So this comment came up recently and I think it’s worth responding to publicly:
> Students keep asking me(probably other mentors too) on private message how > many tasks do they have to have to win. > I don't think we should keep a secret on how we judge. Obviously, we > shouldn't give them the whole algorithm as well, but there is something we > need to do (like an ibot ~ranking command) to tell them the number of tasks > alone is not even close to being the decisive factor. It’s not a secret. I’ve even commented publicly on the GCI mailing list about how all orgs rank. The rules state that it’s based on their "comprehensive body of work” which each org is allowed to interpret, but nearly all take to imply a basic summation over all tasks times their complexity and a valuation factor. It means participants need to complete a lot of tasks and they need to do their best work per task. So in pseudocode, the algorithm applied per participant looks like: for task in tasks ; do ranking += complexity(task) * quality(task) ; done To win, they will need to complete a lot of tasks and almost certainly need to be in the top-5, but only the top-10 may even be considered. In year’s past, the winners tend to average about 1 task per day. They should, however, focus on quality before quantity as that is the dominant variable that changes from student to student. Cheers! Sean p.s. Thanks to most mentors that filled out the spreadsheet. Only a few missing. Remember to give feedback on completed tasks and keep up the great work reviewing! ------------------------------------------------------------------------------ Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server from Actuate! Instantly Supercharge Your Business Reports and Dashboards with Interactivity, Sharing, Native Excel Exports, App Integration & more Get technology previously reserved for billion-dollar corporations, FREE http://pubads.g.doubleclick.net/gampad/clk?id=164703151&iu=/4140/ostg.clktrk _______________________________________________ BRL-CAD Developer mailing list brlcad-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/brlcad-devel
