Reward, Error, and Value: How We Make Decisions

No man ever steps in the same river twice.

We are constantly responding to our environment. At the same time, our environment is constantly changing. Sometimes we behave the same, sometimes we do not. But why?

Imagine you are trying a new restaurant, and you order a dish you have never had before. When it arrives, you take a bite — and it is amazing! It's quite likely you will order similar things in the future.

Now, imagine the same dish arrives; it’s bland and disappointing. You will probably be less likely to order dishes of similar kind again. In both cases, you learn what decision to make in the case of similar situation.

Here, the point doesn't only lie in the outcome but also in our expectations. It’s the difference between them that matters.

Here is how it works in our brain: Two areas of the midbrain, the substantial nigra and the ventral tegmental area. They are responsible for sending dopamine signals to the striatum.

These signals don't just happen randomly. They follow a pattern — normally quiet but bursting with activity when something surprising or rewarding happens. Scientists call this burst of activity a reward prediction error.

In the first case, the surprise creates a positive reward prediction error. Your brain registers that the experience was better than expected and dopamine surges in response.

In the second case, it creates a negative reward prediction error. The experience was worse than you expected. Your brain notes the mismatch between expectation and reality, and dopamine activity decreases.

These reward prediction errors — positive or negative — are crucial for learning. Simply put, it’s the brain saying, “Hey, this didn't go as expected!” It compares what we expected to what actually happened.

The core idea is that the reward protection error signaled by the dopaminergic response can be transmitted to the CBGTC loops — brain circuits that send messages back and forth between brain regions. These signals update the “value” of the situation we are in, helping us to guide our future decisions.

Here, we mentioned an important word: value. But what does it really mean? Is value the same as reward? At first glance they seem similar. However, value goes beyond just reward.

Let's return to our restaurant example. Now, we are deciding between two dishes: One is a spicy curry and the other is a mild, comforting soup. If you are in the mood for excitement, the curry might hold the most value, delivering the thrill of intense flavors. But if you are feeling under the weather, the same curry may seem unappealing, and the soup could feel like the perfect choice.

In this way, value is not only subjective, varying from person to person, but also dynamic, adjusting based on the situation and what matters most to you in that moment. It's not just about labeling things as good or bad. It's about creating a flexible framework that evolves with us.

“No man ever steps in the same river twice, for it’s not the same river, and he is not the same man.” Value, like a river in Heraclitus’s famous phrase, is never the same. It evolves with us, shaped by our experience, expectations, and the environment around us. The river may change, and so do we.

So, the next time you make a decision, whether it is trying a new dish or choosing a career path, remember your brain isn't just seeking rewards. It's about how we assign meaning to those rewards in the context of our lives, reflecting on who we are and what we need at the moment.

What’s Behind Our Decision: Reward, Error, and Value