Rage clicks make excellent screenshots. They also arrive after the person has already decided the interface is hostile. If your weekly report opens with a rage-click montage, you are documenting anger, not locating friction.
In Friction Cartography we treat rage as a late, noisy member of the signal family. It is useful when it clusters on one control after a hesitation window. It is theatre when it is the only clip in the deck.
Hesitation is quieter and earlier. It is also harder to sell in a meeting, which is precisely why it belongs in the atlas.
What we sample instead
A Socket ticket that begins with rage must also show the fifteen seconds before the first extra click. Did the pointer stall on helper text? Did a field clear itself? Did an OTP timer expire while the person was still reading the legal line? Those earlier moves usually name the family: cognitive, temporal, or trust.
We also ask students to count silent backtracks. A backtrack without rage is easy to miss in vendor dashboards because nothing “exploded.” In Thai checkout labs, backtracks from shipping to cart often outnumber rage events three to one. The atlas cares about mass, not volume of emotion.
A practical rule for your next review
If you cannot describe the pre-rage window in one sentence, do not present the rage clip. Go back to the recording. If the vendor UI makes pre-rage hard to find, that is an instrumentation note, not a reason to skip the work.
Rage still has a place: it can justify a fast patch when a control is obviously broken. Just do not let it become the school’s entire personality. UX Friction Analytics is a map of stalls, not a highlight reel of fury.