As I responded in the LinkedIn PM discussion group to this same question, a Monte Carlo simulation would be one way for you to get a range of possible end date outcomes based on the information you have. Of course, this assumes that the actual & variance data you have is applicable to all tasks leading up to the end date. If the degree of variation in the remaining tasks is quite a bit different than that of the tasks completed to date, even an MC simulation is unlikely to help much. The same is true if we change any of the other (assumed) fixed variables such as the staffing or size of the team doing the work.
Depends on how precise your answer needs to be and how good your statistics are.
Probably the easiest is using a normal distribution calculator. For more in depth answers, you need to understand things like the z value. It's not particularly difficult in the example you provided, but most examples will take you at least a page to show how to either calculate it or look up the answer in stats tables. Saving Changes...
Sergio Luis ConteHelping to create solutions for everyone| Worldwide based OrganizationsBuenos Aires, Argentina
I fully agree with @Keith. In fact, is what I am using from years. Data you have is the key. Remember you have samples then parameter calculation on samples has an inherent error. Then my recommendation is to calculate confidence interval for your estimation. If you can make a hypotesis test then it could be great. Obviously you can take the way @Kiron stated above but I think it could be easy to make calculations from data you have without simulate with Montercarlo. Saving Changes...
