Wikipedia:Counter-Vandalism Unit/Vandalism studies/Obama article study

A study to see whether leaving Barack H. Obama open to IP edits hurts or helps the article.

Background

Mr. Obama is in the news daily as a 2008 Democratic presidential candidate. As such, the Wikipedia article gets many views and edits each day. The article is the second result on a Google search for Barack Obama. Because the article is so highly trafficked, it is the target of much vandalism and nonsense edits by anonymous IPs. However, the article is on many watchlists, and vandalism is often reverted quickly. Because of the sheer number of edits though, this is taxing to those who have the article watchlisted. There has been discussion at the article talk page (now archived) about the merits of semi-protecting the article. The article is currently semi-protected (on March 17, 2007 by User:Steel359).

This page analyzes the edits made during the period of January 1, 2007 to March 9, 2007. Edits that are blatant vandalism, nonsense, or otherwise unhelpful (e.g. non-neutral edits, edits that give undue weight, etc.) are noted.

Data

IP edits

Date	Statistics
	Total edits	Unhelpful edits	Time unhelpful edits were visible (hh:mm)	Diff	Notes
March 6	8	5 (62.5%)	00:36	[1]	Article was semi-protected at 16:29
March 5	19	12 (63.2%)	00:55		One edit took 39 minutes to fully revert. Article was un-protected at 08:39.
February 21	6	6 (100.0%)	00:15		Article was semi-protected at 03:35.
February 20	7	2 (28.6%)	00:23
February 19	8	5 (62.5%)	00:06
February 18	2	1 (50.0%)	<00:01		Semi-protection on February 11 automatically expired today at 16:20
February 11	18	14 (61.1%)	00:21		Semi-protected at 16:20
February 10	16	15 (93.8%)	01:12		Lots of vandalism as his formal announcement for the presidency happened today. One edit took 31 minutes to catch; only one edit was not unhelpful. Semi-protected at 21:00 (expired at 00:00 February 11).
February 9	10	6 (60.0%)	00:13		One unhelpful edit, which took 8 hours to catch, was not counted towards the time total because it was so trivial. Another IP editor continually inserted a sentence in the lede, but these edits were not counted as unhelpful since after some tweaks by other editors, the sentence remained intact.
February 8	24	16 (66.7%)	00:54		Lots of vandalism but a fair amount of helpful IP edits as well (some of whom pitched in to remove vandalism).
February 7	29	20 (69.0%)	01:05		Large spike in edits. One edit took 30 minutes to revert. The use of undo instead of full reverts was sometimes problematic as it led to some vandalism slipping through and remaining on the page for longer than usual.
February 6	3	1 (33.3%)	00:09
February 5	7	5 (71.4%)	00:12
February 4	3	3 (100.0%)	00:37		Spam edit took 24 minutes to revert.
February 3	9	5 (55.6%)	01:37		One edit, which was reverted an hour later, could possibly be construed as unhelpful, but its change was too trivial to mark as an unhelpful edit. Another edit was not a clear case of undue weight and was not marked unhelpful, alhough it was reverted an hour and half later. One unhelpful edit inserting unsourced trivia was not reverted for 1 hour and 28 minutes.
February 2	12	12 (100.0%)	00:29		No helpful IP edits today. Almost all of the IP edits were petty vandalism.
February 1	19	14 (73.7%)	01:33		One tiny sneaky edit took 28 minutes to revert.
January 31	5	5 (100.0%)	00:34		Unprotected at 19:07. Four of the five unhelpful edits were made in good faith, but not constructive.
January 22	14	9 (64.2%)	00:19		Semi-protected at 21:53.
January 21	19	15 (78.9%)	~03:01		Article was unprotected at 21:39 on January 20. Contentious edit warring mixed with vandalism made it difficult to determine exactly how long unhelpful edits were displayed on article.
January 3	10	8 (80.0%)	00:09		Semi-protected at 10:22. A removal of a comment, while unhelpful, was not counted towards the total time (it was reverted 4.5 hours later). Undue weight edit, while unhelpful, was not counted towards the total time (it was reverted eight hours later).
January 2	16	13 (81.3%)	01:34		One unhelpful edit took 28 minutes to revert. One unhelpful edit was too trivial to count towards time total.
January 1	6	5 (83.3%)	02:11		One nonsense edit took 27 minutes to revert. One non-neutral edit took 1 hour and 22 minutes to revert.
Totals	270	197 (73%)	-

Non-IP edits

Date	Statistics
	Total edits	Reverts of all unhelpful edits	Unhelpful edits	Time unhelpful edits were visible (hh:mm)	Diff	Notes
Month Day	0	0 (00.0%)	0 (00.0%)	00:00
January 14	34	1 (00.0%)	1 (00.0%)	00:01		A plethora of major edits by User:HailFire to rewrite and improve the article.
January 13	14	1 (07.1%)	1 (07.1%)	00:06
January 12	4	0 (00.0%)	0 (00.0%)	00:00
January 11	7	1 (14.3%)	1 (14.3%)	00:02		The unhelpful edit was an accidental deletion of text by a copyeditor.
January 10	7	2 (28.6%)	4 (57.1%)	00:15
January 9	73	6 (08.2%)	9 (12.3%)	01:02		Flurry of activity as a result of the article's FAR filed on January 8. Unhelpful edits were not related to the FAR.
January 8	5	1 (00.0%)	1 (00.0%)	00:02		Article was nominated for FAR at 18:45.
January 7	10	0 (00.0%)	0 (00.0%)	00:00
January 6	6	0 (00.0%)	0 (00.0%)	00:00
January 5	1	0 (00.0%)	0 (00.0%)	00:00
January 4	11	2 (18.2%)	4 (36.4%)	02:23		The four "unhelpful" edits were made in good faith, but were redundant and detracted from the article.
January 3	23	11 (47.8%)	2 (08.7%)	08:10		Semi-protected at 10:22. The two unhelpful edits were insertion of trivia.
January 2	13	10 (76.9%)	0 (00.0%)	00:00
January 1	13	4 (30.8%)	8 (61.5%)	00:36		All eight of the unhelpful edits were spam by an editor intent on self-promotion throughout Wikipedia.
Totals