3 ms·
I don't claim to know much about California bureaucracy, but in my own state, just in the past decade: - The 787 was years late and over budget - The 520 brid
by ken 8y ago
I don't claim to know much about California bureaucracy, but in my own state, just in the past decade:
- The 787 was years late and over budget
- The 520 bridge was years late and over budget
- The 99 tunnel was years late and over budget
- Microsoft doesn't seem to schedule releases (publicly) or use calendar years for version numbers any more, but from where I sit, Windows hasn't exactly been a shining example of success in the past decade, either
These are public and private, across completely disparate engineering fields. California may be dysfunctional but everybody else is, too. If someone in this country can still "build big", who is doing it?
- curtis 8y agoThere is a big difference between late and not finished at all. Those are all examples of successful big projects.
- craftyguy 8y agoExactly. Humans suck at estimating cost and time for large, complex projects. Measuring the success of a project purely based on whether the estimates were accurate before they started completely ignores any benefit folks have when the project is completed.
- ken 8y agoThat just leads to more questions! If missing estimates of time and resources is a success, then what would be a failure? Just giving up? Can we make any project a "success" by just keeping on throwing time and money at it until it's finished? People (like executives, or voters) made the decision to go forward with these projects based on the estimates. If engineering estimates can't even be accurate to within a factor of 2 (like these), then what's the purpose of providing them? If humans are just bad at estimating, why are these estimates always much lower than actual, and never higher? What about the benefits which could have been achieved by other uses of this money? How can anyone make an informed decision about which possible engineering project to implement, when estimates are this bad? These days, I assume any estimate without a confidence interval is a lie.