People don't think that they hold a particular belief just because they want to hold it. They are generally unaware that the evidence can also be looked at in a different way or there are other facts to consider. The process of getting from evidence to belief itself is biased by our goals.
For example - when the initial evidence supports our theory, we are happy to stop there and ask no further questions. But when the evidence is against it, we go deeper. We attempt to either find evidence to discredit the first hostile evidence or discredit the source of the evidence. By asking more and more questions in such a biased manner, we dramatically increase our chances to discover evidence which is comforting to us.
Reality is complex and any situation you can think of is fraught with complexity and open to alternative interpretations. Look at the situation of Herbalife (HLF). Is it a pyramid scheme ? Will the government step in and destroy its business model ? Bill Ackman seems to think so. But Dan Loeb and Carl Icahn do not share his beliefs and disagree with him. So much so that they are willing to take the opposite trade. Ackman has done his due diligence and the other two managers know about his conclusions but they do not agree with it. Assuming that all three have exactly the same evidence - the difference in conclusion supports the theory that there are alternative ways to interpret the data.
A different way is that people use different criteria to evaluate evidence and come up with a conclusion. For example, people have a very inflated opinion of themselves. They, on an average, think of themselves as better looking, more moderate, less racist, more open minded, better drivers, and better lovers.
A Greek philosopher took on a student to teach him rhetoric so that the student could practice law. In return the student would pay when he won his first case. After his studies the student decided against practicing law. The philosopher sued the student and claimed that now the student has to pay him his dues in either case. If the student wins then he must pay because of the contract and if the student loses then he must pay by the order of the court. The student countered that if he lost then he would not pay because of the contract but if he won then he would not pay because of the order of the court.
Not surprisingly, careful people give more weight to being careful, savers give more weight to savings, beautiful people give more weight to beauty while humble people give more weight to humbleness. Everyone has their own idiosyncratic idea about what criteria needs to reach a particular conclusion.
In an experiment a group of university students were asked to rate the importance of academic skills (e.g., math, public speaking) and personal skills (e.g., creativity, meticulousness) in terms of how important they are for success in college. The students were also asked to rate their own ability on those characteristic. Not surprisingly, the characteristics where the students excelled were also the ones they rated highly for a successful college life.In a paper by Dunning, Meyerowitz and Holzberg, it was found that the predisposition to rate oneself better than the average was reduced when people were required to use specific definitions for each trait. This is probably the most important part for investors. When you have a clearly defined set of definitions and rules - you will have smaller risk of getting biased by your inherent need to “fit data to reach a specific conclusions.”
Juggling criterias is one of the ways in which we get sidetracked into investing into a bad stock. Even after having a checklist of criteria (say circle of competence, financial strength, management, moat, and margin of safety) we make mistakes in reaching to a rational conclusions. Do we prefer strength over management ? Moat over margin of safety ? Or everything is equally important.
Consider the case of a woman who has three equal weighted criteria to select a car - look, price and power. Consider the ratings of the cars as shown in the table below (1 represents best and 3 represents the worst).
We discover something very curious. Let us compare Car A and Car B. Car B is better in Price and Power compared to Car A. It is a no-brainer that she should prefer Car B compared to Car A.
Comparing Car B with Car C she observes that Car C is better in both Look as well as Power. So, she should prefer Car C in comparison to Car B.
But comparing Car C with Car A shows that Car A is better both in Look as well as price. So, she should prefer Car A over Car C.
This paradox is known as the Condorcet paradox.
Summary: B is better than A, C is better than B but A is better than C !
Conclusion: having equally weighted criteria can lead to indecisiveness.
The solution is not very satisfying. The only way to avoid the paradox while maintaining the spirit of the experiment is to have a deciding criteria which supersedes all others (Arrow’s impossibility theorem). For example - she might say that she is going to take the cheapest car.