That’s a great question! Something I didn’t add in the blog post, but is worthy of discussion is how do we reflect on what drove those changes.

There could be a number of ways to do this, driven in part by pragmatism, and what are the questions you are looking to answer.

For the example you provide, such as how to differentiate between a correction and an update, your suggestion might be an approach, or perhaps a relationship of HAS_NEXT_STATE and a relationship property of changeReason to reflect if it was changeReason:'update’ or changeReason:'correction'. This would assume that you one change per new state. All of these will have advantages and disadvantages. For example, if a large majority of your questions were always investigating the reason for a change, and there are many changes — then it would be slow to keep filtering on relationship properties, and therefore another modelling approach may be appropriate if speed is of the essence.