COMP30191: Choose it?
Should you choose COMP30191?
COMP30191 is a theoretical course, as the third digit in its code
indicates. The first part of it is decidedly mathematical in nature,
and there will be quite a few proofs. The middle bit is concerned with
writing programs for finding good strategies, which is where it
becomes most computer-sciency. However, we will only look at general
principles rather than issues that occur in implementation. There is
no code in the lecture notes. Lastly we have a look at the use of
games as models, which shows connections with the social sciences and
biology. Here's a short description.
Pros and Cons
In my experience, students find games a much more interesting part
of mathematics than most others. The subject certainly isn't as dry as
some, and everybody can relate to the examples. Every year I get a
few students who say it's the most interesting course unit they've
That said, we are talking about a theoretical course, which has
examples classes rather than labs. This has a number of consequences
when it comes to how one goes best about learning the material.
- The concepts can be quite difficult to get used to. Hence working
throughout the term is a necessity if you want to do well in the
exams. If you leave it all until the revision period you are likely to
have some idea of what most of the concepts are about, but due to the
mathematical nature of the material, that won't get you very far in
the exam. You've got to give this stuff some time to sink in!
The timetabling effectively tells you what you are supposed to have
studied by then - there are topics published for each examples class
and question session.
- It can be difficult to learn the material by yourself, without
any additional help. The only existing source are my lecture notes,
there are no text books you can refer to if that isn't enough. Hence
attendance of examples classes is vital, and you should make use of
the question sessions unless you understand the lecture notes by
- A number of the exam questions are going to be more similar to a
typical maths question than what you might be used to from other CS
courses. In particular, I will ask you to show that you can apply the
techniques you have learned to examples you have never seen before.
In order to be prepared for this you have to have practiced
answering this kind of question. That means you will have to go
through the exercises in the notes. Leaving that to the revision
period means leaving it too late. Again, working throughout term is
the best strategy for doing well in the exam. On the other hand, if
you do that and understand what's going on, you should be fine in the
exam! If you know your stuff you can score highly on such questions.
- It's easy to convince oneself that the course material isn't that
difficult. However, when it comes down to answering exam questions
students will have to be able to do these things themselves
rather than watch somebody else explain them! Every year I have a few
students in the exam who try to answer such questions, but their
solutions bear little resemblance to the model answers. In the past
every year at least 10 students have failed this course, so don't
- There will be some bookwork in the exam too. That will require
you to intelligently summarize the most important points of some
section, or to explain some concept, or result. Don't underestimate
the understanding required to do in particular the former well.
Every year I get answers where students try to do some kind of memory
dump, going into incredible detail on some issue - but forgetting the
really important aspects of another. They'll lose on the marks
assigned to the second issue!
In 2002/2003, 107 students chose this course. A number of them
told me that they thought it was too hard, and that there was too
much material. Since for the most part this criticism only started to
arise a fortnight before the exam it seems clear to me that many
students did not heed the advice given above: They did not
work for the course during term, and found themselves overwhelmed
when they started revising for the exams. In the last few days before
the exam I received more than 30 emails with questions or requests
for a meeting to discuss the course. On the other hand nobody
had any suggestions for the revision lecture in the last week of
In 2003/2004, 121 students chose this course. This time I made it
clear that I wouldn't be available in January other than for the
revision session. The revision session was fairly well attended and
quite a lot of questions were put to me via email in advance. I also
didn't get so many emails!
Here is my feedback in
ps format, feedback
in pdf format for the exam in 2002/2003.
Here is my feedback in
ps format, feedback
in pdf format for the exam in 2003/2004.
In the CEQs for 2003/2004, the course got a mark of 7.6 for
interest (slightly down from 7.7) and 6 for difficulty (up from
5.71). Quality was marked with 8 (up from 7.8) and suitability with
7.9 (slightly up from 7.8).
Here are some quotes from past CEQs.
- `Interesting course. Maths bit could be dull and boring sometimes.'
- `Some bits tricky to learn.'
- `Just enough for 1 module. A lot to cover...'
- `Interesting to do something new that doesn't involve any Java.'
- `Need more opportunities to win chocolate bars!'
- `Tricky, but challenging - a good thing - keeps motivation.'
- `Excellent. Best course out of all the ones I've done.'
I think it is fair to say that some people enjoyed the course
immensely while others liked the introduction, and then started to
lose interest as it all became more technical.
8 September 2006