A possible inconsistent comment for Figure 12.5

[mcavs]

There is a comment about Figure 12.5 as "There are more negative than positive residuals, which may be seen as a signal of a (local) positive bias of the predictions."

Do you think there is an inconsistency in this comment? Because it has been stated that the negatives are more than the positives, it means the model has a negative bias.

I think that this comment is not valid because the model has a positive bias, so the comment may be "the model has a positive bias because the positives are more in size despite the negatives more in quantity".