Wednesday, November 9, 2011

Tree Decorators and Tree Growers

Genealogists are Tree Decorators or Tree GrowersIt occurred to me the other day that there are two types of searchers on and

I call the two types tree decorators and tree growers.

Their search needs and requirements are different, which explains why some people like Ancestry’s New Search and some hate it. It explains why some people like the more than the current

Tree Decorators

Relevance ranking is a feature of’s New Search and FamilySearch’s record searching. Search results don’t have to match everything you search for, but the more things that match, the higher the record appears in the search results. The more you specify, the more results you get. More importantly, the more things you specify, the more the good results float to the top.

I’m a 5th generation genealogist, the inheritor of a complete, seven generation, largely undocumented tree. (The bug may have skipped one generation; I don’t have any indication that Alma did any genealogy. But I digress…) I spend a lot of my time hanging documentation on my tree.

I’m a tree decorator.

When I search, I search with a complete set of information about my ancestor. I search using every single fact or guess I have about an ancestor. The Search software uses the information to find every single record about my ancestor. Relevancy ranking works spectacularly, even where records have been misindexed, parts of a record were not indexed, or the record creator got some information wrong.

Relevance ranking works great for tree decorators.

Tree Growers

Exact search is the way the and’s Old Search worked. Every result must exactly match everything you searched for. The more you specify, the fewer the results and the better they are.

I split my precious genealogy time working on a brick wall. (Five generations of researchers have failed to break through, I must point out. This is more like a nuclear-bunker wall than flimsy brick. But I digress…) I am trying to grow my tree.

I’m a tree grower.

I am searching for people who are not yet in my tree. When I search, I search with precious little information. With relevance ranking, the less information you know, the worse the software performs. You’ve probably seen the messages: “If you add a name or other information, we’ll give you better results.”

Relevance ranking doesn’t work for tree growers and it knows it.

Exact search does. Tree growers don’t know much about an ancestor, but we do know something and the results we get darn well better match.

Further, tree growers need rich collection-specific search forms. We might know a neighbor’s name instead of a parent’s, or a slave owner’s name instead of the ancestor’s, or a profession instead of a birth date.

Two users. Two search requirements.

Tree decorators and tree growers.


  1. I am both a tree grower and a tree decorator. I need evidentiary records to identify a spouse and children. Often these records are not available on the sites you mention, but then one or the other may have some decorative elements (for recent generations, anyway). The ~really~ decorative items, such as estate inventories, estate sale records listing relatives and neighbors present, the deed where household goods were sold just before a major move, and the will bequeathing to a daughter and her heirs but specifically excluding her husband from any right to property, the record where a Guardian was himself determined to be of unsound mind and requiring a Guardian (thus precipitating another complex series of estate records) -- these are (by and large) not available on the internet at all.

    Without growth the decoration is irrelevant. Each type of search requires its own techniques.

  2. Like Geolover I am also both a tree grower and a tree decorator. The logical conclusion of your analusis is that search engines should therefore cater for both types of researcher (or, if you prefer) for both these approaches to searching.

    So what right have Ancestry and LDS to just assume that all of us fit one type only? It seems to me that often they make changes to their sites without any recourse to what their customers really want. Do they do robust market research, and if they do - do they then take account of the results?

    In this instance both sites have changed their search engines. Apart from a natural 'resistance to change' it's obvious that many existing customers prefer the old search engine to the new one.

    Why can't we have both? if they ever phase out their old search engines, there will be howls of protest from many -including me.

  3. I'm still confused about why the new FS search engine doesn't work for you, Gordon. If you want to use exact search, then click the checkbox to set "Match all terms exactly". (Admitted you need to click advanced search first).

    The only thing about "exact searching" on the new site is that it is, ahem, exact - if the record says "Samuell" and I've asked for "Samuel", then it won't show up.

    Adrian B

  4. As the first two posters said, I do both. I do not have a large tree; and I had no inherited information (if you exclude the dubious complied book "The Stricklers of Pennsylvania). So I must build my tree, but I demand documentation, so I attempt to build a decorated tree.

    Both sites have advanced search with "exact match" but for me "exact match" doesn't seem to work well. Maybe I'm not doing it correctly; but it seems to me that I still get people NOT born in Indiana even though I have made such a specification.

    Can't we have two types of searches (as some else said above)?

  5. I'm trying to document all the children of a couple first found in Pennsylvania in 1765. This means I do both types of searches, decorator for my own line (and what I had wasn't always right) and grower for the allied lines. If the old search on Ancestry goes away, I will be very hampered as that is what I use most of the time. Recently I was looking for Winifred Romine McCaslin after her divorce from Mort McCaslin, couldn't find her, looked for her parents and uncovered a story of sex, lies and murder. You wouldn't have found any of this with the new search as it asks for more information than I usually have.

  6. Ditto Sue's echo, "...Can't we have two types ..."

    And, I must be doing it wrong. Sigh. I make a targeted ("exact") search to narrow the results. I conduct broad or broader search (gimme a suggestion) when I'm looking for a wider range of returns. Truly, though, what defines narrow and what defines broad--I use a different set of qualifiers on my Miller and Clark lines than on Vig, etc. --GJ

  7. Nice analogy. I am also both a grower and a decorator, but I do not like the way Ancestry's New Search works or its relevance ranking. I like to fiddle around with search terms, wild cards, spelling, etc. in my own way, and Old Search works better for this.

  8. I tend to think of the old search being more tailored to those who know the type of items for which they are searching. I frankly don't need Ancestry's superfluous ranking which frequently places results from 1930 high on a list for a person who died in 1833. Yes, I did specify that. Ancestry decided I was too dimwitted to do the 100 year math. Also, I find it annoying to be bounced into the new search when I am looking at results from one area, click a button to see more and then >>bam<< I am in the new search without any warning.
    And I am a tree decorator and a tree grower and, with a great deal of frequency, a tree straightener, by which I mean correcting misinformation.


Note: Only a member of this blog may post a comment.