Here's what they're testing with all the abstract/book knowledge - whether you can properly grow and understand the tradeoffs of common lower level patterns. Ha

Here's what they're testing with all the abstract/book knowledge - whether you can properly grow and understand the tradeoffs of common lower level patterns. Having no knowledge as to why say a hash table would be preferred over a binary search tree will lead to a person using one thing without understanding why. Yes, in general you'll be using the same few heuristics probably 90-95% of the time in real programming. In the cases in which you really need to optimize for max performance (because its company $$ on the line) or max consistency in performance (because its company/service reputation on the line) its worth it to find people who can do that. The cost savings of running just a few servers vs many is large, or the product could not be viable if it doesn't perform to a users expectations (see: anything in gaming). For your average application development shop using tools provided, you can do very well by just applying heuristic techniques towards your domain. If you want to get a real competitive advantage in technology, you need to be able to develop systems that are optimized to a point. Granted, in Jr. or normal dev roles you will not be making many of these decisions - often a more senior or prinicple engineer will be working on the mission critical stuff.