Should an ideal run of K-means clustering produce evenly distributed points near each of the means?

machine_learning
k-meansclustering
data_science

#1

Hi,

While using the K-means algorithm on a set of points, is it necessary that all the means have evenly distributed points in their clusters? What if the situation like the one below?

Is this a good classification or not?

Thanks