“The possible solutions to a given problem emerge as the leaves of a tree, each node representing a point of deliberation and decision.” – Niklaus Wirth (1934 – ), Programming language designer
The decision tree method of logically reaching conclusions has been around for a long time and been the basis for troubleshooting personnel to resolve technical issues. A decision tree is, simply put a descriptive presentation of all possible solutions to reach one decision that is based on certain operating conditions. A decision tree is called so since as a graphical representation it begins with a box or the ‘root’ and then spreads in to ‘branches’, just like in a tree, that offer a variety of solutions.
A decision tree is an effective and helpful method of logically deducing a conclusion. Since it is diagrammatic it is a visual representation of your thoughts and also helps to draw systematic and well documented decision processes. Decision making very often is limited by the fact that we are only able to select from the possible and regular alternatives. A decision tree serves as a brainstorming process that throws up many more possible solutions that can take us to the best ‘decision’ / solution for the pertinent problem. However, it must be understood that a decision tree requires adept handling since the number of solutions it brings forth can appear complex in every day decision making.
For example – you may decide to go out of town for a long weekend and need to plan but there are obviously factors – some known and others completely unknown – that will help to determine what can be done and what you cannot do. This is where a decision tree can help. Start with the root or the decision box that begins the decision tree in this case it would be you going out of town. From that box would be two possible outcomes – either you go or you do not. If you are going – there would be further branches about where to go, number of day and the budget. You would also have considerations of weather conditions and distance. All these become further branches. If you are not going, then you go back to the root to decide what you would do while staying in town – go to the movies, a local picnic, visit relatives and such. Once you put down all the ‘branches’ it would easier for you to decide whether to go out of town or not.
In companies too management are also continuously faced with situations which require them to make some tough and even uncomfortable decisions. There may be some choices that are apparent based on experience or skill but in the current situation the consequences of those choices may not be immediately apparent. A decision tree analysis helps them deal with such situations by throwing up solutions and the possible consequences, some of which may have been hitherto not thought of. So how does is it work:
The diagrammatic representation of a decision tree has nodes and branches, in which the nodes are of two kinds. The rectangular nodes are the ones representing the decision that needs to be made. The branches that shoot out from these nodes represent the different choices the manager has, out of which only one can be implemented. The other kind of node is circular and represent what are known as chance nodes, which means that the possible branches or choices from these nodes have a degree of uncertainty in so much as whether they will happen or not. As pointed out, a decision tree makes it easier to take a call since it is a visual and vivid representation of the various choices faced by the person involved in the decision process – either in everyday life or in strategic management decisions.
In strategic decisions, the manager must analytically assess the current problem to be able to identify the decision that would be most appropriate for it. An analysis of all the possible actions needs to be made that could possibly solve the problem. On a decision tree, these possible solutions are connected to the nodes and go forth as straight lines. The next step for the manager is to take stock of the consequences that each possible solution can bring so as to ensure that the solution does not give rise to a new set of problems. Each consequence would have some element of uncertainty and it would require the manager to evaluate the possibility of each consequence happening or not. The goal would be to understand and bring out the likelihood of each consequence actually taking place. The best way is to take in to account similar outcomes from experiences in the past and if no past experience is available, then gathering and utilizing the wisdom of a set of experts will help to understand what each consequence would lead to. How probable the likelihood is can be decided by the combined consensus and the only requirement is that the total of all the probabilities coming forth from a chance node must be equal to one.
One of the best methods to obtain these likelihoods is to use bygone experience had from similar occurrences. What if no past experiences of similar occurrences are available to the manager? In these cases, the best method would be to utilize the collective skills knowledge of experts as to how likely it is that the particular consequence will occur in the future. Using an appropriate common decision building technique, the panel of experts can be provide combined or averaged views to create a probability of the likelihood of the occurrence of each and every consequence. The only main thing is that the measure of the all probabilities emanating from the consequences from a chance node adds up to one. Post this, an evaluation of each possible choice needs to be made that would result in showing the possible cost for the course of action decided for implementation. All possible choices are evaluated in this manner and so the entire decision tree can be evaluated and the most beneficial alternative can be selected both in terms of money and utility.
With the number of competitors rising and companies constantly struggling to stay ahead and secure their future, management has to deal with a variety of situations and handle think of a number of outcomes to stay ahead. It cannot suffice to make critical and tough decisions based on gut feeling alone and more robust and systematic systems are required to bring about qualitative decisions. More often than not, future scenarios can be narrowly predicted but there is a high possibility of last minute strategic and or operational changes that can be required. Management or top leaders would not have the time to examine and scrutinize each and every aspect and the teams working with them would most likely not be equipped to know the best way ahead. The best and safest method to decide would be through the detailed analysis possible through a decision tree.
A decision tree is a tool that uses a methodical way to think through the possibilities of myriad scenarios and understand the impact of each possibility. This could point to finances and or the risk of a certain choice that could potentially ruin things further. A decision tree allows the user to attach computable value to the choices brought forward and choose the most viable and least costly and risky option. Decision trees are a robust method to evaluate risks and they are also remarkably flexible in the number of outcomes they ‘branch’ out in to. They help managers to use the results and allocate resources – time, effort, money and human – resources to carry the decision forward. These resources can be ramped up or scaled down as and when required.
A decision tree helps people in everyday life and managers in companies make solid decisions by being able to assign measureable value to possible outcomes and more effectively manage future scenarios that would otherwise be completely uncertain and shaky.
