U.S. slips lower in coding contest

TalkBack 19 of 25:

Next »

« Previous

• Thread View
• Flat View

I'd tend to agree

When I was getting my CS degree many years ago, I participated in a couple of the ACM programming contests (never getting above state level). They gave the teams pamphlets of problems to solve, and they could strategize on how to allocate their team resources, and how much they wanted to devote to helping each other out. We were allowed to work on problems together, if we wanted to. The team that won was the one that solved the most problems correctly within the time constraints.

My problem was never with solving the problems, it was solving the problems within the time constraints. It was basically a contest in speed-coding. It didn't matter if the solution was sloppy from an algorithmic perspective, just so long as you got the problem solved in less than an hour or two. The time constraint tended to pressure programmers to the simplest implementation, favoring a hacker approach, no matter what it was. There was no grading of which solution ran faster, just that it worked (though I think the rules said that the program must finish running within 5 minutes, or something like that).

I was never good at speed coding, partly because our own university program was oriented the other way, towards more formalized approaches--abstraction, data structures and the like. For example, when I saw a problem that involved a variable-length data set with multiple atomic elements, requiring back-tracking occasionally, I thought "doubly-linked-list". It took me longer to write (not in on time), but it worked. The winning solution just read in the data set one character at a time until the end-of-line character was reached (not sure how back-tracking was accomplished, maybe with a call to fseek()).

When I was in high school I participated in the American Computer Science League (ACSL) contests. They took a different approach for the local contests. Instead of giving students problems where they had to come up with solutions in an hour or so, they would give us a week to write up our own algorithms (or write code on paper), to basically figure out the problem in our heads, and then at a prearranged time, we were given an hour to code and test on computers, and then our results were tested. We had written tests as well, testing concepts in computer science, like Boolean algebra. We had study materials for that. I did very well with these kinds of tests. Our computer club, which participated in the contest, generally scored in the top 10 for schools in our region, and we went to the national finals every year. I think we might've scored in the top 10 nationally one year, but generally there were a lot of other schools that did better than us there. The national finals were done the same way as the ACM contest. You get the problems, and then you have an hour or two to solve them.

The way the contests are conducted tend to favor people with certain cognitive abilities or technical orientation.

Posted by: Mark Miller   Posted on: 04/09/05 You are currently: a Guest | Members login | Terms of Use

Alert moderator to an offensive message

Subscribe to this discussion via Email or RSS

SponsoredWhite Papers, Webcasts, and Downloads

• Busting the myths about QuickBooks Enterprise Solutions and IBM Smart Business IBM So you already know there aren't actually any alligators in the New York ... Download Now
• ANTI-SPAM SOLUTIONS - JANUARY 2009 Trend Micro IT decision-makers now have a multitude of email solutions to consider. ... Download Now
• Secure Copy Version 5.0.2 ScriptLogic Secure Copy? is a powerful, comprehensive migration solution that ... Download Now