Napoleon As a Military Commander

pugsville

Ad Honorem
Oct 2010
9,658
1. I think you are missing my point here. I was talking about death toll due to logistics, not casualties in general. My point is that Russia suffered as well as the French in that campaign (if not more due to civilian losses).
2. Wounded generally did not returned to the ranks : they usually died after the battle. And remember that "only" 200,000 russians could participate in the following campaigns out of 623,000 according to Beskrovny. And this counting reinforcements…
3. After it’s a battle of sources. Many sources state that only 450,000 men crossed the Niemen (Lentz, Tarle, Markham) which is clearly incompatible with Zamoyski. You are sure about Lieven ?
4. For example, according to Sokolov, the russian army lost 300,000 men, including 175,000 in combat.
5. About the French losses, they are usually overestimated. According to a statistical analysis (the only one I’m aware of) on the French and French allies matricules (they analysed 1 on 500 which is quite robust given the hundreds of thousands), sorry in French only — Pertes de l'armée de terre sous le premier Empire, d'après les registres matricules. - Persée p.45, here are the results for the russian campaign (in thousands) :

KIA or death after WIA — French5
KIA or death after WIA — French allies2
Missing and prisoners that did not come back — French147
Missing and prisoners that did not come back — French allies61
Death in hospitals — French19
Death in hospitals — French allies9
Scratch out in hospitals — French
Scratch out in hospitals — French allies
Total — French171
Total — French allies72
Grand total243
They are stating that figures can be perfected and a good estimation is 270,000 men that did not came back of Russia which is far from 550,000.


The main point is that around 120,000 men "only" participated to the retreat. A big part of the men deserted before and did not say that they returned until a census for a pension in 1853 show it.
On Russian losses, the vast majoirty of caualties on both sides were non battle casualties. Ruusian soldiers who dropped oout of effetcive strength were a lt closer to some sort suitabel place to recover than the french. The death rate for Russian soldiers droppingout of effetcive strngth is going to be a lot less than the French,

1 in 500. How did they make it representative sample There's all sorts of way their sample could be not representative.

Whats the exact methodology.
 
Mar 2014
53
Paris (France)
On Russian losses, the vast majoirty of caualties on both sides were non battle casualties. Ruusian soldiers who dropped oout of effetcive strength were a lt closer to some sort suitabel place to recover than the french. The death rate for Russian soldiers droppingout of effetcive strngth is going to be a lot less than the French,

1 in 500. How did they make it representative sample There's all sorts of way their sample could be not representative.

Whats the exact methodology.
They chose randomly 1/500 on around 2.5 millions of files which make up to 5000 files analysed. It’s more robust than any poll we have today for elections. However, This method tends to be more incertain when the values are extrems. For examples, the confidence interval at 95% for the death (indicated here at 7000) in the russian campaign is from 2000 (-70% not possible) up to 12 000 (+70%). But for the total losses, having 243,000 implies a range from 223,000 (-8%) to 263,000(+8%). Here the author decided to be conservative and use 270,000 (which is thi upper bound for a confidence interval at 99%). It’s only an estimation but the order of magnitude should be correct. And remember that some death are counted as missing in the files.

The methodology is explained from the page 30 to the page 40. The main problem is they didn’t manage to clearly separate missings prisoners and deserters and they put them in the same entry.
They tried to evaluate the number of people that returned, while their files stated their fate as incertain or dead, by two methods (civil registry and pension) and they estimated it at 25% to 30% (300,000 to 350,000 out of 1145000).
 
Mar 2014
53
Paris (France)
Ruusian soldiers who dropped oout of effetcive strength were a lt closer to some sort suitabel place to recover than the french. The death rate for Russian soldiers droppingout of effetcive strngth is going to be a lot less than the French,
I highly doubt it as be the French or the Russian, the land were devastated and there were no efficient medical care at that time (even if it was the early start of hygienism). Once a wounded (French or Russian) managed to go into a real hospital, he had a high chance of surviving. But a lot couldn’t.
 

pugsville

Ad Honorem
Oct 2010
9,658
I highly doubt it as be the French or the Russian, the land were devastated and there were no efficient medical care at that time (even if it was the early start of hygienism). Once a wounded (French or Russian) managed to go into a real hospital, he had a high chance of surviving. But a lot couldn’t.
You assumption that most of the effectives dropping out were wounded is just wrong.

