62 lines
1.6 KiB
Markdown
62 lines
1.6 KiB
Markdown
---
|
|
id:
|
|
aliases: []
|
|
title: Uncertainty in Construction Estimating
|
|
tags: []
|
|
---
|
|
# Uncertainty in Construction Estimating
|
|
|
|
Cross-topic of [[uncertainty]] and [[construction-estimating]].
|
|
|
|
This note is intended to be purely descriptive.
|
|
See [[actuarial-science-for-construction-estimating]]
|
|
and [[risk-management-for-construction-estimating]]
|
|
for strategy based on such uncertainty.
|
|
|
|
%%
|
|
## TALK
|
|
|
|
What do I mean that project cost is uncertain until the last invoice?
|
|
|
|
What distributions exist
|
|
which are known,
|
|
which can never be known and must be estimated?
|
|
|
|
[[bid-price-modeling]]
|
|
|
|
***
|
|
|
|
With perfect information all outcomes are certain.
|
|
|
|
There exists some distribution representing the range of possible costs for a project.
|
|
The remaining uncertainty represents ignorance of information an estimator can not feasibly obtain.
|
|
|
|
***
|
|
|
|
Suppose the estimator has spent years on the bid,
|
|
running simulations, consulting experts, etc.
|
|
|
|
Suppose the estimator has received a message from a time traveler
|
|
which details significant news events and commodity markets
|
|
for the duration of the project.
|
|
|
|
Suppose the estimator is prescient and has observed the final cost of the project
|
|
as reported in the winning contractor's historicals.
|
|
|
|
%%
|
|
|
|
## Expected Value of Perfect Information
|
|
|
|
[[uncertainty#Expected Value of Perfect Information]]
|
|
|
|
%%
|
|
|
|
an organization pays in estimator hours.
|
|
These hours have a direct cost in salary and benefits,
|
|
however the greater cost is the **opportunity cost**
|
|
of tying up an estimator who could otherwise be starting a new project.
|
|
|
|
In many cases, the [[utility]] of doubled throughput
|
|
far outweighs the increased risk of individual bids.
|
|
|
|
%% |