Pattern is a regularity and so should lead to prediction. Observations eventually should lead us to make statements about what we expect to see in future observations. This is a process of modeling and hypothesis generating, although we emphasize the search for predictable regularities (patterns):
(1) Pattern hunting. Looking for a pattern in the natural or social world. Your observations are the the data, so learn to look deeply.
(2) Pattern decoding. Finding rules that produce the pattern in the natural world. The rules are like a code of action that you discover.
(3) Pattern coding. Translating the rules into computer code. This is making a model that should lead to successful predictions.
Without some kind of pattern in the world, we can’t predict or advance science. Without models to indicate what patterns we are to look for, one can look in the wrong place or at the wrong scale, ask irrelevant questions, or perform the wrong experiments. In the early stages, we tend to have mental models that are not yet verified. Our observations may be very informal, like taking a walk in the woods. In the early stages, observation of the world give us clues about patterns, but it takes some effort to translate what we see into language and rules. One should try to formulate statements about what one is seeing or, better yet, specify rules that may generate the patterns. The initial observations lead to a model that predicts what one should find in future observations. A model suggests specific places to look in the future, and often suggests conditions that we can experimentally manipulate before repeating observations. We might manipulate temperature, presence of chemicals, social environments, that we think affects the patterns. If these observations don’t at all fit the predictions, the model may be abandoned (we can formulate an entirely new model). If future observations fit some but not all fit our model’s predictions, as is often the case, we may revise the model. So the process often repeats (a cyclic pattern itself).
Yet the same simple rules that generate predictable patterns can give rise to complex, even chaotic patterns that are very difficult to predict. Patterns of great complexity and beauty emerge from simple rules, as we explore with many games and models. Fortunately, science generally can discern the conditions under which prediction is possible from the conditions which render detailed prediction unlikely or even impossible. Moreover, even when chaotic, unpredictable patterns occur, we often can predict the general types of patterns that will occur.