Again most soldiers who dropped out of efeftcive strength during the cmapaign, Russian or French were not wounded, just unable or unwilling to keep up with the March. And mostly they would recover if given food and rest. The Russian soldiers were a lot closer to freindly food and rest and lot less likely to be killed by peasants and cossacks. In a normal march to the front in Germany in 1813 the Russian would loose about a thrid of effetcives durng the march but would expect to get almost all them back.

Hospital recover rates were generally appalling, better the 7 years war when they were flat out death sentence but still not good.
 

pugsville

Ad Honorem
Oct 2010
9,658
They chose randomly 1/500 on around 2.5 millions of files which make up to 5000 files analysed. It’s more robust than any poll we have today for elections. However, This method tends to be more incertain when the values are extrems. For examples, the confidence interval at 95% for the death (indicated here at 7000) in the russian campaign is from 2000 (-70% not possible) up to 12 000 (+70%). But for the total losses, having 243,000 implies a range from 223,000 (-8%) to 263,000(+8%). Here the author decided to be conservative and use 270,000 (which is thi upper bound for a confidence interval at 99%). It’s only an estimation but the order of magnitude should be correct. And remember that some death are counted as missing in the files.

The methodology is explained from the page 30 to the page 40. The main problem is they didn’t manage to clearly separate missings prisoners and deserters and they put them in the same entry.
They tried to evaluate the number of people that returned, while their files stated their fate as incertain or dead, by two methods (civil registry and pension) and they estimated it at 25% to 30% (300,000 to 350,000 out of 1145000).

This has been aorund since 1972. If historians are not taking it seriously they have their reasons. I do not by there is come consipracy theory and Historians will not reevaluate as a group given new data. If this stuyf holds up there would have been some historians capaitalises on this and writing a new revised history.There isn't so I suspect there is a reason for that.

I would need to read all tehmethodoogoly and understand what they are actually doing. I do not read French. Historical records can have all sorts of baias. 2.5 million files of what, they may be a some sort selection by the file spreesisting before the methodogoly.
 

nuclearguy165

Ad Honorem
Nov 2011
4,822
Ohio, USA
1. I think you are missing my point here. I was talking about death toll due to logistics, not casualties in general. My point is that Russia suffered as well as the French in that campaign (if not more due to civilian losses).
2. Wounded generally did not returned to the ranks : they usually died after the battle. And remember that "only" 200,000 russians could participate in the following campaigns out of 623,000 according to Beskrovny. And this counting reinforcements…
3. After it’s a battle of sources. Many sources state that only 450,000 men crossed the Niemen (Lentz, Tarle, Markham) which is clearly incompatible with Zamoyski. You are sure about Lieven ?
4. For example, according to Sokolov, the russian army lost 300,000 men, including 175,000 in combat.
5. About the French losses, they are usually overestimated. According to a statistical analysis (the only one I’m aware of) on the French and French allies matricules (they analysed 1 on 500 which is quite robust given the hundreds of thousands), sorry in French only — Pertes de l'armée de terre sous le premier Empire, d'après les registres matricules. - Persée p.45, here are the results for the russian campaign (in thousands) :

KIA or death after WIA — French5
KIA or death after WIA — French allies2
Missing and prisoners that did not come back — French147
Missing and prisoners that did not come back — French allies61
Death in hospitals — French19
Death in hospitals — French allies9
Scratch out in hospitals — French
Scratch out in hospitals — French allies
Total — French171
Total — French allies72
Grand total243
They are stating that figures can be perfected and a good estimation is 270,000 men that did not came back of Russia which is far from 550,000.



