User:Teratix/Quantifying instruction creep



Wikipedia has an instruction creep problem. But how severe is it, what must we do to alleviate it, and how will we know if we succeed? To answer these questions, I have sought to quantify both the rise and current extent of creep, examine how Wikipedia's creep measures up to comparable sites, and proposed ways to evaluate our progress in eliminating creep.

Although creep can degrade established editors' experience of Wikipedia, it has its most severe impact on new users still seeking to understand the project's policies.

Which pages to evaluate?
Before we can evaluate the extent of creep, we must decide which pages are to be evaluated. One extreme answer is to include all of Wikipedia's policies, guidelines and essays (which I will collectively dub instructions). However, even the most experienced editor will never need to read every single instruction. Therefore we must distinguish between core instructions and non-core instructions. Core instructions are required reading for most, if not all, editors. Non-core instructions might be helpful for users editing in a particular area, but are not necessary for most users to understand. Obviously it is preferable if all instructions are kept free of creep, but it is particularly vital for core instructions, as they are intended to be understood by all editors, including new ones. Petscan policies Petscan guidelines Petscan information pages Petscan supplements Petscan how-tos Petscan essays (!)