Ancestry.com

What I learned at FGS 2008….

I think the first thing that I learned was that I may be a bit of a history geek. I snuck away from the festivities for a couple of hours and visited the Liberty Bell and Independence Hall. Standing in the same room where the Declaration of Independence and the Constitution were partially crafted and signed was awe inspiring for me. I love stepping through the doorways of history … that may be why I love genealogy, huh?

I also learned that if you can’t enjoy the thrill of breaking through your own brick walls, watching someone else do it is pretty fun. At our booth, we had 4 computers logged into ancestry, and people would come up and do searches. This one guy found a Canadian Census, with what I believe were his great grandparents on it. He was giddy…hopefully all of you know that feeling as well. :-) (And no, I don’t know if he was using old search or new search. ;-) )

The third thing I learned, as I watched people use new search, by far the most interesting and useful piece to most people I worked with was the ability to turn exact on for different types of fields. And as I watched many people do many searches, it would appear that you will get the most mileage on your searching if you choose exact date ranges and exact locations. Toggling between different combinations of what is exact and not exact is also useful in bringing up different sets of information. My anecdotal evidence shows that choosing all exact fields is just not all that great.

Meanwhile, back here at the ancestry.com ranch, we’ve been digging and working on some issues around the new search user interface:

  • Marriage records. Many of our marriage record data sets are, to be technical, quite foobar-ed in the new search user interface, as some of you know. And we know why. Some of these data sets were created many years ago, in ways that we don’t store our data sets anymore. And while we hacked around it back in the old search interface, we don’t hack too much in the new one. And yes, I know what some of you are thinking or muttering, well just go back to the old way and it’ll all be fine. No. Not the answer. I truly believe that fixing the data and standarizing is the way to go. No word on when things will be reformatted…you have no idea what goes into (that would probably make an interesting post), but I’ll keep you up-to-date. And remember, you can always go through the old search ui for now.
  • Soundex and matching algorithms. We are currently working on the back end pieces to bring you that option back..stay tuned.
  • State and country pages. We are working on bringing those up-to-date and giving you a way to easily access those. For example, try looking at the Virginia Database Page or the Italy Database Page. Also check out the Sources pages and the How To’s pages. Now would be a really, really good time to tell me what you think is a must have on those pages and what you really don’t care about. So start voicing your opinions now!

I’m still too tired from all of the traveling and the hot muggy weather to finish commenting on my previous post. I’ll finish it off before the week ends.

And if you have civil war ancestor’s in your past, I recommend the previous post by Jeanie Croasmun : Why So Many Names?. I know I learned something new.

About Anne Gillespie Mitchell
Anne Gillespie Mitchell is a Senior Product Manager at Ancestry.com. She is an active blogger on Ancestry.com and writes the Ancestry Anne column. She has been chasing her ancestors through Virginia, North Carolina and South Carolina for many years. Anne holds a certificate from Boston University's Online Genealogical Research Program, and is currently on the clock working towards certification from the Board for Certification of Genealogists. You can also find her on Twitter, Facebook and Finding Forgotten Stories.

61 comments

Comments
1 JadeSeptember 10, 2008 at 5:13 pm

Anne,

It’s great you had a good time.

You said, “State and country pages. We are working on bringing those up-to-date and giving you a way to easily access those. For example, try looking at the Virginia Database Page or the Italy Database Page. Also check out the Sources pages and the How To’s pages. Now would be a really, really good time to tell me what you think is a must have on those pages and what you really don’t care about. So start voicing your opinions now!”

The State Source Pages are *all* a huge muddle. They include all manner of unrelated stuff.

While it is true that people moved, and you have noted the value of ‘cluster’ research, this should not mean that a Source Page for one US State should include data for all other states.

For example I looked at the Virginia Source Page. There have long been a slew of unrelated listings there. Just for example, I selected the time period “before 1750″ and found the following that do not belong on the listings for that time period, and many that don’t go with Virginia Records at all.

Here is a list of those to **remove** for the pre-1750 view, beginning with the initial page of ‘top pics’ and after that continuing with categories’ full lists:

—-on the main page
Passenger and Immigration Lists Index (listed twice)
U. S. School Yearbooks
1860 Slave Schedules
1770-1790 Census of the Cumberland Settlements
Pssenger Ships and Images (Port of New York, 19th century to early 20th)
Irish Passenger Lists, 1847-1871
Virginia Soldiers of 1776
Remember the Raisin
The Final Disposition, Vol I (remains of Civil War soldiers)
The Final Disposition Vol. II
The Colonial Clergy and the Colonial Churches of New England
Compendium of Early Mohawk Valley Families – this is for central NY, the records begin in the 1730s but have no relevance to VA prior to 1750
Record of Indentures 1771-1773
DAR Lineage Books
Gateway to the West Vol. I – these are 19th century County records and cemetery readings for OHIO. The earliest deed book begins ca. 1798. The actual 1750s claims by Virginians such as George Washington’s family are not included here.

———
Full listing, Virginia Birth, Marrige and Death:
Boston Births, 170-1800
Boston Births, Baptisms, Marriages and Deaths, 1630-1699
Boston Marriages from 1700-1751
Boston Marriages from 1752 to 1809
Revolutionary Soldiers in Kentucky and Roster of the Virginia Navy
Scots in the USA and Canada, 1825-1875
Vital Records of the Town of Middleborough
Vital Records of the Town of Plymouth
Vital Records of the Towns of Barnstable and Sandwich

Marriage Records of Berkeley County, Virginia 1781-1854; this county is now in West Virginia. If you are going to include everything pertaining to present West Virginia there is a lot missing. In any event unrelated to pre-1750 records.

—————
Full listing, Virginia Immigration & Emigration
An Alphabetical Index to Ulster Emigrants to Philadelphia, 1803-1850
Irish Passenger Lists, 1847-1871
Passenger Arrivals at the Port of Baltimore, 1820-1834
Passenger Arrivals at the Port of Philadelphia, 1800-1819
Passenger Arrivals, 1819-1820
Passengers who Arrived in the U.S., September 1821-December 1823

———–
Full listing, Virginia Military
Revolutionary War Records: Virginia
all of the Roll of Honor volumes
The Final Disposition, Vols. I-IV (remains of Civil War soldiers)
The Unpublished Roll of Honor

———–
Full listing, Virginia Directories & Member Lists
Hopewell, Virginia Memberships, 1759-76

————
Full listing, Virginia Court, Land, Wills & Financial
The Burlington Court Book (This is for Burlington Co, NJ)
The Natchez Court Records, 1767-1805

I did not look at “all 7922 Virginia Stories, Memories & Histories”. They should be reviewed as to date and place as well.

2 SteveSeptember 10, 2008 at 5:59 pm

“And as I watched many people do many searches, it would appear that you will get the most mileage on your searching if you choose exact date ranges and exact locations.”

Oh, yes. This worked really well.

I, once again, entered my dad’s name — fuzzy — and exact dates for birth and death and locatons thereof.

Got one hit — an Ancestry
Family Tree. How helpful.

“Toggling between different combinations of what is exact and not exact is also useful in bringing up different sets of information.”

And is just so darned helpful for folks with dial-up connections.

Sheesh.

3 JadeSeptember 10, 2008 at 7:34 pm

Anne,

In your initial post here you say “And as I watched many people do many searches, it would appear that you will get the most mileage on your searching if you choose exact date ranges and exact locations. Toggling between different combinations of what is exact and not exact is also useful in bringing up different sets of information. My anecdotal evidence shows that choosing all exact fields is just not all that great.”

These points are what we have been chewing on for weeks. New Fuzzy does strange things in various combinations of exact and non-exact, and *still* does not retrieve material that readily can be found using Old Search. It is very tedious, time-consuming and even physically painful to do all that toggling, scrolling and clicking, clicking, clicking. Remember that one objective was supposed to be to improve the user experience.

I hope you are not saying that you did not tell the searchers at FGS that they could use Old Search if they wanted to.

I am glad to see “I truly believe that fixing the data and standarizing is the way to go.”

Very sensible. And some of the recently uploaded very large databases are bizarrely indexed as well (England and Wales BMD Indexes, California Voter Lists, to name just a couple).

4 EricSeptember 10, 2008 at 8:49 pm

Hi Anne,

I find it very frustrating to go to a page for a specific location and find things that are common to everywhere. If I were to have an option I would like to see the databases that relate a specific place like Virginia list at the top and the common database later in the page if at all. Or better yet a hierarchy goes from general to specific like is used to drill down into a census image:

Country to Date to State to county to precinct.

Regards, Eric

5 ReedSeptember 10, 2008 at 9:25 pm

Anne,

Welcome back.

For starters, I completely agree with Jade’s posts here (#1 & #3).

