A sealed outcome is the resolution of a matter, with any amount collected and any resolution date, held apart from the file that is scored so that the score cannot see it. The outcome is joined back only at evaluation, after the scored file is fixed.
Sealing is how a score earns the right to be called a prediction. The file that is scored carries no resolution field, no collected amount, no resolution date and no field that encodes any of them. The sealed outcome is stored separately and compared to the score afterward.
The rule is stated about the file, not about the intentions of the person who built it. Status codes that are set only when a matter ends, amounts known only after resolution and durations computed from a resolution date all count as encoded outcomes and are removed.
A sealed outcome is different from a sealed baseline. The baseline freezes the model and its outputs. The sealed outcome keeps the answer out of the inputs. Both are needed before a backtest means anything.
Working a file that turns on this term?
Working through a diligence process?
Institutional partners evaluating a position against this platform's outcome and duration models are welcome to reach out directly.