The main point is that around 120,000 men "only" participated to the retreat. A big part of the men deserted before and did not say that they returned until a census for a pension in 1853 show it.
I can completely believe anything in the 250-300,000 range for Russian losses. That figure of 550,000 for the French-Imperial losses is the figure that I come across most often from various sources I have read on the campaign, which includes Chandler, Zamoyski, Lieven, Nafziger, etc. Even Andrew Roberts states that Napoleon lost about half a million in Russia. However, since this includes 200,000 captured, it explains why those are subtracted from the total number who did not come back, which would probably be 300,000, 30,000 more than your figure if you include likely attrition from disease and wounds among those who were prisoners in Russia until hostilities ended in 1814. These numbers I'm giving are from cross-examinations of different sources I have read, which are likely to be different from French sources.
 

pugsville

Ad Honorem
Oct 2010
9,658
Lieven does not give a comprehensive breakdown on Russian casualties duringt 1812, ,

Page 283
" Of the 97,000 men whom Kutuzov commanded at Trautaino at the Start of the campaign, 48,000 - in other words almost half -- were in hospital. Only 42,000 were in the still in the Ranks"

The other covers the French retreat, and does not cover the other armies involved Wittgenstein and Chichagov's forces, though by all accounts they suffered less than the main Army pursing Napoleon. This is far from a comprehensive analysis. But does suggest most Russian soldiers dropping out of the army effectives made it to Hospital and some hope of recovery unlike the French.

Page284
"Never the less, wrote Krankin, the majority of men in hospital did recover and rejoin the army"

.
 
Mar 2014
53
Paris (France)
Sorry I don’t have much time to answer.

You assumption that most of the effectives dropping out were wounded is just wrong.

Again most soldiers who dropped out of efeftcive strength during the cmapaign, Russian or French were not wounded, just unable or unwilling to keep up with the March. And mostly they would recover if given food and rest. The Russian soldiers were a lot closer to freindly food and rest and lot less likely to be killed by peasants and cossacks. In a normal march to the front in Germany in 1813 the Russian would loose about a thrid of effetcives durng the march but would expect to get almost all them back.

Hospital recover rates were generally appalling, better the 7 years war when they were flat out death sentence but still not good.
My bad you’re right, they were not all wounded. But what I wanted to say is that there were not a lot of facilities during the winter retreat to take care of all the men, french or russian. The Russian losses (and obviously for the french) speak for themselves.
You need to separate men that came back and reinforcement.

This has been aorund since 1972. If historians are not taking it seriously they have their reasons. I do not by there is come consipracy theory and Historians will not reevaluate as a group given new data. If this stuyf holds up there would have been some historians capaitalises on this and writing a new revised history.There isn't so I suspect there is a reason for that.

I would need to read all tehmethodoogoly and understand what they are actually doing. I do not read French. Historical records can have all sorts of baias. 2.5 million files of what, they may be a some sort selection by the file spreesisting before the methodogoly.
I don’t know what make you think it has not been taken seriously. This study is cited in many other works, mainly french but it also used for example by Sokolov.
The 2.5 millions of files in questions are the "matricules" registry, a military registry that men had to fill and keep safely. Not having it was considered as being a deserter.
The main point of the study was to know the french losses for the first empire and it came to a conservative figure of 860,000 which is approximatively the same number as other study had (they discuss it).
About the Russian campaign, it’s not as earth shattering as it seems. When you know that only 100,000 men participated in the retreat, one cannot count 300,000 losses just for it.

I can completely believe anything in the 250-300,000 range for Russian losses. That figure of 550,000 for the French-Imperial losses is the figure that I come across most often from various sources I have read on the campaign, which includes Chandler, Zamoyski, Lieven, Nafziger, etc. Even Andrew Roberts states that Napoleon lost about half a million in Russia. However, since this includes 200,000 captured, it explains why those are subtracted from the total number who did not come back, which would probably be 300,000, 30,000 more than your figure if you include likely attrition from disease and wounds among those who were prisoners in Russia until hostilities ended in 1814. These numbers I'm giving are from cross-examinations of different sources I have read, which are likely to be different from French sources.
The 270,000 include the prisonners. After it’s a battle of sources, but the french demography of that time fit more with a 300,000 losses than a half a million. Now, my hypothesis is that they count "available" people in the army. While the 300,000 count the losses in the French demography. Still half a million is too much as a lot of historians (that I gave) considers that less than half a million crossed the Niemen.
For example, Minard draw "only" 422,000 at the start : French invasion of Russia - Wikipedia
 

pugsville

Ad Honorem
Oct 2010
9,658
My bad you’re right, they were not all wounded. But what I wanted to say is that there were not a lot of facilities during the winter retreat to take care of all the men, french or russian. The Russian losses (and obviously for the french) speak for themselves.
You need to separate men that came back and reinforcement.
I posted a quote from Levin of a Russian General saying almost all the Russian stragglers did reach hosiptal and most returned to service with the Army.

