Let’s assume you’re frying chicken. You will need to defrost the chicken, wash them with warm water, cover with breading, fry them… and magically, you end up with a crispy, golden brown plate of chicken.

An algorithm is exactly that, a set of instructions (defrost the chicken, wash…) to produce a desired output, which in this case, is a crispy, golden brown plate of chicken. Typically used in the context of computer science and mathematics, most algorithms are designed to calculate the value of a function, process data, or even automate decision-making processes.

However, while algorithms are designed to automate and create, in a utilitarian sense, useful tasks and outputs, not all algorithms are deemed of any use. To create an algorithm worth using, your algorithm should be a good algorithm.

What makes an algorithm good?

A good algorithm is very much similar to a good recipe for fried chicken. Not all recipes are deemed good, and some, while still resulting in a very solid plate of golden brown fried chicken, might not taste up to par. Like creating the perfect fried chicken recipe, to design a good algorithm, we must follow the following criteria.

1. Clear and Unambiguous: Each step in the algorithm must be clear and not open to multiple interpretations.
2. Effective: The algorithm should effectively solve the problem it’s designed for. In our chicken frying example, the algorithm should result in a delicious, crispy chicken if followed correctly.
3. Efficient: Efficiency refers to how well the algorithm optimizes resources, such as time and memory in computing, or ingredients and cooking time in a recipe. An efficient algorithm achieves its goal without unnecessary steps or resource waste.
4. Finite: A good algorithm has a finite number of steps. It doesn’t go on indefinitely but concludes after a certain number of steps.