# Ranking: bad joke

#### cge

##### Member
I was under the impression they knew what they were doing, but this is a bad joke….

No other comment to the risk magazine ranking…

NYU Courant ahead of Baruch MFE ---really? I am just curious how this was done…

Really- NYU Courant 2nd and NYU Tandon on 5th? …wait….NYU Tandon has 20 students according to them…20 students instead of 157 students!!

University Class Size Accept. Rate Students accepting Industry affiliated lecturers Employment rate Salary ( US adjusted)

New York University (Tandon School of Engineering) 20 28% 60% 88% 92% 91,200

So if you only get the info for those 20 you can be number 1 – correct?
If the ranking was done based on that they should change it and have them lower than that…but this ranking is a bad joke…not professional either…

Andy- you the man!

If I were CMU I will not choose next year to be in this ranking …really? NYU Tandon ahead of CMU…wait let’s not forget NYU Tandon has 20 students instead of 157 students!

Peter- you get a change to be happy- you got 20 students – this is the data you provided to them?

This is confusing for me…what can I say about the prospective applicants?

CMU with 78% and 93% is ranked number 8? You guys are joking? NYU Tandon with 73% and 86% is ranked number 5?
Are you serious about this ranking or is just for fun?

3.Carnegie Mellon University
Computational Finance
Pittsburgh, PA 98 4.4 78% 93% $117,775 169$86,820 98 FT

9.NYU Tandon School of Engineering
Financial Engineering
Brooklyn, NY 87 3.3 73% 86% $96,961 169.2$70,000 (2 years) 157 FT

#### P. Carr

##### Active Member
Yes, we have an average of 20 students per class. About half of our classes have less students in fact. You seem to be confused about the difference between class size and cohort size. Our cohort size for the first year is 157 as you can see on the far RHS of the 2020 Quantnet rankings.

#### cge

##### Member
Yes, we have an average of 20 students per class. About half of our classes have less students in fact. You seem to be confused about the difference between class size and cohort size. Our cohort size for the first year is 157 as you can see on the far RHS of the 2020 Quantnet rankings.

Thank you Peter – your post just confirmed what I was assuming:

I am not sure how but there is a column which looks like was “tailor” for NYU Tandon.

Does NYU Tandon advertise in the magazine?

I find it a “big coincidence” that this year they added: “class size” and not number of students anymore (you are the only one I know for pushing for this )

see:

Quant Finance Master’s Guide 2019 - Risk.net

see:

Quant Finance Master’s Guide 2020 - Risk.net

#### P. Carr

##### Active Member
Here is a thoughtful post on cohort size for MBA programs:
You seem to be unaware that class size is more relevant for a positive experience than cohort size. I will answer your question when you post your name.

#### cge

##### Member
Here is a thoughtful post on cohort size for MBA programs:
You seem to be unaware that class size is more relevant for a positive experience than cohort size. I will answer your question when you post your name.

I thoughts we were passed by this first grade games: “I tell you if you tell me”

Beating around the bushes again, I will assume the answer is yes and that is why they agreed to add the column that adds more points to the program…

Why we should care about placement when class size is more important...hmm..
watch out folks when you see rankings…do your homework well when you choose a program as well...

Still:

I find it a “big coincidence” that this year they added: “class size” and not number of students anymore (you are the only one I know for pushing for this )
see:

Quant Finance Master’s Guide 2019 - Risk.net

see:
Quant Finance Master’s Guide 2020 - Risk.net

Last edited:

#### Onegin

##### Active Member
C++ Student
many other rankings (US news, etc) use class size as a ranking factor.

the logic is 5 students in a class will get more time and attention from the instructor than 500.

If they added a "Located in Brooklyn" dummy variable, we should definitely be suspicious.

#### cge

##### Member
I am pointing out, (I hope) the obvious, last year there was no such column “class size” and this year, the column: “number of students” was taken out completely and replaced with “class size”… And we know how much Peter likes this and is pushing for it...

Thank you Onegin – this takes me to the below question for Peter:

Are you teaching one class only (20 students) a semester? Elaborate please.

In other words, only 20 students out of 157 sees you teaching? I am not talking about that: “if they need me they can stop by my office thing…I am taking about (only) teaching.

#### Onegin

##### Active Member
C++ Student
@cge - umm, it's not typical for a director of a program to teach a course. There's a lot of administrative work that goes along with such a position. It's possible he teaches something, but it's likely to be an elective or advising individual students on their thesis.

For what it's worth, Peter Carr has a solid reputation on the street. He had a solid gig at Courant, and took on a veritable shit show at Tandon, and is starting to turn it around by all accounts. It is not easy to do something like that, and it takes time. Especially in a university, one of the only sectors left with the preposterous promise of lifetime employment.