Also, please be sure to revisit the previous thread: http://blogs.ancestry.com/ancestry/2008/08/25/finding-levi-s-baker-or-how-to-use-new-search-interface-to-find-him/#comments

There is a lot there—from many blog readers—regarding the need for truly “exact” Exact Searches and improved indexing, results sorting and wildcard functions.

There is also something like a consensus about the need for separate fields/boxes for First, Middle and Last Names as well as for Country, State/Province, County, City/Township/etc. (i.e. the smaller geographic or political divisions).

I think there is also a clear wish for a complete overhaul of the layout of the New Search and New Search Results screens. Less white space, easier navigation (get rid of the mouse-clicks!!!) and less scrolling to view search fields and/or result lists.

Oh, and would you please address—directly and specifically—the comment from Jade (#98) and my followup question (#99)?

Thanks,
—Reed

6 LenSeptember 11, 2008 at 4:45 am

#5 Reed,

I agree 100%

7 MikeSeptember 11, 2008 at 4:55 am

Anne,

I want to echo Eric’s comments about the state/locality pages. Not only are such cluttered with unrelated databases, some of them have so much of such that it looks like an intentional method of puffing up a state that Ancestry really has no coverage on.

Every state does *not* need New England databases listed under it nor Virginia ones, just because most families with long roots in America came through one of those places. Face it, for many states like Hawaii for instance, there just is no coverage and Ancestry should not attempt to hide that fact with unrelated stuff.

Also, too many databases, specifically scanned books that relate to an entire state, are buried under “Stories and Publications” when they should be listed on those state pages. What I mean are early state wide tax lists and such.

8 JadeSeptember 11, 2008 at 9:38 am

Anne,

Again regarding your first post here, “State and country pages. We are working on bringing those up-to-date and giving you a way to easily access those. For example, try looking at the Virginia Database Page or the Italy Database Page. Also check out the Sources pages and the How To’s pages. Now would be a really, really good time to tell me what you think is a must have on those pages and what you really don’t care about. So start voicing your opinions now!”

For the State of Delaware, remove all the general databases: census, passenger lists, naturalizations, military (every war), DAR, Sons of the American Revolution, Genealogical general databases, directories, etc.

Delete all of the specifically irrelevant listings: Natchez court records, Cumberland settlement (Tennessee), New England Town and Vital records of all types, Gateway to the West (Ohio county records).

The Early Settlers of Forks of the Delaware is about Pennsylvania.

Keep the badly done index to Delaware Census enumerations, the two DE newspaper items, the Calendars of Sussex, Co, DE and New Castle Co, DE wills and estates records, and the subcontractor’s awful index of Delaware Marriages.

Add the Calendar of Kent Co, DE wills and estates by de Valenger.

The listings under this state are typical in that they include not only general databases that would be expected to have some DE entries, but also numerous explicitly irrelevant items.

Some of the general databases also lack any DE references, which can be determined by using Old Search for key-word “Delaware” in these items.

I agree with Mike’s #7 comment regarding these listings: “puffing up a state that Ancestry really has no coverage on.”

The myriad irrelevant listings will discourage the new researcher, and do frustrate the more knowledgeable records-seeker who tries to use this avenue to locate resources. All in all the keyword-based Card Catalog is much easier to use for finding any Ancestry databases for localities. That is, if one is searching for records or semblances thereof, rather than for Tree junk.

9 Anne MitchellSeptember 11, 2008 at 10:46 am

Great feedback in here as usual.

What I’m hearing so far is, keep the general stuff out of state specific pages. US Census records go under US, Canada Census records go under the Country pages, and not the state or province. Or as one of our Library/Genealogy experts here told me, store don’t clutter state pages with US level stuff.

I just went through the Hawaii page. Uh, yeah. I see exactly what you mean.

And you would also like to see more listings of related newspapers and books within a page.

Here’s a question, would it be useful to have county pages in the US?

What about the source and how to pages? They are out of date at the moment. Are they worth bringing up to date?

10 ReedSeptember 11, 2008 at 11:56 am

Anne,

Re your #9:

“…keep the general stuff out of state specific pages. US Census records go under US, Canada Census records go under the Country pages, and not the state or province. Or as one of our Library/Genealogy experts here told me, store don’t clutter state pages with US level stuff.” —Agree.

“…more listings of related newspapers and books within a page.”—Agree, AND, furthermore:

NOW is the time for Ancestry’s databases to ‘fess up and explicitly state what is included in a database and what is NOT. (You know, just like your local library catalog will list periodical holdings with “beginning date,” “ending date” and “missing issues/volumes”?)

For example, after two years, many (mostly unanswered) queries and complaints from me–and at least one “major update” to the Historical Newspapers, BMD Announcements, 1851-2003 (about 10 months ago?)–there are many Chicago Tribune BMD items that show as Exact search “hits” on the results page, but for which there has NEVER been a viewable image at the corresponding link (at any percentage size). I have encountered many similar “missing images” for Chicago Tribune BMD hits.

I’ll post more on this separately, as it is a real problem that I’ve never seen properly addressed by Ancestry (even after numerous contacts with “customer service” and a lengthy, detailed correspondence with the Ancestry Insider—back in June, when he still worked for y’all).

“…question, would it be useful to have county pages in the US?” —Perhaps. But here we get into the whole bag of parent counties and their offspring. For example, one of my research interests is in southeast Wisconsin, in the present-day Ozaukee County, which was created in 1853 from Washington County. And, prior to Territory status (1836), Wisconsin was considered part of Michigan Territory (1818) and before that it was part of the Illinois Territory (1808) and before that part of the Indiana Territory (1800), etc., etc., etc. (Here’s a map to explain: http://www.wisconsinhistory.org/reference/maps/pdf/RM018.pdf)

So you can see how this would require some thoughtful attention to detail in the cataloging and indexing of resources at the “County” level… Given the current haphazard cataloging/indexing situation with many Ancestry resources this could either be a big opportunity or a huge headache.

More to follow,
—Reed

11 ReedSeptember 11, 2008 at 12:31 pm

Anne,

This may be slightly off-topic (in which case you or a colleague might want to open a new blog thread), but since you remarked (above, #9) “…And you would also like to see more listings of related newspapers and books within a page…” Well, I see an opening that I can’t resist:

In the “Historical Newspapers, BMD Announcements, 1851-2003″ database there are some VERY large “holes” in the coverage of the BMD data. I had first noticed this when searching for obits from the Chicago Tribune in the 1960s and 1970s, but the problem appears much more widespread.

For example, looking for obits in the “Historical Newspapers, BMD Announcements, 1851-2003″ by using an exact search of newspaper title + location + date. Here’s what happened for one of them. Subject is Minnie Baker Nolton, d. Chicago, 4 Aug 1948.

Searched for all Aug., 1948 obits in the Chicago Tribune and: No obits. None.

So, I searched Chicago Tribune, 1948, obits. Result: No obits. None.

So, how about ANY kind of BMD records in the Chicago Tribune, 1948? Aha! 818 records: 3 [sic] BIRTH announcements and 815 MARRIAGE announcements! It seems that marriage records have decent coverage (though I don’t know if 815 announcements represents anything close to the total number actually published in the Tribune in 1948 or not). Only 3 birth records and no death records indicate something is amiss. So now I’m suspicious, and I ran several searches to see how the BMD records are represented throughout the period 1945-1955.

Parameters: newspaper=Tribune + location=Chicago + date=1950 (+/- 5 years) + record type=BIRTH.
Results: 11 (eleven) notices from the period 30 Sept 1945 to 10 Aug 1955.

Parameters: newspaper=Tribune + location=Chicago + date=1950 (+/- 5 years) + record type=MARRIAGE.
Results: 17,998 notices from the period 2 Jan 1945 to 31 Dec 1955. (That seems more like it.)

Parameters: newspaper=Tribune + location=Chicago + date=1950 (+/- 5 years) + record type=DEATH.
Results: 0 (zero) death notices from the period 2 Jan 1945 to 31 Dec 1955.

At this point I got angry enough (and depressed enough) that I stopped. I now realize that the collection entitled “Historical Newspapers, BMD Announcements, 1851-2003″ is—at best—a serious misnomer. After all, one would assume that—as the home page for the search engine states—this database contains—for example—135 years of Chicago Tribune BMD records:

“Ancestry.com. Historical Newspapers, Birth, Marriage, & Death Announcements, 1851-2003 [database on-line]. Provo, UT, USA: The Generations Network, Inc., 2006. Original data: […] The Chicago Tribune. Chicago, IL, USA: The Chicago Tribune, 1850-1985. […]”

So I decided to click on the “more information” link. Perhaps that link would explain which records are missing for which year? Details? Nope. Just this one nonchalant sentence buried in the paragraph:
‘Note: There may not be records for all three vital events included in this database for each newspaper and year combination.’

