I once noticed a friend of mine from campus who was persistently trying to get close to a girl every single day. He delivered breakfast, bought concert tickets, helped her with assignments, listened to her vent until three in the morning. Basically, his effort was at its absolute maximum, crossing the rational threshold of a normal human being.

One day he vented to me, with a face resembling someone who just got scammed by a fake investment. He asked, why is it bro, I have given everything but she ended up dating an art student who cannot even play the guitar in tune?

As someone who reads API documentation more often than dating codes of conduct, I was actually confused about what to answer. But if we pull this into a mathematical concept, my friend's mistake is very clear: he applied a simple linear regression model to a phenomenon that is clearly a non-linear system.

The concept of linear regression is simple. Y equals aX plus b. Meaning, there is a constant straight-line relationship between variable X (for example your effort) and variable Y (the girl's affection). This model assumes that the bigger your effort, the higher her affection will be.

But in the real world, especially when it comes to human feelings, the system is never linear. Emotions are highly fluctuating, with many hidden variables that are hard to measure. There is noise, there are external factors like mood, timing, and pheromones.

In the early phase of dating, it is true that effort can increase attraction. This is the exponential phase where you just give a little flower and she smiles all day. But humans adapt quickly. There is something called diminishing returns.

If you buy her sweet pancakes every day, eventually the pancakes are no longer a romantic surprise, but a minimum operational standard. The value of the pancakes goes down in her eyes.

Even worse, if your effort crosses the line, the curve does not just flatten, but can actually plunge downwards. This happens a lot. You message her every hour, asking where she is, who she is with, if she has eaten. Instead of feeling cared for, she feels like you are a super clingy electronic traffic camera with no life of your own. Her attraction curve immediately drops to the negative point.

So if it is like this, how do you model a relationship correctly?

Use non-linear system dynamics modeling. You need to understand when to pull the effort lever and when to hit the brakes. There are times when you give maximum attention, but there are also times when you need to give space. Let her think and miss you. This is like machine learning needing regularization so your model does not overfit. If you follow her every whim too closely (overfitting on training data), you will be completely destroyed when faced with a real situation that is just slightly different.

Love needs uncertainty in the right dosage. If you are too predictable about when you will chat and what format you use, the attraction algorithm in her brain will consider you boring.

So for my friend earlier, stop expecting you can buy love with effort coins on a straight upward graph. Learn to read the situation, adapt to the changing functions, and most importantly: do not forget to update yourself. If you can only give sweet pancakes but cannot provide other value like engaging conversation or a clear future, then it makes sense she chose the guitar guy.

The point is, not everything can be solved with one standard formula. Sometimes you have to accept that the random variables are just not on your side.

Khay