Barbarian numbers

[Fritigern]

I have played both Rome and BI extensively and hope to download the platinum mod as soon as I get back home. Reading through the various threads I can see that the folks have done a great job in making the original game much more realistic. One challenge remains though.

Both the originals and mods rely on the 'barbarians' (germans, gauls, britons, etc.) winning generally by weight of numbers. This is fine for gameplay mechanics. Unfortunately it does not match what was really happening some 2000 years ago. In the last fifty or so years historians have stopped taking the roman accounts of barbarian numbers in battles as anything other than gross exaggerations. The current historian perspective (based on archeology, detailed reading of the roman accounts, etc.) have shown that the barbarian hordes were usually significantly outnumbered by the roman armies they faced. Barbarian armies rarely got beyond single digit thousands. One example being Alaric's army that besieged Rome - this is now believed to have numbered, very approximately, 5-8 thousands (and only this many as a result of 'help' from other barbarian bands). A simple rule on barbarian headcount is: divide the quoted headcount by 10 and you have an over-estimate, divide it by 100 and you have an under-estimate - the best estimate being nearer the latter than the former.

Can this be built into mods without wrecking the game? Perhaps. One approach might be to reduce unit headcount and increase the charge bonus. The roman records tend to show the effect of the numbers game - barbarian armies having to rely on a quick shock effect (e.g. at Arausio) before the disparity in numbers began to show (e.g. at Argentorum).

I have little idea how this final major bit of realism could be introduced - but am sure that a way could be found!

[Switch]

Interesting; could you give some articles/books which argue for such low numbers? I was under the impression that while exaggeration of numbers in ancient sources (for whatever reason, be it deliberate attempt to make a threat appear larger or simple confusion and accident) is common for estimates of barbarian strength, Roman armies were still frequently outnumbered by barbarian opponents. Accuracy is, I suppose, impossible when it comes to these things, though how close to the mark a source is depends on the source themselves and the event in question (for example, Ammianus Marcellinus was present at the battle of Strasbourg and, to massively oversimplify matters, seems to be generally alright with details, especially compared to certain other sources). That said, I've seen modern historians offer reasonable theories on barbarian numerical strength based on, as you say, careful use of sources combined with archaeological evidence.

Coincidentally enough I was reading a little about Alaric's escapades in Italy yesterday in Peter Heather's Fall of the Roman Empire. He suggests a figure of 40,000 men under Alaric's command by 408: He had 20,000 on his initial arrival in Italy (10,000 from each of the Tervingi and Greuthungi tribes, united under Alaric. This figure is arrived at based on Ammianus Marcellinus' claim that Valens thought he was facing 10,000 men of the Tervingi tribe at Hadrianople (he wasn't), and that numbers of Tervingi and Greuthungi were similar). About another 10,000 barbarians, drafted into the Roman army from the remnants of Radagaisus' force, defeated in 406, deserted to Alaric after the massacre of barbarian families following Stilicho's death. When camped outside Rome in 408, thousands of slaves left the city to join Alaric, bringing his numbers to somewhere in the region of 40,000 men (the slaves being, probably, more unlucky barbarians previously under Radagaisus). I'm not sure how Heather arrived at a figure for the number of slaves joining from Rome, so I'm withholding judgement on that for now, but his other numbers seem justifiable.

Like you say, though, the sources aren't much help here. Zosimus is terribly wonky when it comes to figures, so he's only of limited help. For instance, he claimed 30,000 troops joined Alaric after the massacre of barbarian families, which is faintly silly. But then I think he (if not him, someone did) claimed that Radagaisus had 400,000 people (not just soldiers) with him when he invaded Italy in 405/6, with other ancient estimates amounting to 200,000-300,000 followers. According to Heather, based on the 12,000 defeated soldiers drafted into the army, the thousands sold into slavery, and the 15,000 Roman troops mobilised to fight him, it seems likely he had about 20,000 soldiers with him.

That is, of course, just what Peter Heather thinks. He's a fairly convincing example of reaching reasonable figures for barbarian troops to me, but I'm open to arguments that revise the figures up or down!

Edit: Oops, almost forgot. Book I was using was
P. Heather, Fall of the Roman Empire: A New History (Macmillan, 2005). In particular, p. 145, p.198, p. 224, p.445 and endnotes 17 and 59 for chapter five and endnote 2 for chapter 4.