That’s putting it mildly! Now, like many Ancestry users, I assumed that this meant the occasional page of records might be missing, or even a few issues of a given paper might not be in the database, especially in the nineteenth-century editions. And, this being Chicago, perhaps the pre-Fire years (1855-1871) will be spotty. But really, an entire DECADE of post-WW2 obituaries (1945-1955) missing? And only eleven birth records from the same period? From the “newspaper of record” of the largest metropolis between the coasts? And who knows how many other gaping holes exist in the database? I haven’t the heart to check.

But it seems to me that Ancestry should check, and should accurately label this database (and ALL databases). After all, if I go to the local library and look up a newspaper or periodical in their collection I expect (and will get) an accurate catalog listing with the start and end dates of the item AND a list of missing issues. From this unhappy customer’s view, it seems that either (1) ProQuest has sold Ancestry a seriously flawed database—which must be fixed ASAP—or (2) Ancestry already knows how extensive the problems are and has been hiding that information from its subscribers—which must be corrected ASAP.

Ancestry needs to speak honestly and plainly to the subscriber/users of its databases so that we know the limitations of each database. The lack of this information is a serious impediment to our use of the data, may cause researchers to reach false conclusions about ancestors with BMD events in these major cities and, most importantly, it veers awfully close to false advertising, something which any self-respecting business should want to avoid at all costs.

Sorry for the rant, but I’ve been using the Historical Newspapers BMD database for almost two years and did not realize how really, really poor the coverage of the BMD image collection is. The gaps in the data must be clearly identified and—one hopes—fixed.

Thank you for your patience. I do appreciate the size and complexity of Ancestry’s operation—and that this may be someone else’s department—but this situation is way beyond what is acceptable.

I look forward to your reply.
—Reed

P. S. Need another—extreme—example?

Check out the “database” for the Chicago Daily News. As the Encyclopedia of Chicago (online) says: “The Chicago Daily News, the city’s first penny paper and the most widely read publication in Chicago during the late nineteenth century, was founded in 1875 …By the late 1880s, when it lowered the price of the morning edition to a penny, it enjoyed a daily circulation of about 200,000, which made it one of the most widely read newspapers in the world…It remained Chicago’s most popular newspaper until 1918…At the end of World War II, when annual revenues approached $15 million, the paper ranked among the top 15 publishing companies in the United States…The Daily News ceased to exist in 1978…”

Wow, what a great resource, eh? So, how many issues of this important daily are in Ancestry’s database? One (1) issue! Thursday, 9 Oct. 1879. (But at least we get all 6 pages!) I ask you, what’s the point? Is Ancestry just “pumping up” the titles and quantites of newspapers “in” the databases?

12 Tony CousinsSeptember 11, 2008 at 12:58 pm

I completely agree with Reed at #5 and all those others who have argued for separate fields on the name and locations. It seems to me that Ancestry has generated simplified indexes consisting of concatenated fields to make searching easier – hence the type ahead in the location field. If you only have to search on two keys for the concatenated name and location it must be quicker, this would also go some way to explain why sometimes there are no hits.

A lot of my research is in the county of Suffolk in England. One of the towns of interest is Stonham Aspal – note only one ‘L’ in Aspal. Try the drop down in the new search for this town, now there are miraculously two – ‘Aspall’.

From memory, which admittedly is in need of an upgrade sometimes, Ancestry used to have both listed one and two ‘L’ spellings. It is clearly listed on the census images as ‘Aspal’. Did Ancestry ‘standardize’ on the two ‘L’ versions to add it into the type ahead location field?

Anyway – go Reed :)

TonyC

13 Anne MitchellSeptember 11, 2008 at 1:07 pm

I don’t think it’s so much a case of “fessing up” :-) but I agree we should make it clear what is in a db, and therefor what is not. No one is hiding anything; I think no one has thought about the best way to do this, or that it needed to be done.

Let’s try and steer clear of discussing specific datasets and what is missing…it really has to be done in a different forum I do believe.

14 JadeSeptember 11, 2008 at 2:25 pm

Anne,

You requested “Also check out the Sources pages and the How To’s pages. Now would be a really, really good time to tell me what you think is a must have on those pages and what you really don’t care about. So start voicing your opinions now!”

The information in the Sources pages badly needs updating and some of the information is wrong. E.g., Pennsylvania does not have statewide recording of marriages; these are still in the Counties and date (with rare exceptions) from 1885. Virginia’s recording of marriages were in the Counties beginning their dates of inception, not 1853. West Virginia’s statewide death records begin in 1909, with gaps varying by county due to a fire.

I think the U. S. States’ Sources pages should be kept and updated by a paid knowledgeable researcher. It should have a link to a table showing changing County formations, giving parent and child Counties.

The ‘How To’s’ are too general to be of a lot of help. Either rewrite with specifics or just leave them. They were quick quotes from Ancestry’s Red Book, which never has been the best overall research source, but did have helpful sections on types of Court records.

You asked if there should be County-specific pages. Ancestry.com does not have enough coverage of most States to make this reasonable. The issue of changing boundaries is a point well taken. However, there could be a page arranging specific resources alphabetically by County, to which there is a link from the State page. This requires the work of a paid knowledgeable researcher.

15 Jerry BryanSeptember 11, 2008 at 8:19 pm

I have a question about the State Source Pages. In her initial post to this blog, Anne made reference to the Virginia Source Page, and references have also been made to the Hawaii Source Page.

In my years of using ancestry, I have occasionally tripped across these locality pages by accident. But I have never been able to find them in the few cases where I went looking for them. How do you find the darn things?

Well, I have discovered that I can find them very easily with Google, but I still can’t find them within ancestry itself. Where do you look?

By the way, I share the views of others that most of the databases that are in the locality pages don’t belong there. Why would I go to the Virginia Source Page or the Tennessee Source Page or the North Carolina Source Page to find the 1930 census? I would go to the census to find the census. If about 90% of the databases were removed from these locality pages, they would be much more useful.

16 Jerry BryanSeptember 11, 2008 at 8:39 pm

Here’s another question about the overall content design of the ancestry Web site (not the look and feel of the searches that we have been talking so much about). Does ancestry have any librarians on staff, and are they involved in the design of the content of the pages?

I’m an IT person, and through the years I have regretfully come to the conclusion that most IT people create pretty lousy Web pages with respect to the organization of the content (again, I’m not talking about prettiness and look and feel, just the organization of the content). I include myself in the list of IT people who do a pretty lousy job of organizing Web content, even though I’m a whiz with HTML and CSS. For that matter, I think a lot (maybe most!) trained Web designers do a lousy job organizing content. Web designers get so involved in “look and feel” and “pretty” and “pretty fonts” and “pretty colors” and animation and “gee whiz” that they forget about organizing content properly.

And who does organize content in a really useful fashion? Librarians!! And I mean professionally trained librarians with MLS degrees. They are trained to organize and index information in ways that make sense irrespective of the medium. The Web is just one medium in which they operate.

For that matter, how many MLS librarians does ancestry have on staff who worked on the design of New Search or who work on the design of the database indexes? I realize I’m being a little negative sounding here, but I suspect that the number is zero. If the number were greater than zero, I suspect that a lot of ancestry’s indexing problems wouldn’t exist, and I suspect that a lot of New Search’s problems of combining place names into a single search box and combining people’s names into a single search box wouldn’t exist. For sure, turn the locality pages over to an MLS librarian to organize.

17 Carol A. H.September 12, 2008 at 12:29 am

To find th “Source pages, Database page, and How To’s, click on Search, click on named state or cartoon shape from a map of US and the they appear in the top right hand side of screen. I didn’t know where they were until tonight.

18 Jerry BryanSeptember 12, 2008 at 5:52 am

Re: #17. Much thanks. They are easy to find once you know where to look. I guess I was looking for them on the Home page instead of on the Search page.

19 45RPMSeptember 13, 2008 at 7:32 am

re: Anne, #13:

you wrote: ” I think no one has thought about the best way to do this, or that it needed to be done.

Let’s try and steer clear of discussing specific datasets and what is missing…it really has to be done in a different forum I do believe.”

Given that users have been complaining about this subject for years, it is most disheartening to learn that no one is listening.

Since you have deemed it appropriate to hijack this forum for your personal “pet” projects, where would you suggest we go to discuss other shortcomings at Ancestry.com?????

20 Jerry BryanSeptember 13, 2008 at 7:52 am

Back to #17, I’m afraid I’m going to have to ask a dumb question again. I found the locality pages in Old Search, but even with your directions I can’t find them in New Search.

