[hnfong]

That's an interesting observation.

Let's say you have a personal belief that something is going to happen with probability x. Would you actually want to tell others that the probability is y, because that's what the data says, without letting people know that for other reasons that are not reflected in the data, you truly believe it is x?

[Dylan16807]

You can give your informed opinion on the probability in addition to what you've derived from the dataset, but you shouldn't conflate them.

Your informed opinion incorporates this dataset, but you shouldn't imply it's "based on" this dataset.