I'm trying to apply the GPS to tear down one of my brick walls in Aalten, the Netherlands.
My ancestor Hendrik Bengevoort married in 1767, had three children with his first wife, remarried in 1778 (no children) and again in 1795. He then went on to have two more children with the third wife, the last of whom was born in 1799. Hendrik had died in 1798, leaving a pregnant wife behind.
The usual places to look for the names of his parents in this area in this period are the marriage and burial records. Unfortunately, the Aalten records are very terse and do not include this information. I've searched census, notarial and judicial records and came up with several more records about Hendrik and his wives, none of which give a clue about the parents of Hendrik or any other relatives, FAN-club or property.
The usual place to look for birth information is the baptism records. I've searched the baptism records of Aalten and all of the surrounding parishes and came up with only one candidate, a Hendrik Bengevoort who was baptized in 1716. That would make him 50 at the time of his first marriage and 82 at the time he fathered his last child.
One of the things I read in the BCG standards manual is that we must assume people behaved normally. To get married for the first time at age 50 wasn't usual at all in that area (or pretty much anywhere else for that matter!). Most men married in their late 20s, early 30s. And only one individual in my database that includes information about over 50,000 people in that area fathered a child after the age of 80.
There is one other loose end: no Hendrik Bengevoort could be found in the 1748 census of Aalten and the surrounding parishes. Since he was probably older than 21 (the age of majority) when he got married, he must have been born before 1746 and should have been in the census. This suggests he may have come from elsewhere.
Personally I don't think I've got a solid proof argument to convince myself or anybody else that the Hendrik b. 1716 is my ancestor. I'm pretty sure I've got the reasonably exhaustive search part of the GPS covered though. So I wonder how you would handle a case like this, when the available evidence doesn't match conventional behavior?