Also, I’ve been thinking some more about the census and the locality pages. I see no problem (and some advantages) in having a single link from each locality page to the main census page (I’m thinking the U.S. census here, but the same concept would probably apply to non-U.S. censuses). But having a separate link for the 1930 census, for the 1920 census, for the 1910 census, etc. on the locality pages makes no sense.

However, many and perhaps most localities have some sort of special census database or census like database that is specific to that locality. Having links to these special censuses on the locality pages makes a lot of sense (for example, Iowa State Census, 1895).

Finally, there are a few census or census like databases that are not really specific to a locality, but that nevertheless appear on the locality pages rather than the main census page. For example, I’m thinking of A Census of Pensioners (which is really an extract from the 1840 census) and U.S. IRS Tax Assessment Lists, 1862-1918.

21 Carol A. H.September 13, 2008 at 1:20 pm

Re: Jerry’s #20 which references his #15, which in turn references Anne’s post.

With just a QUICK search starting from the new search, I did NOT find a way to get to the “Source pages, Database pages and How-to pages” for each state. I went there and copied the addy from the top of my browser and pasted it as an add on link.

http://www.ancestry.com/search/

Now that seems like a dumb thing to have to do! Hear that Ancestry?! Fix it!!

One can do it on an individual basis for each state but that is not practical. Three times 50 equals 150, last time I did the math.

More later if I find a way to get there from the new search. Sounds like driving in LA. “You can’t get there from here!”

Jerry, it was NOT a dumb question. It was a good question since you currently have to be in the old search to get there.

God save the old search!

22 JimSeptember 13, 2008 at 4:12 pm

Just save it as a bookmark or favorite in your browser.

Really, this is not a big deal.

New Search rock on!

23 ReedSeptember 13, 2008 at 7:28 pm

45RPM,

Re your comments (#19) which follow up on Anne’s #13:

You wrote: “Given that users have been complaining about this subject [database errors] for years, it is most disheartening to learn that no one is listening…where would you suggest we go to discuss other shortcomings at Ancestry.com?”

I couldn’t agree more completely. As I (and many other bloggers) have mentioned in previous posts, Ancestry’s corporate “culture of non-responsiveness” is a big, big, problem that is still waiting for a long-term solution.

On the other hand, I think you’re being unfair to Anne when you say: “Since you have deemed it appropriate to hijack this forum for your personal “pet” projects…”

Anne is, after all, an Ancestry employee and from her first post she made it clear that she has been assigned to deal with the specific topics of New Search, its processes and user interface. She has been addressing these topic for several weeks now (in several blog threads, this is the most recent) and—to her credit—she has been vastly more responsive than any other Ancestry.com blog moderator or so-called “customer service representative.”

I too would like to see more blog threads—perhaps an open forum—where we could publicly address the many problems with databases, missing images, index problems, etc., etc., etc…

Meanwhile, let’s cut Anne some slack. She may be overly fond of New Search, but at least she’s been responsive to many of our blog queries and comments so far.

—Reed

P. S. Say, Anne… You wrote (#13) “Let’s try and steer clear of discussing specific datasets and what is missing…it really has to be done in a different forum I do believe.”

OK. So how about getting one of your Ancestry colleagues to open just such a blog thread, eh???

24 ReedSeptember 13, 2008 at 7:50 pm

Jerry Bryan,

A resounding “Amen” to your post #16. Since you put it so well, I’ll just repost a few of my favorite bits here:

“And who does organize content in a really useful fashion? Librarians!! And I mean professionally trained librarians with MLS degrees. They are trained to organize and index information in ways that make sense irrespective of the medium. The Web is just one medium in which they operate.

“For that matter, how many MLS librarians does ancestry have on staff who worked on the design of New Search or who work on the design of the database indexes? I realize I’m being a little negative sounding here, but I suspect that the number is zero. If the number were greater than zero, I suspect that a lot of ancestry’s indexing problems wouldn’t exist, and I suspect that a lot of New Search’s problems of combining place names into a single search box and combining people’s names into a single search box wouldn’t exist. For sure, turn the locality pages over to an MLS librarian to organize.”

I would only add, turn ALL the content over to qualified librarians for proper indexing and filing: locality pages, “Historical Records,” “Stories and Publications,” etc. Oh, and how about Ancestry’s “Card Catalog” as well !?!

This is an especial pet peeve of mine; no self respecting librarian would call Ancestry’s unorganized, hard-to-search hodgepodge of irregularly labeled sources/titles a “card catalog.” There’s a LOT of good stuff there, but—Yikes!—just try and find it! I’ve used small-town libraries in Nebraska and (online) Alabama that have more sophisticated (and logical) search and sort functions.

—Reed

25 WoodySeptember 14, 2008 at 9:57 am

Well, it had to happen sooner or later. I bet most who have been with Ancestry more than a few years had already developed a creeping worry that the rapidly increasing and sometimes poorly structured new content introduced on Ancestry would bloat our previously tidy global search hit lists. We hoped for more care on Ancestry’s part in structuring (e.g. coding and indexing) the content and more interest in supplying better search tools. And so I accept that Ancestry must find a new way to manage the challenge of searching old and rapidly expanding new content.

Although I am addicted to Old Search, because I know it well, I venture often into New Search, because I must get to know it well too. I’ve turned to New Search for simple and challenging tasks. I am consistently frustrated by the slow, often arbitrary, and ultimately unpredictable nature of the New Search experience. I am worried, because even as I learn some subtleties in New Search and as Ancestry corrects major bugs, the search experience does not improve significantly.

Anne says that we should not dwell on specifics of individual databases. Fair enough – she’s not in charge of these databases. But many of the user-provided examples and associated frustrations boil down to problems related to inconsistent data fields and indexes across the individual databases. It is now clearer than ever that the reason Old Search works reasonably well is that at least some of these problems were addressed – albeit inelegantly implemented by various database-specific hacks and programming wizardry we are told.

I do not see how Ancestry is going to solve its problem of managing growing content by ignoring the idiosyncrasies of its individual databases that arise from the fundamentals of content acquisition, organization, and indexing. “Garbage in, garbage out.” If these databases are not restructured to permit useful exact and fuzzy searches, then New Search is doomed.

We are told that revising the backend search engine and database structural problems is a second priority. The first priority is to create and perfect the New Search interface. It has been said (or implied) here over and over – Ancestry got the priorities wrong. I see no hint from the Ancestry team acknowledging the misorder in these priorities. There is a disturbingly intransigent tone in the Ancestry responses to the work of Jerry, Reed, Jade and many others on this blog. It seems that New Search is here to stay, with its visual bloat, poor ergonomics, sometimes arbitrary search logic, and too often unhelpful search hit lists, all built on a creaky, poorly aligned structure of databases with “billions of records.”

26 JadeSeptember 14, 2008 at 11:41 am

Woody, in #25 you present a reasoned and insightful view of the core issues.

However, Ancestry.com and the other TNG components (Rootsweb to some extent excepted) are driven by Marketing, not by Public Service. For just one example, look at what was done to the Message Boards in order to create pages with more advertising space and to create links claimed to be ‘informational’ (note: omitting any links to USGenweb sites) but actually to promote Ancestry Books.

For example: after reading Jerry Bryan’s #15, concerning inability to find the US State Sources and ‘How To’s’ pages, I tried to find them ‘from scratch’ too. I mistakenly clicked on the Learning Center tab.

The Learning Center principally advises starting a Tree, and raves about how there will be automatically computer-found Ancestry Tips concerning one’s ancestors. But this is one of the Famous Ancestry Tricks, because the person who begins a free Public Tree will not be able to view ‘Tips’ documents without subscribing to Ancestry.com.

Lower on the page is a large ad for FamilyTreeMaker – not version 2009, which is available, but version 2008, which is notoriously buggy and defective, and which lacks some of the helpful features of earlier editions (being oriented to link with Ancestry Books rather than to facilitate self-publishing).

The Learning Center is not about how to do genealogical research.

Despite the assertions by Anne, New Fuzzy is indeed based on the Tree-type gedcom format in order to ease searching Trees and, indirectly, to market FTM, Ancestry subscriptions, and even the sick One World Tree to which subscriptions are still being sold.

More directly, the page real estate will be occupied by more advertisements, which will load before any of the search features can be used. (For track record, see the Message Boards and the new ‘Home Page’ setup). Regrettably, no browser allows reversing the loading order unless one sets the browser specifically to not load any graphics. For those of us with dial-up, this will make New Fuzzy even slower and more infuriating to use.

For now, Anne has dropped discusson of New Fuzzy. I hope this means at least that helpful revisions are being made.

27 Anne MitchellSeptember 16, 2008 at 8:56 am

Re #14, I think the idea of having how counties were formed is a really good idea. How to arrange county specific information on the pages will be interesting.

