How to calculate the time for the implementation of a task?

Question

Project management

I am a full stack developer.
I have been working for a company for almost a year where PHP programmers are rare. Or rather, I'm alone.
The team is formed from: 1 PM (project manager), 1 QA, 1 Full-stack, ... .
PM in coding is zero. His task is only communication with the client and the team. Most likely on his shoulders and other important tasks, but I do not know.

I constantly have a problem with estimates. Since, very little time is allocated for the analysis of time, and you have to figure it out by eye.

The reasons for delays are different. For example:
- - Incorrectly selected plugins. Because of what, time is wasted searching for suitable ones.
- - Client edits that are sent in the process of task implementation.
- - Various unexpected bugs. which sometimes take more time.
- ... .

And much more, which ultimately entails a delay in the completion of the project.

I tried to find information about how the time is calculated correctly.
Bob Martin suggests using the "PERT" formula. But my superiors and clients do not agree with the results of this formula. Like, that's a lot.


Actually I have a question.
Who calculates the time for the implementation of tasks, and in the end, who manages to meet the deadline?

Answer 1

But my superiors and clients do not agree with the results of this formula. Like, that's a lot.

And how does reality agree with the results of this formula? If it fits, then "this is a lot" is a purely diplomatic task.
If the situation is such that you say "4 hours" and "that's a lot" to everyone, and when you say "1 hour", and then you still do 4 hours - this is ok, then the real estimate is 4 hours.
And if, in principle, they expect you to work faster than you give out, and because of this, strain, then there are two options:
- either you really do everything slowly
- or the management has overstated and inadequate expectations.
you need to determine which option is yours and act accordingly.
In general, it’s best this way - take some of your estimates from the tasks done, then look at how much it really took, derive the dependency and multiply future tasks by it, adjust according to the results.
Here is another very popular method:

Collected an assessment from the developers, counted. You have to take risks.
- multiply by two?
- No, it's not scientific for two. It is necessary to multiply by the number e or by the number pi, depending on the complexity of the project.
- And how is it defined? "E... how hard" and "P... how hard"?

Answer 2

In fact, there is nothing complicated about this. You need to have several reference tasks that you clearly understand how much time you are doing, for example, for 1/4, 1, 5 hours. Compare new tasks with them and throw from above regarding the area of ​​\u200b\u200bnot knowing part of the task (or it’s better to do research first so that this doesn’t happen) + throw it on all sorts of communications (everything is also taken from experience)

Answer 3

don't worry, everyone has
https://www.youtube.com/watch?v=SQ4NAOU5Jso
the main thing - and this is the main thing that you have to endure from this topic - jump off PHP