Recruiting intelligence

Predicted window

The 80% prediction interval the statistical model expects a program's next opening in, estimated from its past openings, with an expected date inside it.

Confidence score

The confidence score, 0 to 100, measures how much consistent evidence backs a window. It is not the chance that the window is right.

What a window rests on

• Own historyMost of the window's weight comes from this program's own past openings.
• Borrowed timingMost of the window's weight comes from comparable programs (this company's other programs, or similar programs elsewhere at the same level and season), because this program's own history is short.

The percentage is that basis's share of the window's weight, as the model recorded it.

Evidence classes

• ExactThe job board supplied the publication date.
• BoundedTwo archive captures bracket it: absent in the earlier one, present in the later one.
• Observed byVisible by that date. It may have opened earlier.

The three are stored apart and never promoted into one another.

No forecast yet

A program gets a window once the model can compute one: from three recruiting cycles of its own, or from fewer when a comparable program (the same level and season) has timing to borrow. Until then its evidence is listed without a date.

How forecasts are made, and how accurate they are