It's reasonable to ask about new statistics being included in a ranking, and the validity of those stats. It's less reasonable to accuse someone of malfeasance for benefiting from a survey by a respected publication that they didn't even produce.

Risk magazine is a solid publication. Peter Carr didn't conduct the survey. Class size is a reasonable statistic, well accepted for evaluating the quality of teaching. Program directors often don't usually teach.

Now, if you want to argue about the tradeoff between a large (100 person) class size, but the class is a subjected adapted to quant finance vs a 20 person class, but the class is general in nature, I'm in.

#### P. Carr

##### Active Member
In reply to Onegin, thank you for defending me. You are right that it's not typical for a director of a program to teach a course. That said, I personally taught a half-semester course last Fall on static hedging with exactly 15 students and I'm currently co-teaching a full semester course on option pricing with exactly 9 students. As you can see. I'm personally under the Tandon FRE average of 20 students per class recently publicized by Risk. We have 157 students in our first year cohort. To avoid confusion, administrators use only the phrase cohort size'' to describe the number of students who will graduate at the same time and they use the phrase class size'' to describe either the number of students in a particular class, eg Intro to ML. "Class size" is often used as an abbreviation for mean class size, eg. 20 at Tandon when there is no danger of confusion.

With terminology hopefully clear, let me now reply to CGE, sure in the knowledge that he or she will be getting the last word in.
US News uses mean class size in rankings, not cohort size. Quantnet does not use either mean class size or cohort size in rankings, they merely report cohort size. Risk used cohort size negatively last year and replaced it with class size entering positively this year. This benefits programs like mine that have large cohort sizes and small class sizes. The weight was 5%, so I don't think that is the sole explanation for jumping from #13 last year to #5 this year.
You are correct that I think average class size is a better measure of teaching quality than cohort size. I may well be the only person you know who thinks that. but that is purely a statement about who you know. This link
examines the effect of 2018 cohort size in MBA programs. The largest cohorts are at Harvard and Penn, two perennially highly ranked B schools.

You called Andy Nguyen the man so I'm guessing you agree with Quantnet's rankings. If #1 Princeton merged with #2 Baruch, the reported cohort size would double from 29 in each to 58 together. My guess is that the combined program would continue the excellence of each, despite the larger cohort size. Last year's Risk.net ranking would have lowered the ranking of the combined entity. Similarly, if I took my one program with 157 in the first year cohort and split it into 157 programs with 1 student each, I would have had quite a few top-ranked programs in last year's Risk ranking without any change in quality. Last year's ranking was the first one Risk.net did and all that happened is that they improved their methodology IMHO.

#### cge

##### Member
I guess you may be a spokesperson…I think I wrote Peter’s name and was expecting his reply…nonetheless, you did good, you scored points for yourself, Peter and the program, I like it.

The fact stays the same, this year they added: “class size” column and not “number of students” anymore.

Why we should care about placement when class size is more important...hmm…

#### cge

##### Member
In reply to Onegin, thank you for defending me. You are right that it's not typical for a director of a program to teach a course. That said, I personally taught a half-semester course last Fall on static hedging with exactly 15 students and I'm currently co-teaching a full semester course on option pricing with exactly 9 students. As you can see. I'm personally under the Tandon FRE average of 20 students per class recently publicized by Risk. We have 157 students in our first year cohort. To avoid confusion, administrators use only the phrase cohort size'' to describe the number of students who will graduate at the same time and they use the phrase class size'' to describe either the number of students in a particular class, eg Intro to ML. "Class size" is often used as an abbreviation for mean class size, eg. 20 at Tandon when there is no danger of confusion.

With terminology hopefully clear, let me now reply to CGE, sure in the knowledge that he or she will be getting the last word in.
US News uses mean class size in rankings, not cohort size. Quantnet does not use either mean class size or cohort size in rankings, they merely report cohort size. Risk used cohort size negatively last year and replaced it with class size entering positively this year. This benefits programs like mine that have large cohort sizes and small class sizes. The weight was 5%, so I don't think that is the sole explanation for jumping from #13 last year to #5 this year.
You are correct that I think average class size is a better measure of teaching quality than cohort size. I may well be the only person you know who thinks that. but that is purely a statement about who you know. This link
examines the effect of 2018 cohort size in MBA programs. The largest cohorts are at Harvard and Penn, two perennially highly ranked B schools.

You called Andy Nguyen the man so I'm guessing you agree with Quantnet's rankings. If #1 Princeton merged with #2 Baruch, the reported cohort size would double from 29 in each to 58 together. My guess is that the combined program would continue the excellence of each, despite the larger cohort size. Last year's Risk.net ranking would have lowered the ranking of the combined entity. Similarly, if I took my one program with 157 in the first year cohort and split it into 157 programs with 1 student each, I would have had quite a few top-ranked programs in last year's Risk ranking without any change in quality. Last year's ranking was the first one Risk.net did and all that happened is that they improved their methodology IMHO.

