Most people who start tracking their TRT symptoms make the same structural mistake. They log what happened and how they feel as the same thing.
They're not.
What daily state is
Daily state is a snapshot of how you feel on a given day: energy, mood, libido, sleep quality, recovery. One entry per day, scored on a consistent scale, logged at the time. The point is to build a clean daily signal over weeks and months that you can look back at and analyze.
What makes this signal clean: consistency. Same questions, same scale, one entry per day. No interpretation in the moment, just scores.
What events are
Events are things that happened or things you did: a workout, a poor night's sleep, alcohol, travel, illness, a protocol change, a dose adjustment, a stressful day at work. They're contextual, not evaluative. You're not rating how you feel, you're logging what occurred.
Events are unlimited. A day can have five events or none. They exist alongside the daily state entry, not inside it.
Why mixing them breaks the signal
When you log a hard leg session as part of your daily state entry, you've hidden a variable. Your recovery score that day is 4/10. But is that the protocol, or the workout? You don't know, because you didn't separate them.
Run that forward a few weeks. You've been doing hard training sessions three times a week and collapsing them into your daily state. Your recovery scores look consistently low. You start wondering if the protocol is failing. You might change something. But the protocol may be fine. The recovery scores are low because of training load, and you can't see that because you didn't separate the event from the metric.
With event separation, the picture is different. Recovery score is 4/10. You also have an event logged: hard leg session yesterday. Now you can ask whether recovery scores are systematically lower after hard sessions. If yes, that's training load, not protocol. If no, something else is driving it. The separation is what makes the question answerable.
The alcohol example
Alcohol is one of the most consistent confounders in TRT tracking. It disrupts sleep, tanks mood and energy the following day, raises estradiol with consistent use, and elevates SHBG over time. Its effects can look exactly like protocol problems.
Log alcohol inside a daily state score and you get low mood and energy numbers with no apparent cause. Log it as an event and you can correlate. After a few weeks the pattern becomes visible: your worst mood and energy scores reliably follow nights with alcohol. That's a completely different insight than "my protocol isn't working."
Protocol changes as events
Every protocol change should be logged as a dated event: dose adjustments, frequency changes, adding or stopping an aromatase inhibitor, switching injection site, adding supplements. These are the inflection points that later explain why your daily state shifted.
Without them logged, you're trying to reconstruct sequence from memory. With them logged, you can look back at your energy scores and see that the decline started three days after the dose increase, not before it. That's the difference between knowing what happened and guessing.
One entry per day
Daily state works as a signal precisely because it's one observation per day. If you log your energy twice because you feel different in the morning versus the evening, you've created ambiguity that compounds over months of data. Which entry do you use? How do you average them? The analysis gets harder without adding much value.
The right approach is one entry per day at a consistent time, with the ability to edit that entry if your read on the day changes. Editing preserves the single daily signal while letting you correct an early entry with better information. It's a simple constraint that protects the integrity of the dataset you're building.
Most tracking efforts that stall do so because the data stops being analyzable, not because logging stopped being consistent. Structure determines whether months of entries tell you something or just confirm that you logged.