Are there any online repositories of completed real-world projects with their timescales that I can use to callibrate my own development time estimates?
+1
A:
If such a repository existed, how would you expect to correlate your project to find matches?
To expand; every software development project has unique aspects - particularly with regard to project participants - and will therefore have unique dynamics that affect estimates, possibly by orders of magnitude.
To apply past project metrics to future projects and hope they hold up you would need to assume a few things;
- Developers are interchangeable (they aren't)
- Building software is like building a house or a brick wall (it isn't)
- Project risks are negligible (maybe you'll get lucky?)
Finally, if all you need is a ball-park number then isn't "calibration" overkill? Just ask your most experienced developer how long they think - they are usually in the ball park.
Ed Guiness
2009-01-11 17:39:21
Well, presumably thousands of similar projects have been completed for any given new project. I'd use team size, project requirements and all the usual metrics to find a ballpark.
Ben Aston
2009-01-11 17:49:29
Thanks for your thoughts edg. I agree with you; there are an infinite number of variables, which make fine accuracy impossible, but I can't help but think such a repository would be useful.
Ben Aston
2009-01-11 18:03:26
While there are a large number of variables, it's also true that there are some basic factors that largely defy modelling, such as team dynamics, individual performance, and project risks.
Ed Guiness
2009-01-11 18:07:00
I guess it comes down to how much variance there is in the entire taxonomy of projects. My hunch is that *for any given domain*, many projects are very similar in nature (although, of course there will be many exceptions/edge cases) and so useful information could be gleaned from a comparison.
Ben Aston
2009-01-11 18:22:38
Well sure, if you have the same people, the same infrastructure, the same challenges, the same politics, the same budget, the same risks, the same timeframe, the same milestones, the same sponsor and stakeholders, and the same objective then a past project could predict, maybe.
Ed Guiness
2009-01-13 08:30:46
(Although such a project would seem a senseless repetition I guess it could happen in theory)
Ed Guiness
2009-01-13 08:31:39
A:
I don't think relating your project to a set of generic other projects to determine a time est. would be hugely beneficial. You can compare to other projects for defect rates - here's a good starting point for that: http://www.scribd.com/doc/7758538/Capers-Jones-Software-Quality-in-2008
My suggestion would be to look at prior projects in your company with the same relative technology and resources and develop a table (via function points) and then a continual resync. If there is no prior info and/or the technology is new and/or the resources are different - then it's best to use the team's prior experience from prior jobs.
Here's a good book: Software Estimation: Demystifying the Black Art
meade
2009-01-14 17:41:43