Jump to content

Wikipedia:Wikipedia Signpost/2015-03-11/Op-ed

From Wikipedia, the free encyclopedia
Op-ed

Why the Core Contest matters

Casliber is a reviewer, administrator, former member of ArbCom, and coordinator for the yearly Core Contest, as well as, with more than one hundred featured articles and countless good articles under his belt, one of English Wikipedia's most prodigious editors. We featured an interview with Casliber to commemorate this fact last May; here he presents his views on core content on Wikipedia through the lens of the annual Core Contest, and why you, too, should participate.
In 2011 TCO published his findings on progress in the quality of core articles on Wikipedia. His findings, summarized in this PDF and available in full here, were that most of our most important articles have barely improved since at least 2008.

Wikipedia was in many ways a very different place a decade ago, but in some important ways it was also very much the same. In October 2004, user Danny created the page for the first of three eponymous "Danny's contests", one of Wikipedia's very earliest organized content drives. The three iterations of Danny's contest—focusing, in turn, on creating new articles, destubbing, and featured content—would see editors rewarded for their work in key areas of Wikipedia by hundreds of dollars in Amazon vouchers. The third contest, held in September–October 2006, was the most innovative of these, focusing on improving the quality of established broad articles and in particular on trying to correct or influence the flow of featured content. In introducing the contest, Danny presaged many of the changes that Wikipedia would undergo, stating:

It was on these principles that the contests' spiritual successor, the first English Wikipedia Core Contest, was organized in November 2007. Running from November 25 to December 9, 2007, the Core Contest presented its rationale in its introduction, stating that "we all acknowledge the ideals of quality over quantity and the vital importance of core topics - yet how many really key articles do we each know of in really poor shape? ...so to improve [on] this situation we are announcing a two-week-long contest focusing on Wikipedia's most important articles." Danny and several other users had begun development on Veropedia at the time, an early Wikipedia content scraper which solicited recommendations on high-quality Wikipedia articles from editors for the purposes of static re-hosting, a motivating factor in their assistance in organizing this newer, broader effort. Mirroring the negotiations that still take place with broad community initiatives today, the project generated extensive discussion in late 2007, with the greatest topic of concern in particular being sourcing the monetary reward. This was at first to be fulfilled by Danny again, but after a delay in sourcing it (according to speculation, due to the condition of the success of Veropedia) the winners were finally announced and their prize money awarded on November 25, 2008, with the prize money supplied by Proteins. But despite the success of this first iteration of the contest, the bumpiness of actually awarding the winners discouraged future versions, and so the project went on an indefinite—and seemingly final—hiatus.

I have been interested in contests and games as a way of promoting content-building on Wikipedia for as long as I've been an editor here, and in a particularly glib moment in 2008, I started drafting the Flaming Joel-wiki award, a wiki-award offered to editors who improve one of the many subjects mentioned in Billy Joel's eclectic song "We Didn't Start the Fire". When I stumbled across relics of the Core Contest page in late 2011, I immediately saw value in this project and began the process of reviving it. The community scaffold that keeps Wikipedia running had gone through quite a bit of changes and improvements in the intervening time, so I was able to solve the funding issues which were so problematic in the first effort through application to the Wikimedia UK's microgrants program, which provided enough for a modest but sizable prize.

I decided to use vouchers to steer away from a direct cash incentive, hoping that that would lead to more scholarly and Wikipedia-related purchases on the part of the prize-winners, and I chose Amazon again because I suspect that any winner of such a contest could find something of use to purchase through them. I think that a prize on hand as an incentive is an important thing to have: they are a nice concrete gesture for the hours of work that some folks put into the place and a way to move away from sticks and towards carrots in steering featured content quality and focus.

I have run the competition on four occasions since then: March 10 to March 31 2012, which saw £250 in Amazon vouchers shared by six editors; August 2012, which saw the same prize shared by seven editors; April 2013, with the prize shared by three editors; and 10 February to 9 March 2014, with the prize shared by five. Each time the prizes have come from a WM UK microgrant, and buoyed by the success of this program, I resurrected the Stub Contest as well. Each time the contest has run, I have been impressed by the work that has been done, with the top 2 to 3 entries of each contest being particularly memorable. I almost hate singling out a favorite article, as that would mean omitting others I see as being just as important and enjoyable, and I invite readers to take a look at the diffs in the entries section of the contests to see first-hand what it is that I find so exciting about this program.

The featured article process is becoming ever more rigorous, but while this rigour is improving the quality of the articles we generate though the process, it is at the same time leaning heavily in favor of smaller, more esoteric, and more narrowly-focus articles more easily navigable through the straits of featured article candidacy. I continue to be excited about the Core Contest because I see it as a way of encouraging the expansion of broad articles that are typically neglected by our article improvement incentives, a problem that, though it first emerged in Danny's time, has only become more and more stark today. In examining the edit histories of many of the articles brought before the contest, I notice that the majority of our coverage of broadly-constructed topics, those most critical to our success as an encyclopedia, have seen little in the way of substantive community improvement over the years; except in the cases where specific editors make focused drives to bring an article to good or featured status, our core articles as they appear today were mostly written long ago, their content having changed for the most part only cosmetically in recent years. Though the times and the context we edit in have changed, the central principles of the Core Contest remain the same as they were when the contest first ran: to improve the encyclopedia where it matters most yet sees it the least. I see Wikipedia as being at a crossroads: the novelty of being newfangled is wearing off, replaced by the rigour of guidelines, restrictions, and rules that have proven essential in the evolution of Wikipedia. I believe that Wikipedia is traversing a grey area, where the goal is status as an established and reliable online encyclopedia, and we need to strive to ensure our core content is being improved along the way.