Yea that's the fault of the scoring method (Count pipes). Character count would be better... either that or play within the spirit of the game and learn the tools for their intended purpose. I use a lot of awk/sed for quick one-time processing and I use a similar mental model for how I do most of my processing. Build a input stream, use sed to structure the data into records, use awk to process records, massage the output (e.g. sort/uniq). In many cases I could use sed or awk to replace eachother but that makes it take longer to visualize and also harder to re-use.