The engine grades the record, not the name.
The agent or agency graded as an institution on its record, read for the player.
The MLB agency graded from the record.
An agency is an institution, and an institution is a bet too. The engine grades it on auditable terms, the way it grades a player on the record: deals beat-versus-market, the arbitration record, free-agent outcomes against dollars-per-win, and client retention, each carrying its data tier and confidence. The output is a read, not a ranking of names.
The grade is 81 at 70% confidence, and it opens: the agency is MEASURED-strong on arbitration and retention, INFERRED-soft on long-term free-agent timing, and every term names the data behind it. A club negotiating against this agency reads the same record the player reads when he chooses it. The record is the grade. The logo is not a term.
Illustrative engine read on the real institutional grading terms (deals beat-versus-market, arbitration record, free-agent dollars-per-win, retention), each at its own tier and confidence. Composite agency, demonstration figures; no real agency is assigned an invented number.
The amateur draft advisor.
For a draft-eligible amateur, the advisor's leverage is the bonus negotiation. The engine grades him on outcomes relative to slot: the record of signing clients above or below slot for their draft position, the willingness and ability to hold a commitment as leverage, and the read on whether that leverage is real. The confidence reflects the thin, private data.
The advisor grade is 74 at 54% confidence, and the low confidence is the honest part: amateur money is private and thin, so the record is graded against the current rules and carried at a wide band, never dressed up as certainty. A real advisor edge, priced at the confidence the data actually supports.
Illustrative engine read on the real amateur-advisor grading (bonus-versus-slot outcomes, the commitment alternative as leverage), each at its own tier and a wide confidence. Composite advisor, demonstration figures. Amateur-money context flagged v0 and current-as-of.
The player's-eye read.
Turn the same institutional grade around for the athlete. Given his projected path, arbitration-heavy, free-agent-heavy, or draft-entry, the engine reads which representation profile fits and what the institution is actually likely to add or cost across that path. The output is a bet with its confidence, not a recommendation dressed as certainty.
For this player's arbitration-heavy path, the agency with the arbitration record is the fit, likely to add value in the control years and cost a little at free agency, inside a band, with the break-term named. That is a read the player can weigh, not a name he is told to trust. Here is what the record says, here is the band, here is the term that would break the bet.
Illustrative engine read on the real player's-eye institutional read (the projected path matched to a representation profile, the add and the cost across that path, the confidence band, and the break-term). Composite player and institutions, demonstration figures.
The engine will not grade an agency by its name any more than it grades a prospect by his draft slot.
Grade the institution. Read the record, not the logo.
Representation grades an agency or a draft advisor the way the engine grades a player, on auditable terms at a stated confidence, and turns the same grade around for the athlete choosing whom to trust and the club negotiating against them.