GoFuckYourself.com - Adult Webmaster Forum - View Single Post

Nathan · 08-30-2003, 04:38 PM

Choker,

first of all, to answer your question. No, it is not proof of shaving. Why? Because you have the simple problem that a test this short does not tell you anything worthwhile. Run it for 2 months, then it might be a good test.

You also did not give a lot of details on how you tested this. The way which would make the data somewhat reliable would be to send for example 20000 surfers randomly to the one and to the other affiliate code. Sending 10k to the first and then 10k to the next has WAY too many changes you do not account in, the 2 most important being:

a) Time of day of where the surfer is.
b) Network usage/congestion at the time

Your way of tracking if they count well is also very dangerous. Obviously, they will never count as many hits as you did. They will always have around 5% less simply because of network congestion and people closing browsers while all these damn Location headers make the browser hop all over the place.
Then, taking b) this can cause for quite significant changes in the tracking on the sponsors side because you might just happen to hit hickhups on the net.

a) is very important simply because people tend to signup more or less depending on the time of day. And a) of course also changes the basic mix of what countries the surfers are from.

Oh, and btw, [spam on] SexTracker do by far not have the best stats ;) They are buggy, slow, inaccurate and still have the same features as in 1998[spam off]

Give me some more details on how you did that test exactly to maybe workout a more exact setup to really get some good info from this kind of stuff.

08-30-2003, 04:38 PM
Nathan Confirmed User Industry Role: Join Date: Jul 2003 Posts: 3,108	Choker, first of all, to answer your question. No, it is not proof of shaving. Why? Because you have the simple problem that a test this short does not tell you anything worthwhile. Run it for 2 months, then it might be a good test. You also did not give a lot of details on how you tested this. The way which would make the data somewhat reliable would be to send for example 20000 surfers randomly to the one and to the other affiliate code. Sending 10k to the first and then 10k to the next has WAY too many changes you do not account in, the 2 most important being: a) Time of day of where the surfer is. b) Network usage/congestion at the time Your way of tracking if they count well is also very dangerous. Obviously, they will never count as many hits as you did. They will always have around 5% less simply because of network congestion and people closing browsers while all these damn Location headers make the browser hop all over the place. Then, taking b) this can cause for quite significant changes in the tracking on the sponsors side because you might just happen to hit hickhups on the net. a) is very important simply because people tend to signup more or less depending on the time of day. And a) of course also changes the basic mix of what countries the surfers are from. Oh, and btw, [spam on] SexTracker do by far not have the best stats ;) They are buggy, slow, inaccurate and still have the same features as in 1998[spam off] Give me some more details on how you did that test exactly to maybe workout a more exact setup to really get some good info from this kind of stuff. __________________ "Think about it a little more and you'll agree with me, because you're smart and I'm right." - Charlie Munger