Hey,
I don't think you're getting my point which is that it is so dynamic that your results from your process today are not going to be the same as tomorrow and your process is always evolving so you can't make a mathematical equation for it that would hold up any longer than say a couple of days.
Hey brother! I'm risking to be pedantic here, but I'd love to reply again. I got perfectly what you said and I have to repeat and expand what I wrote before. Your post ISN'T an answer to my point let alone to OP's question / demonstration about aff marketing being or not worth your time to make a profit instead of publishing something. In other words, he's asking if it wouldn't be the case to create something -under any circumstance- instead of promoting products made by someone else.
And I already wrote that his demonstration is flawed and why. Hypothetically if he had added a few more considerations and had demonstrated that actually it's statistically less profitable than publishing some own shit HE WOULD HAVE BEEN RIGHT. And this would PROVE that if you make some money as an affiliate, it's just a fluke. Plain and simple. It would prove that you can make money, but it's not a sustainable model. Again, very similar to math models of games, or bets on races where the variables of today are not the same as tormorrow. A lot of this "mathematical debunking" has been made in the past in basically any field of our existence.
Now, everyone saying it's "the most stupid thing" should demonstrate why it's so "stupid", btw... And the point of OP is NOT stupid at all.
Back to your remark. If the variables in a system change dynamically, again, your math model can take care of it and elaborate a way to statistically predict what's gonna happen. OP was simply using a linear system, but I'm comfortable with expanding this to something a bit more complex. Weather forecast is a good example of this on top of my head, I already mentioned gambling on different things, but the list could really be 3 pages...
Even in a system where conditions change (by the hour, the second, the week or the month as you're describing: again it's irrelevant) you can still formulate a reliable analysis. A reliable probabilistic prediction as it happens all the time. Again, that's why the mathematicians dealing with economics predicted credit crunch while the "standard" economists hoped for the "hidden hand of free market" to fix things... Damn you M. Friedman.. lol
So to wrap this up. Again, it doesn't change anything if the system is gonna be different today or tomorrow or in 10 years. It's just a matter of putting the reality you're observing in the right "grid", formulating a model with the right variables, be them set in stone or dynamically changing. Be the system a simple linear one or a probabilistic or stochastic one. The quoted remark still doesn't prove:
a. OP's doubt. Aff marketing is statistically less profitable than being a publisher no matter how many products you're gonna put out and when
b. Your assertion. If the variables in a system change you CANT't really say what's gonna happen in a certain time frame from now.
Both these alleged points not only are not proven, they're false. Be this IM, weather forecast, stock or molecular analysis.
Regards.