Score the list by motivation, with the weighting in the open

1 · Learn the move · Refinement loop

Budget runs out before the list does, so the order you mail in matters. A motivation score stacks the signals you already have - years delinquent, absentee, tenure, how many parcels one owner holds, an estate flag from your scrub - into one number, so the top of the list gets the first letters. Two things go wrong. A black-box score cannot be tuned or defended, and a score can quietly lean on something about the owner as a person that you have no business sorting on. The refinement loop handles both. Round one: you paste your signals and your weights, and the model applies them with every input shown beside every score. Then you read the top twenty and push back. Round two: new weights, the formula restated, and a moved-rows list showing who rose and who fell and why. The weights are yours. The score orders your mailing; it never says an owner is motivated or predicts a sale.

Round 1: apply my weights below to these rows. Show the formula and every input beside every score. Round 2 comes after I read the top of the list: I will change weights, and you will restate the formula and list which rows moved and why. Use only the signals in my rows. No signal about the owner as a person. The score is an order to mail in, not a fact about anyone.