Conclusion

Having done it so many times, I no longer regard optimization as a chore. In fact, I quite look forward to sitting down with something I've developed and exploring the performance bottlenecks.

One great motivator is to record your progress as you go. Write down how many frames per second you're getting at the start, and make notes as you go of the impact of your optimizations. It's fun to see just how much progress you can make. You can also make a little high-score list to see which optimizations have the most real impact.

Above all, keep your eye on the ball. You'll find that most of the time the greatest speed increase will come from rethinking the method you're using, rather than endlessly nitpicking code. Remember, the fastest line of code you can write is the one that never gets called! (Don't think about that for too long; your head will hurt.)