Spreading the risk

Estimating and allowing for risk can be seen to be a perfectly logical approach to dealing with uncertainties. But, it is only logical if you have some means of accurately estimating the uncertainties. What if there is so much uncertainty that you have no way of calculating the extent of the risks involved? What if there are risks that you aren't even aware exist? This is the reality of complex dynamic systems: especially those in a state of chaos.

In such environments, probabilities and the calculations for the element of risk can be no more than intelligent guesses. Values have to be assigned to probabilities on the basis of experience or common sense. But, experience and common sense can also be useless in some environments. Particularly in new environments that are subject to chaos. This is the environment of the Internet where rapidly changing technology and unpredictable competitive strategies afford no precise or justifiable basis for attributing discount multipliers to values or outcomes.

At first thoughts, it might seem that there is no strategy that could be effective in such an uncertain environment. Yet we do know that decisions are being made successfully in the similarly uncertain environment of investment and finance. In this environment successful strategies use a technique called "spreading the risk". In essence this is about making lots of decisions knowing that you cannot be certain that any particular decision is right but that the aggregate outcome is likely to be favorable.

This is similar to the strategy of the professional poker player who will know that he cannot win every hand but will be reasonably certain that he'll win over the course of many hands. However, the big difference between the professional poker player and the professional investor is that the poker player is able to calculate the chances with a fair degree of accuracy because there are only a few variables and unknowns. The professional investor on the other hand has an impossibly large number of variables, most of which are not calculable and many of which are not even known about.

The professional investor therefore has to have a more pragmatic approach to risk taking which might best be understood from another scene from the CD-ROM "How God Makes God":

"How is your latest business venture going?"

"It has gone bust."

"Broke again?"

"Yes. Stoney broke."

"You remind me of the professional gambler who keeps losing, and his friend asks him why he still keeps on playing when he never wins."

"What was his reply?"

"He said he couldn't give it up because it was his living."

"How very droll."

"But, why don't you give up the idea of going into business for yourself and get a sensible job?"

"Because statistically I stand a good chance of being successful in making money even though the odds would seem to be against me."

"That doesn't make a lot of sense?"

"Yes it does. If there are many opportunities and if I try enough times it is highly probable that I will succeed."

"If at first you don't succeed try, try and try again?"

"Mock me if you wish , but, the way I see it, just trying to start a business venture is more profitable than getting a job."

"Even if you do not succeed?"

"I did not say failed attempts have value I said the attempt itself has value."

"What value is there in an attempt?"

"In a mathematical sense there is a very real value. It is the value of the business if it is successful, multiplied by the probability of success."

"What is the value of your non existent business?"

"I base most of my business attempts on building up a business which will be worth half a million dollars."

"And what is your probability of success?"

"Each time I make an attempt I estimate that I have a twenty percent chance of succeeding."

"So how does that give you a value for each attempt?"

"To my mind each attempt is worth twenty percent of half a million dollars; that is one hundred thousand dollars a try."

"But that is not a real value?"

"In a mathematical sense it is very real. Quantum mechanics and the new physics are based upon such propositions and look at their success in making sense of the world."

"How can you earn money by keep trying out ideas and not succeeding? You could go on failing for years and years."

"That's highly improbable. Supposing ten people tried out ten different business ideas each with a twenty percent probability of success; what are the odds against every single one of them failing?"

"They could all fail quite easily."

"How easily? Come on, work it out."

"You are the mathematical expert. You tell me."

"Ah! This requires a conceptual twist that is not intuitive. This is the calculation that financiers and professional investors use. Try thinking about multiplying the odds of them not succeeding."

"You mean the odds of them succeeding?"

"No. The odds of them NOT succeeding. That's the trick. If you use probability theory to work out the chances of success by using the probability of succeeding you get the wrong answer because you have to multiply probabilities. This will only allow you to work out the probability of them all succeeding. So, you have to work out the probability of them all failing and take that answer away from one"

"That's confusing. You'd better explain in more detail."

"If there were only one man, the chance that he would fail would be four chances in every five. If there were two men, what would be the chances that both would fail?"

"Mmmmmm......

That's one of those probability things. You have to multiply the probabilities together. Four fifths multiplied by four fifths; let me get my calculator. I make it nought point six four "

"That's right. There is an eighty percent chance that one would fail but only sixty four percent chance that both would fail. Now tell me the probability of ten people failing; each with an eighty percent chance of failure?"

"Multiplying four fifths together ten times gives me an answer of just over point one."

"Yes. The chance that every one of them would fail is about one chance in ten. Now, you can turn this around to say it also means that the chance of at least on of them succeeding will be around ninety percent certain."

"A hundred percent less ten percent. That's right, but, what does that prove?"

