Two drivers get cut off by the same car. One mutters something and is over it inside a minute; the other is at an eight, composing a speech nobody will ever hear. The event was identical, so the event didn't set the number. This session is about what did: the read.
Between every trigger and every surge sits an appraisal: a fast, automatic read of what just happened. You never experience the read itself, only its output. The lane-cutter through one lens ("they didn't see me") sends the curve to two. Through another ("they did that deliberately") it goes to eight. Same event, different bill.
Any of the three can be accurate. The alarm isn't malfunctioning; it's answering, honestly, the read it was handed.
Decades of research show that people who struggle with anger tend to read ambiguous behavior as deliberate. This has a name, hostile attribution bias, and it sorts everyday ambiguity in one direction: the neutral face files as contempt, the slow reply files as a message. Not because you're foolish — a threat-detection system shaped by hard experience would rather ring falsely than miss once, and yours has had practice. But the word "deliberate" is expensive. It turns an inconvenience into an insult, and insults go up the curve. The same late reply, from the same person, lands at three as "they're overloaded" and at eight as "they're ignoring me on purpose."
We're not calling your reads false, and we're not arguing the alarm down. Some reads are right. The point is only that a read is a claim rather than a fact, and in week four you'll get a proper procedure for checking claims. For now it's enough to see that there's a lens at all.
One new line in the log: the read. When you note an episode, note what you took the event to mean. Don't fix anything yet. Be a good reporter and file three fields, event, number, read, and let the record accumulate. Seeing your own lens in writing turns out to do some of the loosening on its own.