The Value of Coding Challenges?

[DorkRawk]

So in looking toward next year's applications I'm thinking about ways to beef up my apps and show CS prowess. I was looking at the Google AI Challenge ( http://www.ai-contest.com ) and it seemed like it would be a lot of fun to play around with. How would participation (and maybe success) in something like this be viewed if mentioned in a personal statement for someone looking to work in AI?

[the poisoned pawn]

So in looking toward next year's applications I'm thinking about ways to beef up my apps and show CS prowess. I was looking at the Google AI Challenge ( http://www.ai-contest.com ) and it seemed like it would be a lot of fun to play around with. How would participation (and maybe success) in something like this be viewed if mentioned in a personal statement for someone looking to work in AI?

It will definitely be a boost to your application, but I personally believe many people will not equate that competition to a research experience.

Edited April 1, 2011 by the poisoned pawn

[edvolkov]

I know a guy who had a first place at ACM-ICPC. And he was accepted to UCB without any publications or GRE subject. And his major was not CS (he switched to another program at Masters because of conflict with faculty)

[nobody2008]

First place at ACM ICPC is something that even the best Computer Science majors will find extremely hard to secure. No wonder UCB was more than willing to accept him.

[the poisoned pawn]

The point is there is no guarantee that you will get one of the top prizes, so, it is a little bit risky to put all effort in this contest. A research project with a professor will get you a good recommendation letter and even if you are not able to produce a really good result, it is still counted as a research experience in your application.

Joining the contest will be very interesting, but I will do a research with a professor as well. That's a safer path to make a good application.

Edited April 2, 2011 by the poisoned pawn

[DorkRawk]

Thanks for the answers. I've been working on some some research as well. I was just curious if something like this would be worth mentioning in a personal statement if I were to participate.

[BKMD]

I think it would help, but mainly if it gives you research experience in the process - coming up with something innovative is more important than doing well in the contest. Saying "I came up with a new and unique algorithm, and even though it did not perform as well as existing methods, I learned a lot about the research process" would be more convincing in your essay than "I tried every algorithm I learned about in class, and this one ended up working quite well."

[Azazel]

Don't expect to impress grad schools by participating in these competitions. My experience is that admission committees care little about them. You will be better served putting time and energy into doing good research, and acquiring strong letters of recommendation. Of course, good research and coding challenges are not mutually exclusive, but you probably only have so much time...

So I'd advise you to compete in these challenges only if you find them inherently fun. Don't expect them (even if you win) to work magic on your grad school application.

Edited April 4, 2011 by Azazel

[Azazel]

I think it would help, but mainly if it gives you research experience in the process - coming up with something innovative is more important than doing well in the contest. Saying "I came up with a new and unique algorithm, and even though it did not perform as well as existing methods, I learned a lot about the research process" would be more convincing in your essay than "I tried every algorithm I learned about in class, and this one ended up working quite well."

That's true. I'd offer two thoughts, however:

It's unlikely that your solution will be "innovative" in any non-trivial way. When you consider that all the coding challenge problems already have known solutions, how novel can your algorithm be? And the time constraints in these competitions tend to bias answers towards hackish solutions, or solutions isomorphic to well-known algorithms, not new theoretical contributions.

Even if you do come up with an "innovative" algorithm, how will your work be validated? When you talk about research results in your application, what you say will be backed up by a letter writing professor or (even better) a publication. Who (besides you) will speak to the originality of your work in a programming competition?

Just something to think about.

Edited April 4, 2011 by Azazel

[csperson]

Actually, I had never heard of google coding challenges...

But as far as I know, most people who got into top programs did not do google coding challenges.

It seems like the best way to get into a top program is to have a recommendation from a professor that the admissions committee trusts. That recommendation needs to talk about some research you have done (kind of in details). So I would just ask a professor if I can do some research for him instead.

Edited April 4, 2011 by csperson

[DorkRawk]

Ok, thanks everybody! The coding challenge just seemed seemed like it would be fun and I was wondering if I could kill 2 birds with one stone. I might still work on it, if I have time, just because I think I'd learn something but I'll stick to mainly research related stuff in my SoP.

[adinutzyc]

You can definitely mention it in your SOP if you have space, but just don't base your SOP on it. I did mention mine, but only as a line in my resume....