Re #16, BTW, yes we do have professional genealogists and librarians on staff; some of them have alread been in contact with me about these pages. And I will seek their input on this.

Re #19, 45pm, I suppose these are my “personal pet projects” as I am the the Product Manager for ancestry.com. It is also impossible for me to have a useful conversation about multiple things at the same time — I’m just not that talented. And I have to say that I think working to understand the needs of search users to create a better search is a reasonable pet project. Your mileage may vary.

Re #20, I think having a general link to the census off each state, etc page is a good idea. But I agree that these pages should be set up to help researchers uncover datasets they may not know about.

Re #21, you can’t get to the state/country pages from New Search, yet. I’d bookmark it, or add it to your quicklinks.

Re #23, I think it may be me that needs to open that thread, but one thing at a time. It won’t be a quiet thread, and I’m about to go on vacation
until Monday…so I don’t want to leave unattended that long. :-)

Re #26, we aren’t driven by either marketing or public service. We are a for profit company … as most of you are aware. And our customer’s
opinions matter. And advertisements that you see in some places on the site, help keep subscription costs down. And yes, we are working on some revisions.

28 ReedSeptember 17, 2008 at 10:33 am

Anne,

Thanks for the update (#27). One comment of yours got me thinking. You wrote:

“Re #23, I think it may be me that needs to open that [new-topic] thread, but one thing at a time…”

OK, fair enough. But why should you be the only one to do the heavy lifting on product-improvement communications for Ancestry.com? As this blog has often noted (here and in other threads), Ancestry has many content and functionality issues that need addressing. And Ancestry also has a big, ongoing problem with (in)famously unresponsive customer service.

So, a proposal: The Ancestry blog needs a moderated, open-topic thread where blog readers can ask questions of all sorts and propose new blog threads for further discussion, questions and answers.

You know how some newspapers have a “Public Editor” or “Readers’ Advocate” (e.g. the New York Times), someone responsible for fielding reader questions and finding answers from the right people in the organization? Ancestry needs something like this, an advocate who could field questions, get new threads formed, and then find the right Ancestry employee to moderate each thread.

Now, I’m sure that moderating an Ancestry blog is sometimes less-than-blissful. But when you look at the big picture, all those (occasionally cranky) blog posters are really a huge resource for Ancestry. Think of the hundreds of comments and detailed explanations and suggestions you’ve received in the just last few months on the New Search blog threads.

The readers of the Ancestry blog are a kind of transparent focus group, comprised of men and women who use the product regularly, who care about improved functionality and resources AND who are willing to donate hours of their time to investigating issues and communicating in writing to each other and to Ancestry.

And let’s emphasize the word “transparent.” Ancestry would have many more satisfied users if only the company would take the time to explain, publicly (much as you have tried to do with the New Search blogs):
• each problem/issue
• the plan for improvement
• what’s happening
• when we users might expect improvement

While I still don’t like using New Search (bad GUI, poor search results), I do appreciate how you have moderated the New Search blogs. If only we Ancestry users didn’t have to wait for the company to begin each blog thread that is relevant to our concerns.

So. How about it?

—Reed

29 Anne MitchellSeptember 17, 2008 at 2:33 pm

Reed, I think Michelle, over on the FTM entries is doing plenty of heavy lifting these days, and keeping up quite nicely I’d say.

I think anything that deals with search and datasets is my domain to discuss. And you do see Kenny and David on the blogs as well as others.

There are message boards for ancestry.com proper, but I know I don’t monitor them on a regular basis, and it’s a time problem. I could spend all day chatting with people about this problem and that, and lots of new ideas, but then I’m spending all day chatting and not working to solve problems and improve search. It’s finding the right balance.

Blogs work best for me, because I can bring up one issue at a time, read the responses and take that information back to work on the things I need to work on. Different Product Managers will have different styles. Yes, I’m guiding the conversation, but for me it is the best way to get input from this group of users. And so far, it’s been very educational. :-)

A “Readers’ Adovcate” is an interesting idea, but I’ve seen these tried before and it takes some time and thought to get it going and make it actually work. But it is worth thinking about.

I won’t be back on the blog until next Monday — I’m off to take my son to college. Hopefully every one will have a lovely weekend.

30 Andy HatchettSeptember 18, 2008 at 7:02 pm

Anne,

I think I may have stumbled onto something that will impact your work on the search feature.

I’ve joined the World Archive project, working on the 1866 Calhoun County, Alabama census, and find a strange thing.

When keying this project the only info to be keyed in is:

County
Race
Surname
Given Name

What puzzles me is that the census form clearly states at the top:
“White Population of Fractional Township No.15, Range 12, in Calhoun County, Alabama.

Am I to understand from this that, since we are not keying the township, this particular census will not be searchable by Township?

You might want to check this out.. and what other fields are not being indexed on other World Archives Projects?

I am correct that search is based on indexed fields-right?

Andy Hatchett

31 TonyCSeptember 20, 2008 at 9:32 am

Anne,

I know that a number of people have asked when Ancestry will allow users to make changes to fields other than the name – when will that be.

The reason I ask is that I do a lot of research in Suffolk, England and have noticed towns spelled incorrectly.

Two examples are Stonham Aspal and Saxtead.

Stonham Aspal is the correct spelling but Ancestry and FTM2008 drop down has it as Stonham Aspall, even though the images are clear and a quick search with any mapping application gives the ‘one L’ spelling.

On Saxtead, Ancestry researchers or others have inserted an ‘S’ so it is listed as Saxstead. Again, all mapping applications list the place as Saxtead.

Obviously this makes it impossible to use these as either residence or birth location.

Where can these glaring errors be reported and what chance is there of any one at Ancestry listening and dare I ask – doing something about it.

TonyC

32 Terry ReigelSeptember 20, 2008 at 3:33 pm

Hi, Anne,

I just found your blog, and have been reading all the comments about the new search. I normally use the old Advanced Search, and wish my home page at Ancestry would “remember” that I always want the advanced version.

I’ve tried the new search a couple of times, and find the autofill of places is nice, except it fails to understand many places have “moved” over the years. For example, it thinks Greenbrier Co. is in WV. Maybe it is now, but for the period I care about it was in Virginia.

But, more problematic, I generally get no results returned. To cite the latest example, I entered:

Name: Thomas Arbuckle
born: 1780 Greenbrier Co. Virginia USA
died 1838 Clinton Co. Missouri USA
spouse: Jane Davis

Result – not a single hit.

Entering exactly the same data in the old advanced search yields a lot of hits, the first two being him. When I use the “see others looking for him” link a see a page full of trees about him listed.

It’s hard to like a new feature when it doesn’t work. :-(

33 Jerry BryanSeptember 20, 2008 at 6:38 pm

Re: #32. I tried the search as exact with both Old and New Search, and I got no matches either way. I then tried the search as fuzzy with Old and New Search, and I got lots of matches either way, with the first two being him with both Old Search and New Search. It seems to me that with the exception of the typeahead for the place name, New Search is working just as well as Old Search for this particular search.

In the case of New Search, I typed in Greenbrier County, Virginia myself and ignored the New Search type ahead for place names – in my opinion one of the most poorly designed areas of New Search. Well, the typeahead seems to me to be more of a symptom than being the real problem. The underlying problems are the fact that New Search doesn’t break places names down into separate fields for country, state, county, town or city etc., that places names specified for searching individual databases are not specific to the locality implied by the database, and that there is no historic understanding that places names are time dependent (e.g., your example that Greenbrier County was in Virginia prior to the Civil War).

34 JadeSeptember 21, 2008 at 5:22 am

Re: Terry #32 and Jerry #33, this is a real flaw in New Fuzzy.

Old search does not have a problem in global search with differentiating between where places were, geopolitically and where they are now.

The gedcom standard is “where a place is now”, which often does not match up with place-names entered in databases.

This is a big gap between the ‘Tree’ approach and the search-database approach.

New Fuzzy’s place-orientation appears to work against the search engine’s ability to find, say, Greenbrier County, WV entries for Census returns prior to 1870.

The problem with database place misspellings works the same way. Who would ever think that an exact search for ‘Cogle District’ would be required to find ‘Eagle District’ in the 1890 US Census Veterans’ Schedules? :-(

35 Mary JochimSeptember 21, 2008 at 8:41 am

Anne, you see like a nice person and good go to person. I’ve been looking to see if there is a discount when order muliple images from Ancestry.com …

If I order multiple images do I get a price break?

Thanks Mary (please respond to sterlingworld@aol.com

36 Terry ReigelSeptember 21, 2008 at 11:14 am

Re: Jerry #32

By “fuzzy” you mean without checking “Exact,” right? I normally use the default mode, with exact off, and that’s what I did in both cases in this example. Are you saying that Exact is on by default in New Search?

