Scientific Method and Porn Sites

I wake up, brew a pot of coffee, and remember that earlier in the week I was wondering why slow-roasted coffee is better than other methods, off to Google where I always go, the Oracle at Delphi of our society. But much better. Turns out slow brewing's not better. Just different. Real afficianados prefer other methods anyway. Should have known.

The oracle priestess is said to have always answered in cryptic answers. The Web again, but in Ancient Greece. I check my stats. Is this better than yesterday? Is this better than last week. Success, for me, is when today is better than yesterday. For some people, it's when today is as good as yesterday, but you have to be really big to have that kind of attitude. You have to be an Isaac Newton .. or a Babe Ruth.

Earlier, I discussed gaining one percent a day and exponential gains. How we spend so much of our time in the early part of the curve, waiting, waiting, for Father Time to spin the world a few more times. But the Universe doesn't give out one percent gains every day. She works in stop's and go's .. we might gain 25% in one day, and then stagnate for a month, and then EXPLOSION. 25% is a damned good day. A revelation of revelations.

The one percent solution is a model, it's not the real world as is, but it helps us to understand what's going on. A real graph of dollars per day is choppy, reflecting moments of revelation and heartbreak, discovery and accident. And in the end, an S curve is an even better model than an exponential one. As you get larger, it gets more difficult to achieve those gains. The curve will flatten out in the end.

The real world is usually quite different than how we've modeled it. In my notebook, I have a page called "Principles of Circle-Jerking." There are a lot of crossed out and edited remarks. The principles are getting better, more accurate, applying to a wider domain of situations. But the empty spaces in between the lines are just as interesting as the lines themselves.

I use the scientific method in my attempt to understand how things work. Why? Because it's too easy to get biased and think one understands what's really going on, when in reality, maybe today being a good day was just an accident. The aim of the scientific method is to remove bias and prejudice.

A simplified school-text book summary of the scientific method might be as follows:

1. Observe a Phenomenon
2. Formulate a Hypothesis
3. Test the Hypothesis
4. Reformulate the Hypothesis

Nothing is ever really that simple. But it's a good guideline.

The basic idea is this, one must have a good problem, something that bothers and annoys you. Why the HELL does THIS work THIS WAY? An example of a problem is this: Say I run a TGP and think that my site traffic increases fastest when I have a few big trades rather than a bunch of smaller ones. Now there are a lot of people that believe this. There are a lot of messageboard posts where people repeat this dictum and they repeat logic as to why it is so. The logic being that sending hits to smaller trades takes away from sending hits to your larger trades, and therefor your site traffic will decrease. But I don't believe anything I read on a messageboard. Psychologically, one is more likely to believe something when the word "because" is used in a sentence. There are good studies that report to show this. So you have to be really careful when your read "such and such is true because .. LOGIC." Reading someone's post on a messageboard like that, it doesn't fit the criteria of the scientific method. The poster might not have been careful in their analysis, they might have read it somewhere else. It might be an often repeated myth.

Now I don't KNOW whether the statement is true or not. I haven't analyzed it properly. I find it an interesting question. I have an intuition about it, based upon some mathematical properties I know, but I don't know if I'm right or wrong. I've witnessed a phenomenon and hypothesis, people saying that site traffic increases faster with a few big trades rather than with a bunch of smaller ones.

So I know now that I have to create an experiment to test the hypothesis. I haven't actually done this, so I don't know what really happens, so the actual results I mention now are all fictional. So very simply, I might kill off all my trades under a certain size and see what happens. And say my site traffic goes up that day. Now what? Well, here is the most difficult part. That doesn't necessarily mean that the hypothesis was right. Maybe my traffic NORMALLY goes up on that day of the week. Maybe the server gods were nice that day. Maybe I didn't have enough traffic for my data to be reliable. Maybe those people who's trades I killed, sent ME some traffic, but I didn't have to return any, so that traffic went to other trades, free traffic. There are a lot of possibilities.

So after I think about all this stuff for a while, and maybe design some better experiments, maybe I feel strong enough about it to elevate it to the status of theory. Maybe I would state the theory like this: Traffic increases faster when I have a few big trades rather than a bunch of smaller ones. That's not a very well stated theory. I haven't said what "few" and "bunch" are or for that matter, what "big" and "small" are. So if you think about that, it's easy to appreciate how little information is actually being conveyed in posts where people state something like this.

So now I try this idea on 3 of my TGP's and call up a friend and ask him to try it on his. After a week, my sites are increasing in hits but his shrink. "Ah ha!" Maybe this theory works only under certain conditions. In physics for example, Newton's laws have a domain of application. They work pretty good when you're not traveling really fast, or in a strong gravitational field. Similarly, our hypothetical theory might also have a domain of application. Maybe it works on sites above 12k. Or maybe the opposite. Maybe it works when the site pages are less than 100k in size. Or again, maybe the opposite. Maybe it only works on Sundays, hell I don't know. The point is that if you're using "the scientific method" the next thing you do is reformulate your hypothesis (or form a new one). Anyway, now you start the whole cycle all over again.

It's easy to get caught up in the chase for greater understanding, but you must also spend a good deal of time applying what you know. You could spend years trying to understand all the subtle nuances of this business and not get anywhere. The best in this business blend an excellent combination of real understanding with hard work - actually doing it.

It's also important to remember that this IS a business. Science lays out some good guiding principles that one may use to attempt to remove bias from one's observations and give a better understanding of certain phenomena, but there are many other factors that come in to play in business success than just these. Personal relationships with other webmasters, money management, some luck, and a gamut of other factors also play a large role. Each of us has an original blending of these factors and they make us who we are, what we are, and how much we make.