Wrote an article - score it
 

Why dont you post it? I dont wanna click linkies.

It's friggin long tho, like 2k words.

Funny cuz i wrote an article with similar reference to Poisson distribution
http://www.buildinganempire.com/poisson2.html

Well, actually, you seem to missunderstand the math.

This is actually false. lambda varies with the number of visitors, k varies with the number of sign-ups, the probability depends on both.

Obviously, things fall apart from there on,

Which is patent nonsense.

Affiliates have to make the following business decission : I've sent X visitors to a sponsor with 0 conversions so far, how long do I give it before I pull the plug ?

With the connoted question of, how likely is this to be a mere statistical artefact, and how likely is it to be a bad fit for conversions ?

So, it's nice that you've tried, but I think you should try re-reading on the subject.

PS. Do the exercises too, it will help.

Priceless, your article, it turns out. Thanks for linking.

I always thought standard deviation is somehow related to the statistical variance, for whatever reason. But I guess we might as well re-christen variance and call it the average. Altho I always thought it's a mean not an average. But I guess we could re-christen means to be averages too.

Bloody hell I haven't laughed this much in months.

68%.

As this page says - "Under the conditions where the Poisson distribution is applicable, the standard deviation may be approximated by the square root of the mean." Is it the word "approximately" you felt is missing? Or maybe you would prefer "mean" over "average" but I think most people would understand average better.

Of course you have to make a business decision which is exactly why i stated that "No mathematical distribution can make a decision for you, but they can give you a very good idea of when enough is enough. The question then becomes, what probability level is enough for you?" with further elaboration both before and after.

solid F as in "fucking fat idiot". congrats

A mean is not the same as an average Colin.

Other than that, you're confusing lambda with k. in my formula, or e and m in yours, or a and x in the formula on this latest page you quote.

Let's get them clear :

lambda, e and a equal the EXPECTED number of occurences. Thus, for 1:100 conversions, and 900 clicks, lamba, e and a equal 9.

k, m and x equal the ACTUAL number of occurences. Thus, that'd be the number of actual sign-ups.

Now, the standard deviation, which is the root of the the variance, can be approximated as the square root of the mean of lamba, e or a, depending on which formula you use.

Which makes it the square root of the number of expected signups, not actual signups, as contrary to

Much more interesting are the points you don't answer to, namely

1. The false statement that signup probability is independent of the number of visitors,
2. The false statement that one should not think "I'll send X visitors" but instead "I'll send X sign-ups"

The entire point of using this math is to establish how many visitors to send, and under which feedback conditions.

And while you claim math can not make a decission for you, that is only because you don't quite know how to use it.

Properly applied, this bit of math will tell you, after having sent x visitors, if it's worth trying to send k more, or not.

You sure have a hard-on for me dontcha.

meds?
 

great article thanks:thumbsup

Whoa I can't believe somebody posted something nice in a thread of mine ?

Isn't this like in the prophecy of the apocaplypse and all ?!

Worthless as usual. Every word you type just sinks into an abyss.

Give BradM a stick, he'll beat himself silly with it.

bump for fat ass retard and more useless shit.

You're most welcome :-)