I just went to the new search advanced screen, which shows “exact,” which I left unchecked, and entered the same data again. Again I “no matches.” What did you do to get any matches with new search?

37 Sheila SkeneSeptember 21, 2008 at 5:07 pm

I am a little bit lost here. Can you please help.

38 Jerry BryanSeptember 22, 2008 at 6:31 am

Re: #36. I wish I could easily post a screen capture to this blog, but I can’t, at least not easily. In any case, I just repeated the experiment. I went to New Search, advanced mode, and unchecked the exact option.

(I repeat for the umpteenth time that New Search’s advanced mode is a terrible design. Advanced mode is required in order to see the exact option. The design shows a very strong bias towards fuzzy search, especially since the only thing advanced mode does is to display the exact option.)

The exact option that I unchecked was the master switch at the top of the search box that then turns off all the other exact options throughout the search box.

Here’s the data I entered, as displayed in the Refine Search box.

Name: thomas arbuckle
Birth: 1780, greenbrier county, virginia
Lived In (Residence):
Death: 1838, Clinton County, Missouri,
Family Members: jane davis
Migration:
Military Service:
Marriage:
More:

You can tell from the Refine Search box that the search is fuzzy because none of the items are enclosed in quotes. Any items that are exact would be enclosed in quotes. I got bunches of matches as described in #32. I don’t know why you are not getting any matches. How does your Refine Search box look as compared to mine?

(I repeat for the umpteenth time that New Search’s Refine Search box is a terrible design. It is intended to keep your search arguments visible on the screen and to reduce the amount of scrolling that you have to do. In practice, it greatly increases the amount of scrolling you have to do. Also, it takes away very valuable real estate on the screen that should instead be used to show search results. And it shows your search arguments in a way that is very incomplete, misleading, and hard to understand. Many times while using New Search’s Refine Search box I would kill for the ability just to scroll down to the bottom of the screen (I know, I know!) and to see the whole search template all on the screen at the same time, with all fields displayed including the empty ones. The only way to see the whole search template at the same time is at the beginning of the search, not while you are refining at the search. And at the beginning of the search the search template is so full of air and white space that the whole thing won’t even fit on the screen. You still have to scroll to see it all.)

39 Terry ReigelSeptember 22, 2008 at 7:52 am

RE: #37 – Thanks for the reply Jerry.

I’m puzzled by your comment that you unchecked the Exact option at the top – it comes up unchecked for me.

I can’t tell you what’s in the Refine box because since when there are no results, there is no Refine box. :(

I copied the search terms from your post this time, and still get no results.

I even tried checking and then unchecking the Exact box, and it makes no difference.

Interesting thing – when you get no results from New Search the Old Search screen is presented on the “no matches” screen. Does somebody know something? :)

40 Jerry BryanSeptember 22, 2008 at 8:38 am

Re: #38: I’m puzzled by your comment that you unchecked the Exact option at the top – it comes up unchecked for me.

New Search seems to remember whether you are in Exact or Fuzzy mode (where Fuzzy simply means “not Exact”). So the last time you searched, you must have been in Fuzzy mode and it remembered it for you by leaving the Exact option off. “Remembering” your Exact vs. Fuzzy mode in this manner is a good design.

I search in Exact mode 99% of the time, so it usually comes up in Exact mode for me and I have to turn Exact off if I want Fuzzy. Also, if ever in New Search you turn off Advanced search, it also turns off Exact Search (another symptom of the bias towards Fuzzy search – this is not a good design). So you may have turned off Advanced Search which unbeknown to you also turned off the Exact option. I never turn off Advanced in New Search because I always want to be able to see the Exact option.

I can’t think of any good reason why the Advanced option in New Search even exists. The Exact option should always be visible. Or if ancestry wants to keep the Advanced option more or less in its current form, then when you are not in Advanced mode searches should revert to Exact mode.

Old Search also seems to remember whether you are in Exact mode or not, except that the mode information often seems to get lost when I switch between New and Old (which I do a lot, like every time something seems not to work correctly in New Search). So when I switch between New and Old it seems like I usually have to turn Exact back on.

41 Jerry BryanSeptember 23, 2008 at 9:26 am

Here’s an example search that’s not really about New Search vs. Old Search, but rather is about how New Search could and should be a lot better.

I’m looking for the 1910 census entry for Thomas Calvin Scarbrough, born 16 Dec 1850 in Tennessee, and I have a very high expectation that his 1910 census entry will be found in Anderson County, Tennessee. It’s really a very easy search with either Old Search or New Search.

Most records list his name as Thomas C. Scarbrough, so we basically search for Thomas Scarbrough. A major defect of New Search is that if you exact search for Thomas Calvin Scarbrough, it will match on Thomas A. Scarbrough, Thomas William Scarbrough, etc., as well as on Calvin J. Scarbrough, etc. And if you exact search for Thomas C. Scarbrough, it will match on John C. Scarbrough, William C. Scarbrough, etc. An exact search for Thomas Calvin should really only match on Thomas Calvin (and maybe on Calvin Thomas), not all the other variations it now matches on, and similarly an exact search for Thomas C. should match only on Thomas C. Hence my decision to search only for Thomas Scarbrough and to ignore the middle name.

Records for the Scarbrough family often spell the name as Scarbro, Scarbrough, Scarboro, and Scarborough, even for the same person. Indeed, it turns out the 1910 census entry for our guy spells his name as Scarborough. I think exact searches should take such spelling variations into account using the name authority, and that you shouldn’t have to get into all the craziness associated with fuzzy searches just to search for common spelling variations. Another way to say this is that the name authority should be primarily on the exact side of the house rather than on the fuzzy side of the house (that’s the way the LDS familysearch Web site does it, for example). In this case, I can get around the problem by searching for scarb*, but I’m not always so lucky because the wildcard often needs to go into the first three characters. Of course, if exact search did work in the way I described, you would still need a super exact mode that would only match precisely what you entered (again, the LDS familysearch Web site has this feature).

So I did an exact search in the 1910 Anderson County, Tennessee census for thomas scarb* born in 1850+/-10. Normally, I would have included the birth place of Tennessee, but I’m so confident about being in the right county, I figure that any additional information I put would just be an opportunity for things to wrong. Usually, the less information you put, the better. For example, everybody in the 1860 census for Anderson County, Tennessee who was born in Tennessee is incorrectly indexed as having been born in Pennsylvania (go figure!). Also, I normally would have specified 1850+/-5, but I wasn’t expecting very many matches and the further you get from a person’s year of birth the more likely it is that their age is wrong. So +/-10 gave me more chances for a match than +/-5.

There were no matches. With Old Search, I solved the problem by omitting the first name altogether. There were 7 matches, one of whom was Tom C Scarborough, who is my guy. As I said, it’s an easy search. But I only checked Old Search after completing my work with New Search. With New Search, my first refinement was to leave the first name as Thomas but to make it fuzzy. I thought this was a good chance for New Search to really shine.

Instead of New Fuzzy you can exact search for tho* which will get you abbreviations such as Tho. and Thos., but which will not get you nicknames such as Tom or Tommy. I thought making Thomas fuzzy was a better choice. And indeed, it worked fine. I got the same 7 matches I did by leaving out the first name with Old Search, and my guy was at the top of the list. So New Search did shine a little bit. But New Search could shine even better.

Just to repeat, the best solution for this particular search would be for New Search to use name authority on exact searches for matching spelling variations on both the first name and last name. So an exact search for Thomas Scarbrough should have found Tom C Scarborough. I anticipated the Scarbrough vs. Scarborough problem by searching for scarb*, but it’s harder to anticipate the Thomas vs. Tom problem. I easily found my guy with fuzzy search, but I think the results are strange.

Tom C. Scarborough was a 3.5 star match. It should have been a 5 star match. Eliza J. Scarborough, Charlie W. Scarbrough, Marry E. Scarborough, John Scarborough, Vistie Scarborough, and Melvin Scarbrough were all 3 star matches. It makes no sense to me for there to be only a half star difference between my guy who was only off by a spelling variation and the other so-called matches that aren’t even remotely close. The others should be 1 star matches, or maybe even 0 star matches. Now I grant you that Melvin’s full name could have been Thomas Calvin Melvin Scarbrough, and maybe we don’t want to miss matching on him. But that kind of thinking is what leads to the craziness of fuzzy searches yielding millions of matches.

I think that a properly functioning search should have yielded only one match for this particular search. And one match should have been the result whether you were using my suggestion of an exact search that includes name variations, or whether you were using fuzzy matching on the first name. Even with a fuzzy search it’s hard to justify thinking that Thomas was the same person as Eliza, Charlie, Marry, John, Vistie, or Melvin. (Vistie is misindexed, by the way. Her name was Vestie and that’s the way it was enumerated. Because the mistake is in the second letter, she could have been hard to find.)

