File used: 400 people 500 media 2000 facts 1000 citations in FTM.
FTM gedcom to RM8 (get 1000 events 2000 citations) and upload without media to ancestry (RM8 bug prevents media upload). As I work hints I ignore them after downloading any good finds.
On Ancestry the FTM tree shows no new hints 6 generations back (all worked & ignored with 2000 left). However the RM8 new tree has only 200 hints but about 1/3 are new hints not showing in FTM.
Apparently RM8 is getting new hints better than FTM. The only obvious difference is the lack of media in the RM8 tree. Fewer hints but more new ones makes RM8 quite useful to add data to my FTM database.
Two different Ancestry Trees. One old, one new. The new one knows nothing about what hints were accepted or ignored by the old. Hint generation is neither instantaneous nor comprehensive and is influenced by user activity. I expect you will get lots of hints in the new tree for Ancestry sources you have already uploaded because they are now Other Sources.
Yes and no. The two new trees (one with media in RM8 and the other with none and both with none uploaded to ancestry) are getting more hints over time. However the old FTM tree with media uploaded to ancestry is not getting new hints. The presence of media in the source database seems to make a difference in how many and which hints you get.
RM8 currently prevents me uploading media to a linked ancestry tree and this seems to help with new hints but they are buried among many older worked and ignored ones.
None of us outside Ancestry’s developer team know how its hints generation algorithm works but my experience suggests:
Trees on which the user has been inactive are touched occasionally to find something to incite the user to appreciate how valuable the service is and to keep them as a subscriber or induce them to upgrade as a paying member.
Trees on which the user is actively searching, adding, editing draw more of its attention and consequent hint generation.
Possibly, Public Trees that are more popular may draw its attention.
So I would suggest that if you get onto your FTM-linked Ancestry Tree and work on it in Ancestry, you will see more hints pop up and they will respect (filter out) the already selected Ancestry Sources and Ignored Hints. Whether the presence of user contributed media has any role in the algorithm is moot.
Makes sense that they would favor active and popular public trees. However past behaviour (dumping their FTM customers in 2015 due to incompetence) suggests a certain element of dumbness is also involved. Have been active in the FTM tree to see if that helped (not so far) and might try a public tree also. So far time seems more important than activity.
FTM tree is getting new hints as is a new no media RM8 tree slowly over time.
Waiting over three months for a serious mac RM8 bug fix update has proved a boon. My FTM file uploaded to Ancestry with media generates 2000 hints. The same file imported to RM8 as a Gedcom and uploaded without media since RM8 bombs with media gives only 500 hints. Working my FTM hints has proved useful in 10% of trees but 90% of records although many of the latter duplicate or overlap. These finds are scattered over the timeline with some back to the mid-1500’s. Ancestry makes it easy to download the document image, transcript and source in one pdf file and then ignore the hint. I only sync up to Ancestry never down.
As TomH has pointed out, working hints seems to trigger new ones (squeaky door syndrome). Despite the daunting total I have started working all hints for one person at a time and have finished 10%.