|
I've also taken part in the IOI - twice - and like you, am a self-taught programmer (I was when I took part in the IOI, I've since also studied computer science). In my opinion the IOI problems, at least for a competition, are pretty good as they focus on using correct algorithms and data structures as opposed to hacks or mathematical tricks. Brute force solutions almost never produces good results for IOI problems. Edge cases must usually be considered as the test cases used to grade the solutions try to be varied and cover as many cases as possible. IOI does nothing to teach good software architecture and I don't think any similar competition could ever hope to do that. It does, however, teach algorithmic thinking, use of data structures, understanding of time and space complexity and coding under pressure. (Well, all of these to a degree - like you said, recursion, branch pruning, trees and graphs and the like. more to be desired? Sure! But not a bad start, especially for high school students who are unlikely to have much exposure to problems which would otherwise have forced them to learn these things - I know people even now who, having only ever done business applications, would fail miserably at solving any IOI-like or algorithmically-heavy programming problem because they were never sufficiently introduced to the things the IOI focuses on.) All great lessons, even if there is more left to learn to become a great programmer. I personally dislike PE because the problems I attempted seemed like they could often be better solved by employing a mathematical trick or insight as opposed to clever use of algorithms, data structures or programming ingenuity. My programming skills are much greater than my math skills, so I guess ultimately, because of that, PE just doesn't do it for me. |