Average loss rate of the Roman army 1st BCE to 1st CE

tomar

Ad Honoris
Jan 2011
13,557
#1
Any idea on what was the average loss rate of the Romans per year during that peiod as expressed either in total men or in % of the total legion strength ? And how it evolved over time (got better or worse ?)
 

Ichon

Ad Honorem
Mar 2013
3,622
#2
The legions would be relatively steady numbers unless on campaign. As each group of men reached retirement there were new recruits incoming to replace both those and natural losses but outside of wars or active patrolling in border areas most of the losses would come from disease and accidents.
 
Sep 2013
619
Ontario, Canada
#3
Pitirim Sorokin (from Social and Cultural Dynamics, volume 3) estimated about 885,000 casualties through the nine centuries of history that the Roman legion was in operation, from about 400 BCE to 500 CE. Another author, VD Hanson (from Carnage and Culture) estimated 500,000 in five centuries after Hannibal.

So losses were pretty steady, from 100 BCE to 100 CE, probably roughly about 200,000 legionaries lost right on the battlefield to the enemies of Rome. This is likely a minimum as a similar number were killed by disease or through accidents over the same time period.
 

caldrail

Ad Honorem
Feb 2012
5,296
#4
Any idea on what was the average loss rate of the Romans per year during that peiod as expressed either in total men or in % of the total legion strength ? And how it evolved over time (got better or worse ?)
You won't see many estimates of this because those sort of statistics aren't available to us. Some people attempt to extrapolate from losses mentioned in the sources in specific instances compared to expected troop strengths (since we have a pretty good idea of how many legions were around at any time).

However, Roman defeats were pretty much slaughters, losing anything from a third of the force up to almost all of them, though in fairness this would include executions after the event, or if they were lucky, enslavement and removal from the records. The losses of a Roman victory are much harder to gauge. The Roman themselves don't dwell on the subject of losses generally unless there was an unusual result (such as the loss of a bare minimum by Crassus's troops defending the wall against Spartacus in the toe of Italy) thus it's hard to escape the conclusion the losses were normally relatively small if a victory was won.

It is true however that in siegecraft a general might well be tempted to launch an assault rather simply wait it out, both for strategic and political reasons, and for that reason were perfectly willing to contemplate a higher casualty rate if it guaranteed an early victory.

Personally I would treat any statistical analysis as a little suspect. Casualty rates varied considerably according to circumstance, and we do know that desertion was a problem that never went away (pun intended)
 

tomar

Ad Honoris
Jan 2011
13,557
#5
Pitirim Sorokin (from Social and Cultural Dynamics, volume 3) estimated about 885,000 casualties through the nine centuries of history that the Roman legion was in operation, from about 400 BCE to 500 CE. Another author, VD Hanson (from Carnage and Culture) estimated 500,000 in five centuries after Hannibal.

So losses were pretty steady, from 100 BCE to 100 CE, probably roughly about 200,000 legionaries lost right on the battlefield to the enemies of Rome. This is likely a minimum as a similar number were killed by disease or through accidents over the same time period.
This does seem rather low does it not.. Both estimates come to 100,000 per century or 1,000 per year on average.... Even if we triple it as you suggest for disease etc.... 3,000 / year would still seem pretty low... Also I would expect the numbers to grow as the roman empire extended.... Even in periods of extended piece there was the occasional raiding, "police duties", epidemics etc.....
 

tomar

Ad Honoris
Jan 2011
13,557
#6
Personally I would treat any statistical analysis as a little suspect. Casualty rates varied considerably according to circumstance, and we do know that desertion was a problem that never went away (pun intended)
Indeed, but I would expect them to even out over time (so for example a period of active campaigning/battling migh be followed by a period of extended peace)... That's why I am looking at a longer period....
 

caldrail

Ad Honorem
Feb 2012
5,296
#7
No, they wouldn't even out, because the rate of conflict and confrontations would also vary. I don't doubt you would be able to take an average, however misleading that would be, if only we had the information to extract one.