Ammianus Marcellinus on the battle of Strasbourg: 16.12

[Fritigern]

I am currently away from home at the moment and cannot therefore give full references. However, the most direct book on this subject is: 'The Barbarians - Goths, Franks and Vandals' (Malcolm Todd) - 'Armies of several thousands would have joined battle only in the course of a mass migration......'. The archeological sources that evidently led to this to this thinking comes from village sizes, a typical village having 80-100 inhabitants, admittedly somewhat more if near the coast - This comes from the book 'Die Ersten Deutschen' (The First Germans - S. Fischer-Fabian). and also e.g. Wikipedia ('Germanic Peoples') - 'Germanic settlements were typically small, rarely containing much more than ten households'. Clearly the number of fighting men a village could raise was not headcount divided by two (lots and lots of children, women, aged, unfree, etc.). I can have a more thorough rummage when I get back home!

[pannonian]

I am currently away from home at the moment and cannot therefore give full references. However, the most direct book on this subject is: 'The Barbarians - Goths, Franks and Vandals' (Malcolm Todd) - 'Armies of several thousands would have joined battle only in the course of a mass migration......'. The archeological sources that evidently led to this to this thinking comes from village sizes, a typical village having 80-100 inhabitants, admittedly somewhat more if near the coast - This comes from the book 'Die Ersten Deutschen' (The First Germans - S. Fischer-Fabian). and also e.g. Wikipedia ('Germanic Peoples') - 'Germanic settlements were typically small, rarely containing much more than ten households'. Clearly the number of fighting men a village could raise was not headcount divided by two (lots and lots of children, women, aged, unfree, etc.). I can have a more thorough rummage when I get back home!

At the Org, there was a thread looking at the impact of WW1 on the combatant nations. Someone did an analysis of what the numbers meant, ending with the interesting find that a 5% figure which, while not looking that significant alone, when factored into recruitment choices and age groups, would mean that 1/3 of the 20-30 age group was killed.

Here it is. 3.6% of the population lost meaning around 1/3 of the 20-30 age group killed.

[Fritigern]

Still not home yet, but found this link which appears to be well researched:

http://books.google.co.uk/booksid=78oh3_OUoeoC&pg=PA23&lpg=PA23&dq=germanic+tribes+population&source=bl&ots=_lDepdVQqW&sig=lqItAB3WvFFJMDI1cRLAnBo2VqA&hl=en&ei=k4HlSbuLKtirtgfX9dGXAg&sa=X&oi=book_result&ct=result&resnum=1#PPA23,M1

The book is by Hans Delbrueck and the link takes you to the bit where he reviews the various estimates on population statistics generated by studies in this domain. I can't cut and paste bits (copyright restrictions, I guess)

'Today it is recognized that the figures given by the Romans for the peoples of Germany, which until recently were repeated without any questioning, are worthless.'

The various researchers quoted on this book hover around population densities of 4 to 5 per square kilometer. The author estimates that a tribe may have numbered up to 25 000 people - but he admits that this has been seen as too high a number by another researcher. But sees no real clash in this by stating: 'First of all, let it noted that is the maximum, the upper limit.'

Interestingly, and perhaps coming full circle, he says, understandably:

'The military accomplishments of the Germans against the Roman Empire and its battle hardened legions were so great that they appear unthinkable without a certain mass of men, and in comparison even 5000 warriors in each tribe seems such a small number that no one will be inclined to lower it even further'.

Jumping to another series of historical events, that of the migration of the Teutones and Cimbrii. These people migrated from a Jutland peninsular containing a population of some 80 000 to 100 000 people (can't find this particular reference, and will have to remember where I found it). Clearly not everybody left the area, the Vandals who simultaneously left the same region went elsewhere. These tribes were individually defeated by Gaius Marius at Aquae Sextae and Vercellae respectively. According to Wikipedia Marius had some 40 000 and 50 000 men at each of these battles respectively (and I hope that the Romans estimated their own numbers correctly!). Pulling this all together, the Romans fighting these battles with odds of 10:1 or better provides a consistent picture with that painted by Hans Delbrueck.

So, going back to my earlier point, it is a shame that Rome Total War mechanics is built upon hordes of barbarians fighting outnumbered Roman legionaries.

[Quinn Inuit]

Please do. I'm very curious about this.

