User:Jeblad/social Kanban

From mediawiki.org

It is possible to make a kind of Kanban-like social/cultural crediting system by creating a limited report of contributions in time and context. The report can for example be limited to a couple of weeks time and the topmost 5-10 projects within some context. Such contexts can be the topmost articles where the user have contributed the most, categories where the user have contributed or articles linked to from some project page.

The limited timeframe will force the user to completes the tasks instead of leaving them unfinished to get a high score, if he leave tasks unfinished he will get a lower score. This leads to a kind of Kanban system where only focused contributions gets the highest social score. This isn't really about the finished state but rather a measure of focused contributions.

There are several variations on this scheme. One approach is to calculate a mean or median for the upper fixed number of mostly edited articles and normalize against other contributions. A more focused behavior will get a higher overall score and the user will be ranked higher. Another one is to remove sporadic contributions and estimate the standard deviation and compare that to a fixed limit on the topmost projects.

By manipulating the scoring mechanism various facets of social an cultural interactions can be measured and produce feedback. It is for example possible to score according to retention rate for newcomers, were a user gets a score according to his ability to transform newcomers into good contributors. His part in each newcomer could be measured by a coincidence analysis of subject and talk pages.

If a user keeps adding welcome templates without actually helping the newcomers the effort will lead to a pretty low overall score as many newcomers will leave. Those involved in dialogue with newcomers that stays and starts to produce content will rank better.