Why do we have to re-run the model if the order changes?

Good morning. Why do we have to re-run the model if the order changes? Do you mean the order of the data in the dataset? Do we have to expect a different decision tree if we just shuffle the data? Thank you