[Pat89]

Now I am curious if this is for in the period that ExRM and Platinum play (ie a starting date around 280-270BC), or that this is only for the later period of Barbarian Invasion...

I ask because you refer to this book: "The Barbarians - Goths, Franks and Vandals". I would read from that title that it is about the Barbarians who did massive invasions in the Roman Empire (divided in East and West)... With huge migrations and such...

This kind of invasions can't be compared with the invasions/attacks the Romans had during 270BC. Also the barbarians of 300AD or whatever are not the same people as those in 270BC. Both periods are way too different to make a fair comparison and this way come to conclusions for ExRM/Platinum.

Imho...

[Fritigern]

I can't really comment on the Celts with any great pretence of expertise - my interest is in the early Germanic tribes (in part as these had a much bigger impact on history than the Celtic tribes).

With respect to numbers, the references I quote above relate to the period around 113BC to about 478 AD and later - this covers Rome Total War II and Barbarian Invasion - essentially from the point at which the Romans and early Germanic tribes first meet to the point where the Western Roman Empire comes to an end. They paint a consistent picture across the entire time spectrum - massive exaggeration of barbarian numbers and no real evidence that the Romans ever fought a battle against their Germanic foes without the benefit of significant odds in their favour.

The very laudable work done by the team to make the units, etc. more realistic has, in my mind, not addressed the one remaining key inaccuracy of the original game concept - that of barbarian numbers encountered. I believe that somehow it must be possible to address the numbers issue in a playable and balanced way.

You are right to say that 'the barbarians of 300AD or whatever are not the same people as those in 270BC' - there had been many changes. For the Germanic peoples this was a move towards better leadership (an evolutionary process?), better armament (greater access to and use of Roman resources), re-shuffling of tribes, etc. Many of these changes can be captured in the unit details. However at the end of the day we are still stuck with a battlefield picture in the game which grossly misrepresents the number of bodies on the ground.

[Pat89]

Still 113 BC is a long time to go from 270 BC... You can pretty much say that many people will never ever reach that point in any game... So if we want to balance the Germanic units in such a way that they are capable of fighting roman armies even when seriously outnumbered, then this would in no way represent the situation in 270BC, the period that has been chosen as starting date. It is possible to implement reforms, but still they cost unit slots. And units slots in one of the things that are limited by the RTW engine. So for that reason, it would be:
1. Have reforms for the Germans and this way reduce the number of unit types they have access to pre- and after the reforms
2. Have reforms for the Germans only and leave the other factions
3. Have no reforms and live with the fact that there might be an unhistorical situation if the player ever reaches the 113BC period.

The RTR team of RTR 6 has chosen for the third option. And in my opinion this was a very wise choice. It is very difficult, not to say impossible, to predict in what way the game will progress. So you have to leave it open ended... After all, we don't know if the Roman Empire in RTR will have expanded the same way as they did in history. If they don't expand in the same way, but instead go East recreating Alexanders empire for example, then the Germans will never get into contact with the Romans... Would that mean that they get the reforms anyway? The better units/armour they stole from the Romans?

All I actually want to make clear, is that the period you are referring to, is waaaayyy to far on the time-span of the RTR campaign and thus we can't balance things as it would have been historically during that period.

[Switch]

The Google Books link doesn't work for me.. What's the title of the book? Sounds interesting, as does The Barbarians - Goths, Romans and Vandals, though at the moment I'm having trouble finding it at my university libraries. You say the Google Books one is by Hans Delbrueck? Not that it would invalidate anything he says, but he is a historian/archaeologist of a much older generation, isn't he (late nineteenth/early twentieth century)? Have his ideas been developed/reconsidered since then?

It's definitely an interesting topic. However it would appear that many historians even today still believe that barbarian numbers could be tens of thousands in battles (I'll find some examples later), so if the matter is still up for debate then it's going to be impossible for the RTR team to reflect both possibilities.

Cheers for bringing this up by the way .

[Caligula Caesar]

I would just like to mention that the Romans first encountered the Germans in 113 BC (that would be the Cimbri and the Teutones), and since the Greeks never met the Germans, there is no written evidence of Germans even existing before then. I'm not saying they didn't. I'm just saying that PatricianS' argumentation isn't really valid, because we have no idea of how the Germans fought before then. However, one must assume that there were no great changes. I especially can't see them having cities and then suddenly losing them (and so their numbers) at exactly the time they met the Romans...

