vault backup: 2026-02-25 08:15:32
This commit is contained in:
@@ -0,0 +1,26 @@
|
||||
# 2025-12-18 14:18:??
|
||||
|
||||
### Estimating Golf
|
||||
|
||||
What interests me most in [[construction-estimating]]
|
||||
is an idea you might call "estimating golf":
|
||||
the goal is to produce a satisfactory[^1] estimate,
|
||||
but the estimator can not view the project documents
|
||||
and must instead ask questions about the job
|
||||
(answerable in a sentence or less)
|
||||
of a neutral party ("the reader").
|
||||
The estimator fails if the estimate is unsatisfactory[^1],
|
||||
but otherwise is scored by number of questions asked.
|
||||
|
||||
[^1]: satisfactory in terms of accuracy and precision,
|
||||
according to the standards of the organization.
|
||||
a control estimate must be prepared accordingly
|
||||
by a neutral party ("the control").
|
||||
|
||||
You could further imagine different brackets
|
||||
for required accuracy and precision,
|
||||
whether organization historicals are freely available
|
||||
or must be questioned like project details.
|
||||
|
||||
The most interesting part of this problem is choosing when to stop,
|
||||
since it requires one to estimate their certainty of their estimate.
|
||||
Reference in New Issue
Block a user