On Wed, May 16, 2012 at 10:58 PM, Gwern Branwen gwern0@gmail.com wrote:
On Wed, May 16, 2012 at 10:49 PM, Anthony wikimail@inbox.org wrote:
First shouldn't we guess as to what percentage of the links were actually good in the first place?
I must say, I didn't expect to see someone rationalizing the results even *before* they happened.
No need to get personal, I wasn't rationalizing anything.
But no, you don't need to guess: you edit Wikipedia
I do?
If you don't, you can go click on Special:Random 10 times and ask yourself, 'would I delete the last link in the External links section?' If you think 2 links are rotten, then perhaps you should be predicting that - since everything is well, and any result is acceptable, and the status quo is perfect - only 80% of the edits will be reverted.
I certainly wouldn't try to make a prediction about the percentage of links which are bad based on a biased sample where each link was the last one in the External links section.