[HamilcarBarca]

I would just like to mention that the Romans first encountered the Germans in 113 BC (that would be the Cimbri and the Teutones), and since the Greeks never met the Germans, there is no written evidence of Germans even existing before then. I'm not saying they didn't. I'm just saying that PatricianS' argumentation isn't really valid, because we have no idea of how the Germans fought before then. However, one must assume that there were no great changes. I especially can't see them having cities and then suddenly losing them (and so their numbers) at exactly the time they met the Romans...

I hate to be a pedant but... note that the fasti triumphales record a Roman victory in 222BC by the consul Marcus Claudius Marcellus over Insubrian Gauls – and Germans. This reflects the fact that the Insubrians hired a large number of mercenaries from north of the alps in order to combat the Roman invasion of that year; these 'germans' were likely Transalpine Gauls, but nonetheless it is interesting to consider that (1) the Romans considered all inhabitants of the lands north of the alps as 'germans' or (2) that a large number of germanic warriors did indeed take service with the Insubrians.

H.

[Fritigern]

Apologies re the non-functioning link in the earlier posting. The Hans Delbrueck book is available on Amazon:

The Barbarian Invasions: v. 2: History of the Art of War (1 Jun 1990) by Hans Delbruck and W.J. Renfroe

I couldn't find the Malcolm Todd book there though (it must be out of print), making me a proud owner of one of a few copies.

Some thoughts on the earlier postings:

The distinction between Celts and Germans will have developed/sharpened in the centuries prior to the arrival of Julius Caesar. For example, along the rhine, where, in a number of cases one cannot clearly ascribe tribes as belonging to one family or another. The Romans, when fighting against Spartacus, did notice that the northern barbarians split into two groups - a gaulish group and a group consisting of the freed Ambrone, Cimbrii and Teutones. By the time of Julius Caeser's Gallic campaign the Roman world had recognised and defined the distinction. However, even at this time they were not seen as radically different to the Celts, being labelled as purer Celts, or - super-Celts ('bigger, blonder and fiercer'). So, this is consistent with the comment from HamilcarBarca

I can't really follow the comments from PatricianS re 'reforms'. From 270BC to well past 100AD the German fighting style, armament, their headcount, etc. remain largely unchanged. So no 'reform' is needed. My suggestion, unencumbered by any serious knowledge of game mechanics is to reduce the number of bodies in a unit, but compensate by other means. No changes in the armament, or any of the other work done to make the units more realistic.

Re comment from Switch that 'many historians even today still believe that barbarian numbers could be tens of thousands in battles' - who can blame them! There is only one documentary source - that of the Romans - your modern historian can only say, like Todd, 'several thousands' or similar - which is very vague. The introduction of archeological evidence from Delbrueck, some time ago, through to Todd led to the re-think. This was also helped by a re-reading of the Roman records - hundreds of thousands of people migrating along and then circling their wagons for the night as in an old Western! The Roman authors did not make many efforts to maintain internal consistency in their accounts.

[Pat89]

I can't really follow the comments from PatricianS re 'reforms'. From 270BC to well past 100AD the German fighting style, armament, their headcount, etc. remain largely unchanged. So no 'reform' is needed. My suggestion, unencumbered by any serious knowledge of game mechanics is to reduce the number of bodies in a unit, but compensate by other means. No changes in the armament, or any of the other work done to make the units more realistic.

Well... I'm no historian at all... But if there is no information about the Germans prior to 113 BC (except for that one battle with the Romans), then I don't see why it would make it would be realistic to give them the same units in 270 BC as they had in 113 BC... Many things could have changed in such a period of time...!

You say: "From 270BC the German fighting style remained largely unchanged". How do you know this if there is no information about the Germans prior to 113 BC? Of course... If you have any evidence of this, then it's ok, but if there is no evidence then this doesn't seem like a valid statement to me.

HB: Are there any numbers about the battle you mentioned? It would be very interesting to look into... This way we could decide who was the "stronger"... Germans/Romans?

I would happily accept it of course if there is some decent prove of it It just seems so very odd to me, if the Germans were really that outnumbered as has been mentioned... like:

the barbarian hordes were usually significantly outnumbered by the roman armies

And how about the victory/defeat ratio? If they were always outnumbered, but always lost the battles, then there's nothing special about this fact

Like I said before, I'm no historian... So if you come with good prove, then you've got me convinced

[mcantu]

i believe i have that Delbruck book. iirc, it was first written about 100 years ago, so not really 'new' research...

