Death Takes a Holiday
September 23, 2016 2:27 PM   Subscribe

Is there any serious confirmation or statistical corroboration for the idea that they changed the dates on death certificates in some wars, on certain dates?

Snopes says this is a myth, but it also has a comment that someone in the discussion came up with 52 names listed as dying on Dec. 25th. This is 52 names out of 58,220. When I divided 58,220 by 365 (days in a year), I got the figure 159.51 as average deaths per date of the year.

A couple of other links I tripped across while trying to research:

Vietnam War Statistics
Vietnam War: Facts, Stats & Myths

The Snopes article also has comments indicating they changed the date of death for at least some soldiers who died on the final day of WW1, the day the treaty was signed. I tried to google that and found no corroboration.

I am interested in anything that casts light on whether it is true or false that dates were changed on some soldier's birth certificates in some wars, plus any data as to what would account for the "low" count of Vietnam death's on December 25th (such as routine cease fires for the holiday), as well as number crunching by people who know more stats than I do that casts light on how within normal and expected variation that number is, etc.

posted by Michele in California to Society & Culture (5 answers total)
some soldier's birth certificates

Obviously, that should be death certificates, not birth certificates.
posted by Michele in California at 2:31 PM on September 23, 2016

Not a war or records expert (and hardly an expert on statistics but of the three things its the one I know the most about) but:

Knowing that "on average" there were about 160 deaths per day doesn't actually tell you whether the 52 recorded on Christmas is aberrant or not. You make an unspoken assumption that the probability of soldiers dying is equal on any given day which is probably not a valid one - incidents like offensives, attacks, etc which surely contribute to soldier fatalities are not equally likely on any one day.

In order to know whether the 52 is even notably different from how many soldiers you would expect to die on any given day you would need to know the standard deviation of deaths by day in addition to the average. Forgive me if this is too basic of an explanation but SD measures the relative variability in the data - I would expect to find a fairly high standard deviation following my logic above that the number of soldiers who die on any one day is not justifiably equal (and that on some days lots of soldiers died and others barely any). Given that, it could easily be said that the 52 in the comment is perfectly normal and that there is little to no evidence in it that death dates were toyed with.
posted by Exceptional_Hubris at 2:39 PM on September 23, 2016 [4 favorites]

I don't know a ton about the battles of the Vietnam war, but there is no reason to think that deaths should or would be evenly distributed in any war - they would largely be concentrated on days of big battles or during big events. My quick wikipedia-ing suggests large casualties during the Tet Offensive which ran on and off from January to October, and the Easter Offensive which happened between March and October.

The person in that snopes discussion linked to a searchable page - neither December 24 nor December 26 have a disproportionally large number of casualties.
posted by brainmouse at 2:42 PM on September 23, 2016 [1 favorite]

Is there any chance that some dates were changed to match a US time clock? The 21st in EU might be the 20th in the US?
posted by Lil Bit of Pepper at 5:51 PM on September 23, 2016

There were often stand downs during Christmas as well so it would make sense that a lack of Offensives would lead to fewer casualties as well.
posted by bowmaniac at 8:01 PM on September 23, 2016

« Older I need a mini amp for electric guitar   |   When is it wise to crowdsource for info... Newer »
This thread is closed to new comments.