If you’re anything like me, you would have over thought something in life. Overthinking rarely ends well. I come up with spurious correlations, irrational tangents, and my ability to make any meaningful predictions becomes clouded. Come to think of it, we fill our days with predictions to such an extent; I’d question the honesty of someone who doesn’t admit to overthinking on a weekly basis.

If you don’t see the predictions you make you’re overthinking the concept of prediction. You don’t have to have reams of data, graphs, and equations to make predictions. When you type an email, you are predicting how the person will respond. You’ll be predicting outcomes of daily events such as bookings, traveling etc. In fact, making predictions is so woven into your daily life you usually only notice that you’re trying to predict something when you overthink.

So is overthinking a human flaw? A result of fragile emotions? A personal defect in being able to deal with pressure? Whilst some of these factors definitely play a role in the prevalence and outcomes, the inability to draw conclusions from over thinking isn’t. It’s mathematical. A computer will be able to process for more variables than you, but it’s ability to make predictions can still be hindered by overthinking. In this case, we call it over-fitting.

The above image shows the same data. If we try and use a math formula that covers all data points we come up with clearly flawed predictions at the end. The data trend doesn’t drop, however, the over-fitted equation predicts a decline. Our simpler model equation misses data points but comes up with a more accurate model of the bigger picture. Even when we’re dealing with big data and machine learning, over-fitting is still something professors warn about.

So next time you’re struggling because of over thinking don’t take it personally. It’s not your brain. Computers can have this problem too! Relax the parameters and paint with a broader brush.