[Caligula Caesar]

But if there is no evidence before 113 BC, and you are not going to take what we know about for then, then we could just as well use our imagination...

[Pat89]

I don't know... It's not really my department... But I guess you could also base your ideas on combinations of facts... For example a combination of what is known about the Celts and about the Germans of 113 BC etc etc...?

But it doesn't look easy to me to make something historical if there is no info about what actually *is* historical. So all we could use is some logical thinking... And it doesn't seem logic to me that the Germans of 270 BC would beat a roman legion while seriously outnumbered, while the Gauls would not be able to do something like that...

This is an example were the Gauls of Cisalpine Gaul were defeated by an outnumbered roman army:

Following the battle between Marcellus and the king of the Gauls, the outnumbered Romans broke the siege of Clastidium, won the battle and proceeded to push the Gallic army all the way back to their primary headquarters of Medioladum.

Found on wikipedia, http://en.wikipedia.org/wiki/Marcus_...e_spolia_opima

Why would the Germans be so way superior to the Gauls?

I can't state it often enough though, that if you have any evidence against what I'm saying, that you would have me convinced

[Fritigern]

Okey Dokey! Here goes - an attempt to address the questions/concerns raised by PatricianS

First the 113 BC vs 270 BC question. I quite agree, we do not have detailed documentary evidence of what the ancient Germans were equipped with in 270 BC. We do have to use a bit of conjecture. A number of authors on this period do highlight the ancient Germans very conservative approach to warfare (as well as agriculture). If there was no significant change in weaponry and tactics across centuries of contact with the Romans then a bit of extrapolation back to 270 BC is not a huge leap. Reading across the texts there is also no change in grave goods across the period (and beyond). A bunch of people armed primarily with shields and spears remained a bunch of people armed primarily with shields and spears across the Roman period. Only the proportion of swords and body armour increasing in the very late Roman period, and then only thanks to Roman contact (as well as certain East Germanic tribes being able to 'delegate' the hard grind of farming, etc).

I guess that this would be 'small beer' compared to extrapolation and assumptions necessary elsewhere.

As for big battles the one at Arausio/Orange is reputedly the biggest military disaster the Romans faced (worse than Cannae). Though to be fair the huge loss of Roman life in this battle may have been exaggerated by the contempory authors. There were a string of Roman defeats against the same tribes leading to this. The disaster at Arausio/Orange led to the reforms introduced by Marius. Augustus, according to a near-contempory chronicler, suffered only two military defeats in his lifetime and both against Germanic tribes. I hope that this addresses the victories/defeats question.

From Julius Caesar's time it was a given that the Germanic tribes were tougher opponents than the Celts. Lots of examples, some of them comic! One example is Tacitus himself. He writes quite disparagingly about the Celts living in Britain (unlike his excessive praise of the Germans). Julius Caesar's Gallic campaign has the episode where a tribe in/around Belgium claims to be part-Germanic - as a threat to Julius Caesar - something that he disbelieves.

Re the Romans defeating the siege of Clastidium, yes, it is not at all surprising to find such an account. If we believed the Roman accounts we would find that they only fought against odds and therefore won herioc victories by the bucketful.

We will never know how many people there were on each side in a battle. However, treating the Roman accounts as being accurate is not the right approach either. We would end up believing that Julius Caesar defeated 500000 Belgae in a single battle during his Gallic campaign and that Boudicca turned up with 230000 men at her last battle. That is why the approach taken by Todd and Delbrueck and their ilk is more sensible - work out what the population from archeological was and use this to make a best estimate.

The one defining attribute of the Roman empire was the ability to raise large armies (and clearly having the social and political systems to support this).

I am no professional historian either (and therefore no mental bibliography ability to recall what came from which book). If you are interested in tracking down some of the above then I can rummage and rummage.

PS. Wikipedia is not perfect in these areas - some of the discrepencies when one compares accounts of the same event in different languages are quite amazing.

[Caligula Caesar]

I think that what he is saying also applies to the Gauls, if it is true.

[Pat89]

I know about wikipedia... I'm not a real fan of it myself, but usually it is good for a quick search of things...