Kendall is very proud of the millions of matches approach, and he bragged in one of his Webinars about matching on birth dates that were off by 100 years. That’s all well and good, except you are never going find anything in a list of millions of matches. If you really think that Thomas Calvin Scarbrough might have been named Thomas Calvin Melvin Scarbrough, then you need to search separately for Thomas Scarbrough, Calvin Scarbrough, and Melvin Scarbrough, or you need to leave the first name blank. You shouldn’t expect a search engine to be able to figure that out for you without getting into the millions of matches problem. The existing philosophy of fuzzy searches yielding millions of matches for fear that you might miss one effectively renders it pretty useless.

42 Terry ReigelSeptember 23, 2008 at 1:21 pm

Re: #40: New Search seems to remember whether you are in Exact or Fuzzy mode…

Thanks, Jerry, that helps.

43 Terry ReigelSeptember 23, 2008 at 1:30 pm

I recently found that when I use Old Search, then in the results page click the “Census” link, the refine search screen on that page no longer allows you to specify the census location – just the birth and death locations. How are you supposed to tell it which state to look in when you expect the person to be in a different state than they were born or died in?

Thinking this just has to be a mistake, I sent a note to support who just responded that this part of the changes being made to search. I know you said you wanted to reduce the number of search templates Anne, but this is ridiculous. Please tell us this is an error.

BTW, I said this was reached from Old Search. I have no idea what happens from New Search because I’ve never yet gotten a single match from it, so no way to see the “refine” screen.

44 JadeSeptember 24, 2008 at 4:22 pm

Terry, re #43, if you want to globally search the US Census, the page here:

http://www.ancestry.com/search/rectype/census/usfedcen/default.aspx

allows entering place of enumeration (not birth place or birth dates), parents’ names and birthplaces spouse’s name.

You must be using the more general ‘Census’ search.

45 JadeSeptember 26, 2008 at 10:26 am

Isn’t anyone checking the basic Ancestry ‘results’ web page tables for minimal accuracy?

The word ‘cemetery’ has no ‘a’ in it.

But see

http://search.ancestry.com/cgi-bin/sse.dll?rank=1&gsfn=Clifford&gsln=Jump&=&f5=&f6=&rg_f7__date=1927&rs_f7__date=5&f9=&f11=California&f20=25&f21=Dec&rg_f22__date=1944&rs_f22__date=0&f24=&f28=&f29=&f26=&f14=&f15=Private&f18=&f16=&f31=Purple+Heart&f35=World+War+II&gskw=&prox=1&db=abmc&ti=0&ti.si=0&gl=&gss=mp-abmc&gst=&so=3

where the table heading spells it ‘cemetary’. Grrrrrr.

This is the results table for U. S. Korean, World War I and World War II casualties.

46 Terry ReigelSeptember 27, 2008 at 6:32 am

Jude, Re: #44,

I generally start my search for someone with the global search, looking for all types of records.

When I don’t find the person in a specific census where I think they should appear. I click on “Census and Voter Lists” to focus the search on census records. If that fails to find them, I start refining the search terms to look for alternate names, names of others in the household, etc. But by now I’ve lost the ability to specify the census location to search, making the search much less useful.

I note that if I go from there to a specific census the location field is back. Why is it present on individual censuses and not for all censuses?

47 JadeSeptember 27, 2008 at 8:14 am

Terry, re: #46,

I get too many irrelevant results with global searches, and usually am looking to solve problems as to particular time/place, so hardly ever use the global searches for Census enumerations.

I think not being able to specify locations is silly. Ability to do that would not encourage me to use global searches, however, since the search engine tends to disregard them anyway (in ordering results) as well as the time-frame parameters.

48 Terry ReigelSeptember 27, 2008 at 8:28 am

Jade, re: #47,

I suppose my recent usage may not be all that typical in that I’ve largely been trying to confirm and flesh out information on cousins that I’ve received from others. So I have quite bit of information to start with. I always use the Advanced mode of Old Search as the starting point. That generally gives me good results, but often with a few expected records missing, which leads to the census issue I was talking about here.

I go directly to a census search only when I’m looking for a specific census for someone I’ve previously done a lot of work on.

49 Teddy BrockSeptember 29, 2008 at 9:55 am

Regarding the changes at Ancestry’s homepage and ‘new search capabilities’—I have read hundreds of comments with constructive criticism and many have even taken the time to cite extensive examples of what is NOT working and yet the non-functionality of the new search vs the more reliable, ease of use, of the old search and homepage, seems to fall on deaf ears. Ancestry made an arbitrary decision to revamp our entire search experience without ANY notice OR prior input from subscribers and we are now essentially told to like it or take our business elsewhere.

As a subscriber the past nine years, I have found Ancestry to be an invaluable tool for obtaining information to flesh out my knowledge as it pertains to this wonderful quest of my ancestors and genealogy research. And yet, I have come to a point where I have decided to take a stand based on my search experiences of the past 3 months and my general dissatisfation of same, and concluded I am not inclined to renew my subscription that expires tomorrow, Sept. 30th. I placed a call this morning to Ancestry and have officially cancelled membership. This was a very difficult decision and no doubt will in some ways impact me adversely, but sometimes one has to act on principles irrespective of the personal inconvenience. As for Ancestry, they will neither care nor notice the absence of this subscriber, but hopefully continued input from their conscientious and knowledgeable users who offer insight and suggestions will lead them (Ancestry) to re-examining their autonomous position and improvements will result.

Teddy

50 Carol A. H.September 29, 2008 at 10:38 pm

Anne, Are you back from getting your son in college? There are lots of posts here. I would like to see some responses from you. I read these things almost every day. It’s time we heard from you. You started off so well and actively participating. Don’t stop now. What is Ancestry doing to address our concerns?

51 Jerry BryanSeptember 30, 2008 at 11:49 am

Re: #50, I share your eagerness to hear from Anne again on plans and progress with respect to solving problems that have been identified with New Search.  Postings on the blog have slowed down considerably, I suspect in part because of Anne’s temporary absence and in part because of a desire to avoid repeating problems that have already identified.

In the meantime, I would like to take a slightly different tack.  I try to use New Search as much possible.  I figure I might as well learn to use it as effectively as possible because Old Search is going away.  Plus, using it on a regular basis helps in making informed suggestions for improvement.

But New Search still drives me crazy on a regular basis, and it still frequently drives me back into using Old Search.  It’s just not the sort of thing “you get used to”.  Office 2007 is radically different from its predecessors, and I got used to it in about an hour.  I’ve used New Search for dozens of hours, perhaps over a hundred hours.  I’m not naturally resistant to change, but I am definitely not “used to it” when it comes to New Search.  So I thought I would just sort of summarize the things that still drive me crazy.  I think all these items are already in the blogs, but this is my attempt to summarize them.

I revert to Old Search anytime I need to search on the basis of the date of a record rather than on the date of an event.  This is the standard, non-advanced mode of Old Search, and New Search has no equivalent.

I revert to Old Search anytime I need to search on the basis of Soundex.  Anne has stated that a fix for this is coming.

I revert to Old Search anytime I need to search for two people at the same time (usually spouses).  New Search’s handling of the second spouse is very poor.  My frustration happens most often for marriage indexes, but it also happens when searching censuses from 1880 to 1930.

I revert to Old Search anytime I need to search the U.S. Public Records index.  New Search’s ability to refine searches for this database is very poor.  I most typically need to add an address to the search and remove a first name.  Removing the first name is easy, but adding an address is an incomprehensible exercise with New Search and is trivial with Old Search.

I revert to Old Search anytime I need to search for a first and middle name, or for a first name and middle initial.  If I search for William A. Smith, New Search will find John A. Smith because of the A.

I revert to Old Search anytime I need to search the SSDI.  New Search does not allow me to specify dates accurately enough or flexibly enough – e.g., a complete month/day/year, or a month/day without the year, etc.

This is not quite a reversion to Old Search, but New Search’s refine search dialog is so awful that I frequently give up on it and just click on Home, starting over again rather than fighting with refining the search.  Just changing a name or something like that is not so bad.  But if you get into changing much more than the name, and especially if you get into the “More:” part of the dialog, a series of revisions and searches quickly gets so confusing that you can’t tell what’s going on.  It’s essential to be able to refine a search with a complete search template all on the screen at the same time.

There are lots of other things in New Search that still drive me crazy, as well.  So the above list is not an exhaustive list of all the problems in New Search.  But it’s pretty much the list that drives me over the edge to the point that I have to retreat back to Old Search.

