Colleges - Jesus or St John's?

Oh god you really don't understand statistics.

You would need to analyse the statistics for many years, and account for the bias of what type of people apply to which college.

Repeat after me:
Correlation does not imply causation.
Correlation does not imply causation.
Correlation does not imply causation.

PRSOM.


Also you need to know exactly how the system behind those numbers works.
This is why statistic data is often very misleading and is misinterpreted by people who can only see the numbers without sufficient knowledge of what's behind the scene which led to these numbers.

Yeah there's absolutely no way anyone without access to far more detailed data could provide a useful statistical analysis. Even with all the data it would be a daunting job, which is probably performed by an experienced statistician.

Funnily enough, they happen to have a lot of people who are quite smart and rather good reading data and analysing it. And they've been been doing it year after year.
Lucky them.

Funnily enough, they happen to have a lot of people who are quite smart and rather good reading data and analysing it. And they've been been doing it year after year.
Lucky them.

Yeah. Lucky them

Indeed.

Which also means that public statements by the university such as...

"Some applicants think, or are wrongly advised, that choosing a College that attracts fewer applications or making an open application will increase their chance of being made an offer. In fact, careful ongoing analysis of our admissions statistics shows that, for equally well-qualified applicants, making an open application or applying directly to a College does not affect your chance of being made an offer of a place."
Source: http://www.undergraduate.study.cam.ac.uk/colleges/choosing-a-college

...would get them into trouble if they weren't actually true. So, as a budding lawyer @Abstract_Prism, I think you do have to accept this is beyond reasonable doubt.

The thing is, what's written there is pretty much correct. But that's not the point I was making.

Applying to a college that gets fewer applications won't necessarily increase your chances of getting an offer. A college that gets 10 applications and makes 1 offer is no better for your chances that a college that gets 100 applications and makes 10 offers. You have to do a more detailed analysis, where you examine their applicant : offer ratio, and factor in the number of people that get offers after being pooled.

I think a big cause of the misunderstanding over this is because of a belief in a mythical 'good enough' standard, that means that if you are suitably qualified you will get into Cambridge regardless of the college you apply to. It can be summarised as 'If you're good enough for Cambridge, you will get in.'

But this isn't true.

What's 'good enough' for one college is not good enough for another. One of the main things that will affect whether you are 'good enough' is how many applicants they get and how many offers they are prepared to make. Where a college receives a ton of applications but are prepared to only admit a few students (and it is true that colleges have a rough idea of how many offers they want to make), the college will be forced to be much more scrutinous in assessing their applicants, and naturally the calibre of students at that college will be higher simply because they have to be exceptionally good to be given a place over the army of other applicants. For a smaller college that receives less applicants and makes more offers, the college will be more willing to accept applicants who wouldn't have made it at the larger college that doesn't make many offers. (Of course colleges are not obliged to take anyone, but, yunno, they do anyway, so it's reasonable to assume that smaller colleges - to some extent - adapt their standards based on the general quality of that year's cohort.

What Cambridge said in that link you posted is probably true. There are certain - note how they specified 'well-qualified' applicants - applicants who will probably be made an offer regardless of what college they applied to. But for fringe candidates like me, the college I apply to (based on how many applications and how many offers they make, as well as the general quality of their cohort) will make all the difference.

It's not like this should be the only factor in picking a college though; there's no point in applying to a college you hate just because your chances are better there. But I think it should be a factor. People should come to a decision on what college to apply to through both the admissions statistics and their personal preference. (Or rather, they shouldn't, because then it would make it very difficult for me to make consistent predictions )

Your time is much better spent on trying to strengthen your application so that you won't be a fringe applicant than spending hours trying to analyse their statistic and come to a wrong conclusion.

I do agree that it's better to spend time focusing on one's subject and school work than on trying to analyse admissions statistics to try to work out one's best chance.

While 'popular' and 'less popular' colleges don't tend to change quickly, individual numbers within subjects at colleges can change quite radically from year to year. The strength too can change radically. Two years ago at Christ's we had 3 of the top 4 ranked students in a certain subject & 7 of the top 28 in the university. We took a high number and a load of others got into other colleges via the pool. Last year in the same subject, we had a similar number of applicants but nothing like as strong and we made a lot fewer offers and not many got in through the pool.

Even overall numbers at individual colleges can go up or down significantly. Having been at a meeting of ATs on Friday, I know a number of colleges who are significantly down compared to this stage last year and others who's numbers are up (and also quite a few who haven't even looked at numbers yet!).

Honestly, the CAT did his own detailed analysis a couple of years ago and was satisfied that overall there's no point playing the stats game.