Interesting how much you write to explain something else when the answer to my question is not really there…or it there is kind of an answer, is very vague: Last year's ranking was the first one Risk.net did and all that happened is that they improved their methodology IMHO.”

My questions to you: do you advertise in the magazine?

Did you suggest to the magazine in any way to add this column: “class size”?

#### Michelle Kuang

##### New Member
In reply to Onegin, thank you for defending me. You are right that it's not typical for a director of a program to teach a course. That said, I personally taught a half-semester course last Fall on static hedging with exactly 15 students and I'm currently co-teaching a full semester course on option pricing with exactly 9 students. As you can see. I'm personally under the Tandon FRE average of 20 students per class recently publicized by Risk. We have 157 students in our first year cohort. To avoid confusion, administrators use only the phrase cohort size'' to describe the number of students who will graduate at the same time and they use the phrase class size'' to describe either the number of students in a particular class, eg Intro to ML. "Class size" is often used as an abbreviation for mean class size, eg. 20 at Tandon when there is no danger of confusion.

With terminology hopefully clear, let me now reply to CGE, sure in the knowledge that he or she will be getting the last word in.
US News uses mean class size in rankings, not cohort size. Quantnet does not use either mean class size or cohort size in rankings, they merely report cohort size. Risk used cohort size negatively last year and replaced it with class size entering positively this year. This benefits programs like mine that have large cohort sizes and small class sizes. The weight was 5%, so I don't think that is the sole explanation for jumping from #13 last year to #5 this year.
You are correct that I think average class size is a better measure of teaching quality than cohort size. I may well be the only person you know who thinks that. but that is purely a statement about who you know. This link
examines the effect of 2018 cohort size in MBA programs. The largest cohorts are at Harvard and Penn, two perennially highly ranked B schools.

You called Andy Nguyen the man so I'm guessing you agree with Quantnet's rankings. If #1 Princeton merged with #2 Baruch, the reported cohort size would double from 29 in each to 58 together. My guess is that the combined program would continue the excellence of each, despite the larger cohort size. Last year's Risk.net ranking would have lowered the ranking of the combined entity. Similarly, if I took my one program with 157 in the first year cohort and split it into 157 programs with 1 student each, I would have had quite a few top-ranked programs in last year's Risk ranking without any change in quality. Last year's ranking was the first one Risk.net did and all that happened is that they improved their methodology IMHO.
Awesome,Peter likes to talk.

#### longgamma

##### Active Member
C++ Student
Interesting how much you write to explain something else when the answer to my question is not really there…or it there is kind of an answer, is very vague: Last year's ranking was the first one Risk.net did and all that happened is that they improved their methodology IMHO.”

My questions to you: do you advertise in the magazine?

Did you suggest to the magazine in any way to add this column: “class size”?
It is quite an accusation to make, even without basic fact checking. Even if lets say NYU advertised the publication, do you legitimately think that Risk.net would engage in pay-for-play kind of behavior? If so, then you must have a dim opinion of the publication, so why even bother with their rankings?

Here is the methodology they employed:
• 5% – Average class size;
• 10% – Acceptance rate;
• 10% – Percentage of offer-holders who enrol;
• 5% – Ratio between lecturers and students;
• 10% – Number of industry-affiliated lecturers over the total number of lecturers;
• 30% – Employment rate in finance sector six months after graduation;
• 5% – Number of citations for the five most cited lecturers in the past four years;
• 25% – Average salary six months after graduation, adjusted for the purchasing power conversion factor provided by the World Bank.

As you can see 55% of weight is to employment statistics. NYU Tandon is in NYC and obviously it helps tremendously with placements and internships. You can take the subway and meet a MD for coffee within 30 minutes who might give you a job. Not a lot of places have that kind of access and that is the draw of programs in financial hubs.

I actually see some issues with the "Number of industry-affiliated lecturer.." having 10% instead of 5% weight for "Number of citations...". But anyways, its an issue with the people employed at Risk.net who came up with the methodology. Obviously, NYU tandon would do well here as its in NYC and it is easier to get Wall St folks to give lectures. So programs near Wall Street, Bay Street or Canary Wharf will excel in this ranking and this is what we see (UoT, NYUs, Imperial and Columbia)

#### gradschoolhopeful

##### New Member
In reply to Onegin, thank you for defending me.
This is not relevant to the ranking post, but since some people mentioned that NYU Tandon has good placements, may I ask you something Mr. Carr?

Obviously the program prepares students for a career in quant finance. However, if that doesn't work out, does the program help students land investment banking or management consulting jobs? Is joining finance/consulting clubs at NYU Stern a possibility? If I have prior experience in finance, will I start out as an Analyst or Associate? I don't want to attend an MBA since most are not STEM-designated.