I don’t know what make you think it has not been taken seriously. This study is cited in many other works, mainly french but it also used for example by Sokolov.
The 2.5 millions of files in questions are the "matricules" registry, a military registry that men had to fill and keep safely. Not having it was considered as being a deserter.
The main point of the study was to know the french losses for the first empire and it came to a conservative figure of 860,000 which is approximatively the same number as other study had (they discuss it).
About the Russian campaign, it’s not as earth shattering as it seems. When you know that only 100,000 men participated in the retreat, one cannot count 300,000 losses just for it.
But French records are not going to be a good guide. About half the army were not french. And do we have all French records, if some did not sruive are the ones the survived a representative sample. The Devil is in the detail.

And While only 100,000 started the retreat signicnat numbers joined the retreat, and the French had large numbers of stragglers in there rear areas that were never evacuated. Not sayng 300,000 perished in the retreat, but that more than started the retreat from Moscow were certianly available to become casulaties.

The 270,000 include the prisonners. After it’s a battle of sources, but the french demography of that time fit more with a 300,000 losses than a half a million. Now, my hypothesis is that they count "available" people in the army. While the 300,000 count the losses in the French demography. Still half a million is too much as a lot of historians (that I gave) considers that less than half a million crossed the Niemen.
For example, Minard draw "only" 422,000 at the start : French invasion of Russia - Wikipedia
Significant reinforcements arrived after 422,000 started. 140-180,000 arrived at various stages.

Some 80,000 left Russian (but about half Austrian and Prussian, (which while survivours are going to fight against Napoleon in 1813) and mostly from ther northern and southern flanks) Generally the figures given for prisoners is 110,000 , that's around 200,000 survivors, something like 400,000 dead. (roughly 500,000 causalities)

The dicussion is interesting Numbers are important populalry quoted ones can be wrong. I have not got my hands on Nafziger, maybe he's got a good numbers discussion.
 
Last edited:
Mar 2014
53
Paris (France)
I posted a quote from Levin of a Russian General saying almost all the Russian stragglers did reach hosiptal and most returned to service with the Army.
If that is the case, how to explain :
1 - the important russian losses
2 - the fact that the coalition (including the prussian) numbered "only" 200,000 men at the start of the german campaign in early 1813 ?

But French records are not going to be a good guide. About half the army were not french. And do we have all French records, if some did not sruive are the ones the survived a representative sample. The Devil is in the detail.

And While only 100,000 started the retreat signicnat numbers joined the retreat, and the French had large numbers of stragglers in there rear areas that were never evacuated. Not sayng 300,000 perished in the retreat, but that more than started the retreat from Moscow were certianly available to become casulaties.
This are actually good question. The files contains every man that was under french flag, be it french or not. The exception are foreign armies like the prussian or the austrian for example.
Please note that Sokolov recently try to improve the work of Houdaille (with 8432 files analysed) and come up with very similar results : (in russian) https://imo.sgu.ru/sites/imo.sgu.ru/files/2019/09/o._v._sokolov_337-343.pdf and a google translate : Google Translate

Significant reinforcements arrived after 422,000 started. 140-180,000 arrived at various stages.
The only reinforcement I’m aware of are the garrison in Vilnius and the northern army. I never read of reinforcement from France or Germany.

Some 80,000 left Russian (but about half Austrian and Prussian, (which while survivours are going to fight against Napoleon in 1813) and mostly from ther northern and southern flanks) Generally the figures given for prisoners is 110,000 , that's around 200,000 survivors, something like 400,000 dead. (roughly 500,000 causalities)

The dicussion is interesting Numbers are important populalry quoted ones can be wrong. I have not got my hands on Nafziger, maybe he's got a good numbers discussion.
Yes that’s the numbers that are "traditionnal", but the primary sources (not memoirs but civil and military registries) and the demographic analysis tends to prove that it was not that huge. But again maybe the discrepancies are due to the difference between people not available in the army and people definitively irretrievable.