Unconditional apples and conditional oranges: can we compare different styles of attribution?

In the aftermath of an impactful heatwave, storm or flood, attribution studies often ask whether events like this are becoming more intense due to climate change. Answers to this question can be deduced both from unconditional extreme value statistics and from approaches like storylines or circulation analogues, which analyze the event conditional on the dynamical weather conditions. But can we compare these two kinds of statements? How should we interpret situations where they disagree? This talk takes a systematic look at different notions of intensity changes. We see examples where conditional and unconditional approaches are fundamentally incomparable and find special cases where they can reasonably be compared and potentially combined into a single attribution statement.