WebeleonGive every take two ratings. One from inside the performance. One after listening back.
They will disagree. Often. The felt rating says the hard-won take was a 4. The verified rating says the easy clean one was a 3. The audience would have ranked them the other way around.
That disagreement is not noise. It is the most useful thing a practice session produces. The felt rating measures effort. The verified rating measures sound. Only one of those reaches the back of the room.
The felt rating
The felt rating is the one you give with the instrument still in your hands, seconds after the last note. It answers a real question: how did that feel? It is worth keeping. A song that felt like a 2 last month and feels like a 4 tonight is improving under the fingers — even on a rough night.
But other things ride along in that number. A take you fought for feels better than an easy one. The first run-through after a break feels worse than it sounded, because your hands remember the rust. Effort, relief, and recency all leak into the score.
None of them are audible.
The verified rating
The verified rating comes after the listen-back — ideally in the same session. It answers a different question: what would that have sounded like from the audience?
This is the rating tied to reality. It is also the rating you will resist. The microphone is immune to the story you told yourself while you played. I have caught myself hesitating a full minute before tapping a 2 on a take that felt like a 4. The hesitation is worth more than the number. The hesitation means you heard the gap.
Why verified counts double
The weighing rule I use is simple. Felt counts once. Verified counts twice.
A felt 4 with a verified 2 lands at (4 + 2 × 2) / 3 = 2.67. Woodshed territory — exactly where it belongs, because that is the take an audience would have heard. An easy felt 3 with a verified 4 lands at 3.67, the gig-ready side of mid.
The arithmetic is not the point. The asymmetry is. When the post-listening verdict counts double, effort cannot talk its way past sound. Optimism gets a vote. Not a veto.
One more reason ties break toward listening. The felt rating exists only in your head, on the day you practiced. The verified rating is anchored to a recording that sounds the same every time. Two people — or one person on two different days — will feel the same take differently. The take does not change.
One number, never blended across sessions
The weighted score belongs to exactly one session: the last one.
Last week's 3.8 does not dilute tonight's 2.67. Tomorrow's plan reads tonight's verdict. Full stop.
The history lives in the journal that writes itself — a trail of past ratings and dates. That trail is where trends show up. A song that has sat at 2.7 for a month announces itself as stuck. But the score itself stays per-session. A blended, ever-smoothing number is how a slow decline hides inside an acceptable average.
You played it. You heard it. You know exactly where the number comes from. That traceability is the whole design.
The notebook version, and why it fails
None of this needs an app. Notebook. Two columns per take: felt, then verified after the listen. The system works on paper for about three weeks — longer than most practice inventions last.
Then it meets its failure mode: the tired Tuesday.
On the bad night — the night the take was rough and writing the 2 down feels like filing a report against yourself — you skip the entry. On the good night, you write faithfully. Two months later, the notebook has quietly become a record of your good days. The gap between felt and verified — the data you most needed — is exactly the data you stopped collecting.
The fix is structural. Rating has to cost less than skipping it. And the rating you resist most must be the one the system cannot run without. In Edgars, the two ratings are the session. There is no instrument-free way to skip them.