Wikipedia talk:Wikipedia Signpost/2015-07-29/Recent research

Average Account age by year
If we weren't recruiting any new editors the community would be getting a year older each year, so six months means we are still getting new editors. On the face of it this looks very healthy, after our initial period of exponential growth when the average editor had been here less than a year we are now getting a bit more experienced. We are still getting lots and lots of newbies, but with community size broadly flat we are only keeping as many as we lose. My suspicion is that if we combine this with other measures we would find that despite a steady inflow of newbies we are broadly stable, but as we have no way to work out our twenty let alone fifty year retention it is hard to know how healthy this is. Logically a new volunteer endeavour whose founding generation skewed very young will continue to get older and more experienced for decades, at some point if all goes very well indeed, the experience gained annually by the remaining community getting a year older will balance the experience that is dying, retiring or being blocked, and the number of experienced editors lost will match newbies joining. But an organisation barely 14 years old should be decades from such stability, even if we knew editors ages, or which former editors are still alive, or which "newbies" are anything but. But nice work, thanks for doing this.  Ϣere Spiel  Chequers  22:19, 3 August 2015 (UTC)


 * how charmingly Hegelian. a darker view would be that there needs to be twice as much new editor retention to get to a steady state. this is a long term trend that no amount of happy talk or software change has budged. i don't see any measure of increased productivity of the remainders compensating for the decreased numbers. when the work load and backlogs are growing, it is hard to imagine any reasonable person calling it healthy. actually fixing cultural problems is too hard, and so the community is resigned to the status quo. Duckduckstop (talk) 19:41, 5 August 2015 (UTC)
 * Errh, aside from the issue that edit filters, faster vandalism reversion, the move of intrawiki links to wikidata and indeed the rise of wikidata all mean that we can't compare current editing levels to past ones and we don't really know when the true peak was; If you do worry about raw edit count, the first 6 months of this year have all had more editors doing more than 100 edits that month than the same month a year earlier. That doesn't mean there aren't hard cultural problems embedded in the community and in the environment in which we now operate. One, two or three months increase year on year could still be consistent with a long term trend - a single month could easily be down to comparing one month with 5 weekends against one with 4. But 6 months in a row showing an increase is not consistent with the theory that "there needs to be twice as much new editor retention to get to a steady state".  Ϣere Spiel  Chequers  19:20, 13 August 2015 (UTC)