Wikipedia talk:Votes for deletion/Gag on my Cock

From Wikipedia, the free encyclopedia

[edit] Alexa comparison

Last link in alexa's top 500:

(Insert massive gap here)

So either this website is slightly less important than the most active alternative new source on the internet and slightly less important than the Unites States Department of Justice, or Alexa doesn't work well for porn. You decide.
brenneman(t)(c) 15:25, 4 September 2005 (UTC)

I think that's a false dichotomy. Alexa suggests that this is a rather popular website. A check on the Alexa top 500 tends to refute the suggestion (which I agree was plausible) that half the top sites by alexa count are porn sites--indeed very few of them are. All we're using it to decide is whether it's a site that lots of people visit. Clearly it is. Whether it's more "important" than any other site is not a question Alexa can answer, and not the question that is asked. --Tony SidawayTalk 18:05, 4 September 2005 (UTC)
  • Sampling the top 500 does not refute DS's theory, as website popularity has a pareto distribution. As it it not (to my knowledge) possible to sample based upon Alexa ranking (i.e. choose rank 5,178) a sensible alternative is to compare the Alexa of websites whose popularity we're familiar with. There is a clear dichotomy. To state this another way: What evidence would be enough to convinve you that Alexa was not a reliable indicator of pr0n popularity?
    brenneman(t)(c) 23:55, 4 September 2005 (UTC)
  • It doesn't totally refute Doc's theory, but it blows so many holes in it that really we need some strong external evidence to support his thesis that 50% (or 90%, depending on which statement by Doc we're discussing here) is correct. Since the Alexa evidence does show that the website in quesiton gets a lot of hits--some 200-400 per million over the past two years and fairly consistently in the top 4000-6000 websites by daily traffic rank and with a fairly consistent 5-15m page impressions per day. In short, this is a mind-bogglingly successful website.
  • I think it would be nice if, in attempting to support the claim that this isn't a very popular website, you could demonstrate that the Alexa figures are inaccurate and in fact far fewer people visit the website than Alexa claims. Because that's what it would take. --Tony SidawayTalk 00:44, 5 September 2005 (UTC)
    • Firstly, let us try to avoid any straw men: The fact that 50% of the websites with Alexa ranks higher than GOMC are porn is not the question.
    • The question is: Is Alexa valid for porn?
    • Alexa is a sampling method. It is not on every browser, it attempts to model larger trends from those computers who have the toolbar installed.
    • And it's not terribly accurate for sparse populations. [1] Good enough?
brenneman(t)(c) 01:16, 5 September 2005 (UTC)

Well this is very odd, I must say! Doc's argument was based on the claim, which doesn't seem to be supportable, that at least 50% of all websites more popular that this one are porn sites. And now you tell *me* it's a straw man? Have you told Doc? Have you, indeed, removed your claim that Doc is correct on this?

On your point that Alexa isn't valid for porn, well on the evidence given it isn't valid for *anything*. Thanks, I'll be sure to refer to this next time someone goes on about Alexa rank. --Tony SidawayTalk 02:25, 5 September 2005 (UTC)

  • If you choose not to argue that For a porn site an Alexa rank of over 4000 does not indicate notabilty I'll take it you agree and urge you to change your vote to Delete. ^_^;
  • Please understand that this still leaves a wide range of subjects for which Alexa rank is clearly an excellent indicator.
  • Just say saying "Delete - low Google" is incorrect, so in "Include - low Alexa". There are shades of grey, problems with systemic bias, etc. I'm sure that we agree that if this website is notable, it should be included.
    brenneman(t)(c) 02:49, 5 September 2005 (UTC)
  • Sorry if this is a private quarrel but, since I started it, my 2c. My comment was just that, a comment - not a theory based on any research (I did none - and still have no desire to count porn on Alexa). My wider point still stands, if a double glazing company had an alexa of 4,000 odd, that would be impressive - and perhaps the most popular site of its kind on the net. But a porn site will surely have a great deal of alternatives with better ratings (I don't think Tony would refute that) - thus this site is not in the top 50, or perhaps even to 500 pornsites. Further, the article's only claim to notability is its site: my mythical glazing firm is an established business, which happens to have a notable website too. Personally I'd need a very high alexa reading, or some other claim to importance ('the first of its kind', or notably contravertial, or endorsed by G.Bush) before I'd think it worthy of an article - thus my delete stands. (I'm confused as to what point Aaron is making - if alexa is unreliable for porn, then the rating would be irrelevant, and my 'theory' would surely fail.) --Doc (?) 09:03, 5 September 2005 (UTC)
    • We're just attacking it from different angles. I say "Alexa doesn't apply here for reason X" while you're saying "Alexa doesn't apply here for reason Y". Same outcome. ^_^
      10:30, 5 September 2005 (UTC)

OK, I get it. If you're right - then I'm wrong. But it's academic as in either case nn so delete. --Doc (?) 10:34, 5 September 2005 (UTC)

  • No, I think both of you are grasping at straws. Yes, Alexa numbers are ropey (as are google numbers, for different reasons), so we'd be silly to rely on them for scientific purposes. I don't have a definitive answer to the question: is a site that shows a four-digit Alexa rank a highly popular website? From the look of the figures over the past two years (pretty consistent) I'd say that most probably it is, but that's just a hunch. So that's it, I cannot use Alexa figures to convince anyone that a site in that range is or is not popular. It doesn't matter whether the site is a porn site or any other type. The sampling method gives unreliable results.
  • Does this affect my vote? Not really. This is apparently a very popular website and I'm prepared to accept the Alexa figures as a sanity check on my commonsense estimation. Given the popularity of web-based pornography, our coverage of the subject is lamentable, possibly because of misplaced puritanism, so I don't see any problem with voting to keep this site. If I encountered the name I think I'd like to be able to look it up on Wikipedia and get a neutral description, rather than visit some porn site and get a commercial schtick, though that name, I admit, leaves little to the imagination. I suppose it's useful at least to know whether the title has, as at first I thought, gay connotations. --Tony SidawayTalk 14:20, 5 September 2005 (UTC)

Comment. Brenneman's statement got me curious: just how many porn sites are there in the Alexa 500? I couldn't find out the answer directly from their website (although they offer a list of the top 10,000 sites for a low $999.--), but here's some anecdotal results from reasonably informed querrying on their website.

  • In the top 100 for English-language sites, the only porn site in the top 100 is "pornaccess" at 98; overall, it is ranked 201, so unless there is a porn site in the top 10 or 20 for German, French, Japanese or Chinese, I'd guess this is the #1 ranked site. (From Alexa's list of related sites, "proadult" is ranked #11,598, "sweetvideos" #161,603, "pornbooth" #525,421, & I lost interest at this point.)
  • For likely candidates, here are four results:
    • Playboy - #859
    • Penthouse - #8424
    • Hustler (dba online as Larry Flint publications) - #4084
    • Playgirl - #45,357
  • Let's try some famous names:

Sex might sell, but I'm not finding solid evidence that it helps your Alexa ratings: in fact, from this admittedly unscientific survey, I'd say either (1) any porn site that cracks the top 20,000 is notable, or (2) Alexa ratings aren't useful for establishing notability. Me, I'm leaning towards conclusion #2. -- llywrch 20:57, 5 September 2005 (UTC)