If you are going to estimate dates for someone you have to have a reference point that is a known date.
If you have a dated document where someone states his age, simple subtraction gives an estimate of the birthdate, but you need to keep in mind that the person’s statement could be not correct for a variety of reasons. Thus the birth date is an estimate, so a qualifier such as “est” or “abt” could be used.
If you know, for example men married at age 20-25 and women’s at 15-20, and you have a marriage date, by subtraction you could estimate a 5 yr range for their respective births. If it is a second or third marriage, the logic fails.
If somebody signed a will on a certain date, you know they died after that date.
Once the British discovered Hawaii in 1778 a reference point was established. If a Hawaiian ruler stated his age, or his age was estimated by someone and recorded, then his birth can be estimated, again subjected to the uncertainties of a stated age.
If a culture used a calendar, and recorded events relative to that calendar, and there is a correlation with the Gregorian calendar, then you have your reference points.
Going back, the length of a generation can be estimated, and approximate dates for someone can be estimated. But, these are more uncertain. The further back you go, the more uncertain the estimates become.
You need a little appreciation for biology: men can father children at older ages than women can bear them. A 70 year old man can be a father, but a 70 year old woman is well past her childbearing years.
If you can say someone lived at the time of a known historical event, you can place the person in time, but birth/death estimates become somewhat meaningless.
You need to know something about the culture and the history of the area they lived in.
In short, there are ways, but not everyone agrees on the result. There is nothing wrong with an estimate if you can state the reason why you made that estimate.