I was reading about protocol design and implementation, I came across this guideline. (This was one of the RFCs. If someone knows which one, please let me know. I will add the reference.) The packets that you send must follow the protocol strictly. But when you receive a packet, accept packets even if the sender has not followed the protocol strictly. This is, perhaps, the most important life lesson from the Computer Networks course. Always say the right things, speak the truth. But when someone else is speaking, don't cutoff communication because the other person has said something which shouldn't have been said. Be tolerant of other people's mistakes.

While discussing congestion control, we talk about reducing our own demand as long as there is congestion. An aggressive behavior will result in prolonging of congestion. It might help that particular sender, but it does not improve network throughput. The idea that we must optimize for social good and not for individual good. Of course, given that some persons will always be aggressive, we also want to study systems in which behaviors favoring social good are rewarded (or anti-social behavior is punished) so that it also becomes a behavior for individual good.

While discussing different queuing strategies, we explain that there is no unique definition of fairness. The max-min fairness appears to be a reasonable choice in this particular context, but change the context and something else could be fair. So in any negotiation, don't assume what you think is fair is necessarily fair and if they are not agreeing on what you are asking, it is not because they are selfish or stupid. Try to understand fairness from their perspective.

As I think of more examples, I will add some more. But studying networks has been a fascinating journey because it is so closely linked to how you communicate in real life.