At a high level, solving a problem involves two phases.
Figuring out
Fixing it
In a perfect world, we should spend very little time ‘figuring out’ the problem and devote more efforts to resolving it. In reality, that's a different story. A complex distributed system with so many moving parts makes it difficult to understand the epicenter of the problem itself.
During the “figuring out” phase, One of the common misconceptions is the incorrect interpretation of two events, totally unrelated but similarly behaving, thinking that one causing the other.
This brings us to the popular maxim,
“Correlation doesn’t imply causation”
One popular example is observing sunglass sales against ice cream sales. The below chart shows that how sunglass and ice cream sales are related.
The above observation could lead us to incorrectly conclude that increasing ice cream sales increases the sunglasses or vice versa. But the reality is that sales of ice cream and sunglasses were primarily driven by the arrival of summer.
Correlation is just the indication of whether two elements share similar traits or tendencies. Whereas causation is a phenomenon of cause and effect. For instance, the thing X changed the thing Y.
Confusion over these terms leads to mistaken assumptions about how the system works. We then act upon that wrong conclusion and make a decision that can be very expensive.