Review
(There have been many changes to this site since ther review was writtne)

This site provides access to a number of indexes and data bases.

Parish Registers

FindMyPast have digitised some of the Hertfordshire Parish Registers and Bishops Transcripts held at HALS up to 1910 for baptisms, 1928 for marriages and 1990 for burials.

General Registrar Office indexes of Births, Marriages and Deaths.

The indexes are large books, the earlier ones manuscript, later printed. For the normal records there are separate sets for births, marriages and deaths, with entries in alphabetical order by quarter.  The pages have been photographed and can be displayed. A search for a birth covering 10 years would involve downloading and visually searching 40 pages.

The same facilities are available on Ancestry and there is little significant difference in the way the pages are selected. I examined the quality of the findmypast image and found it was a little better than the Ancestry image for a number of pages, but for one item I looked up where there was a difficulty with the original, the Ancestry image was the most easily readable. However the presentation of the pages, particularly with ancestry's magnifier, meant that ancestry was preferable for the pages I examined.

Census Returns

A recent advertising email claimed "full census data - best transcriptions online'" On visiting the site you find that it offers the 1841, 1861, 1871 and 1891 censuses. A FAQ states that the 1861 census  is complete. There are no equivalent information on the other censuses - and testing quickly revealed that the 1871 census was far from complete - and that there were no entries for Hertfordshire. In view of the wording of the advert, and the complete absence of any warning message that the 1871 census is incomplete, I have carried out the tests on 1871 as if it was complete and recorded negative results.

I repeated the tests An Assessment of Ancestry for Census Searching. The interface of findmypast was far more awkward to use and with no ranking system was far less selective. The results of the tests were:

If one ignores cases where the census had not been indexed findmypast was more successful in correctly transcribing names, the only simple transcription error listed Walter Rendell as Rondell in 1891. However something went drastically wrong with the entry for Francis Reynolds in 1861 - the original is very clear and the rest of the line (and the household) is correctly recorded but the name is recorded as Eagleton. The transcriber's eye must have strayed as there is also a Frances Eagleton on the same page. There is no way such a serious error could have got past any verification process.

I decided to carry out another test. About 20 years ago I published a one name study of the Phipson family. It is a rare surname with a reasonable geographical spread, and my searches in the GRO indexes, unindexed census books, etc., meant I was certain that there would be less than 100 countrywide.  A search for the surname in the 1861 census produced 74 hits using Ancestry and 84 hits using findmypast. I started to compare the two lists:

At this point I decided to give up a full check - as it was clear that both lists contained most of the known Phipsons and many of the later discrepancies were "ditto" family members repeating the faults listed above. However two additional "bad" errors by findmypast were immediately apparent. Irene Jane Phipson had been indexed as Jane Jane Phipson and Ellen Emma Phipson had been indexed as Ellen George Phipson.

It is hard to be polite about the performance of findmypast. In about 100 tests I have found a "Reynolds" indexed as "Eagleton", an "Elizabeth A Phipson" indexed as "Sarah Heney", an "Irene" recorded as "Jane" and an "Emma" recorded as "George". In each case the original is quite easy to read and such complete fiction errors indicate an appallingly slipshod approach by the transcriber and a complete lack of verification or similar quality control measures. The combination of advertising claims coupled with the undeclared gaps in the 1871 census comes very close to mis-selling. By comparison with these glaring problems the serious weakness in recognising leading capital letter errors in the Phipson search - a well know difficulty in transcribing such texts - seems a comparatively minor fault.

If you have used Ancestry (with all its faults) and had negative results there is a small but positive chance that a search of findmypast will produce something. I can think of no circumstances when name searching using findmypast's censuses would be my first choice. However the ability to look for occupations, and extended address searching may sometimes be useful.

Other Facilities Offered

There are a small number of other, mainly much smaller lists, most of which have limited relevance to Hertfordshire. Visit the site to see what is currently available as the number of items on offer is expanding. However you should be aware that if you search for a surname and the response is a number of pages you will find each page contains an alphabetical list to visually search which may not contain the name you seek.

See also Problems with finding census returns

If you find a transcription error in the census returns they are good at correcting it - so if you find any errors report them.

Page updated March 2007
December 2009 - Site URL changed from findmypast.com to findmypast.co.uk 2013

March  - note on transcription errors