BEF 1914-1918: Intensity of Warfare - Concentration of Fatalities.

[Guest]

While doing some research on the 12 battalions* of the Guards Division, I noticed that their casualties - and in particular the fatalities - were highly concentrated into a few dates. While not unexpected, the degree to which the fatalities were concentrated into a small number of days did surprise me. I thought it would be interesting to explore the following by battalion

Highest day's fatalities

Highest day's fatalities as a % of total fatalities

Number of separate day's that accounted for 50% of total fatalities.

The results were quite surprising. At the extreme the 4th Bn Grenadier Guards suffered 15.7% of all its fatalities on a single day. 50% of its fatalities during the whole war occurred on just 7 separate days. This battalion was on the western front for 1,180 days up to Armistice day, meaning half its fatalities occurred during just 0.6% of its time in Theatre. That surprised me. A lot.

While this battalion had the most extreme data within the small sample, it was not a particular out-lier. The number of days accounting for 50% of all fatalities averaged at just 18 days across the 12 Battalions. The bar chart of fatalities for the battalions forming the Guards Div is also useful as a visual illustration.

I have little doubt that some battalions that suffered severely on 1st july 1916 will show more extreme characteristics, however I thought this sample was interesting given they largely served from the beginning and saw continued, intense fighting throughout. In addition, from Aug 1915 the formation was largely unchanged to the end of the War, I will cast the net further.

MG

* The 4tn Bn Coldtream Guards was excluded. As a Pioneer battalion it was used in a significantly different way to the 'front line' battalions in the three Guards Brigades.

Note the total for the 12 battalions will differ slightly from the Guards Div as roughly 5% of fatalities are annotated as being in Reserve Battalions and do not have designated regular battalions they they died fighting in.

[Guest]

4th Bn Worcestershire Regt (annihilated at Gallipoli)

19.6% of all its fatalities during the whole war occurred on a single day - 6th Aug 1915 at Gallipoli during an ill fated diversionary attack. The battalion subsequently served on the Western Front. 50% of all of this battalion's fatalities during the Great War occurred over just 19 separate dates.

10th Bn West Yorks (annihilated on the 1st day if the Battles of the Somme)

27.8% of all its fatalities during the whole war occurred on a single day - 1st July 1916. Some 50% of all of this battalion's fatalities during the Great War occurred over just 14 separate dates although this is heavily skewed towards the 1st Jul 1916. Stripping out this data and focusing on the other dates (72.2% of total casualties) 50% of the fatalities (ex 1st Jul 1916) occurred over 31 separate days, which still suggests, despite the heavy distortion of 1st Jul 1916, there was still an underlying skew in the data for the rest of the war.

[David Ridgus]

Martin

I would not normally pitch in on something like this as your grip on statistics and minutiae is way beyond mine. However glancing at your graph above I'm not sure that I can see the skewing you allude to. Taking admittedly only a rudimentary look,the spikes in fatalities seem to match the major blood letting of the war: Mons, the retreat, the race to the sea, First Ypres, Loos, the Somme, Third Ypres, the Spring Offensive and the Hundred Days. Wasn't the nature of much of the war on the Western Front, long periods with not a lot happening and with only a steady drip of casualties, then punctuated by horrendous battles with high attrition rates.

As I say, only a quick look late on a Sunday night, so fully expecting to be put in my place!

David

[Guest]

Martin

....I'm not sure that I can see the skewing you allude to. Taking admittedly only a rudimentary look,the spikes in fatalities seem to match the major blood letting of the war: Mons, the retreat, the race to the sea, First Ypres, Loos, the Somme, Third Ypres, the Spring Offensive and the Hundred Days. Wasn't the nature of much of the war on the Western Front, long periods with not a lot happening and with only a steady drip of casualties, then punctuated by horrendous battles with high attrition rates.

David

David

I guess we are all generally aware that trench warfare involved a lot of down time between intense periods of battle, so there is a systemic 'skew' in all the data. My point, badly made, was that the magnitude of this skew really surprised me. Some battalions saw 50% of all their fatal casualties occur in just a few days' fighting, despite nominally being on the Western Front for over four years. Against this was a backdrop of low level attrition with very small numbers of daily casualties over hundreds of days. If the data is re-arranged and ordered by fatalities we see a few days of massive spikes followed by a very sharp decline and a long, long trail of hundreds of days of single digit fatalities.

It might indicate that the intensity of warfare varied considerably and the difference in being in a deliberate assault v a supporting role was often very extreme.The data has an almost binary nature in the sense that there were either a lot of fatalities or very, very few, and not much in between.

Interestingly the data of all the Guards Regiments and battalions differ despite being largely in the same formations throughout. The Scots Guards is an interesting example: Of the regiment's 10 worst days in battle, four occurred in 1914, and three in 1915. The regiment lost more men in 1914 than in 1916, the peak year of losses fro most battalions in the Guards (and indeed for most battalions that served throughout the whole war on the Western Front). It lost more men in five months of 1914 than the whole of 1917 and lost more men in the five months of 1914 than the eleven months of 1918. That surprised me.

The big swing factor was First Ypres. again it seems to indicate that the intense period of 1914 and particularly First Ypres was extremely fatal and when measured as a per cent of numbers involved, was arguably the deadliest period of the war for some units.

I have the same data for all 104 battalions of the regular BEF - the cohorts that saw the longest periods on the Western Front. It will be interesting to see the distribution of fatalities by Battalion through the war. I strongly suspect Gheluvelt will be a surprisingly large out-lier. We shall see. MG

[David Ridgus]

It might indicate that the intensity of warfare varied considerably and the difference in being in a deliberate assault v a supporting role was often very extreme.The data has an almost binary nature in the sense that there were either a lot of fatalities or very, very few, and not much in between.

Martin

I take your point here. I saw the spikes but didn't really grasp the almost flatlines in between. All very interesting. I look forward to your further analysis

David