A scorecard helps until it becomes the goal
💎CORE ARGUMENT
A scorecard makes a behaviour visible, and for a while visibility is all I need. Then, quietly, I start working for the mark.
Franklin ruled a little book with a column for each day and a row for each virtue, and by his own account never arrived at the clean page. What strikes me is that he kept the book long after he knew that. The marks had become their own activity.
I have three scorecards and the same drift in each.
- Franklin's chart, borrowed for three weeks in June. I stopped, and the experiment note says why.
- The habit tracker. The minimum version of the writing habit is one sentence. It exists so that a bad day still counts, which is right. It has also produced sentences written at 22:50 for no reader, to keep a box from staying empty.
- Carrier scores, the one that is not about me. A carrier scored on on-time delivery will learn what counts as on time.
The turn comes at the same point every time: when the person being scored can move the number more cheaply than they can move the thing it stands for.
This is still a seed because I only have the warning and no rule. I do not know how to tell, from inside, a tracker that is helping from one I am serving. A candidate test: would I do it today if it could not be recorded? It belongs beside 💎 Extreme results drift back toward the average on their own, since one bad month on a scorecard is mostly noise, and 💎 An interval is more honest than a point, since a score shown as a single number hides how unsure it is.
🧱BUILT FROM
🌍WHY IT MATTERS
For my own trackers: keep the box, distrust the streak. For the scorecards work: every measure we propose gets one extra question, how would a sensible carrier game this, asked before the prototype and not after.
☑️NEXT STEPS
- Ask of each proposed scorecard measure how a sensible carrier would game it
- Stop counting streaks on my own trackers, and keep the boxes
- Ask Bram where a number on his whiteboard has ever misled him
📖 Franklin on the thirteen virtues: the book of marks, and his own admission that the page never came out clean.
💎 Extreme results drift back toward the average on their own: a carrier's worst month will improve whether or not anyone scores it. 💎 An interval is more honest than a point: a score with no range around it claims more than it knows.
A score works by making something visible. It fails once the scored person finds it cheaper to move the number than the thing. Nothing announces that moment, and from then on the score measures skill at scoring.
Bram would say the opposite and has a whiteboard to prove it: what is not written down does not get lifted. For a barbell the number and the thing are the same object. That may be the whole difference.
Tried on myself in 🧪 Franklin's chart for one value a week. It matters far more in 📂 Carrier Scorecards discovery, where somebody else's contract may depend on the number.