"Well, isn't this the same as saying that if one man tries ten times he has a ninety percent chance of succeeding?"

"Even if at each try he only has a twenty percent chance of success?"

"Yes. He'd be ninety percent sure of winning through within ten tries.

"Oh! I see it now. But you are still broke?"

"Yes, but if I make a new try every year, I'm effectively earning one hundred thousand dollars a year."

In this dialogue, notice in particular how you have to work out the probability of all failures before you can get the probability for succeeding. This is vitally important for Zen-ness because, if you need to allow for failures, you have to have an idea how many failures you are likely to be able to absorb before you can be reasonably certain of success. This is not intuitive and needs to use this little trick.

The trick was taught to me by a scientist at the research establishment at Great Malvern when I'd been a student. I'd been really disappointed at being allocated to spend three months in his department because it appeared to be the worst possible placing. Everyone else was working on missile control systems, advanced radar projects and I had to spend three boring months testing components. I had to put hundreds of components in heated damp ovens and onto shaker tables for hours on end, then test every one to see if it was still working. How more soul destroying can life get?

Yet, this three months taught me a lesson that was to be of inestimable value in my entrepreneurial career - because I'd asked a naive question. I'd asked what was the purpose of the testing. I'd expected him to tell me that it was to help improve the design of the components. It wasn't. He replied simply, "To find out how many radio sets a pilot has to take with him on a mission"

I thought at first he was just joking, but, he was perfectly serious. It would be disastrous for a pilot to lose contact with base because in those days all radar was ground based and the pilot needed to be warned as to where the enemy were and if they were approaching. Because components at that time were not as reliable as they are today they had to work out how long the radio lasted when the life depended upon the failure of a single component.

If they knew the probability of the worst component failing they could then work out the probability of a failure of any one of many similar components failing. This would allow them to work out the probability of a radio lasting throughout the length of a mission. In certain cases, it was necessary to take two, three or even more where conditions were particularly bad. It was in the explanation of this that he taught me the trick of working with the probability of failures to get the chances of success.

In the CD-ROM "How God Makes God" I used a nineteenth century drawing of a mass execution to illustrate this point. It depicted fifty-four people being hanged on a multi storey gallows (it was actually from a black humorous cartoon). I'd animated this to allow one of the characters at random to suddenly stop wriggling (he'd died). The game was to keep running this animation and each time to guess which one of the hanging men would die first.

The problem was then to work out how many times you'd have to run the animation and make a guess before you could be ninety percent certain of getting a right answer. This problem cannot be solved unless you use the trick of using the probability of any of them NOT being the first to die. In the CD-ROM I could animate the formula to show how this calculation was made. In this book I'm afraid you'll have to experiment with it yourself. It is worth the effort though because it will greatly assist in Zenning winning strategies in conditions of uncertainty, change and competition.

The dialogue of the man discussing his business strategy also makes a number of other important points. Firstly, the spreading of the risk over a number of ventures greatly increases the chances of a success even though each individual venture might be quite risky. From a bankers point of view, the financing of a single venture could be prohibitively risky. The interest return, to cover the risk of an eighty percent chance of failure, would need to be in excess of four hundred percent for it to be viable to make a loan.

Looking at the financing of ten ventures. If one success could at least cover the cost of ten attempts then the banker would need to discount only ten percent of covering the cost of ten attempts. This transforms the financing of risky ventures into a reasonable proposition. Venture capital companies work on this basis. They will risk capital knowing that there might be a high failure rate but those that do succeed would cover the losses of those that fail.

Mapping this kind of thinking across to e-business and e-commerce, it is easy to see how the uncertainties and unknowns of the chaotic environment of the Internet can be allowed for. The risks can be spread across a number of decisions rather than banking on a single one. This applies not just to finance but to time and effort.

In the conventional physical world it is not usually practical to think in terms of multiple projects or multiple approaches to projects. In the world of the Internet this is not only practical it is the only sensible way to proceed. As already discussed, it is impossible for any single person or even a group of people to have complete knowledge of all the developments that are constantly occurring in the digital communication environment. This means that any company or entrepreneur who invests in a single approach is going to run a greater risk of making a wrong move than those who invest in several.

The final statement in the dialogue between the two men - "I'm effectively earning one hundred thousand dollars a year." - would seem to be a nonsense. However, in a theoretical sense that statement is quite correct. The man is effectively earning one hundred thousand dollars a year. It is not possible to see this as being a truism by thinking solely of this one individual, however, if one thousand individuals were using this strategy their average earnings would be one hundred thousand dollars a year.

It is this statistical approach to thinking that will be important in the chaotic environments of e-business and e-commerce, as shall see when we cover game theory in the next chapter.