52 JadeSeptember 30, 2008 at 2:20 pm

While trying to use New Fuzzy to locate records for a particular state via New Fuzzy’s version of an interface with Card Catalog (I really don’t recommend trying to do this, the filters don’t work and you get completely irrelevant results in huge numbers), it came to my attention that Ancestry has newly added the U. S. Compiled Revolutionary War Military Service Records, 1775-1783, from the National Archives.

I eagerly clicked on this to search. After several minutes’ wait for the search page to load, I found that the only search parameters available in New Fuzzy for this database are “Lived In” and “More” (keyword).

What? I could get a list of, say, 50,000 unidentified pages for Virginians? Unsortable by any handy tool such as *name* or *regiment*?

Is this another Famous Ancestry Trick, like the non-indexed Miscellaneous Revolutionar War papers images that were added at the time of the Revolutionary War Database Rollout (the National Archives has microfilmed indexes to these papers, but these were not added by Ancestry of course).

So what is the use of this?

53 Yolanda SimpsonOctober 1, 2008 at 7:29 pm

Hello. I am looking for my father by the name of Leslie George Simpson. He was last seen at a gas station in either Ozone Park or South Ozone Park,Queens,New York. I have not seen my father since 1994 and I desperately want to reunite with him. If you know any info about his whereabouts please contact me at bswtsuga@aim.com with a possible pic and location of him. His b-date is 1/15/47. Thank you and God bless you.

54 Nancy RogersOctober 3, 2008 at 12:23 pm

Believe it or not last night I got to be one of the lucky ones who was selected to give my feedback on ancestry.com. My responses mostly ranged in the negative side, and then when asked about the scrapbook, my face type of environment I responded with never will use.
I think we have been left again in the dark hole of no where. It has now been since September 17th that we have had any response from ancestry.com. I wonder what would happen to us if we set our credit cards to pay in this manner?

55 Anne MacDonaldOctober 4, 2008 at 8:09 am

I am searching for my grandmother “Beatrice Mae Noel,I obtain some information on her from this web,now i can’t get the information back.how i got the information before was through 1901 cenus in Montreal, Quebec. She is listed.My grandmother last redience was:5155 St.Catherine ST. East Montreal.Date of death:July 12,1965 at Grace Dart Extened Care.Montreal.Why is it i can’t come up with the information again.Please respond.Anne MacDonald. n44@telus.net

56 ReedOctober 7, 2008 at 9:45 pm

Hello?…

Anne? Anne Mitchell?…

Anybody home?…

******************************

Well fellow bloggers, I guess it’s over. It’s been 3 weeks since Anne’s last post. Apparently, she has been sucked into Ancestry’s infamous Customer Service Vortex Of Oblivion.

And with her absence—and no regular blog feedback from anyone at Ancestry—this blog devolves into another random message board. Better than nothing, perhaps, but as Anne said in her last post (#29, Sept. 17):

“There are message boards for ancestry.com proper, but I know I don’t monitor them on a regular basis […] I could spend all day chatting with people about this problem and that, and lots of new ideas, but then I’m spending all day chatting […]”

Exactly so. There is really no reason for any of us to frequent an unmoderated message board. We Ancestry users/customers already know what’s going wrong with New Search, databases, indexing, etc.

We need to know that:
• Ancestry understands the issues,
• Is working on solutions, and
• We get some reality-based estimates of when problems will be corrected.

Without a commitment from Ancestry to interactive, ongoing, customer communication and service, this blog feels pretty pointless, and I’m sure I’m not the only one who feels that way. (Although I continue to enjoy and profit from the interesting observations of many of you fellow bloggers! Thanks, y’all.)

Well, it was nice while it lasted…

—Reed

P.S. Now, Anne seemed quite likable and perhaps she still has the customers’ (our!) interests at heart. Perhaps she is, at this very moment implementing wonderful improvements to the site. But who knows? The net result is—once again—we users are left in the dark…

All of which, I think, reinforces the need for an ongoing Ancestry.com “User’s Advocate” (see my post #28, above, for details), a one-person “clearinghouse” who could field queries,suggestions and complaints from blog readers and find answers from the most appropriate, in-the-know Ancestry employees and pass them along.

It would be a win-win for company and customers. Ancestry gets an ongoing, passionate, experienced and informed “focus group” and we customers get some sense that our concerns are heard and will be addressed.

The ball—once again—is in your court, Ancestry.

57 Carol A. H.October 8, 2008 at 10:18 pm

Well Reed, I think you are right. This blog thread is just going to expire from lack of attention. Since we don’t know if there is a good reason for Anne’s absence, perhaps we could email her directly at Ancestry. She did respond to one of my emails (very early in this game) personally and said to email her.

Remember, she posted her addy at the begining of this thread.

“About Anne Mitchell
Anne Mitchell is a product manager for search at Ancestry.com. Feel free to contact her with your thoughts about search. She can’t help you find your long lost ancestors – she’s still searching for her own, but thoughts on how to improve search functionality are always appreciated. Her email is amitchell at tgn.com.”

Actual addy is:

amitchell@tgn.com

I have gotten some good help talking to live humans at ancestry recently by telephone. That is good because I did get the help but not so good because I have problems with the site.

I’m not giving up…yet, I’ve been busy doing real research using OLD search!

58 Andy HatchettOctober 8, 2008 at 11:40 pm

Well, I just wrote her so let’s see if it goes through or bounces.. that will tell us something anyway.

59 JadeOctober 9, 2008 at 11:55 am

Anne has had plenty of feedback from her questions and regarding things she did not specifically ask.

The search engine interfaces are not her only responsibilities.

If Anne wants further feedback I am sure she will post something. Some of us will check back occasionally.

60 Andy HatchettOctober 10, 2008 at 4:48 pm

Well.. it seems Anne considers this particular bolg closed…

She does have a new one on the October Blog.

61 KarenDecember 17, 2008 at 9:26 am

Hi! There are many, many problems with Ancestry search engines. My problems with the marriages are as follows:
1) I already have the marriage date, spouse, and county for the marriage entered.
2) I SIMPLY want to attach the marriage record.
3) The name is misspelled by one SIMPLE letter (Exa: the second letter in the last name OR the fifth letter in the last name,,,,which occurs on at least 30 percent of the records)
4) To accomplish this SIMPLE task of attaching the record, which I already know exists and have entered all the information on, I must go through the following TIME CONSUMING ordeal:
a) Do a search on my ancestor
b) Hit the button to switch back to the old search as the new search won’t even bring up the data bank and only lists some strange data bank for another state or country as a hit.
c) Hit the button to go back home and bring my ancestor’s page back up d) Hit my ancestor’s profile page and get out of the BETA (no option to just go to the profile page if you are not loving the BETA test thing)
e) Hit research historical records again.
f) Hit the marriage/death catagory button to narrow the search.
g) Hit the 50 results button because why would anyone in their right mind want to page through in 10 hit increments when they can look at 50 hits on a page?
h) Select the match category instead of alphabet.
i) Scroll down through data banks for other countries and states until I find the right data bank and hit on it.
j) My relative doesn’t come up, because his name is wrong on the second letter and I MUST have the first three letters of the last name right or I am stuck paging and paging.
k) Make a choice to either scroll to the bottom and try a new search under just the year and county, which will lose my direct connection to my ancestor OR scroll through page after page of records in 5 page jumps to get to the Ws, because my relatives last name starts with a W!
l) Finally, there is the record and I need to make a correction to the record for your record purposes. That’s two more hits. Three more hits to get back to the original record. And finally AGAIN, I get to click and hit the save record button. BUT WAIT, theres more!!!,,,Now, I wait for the page with the actual record to load so I can attach it. If I am lucky, my ancestor only has ONE marriage, because if they have more than one, which again, is already entered on my ancestor’s profile with names, dates, state, county, there is a good possibility your program will try to attach it to the wrong marriage and date. There is no way out of this unless I hit cancel, page back, remove the marriage I don’t want and go through the whole process again, JUST to attach the record.

Do you see where I am going with this? I know you all are working hard, so please don’t take this personal. But, think of me. I am traveling at 100mbs. Everytime I attach a record, your site gets progressively slower. Actually, once I have been on awhile, I can clean a room and do laundry while waiting for pages to load. The heaven it would be, to just have your program read my date, county, state on the marriage and bring up all the last names that start with a SIMPLE W if there is not a direct match!!!!!!

About the Ancestry.com blog

Here you will find informational, and sometimes fun, posts from the folks behind the scenes here at Ancestry.com. We hope you’ll notice just how passionate we are about family history and about the products we’re building to help connect families over distance and time.

Visit Ancestry.com
Notifications

Receive updates from the Ancestry.com blog Learn more