+/- 20% per repeat

[formerly known as hf]

Last night must have killed off more of my neurons than I realised for me not to know this for sure...


If it is a +/- 20% per repeat, than the more repeats a project has, the less likely the final amount will deviate from the standard rate.

i.e: the more repeats there are, the closer the final amount will be to +/- 0%




My recollection of probability mathematics does serve me right?

[Talapus]

Yes. Think of it this way: you have one project that is ten days worth, and 10 projects that are 1 days worth. If the big project is a +20% project, then each of the small projects would have to be +20% projects to match the output. The larger the number, the closer that the average is close to the daily rate.

[Jos Elkink]

Of course this was already the case for having many projects ...

[docent]

Hmm. Looks like when you repair mechanized projects you also remove randomization of others.. I just end two wood colections and both have value exactly typed by me (375 and 600)

???

[Chris Johnson]

The final fix was committed but not rolled out - That should be fine now

[docent]

Ok understood