Collections: Teaching Paradox, Imperator, Part I: Divisa in Partes Tres (2024)

This is the first part of a three-part (if I can keep it) series, examining the historical assumptions of Imperator: Rome, a historical grand strategy game by Paradox Interactive, set during the rise and collapse of the Roman Republic from 304-27 BC and covering the broader Mediterranean world and South Asia. This is also the continuation of a larger series on Paradox’s historical grand strategy games, where we have already discussed Europa Universalis IV, Crusader Kings III and Victoria II.

I should note at the outset that this is not a review: we are examining Imperator‘s historical assumptions, not its quality or enjoyment as a game. That said, let me say that I think Imperator, especially in its current form, is quite a bit stronger of a game than it is generally given credit for. In particular, the current version (2.0.4) features a lot of changes from the 1.0 release version which I think make the game a lot more fun to play and more historically interesting – and perhaps most importantly gave the game a lot more of its own personality that makes it feel different and distinct from the rest of the Paradox catalog. This look will focus on that current version. So while this isn’t a review, if you want my opinion: Imperator: Rome is fun and worth your time. It is a good strategy game.

More importantly, this post is part of my ongoing campaign – which, I will note, Paradox is aware of – to bully Paradox into revisiting this period, either in a sequel or a new IP in the same setting. Now, you may say, ‘Bret, that’s crazy talk, you’re not going to bully Paradox into announcing Imperator II!’ But let me remind you, that I announced my intention to bully Paradox into green-lighting Victoria III in April of 2021 and then Paradox went and announced Victoria III in May of that same year. I have a track record on this.

Somewhat more seriously, I should note, in the interest of full disclosure, that I have had a chance to meet some of the developers who worked on this game when I spoke at Paradox’s convention (PDXCON) back in 2022, including a chance to actually talk over Imperator in particular with the game’s Creative Direct, Johan Andersson. I don’t have any ‘inside information,’ so to speak, but the connection seems worth noting.

But first, if you like what you are reading here, please share it around, as I rely on word of mouth for all of my new leaders. If you really like it, you can support me, my academic research (on the Roman Republic!) and this project on Patreon! I cannot promise I will use your donations to buy replica Roman swords, but I also won’t promise not to waste use them to buy swords. If you want updates whenever a new post appears, you can click below for email updates or follow me on twitter (@BretDevereaux) for updates as to new posts as well as my occasional ancient history, foreign policy or military history musings, assuming there is still a Twitter by the time this post goes live. I am also on Bluesky (@bretdevereaux.bsky.social) and (less frequently) Mastodon (@bretdevereaux@historians.social).

Divided Into Three Parts

Normally when I start one of these analyses, I arrive pretty early at a statement of a game’s core design focus. Europa Universalis IV is a game about states, Victoria II is a game about pops (whereas Victoria III is more a game about the intersection of economics and politics an interesting distinction), Crusader Kings III is a game about personal rule. We haven’t gotten to it yet, but of course Hearts of Iron, all of them, are games about modern warfare. Absolutely, these games have many other systems and many other concerns, but their most developed, strongest mechanics swirl around a single theme, a single point of focus.

Imperator simply lacks such a singular pillar, which I think is part of why it struggled so badly to find an audience and an identity. Instead, like all of Gaul, it is divided into three parts (Caes. BGall 1.1): war and conquest, monumental urbanism and the related pops system, and politics, particularly an interest in political instability. At release, it was easier to see all of these as simply mediated by the state and to say Imperator was, like EUIV, a game about states, but this became less and less true as these design pillars were fleshed out in post-release patches. These pillars interact, of course, but they do not quite fuse the way Victoria III‘s twin pillars (economics and politics) do. Instead, the player can often choose to engage in these systems more or less intensively, choosing objectives as they will.

And this tripartite structure, I should note, mirrors the way the public tends to understand this period and the things that draw the most interest. If we were to convert our three parts from gameplay pillars to historical topics, the whole thing suddenly makes a great deal of sense – the game is interested in spectacular Roman conquests, the monumental ruins of great cities, and the fall of the Roman Republic.

In truth, I think the fact that this game is structured on three interwoven but not quite fused pillars is part of why it struggled so much to find a distinct identity after its launch. But it also makes it a bit tricky to write about, because these pillars interact, but they don’t share a single overarching theme or theory of history. Instead, Imperator has a number of largely disconnected historical ideas about the period that it is trying to convey through its mechanics. Of course that is itself a part of Imperator‘s theory of history, that all of these themes are important and central and while they’re not yet fully integrated with each other, the points at which they connect are likely to be the points that matter.

So we’ll deal with those ideas in sequence, beginning this week with war and conquest, then moving next week to urban development and population and then finally in a third part to what Imperator thinks about the internal politics of Mediterranean empires.

Passive-Aggressive Expansion

Imperator borrows a good deal of its diplomacy and warfare systems from Europa Universalis IV and that includes how it understands the interactions between the polities in the game (‘countries’ in the game’s parlance) as well as the polities themselves. Now we’ll get to the latter point – that Imperator, in borrowing from EUIV has borrowed a state-centered model of history and then tried to bolt on mechanics to simulate the many non-state polities it covers – in a later part of this series. Instead, I want to focus on how Imperator adopts and adapts EUIV‘s model of interstate anarchy we discussed in that series.

To recap briefly, EUIV simulates – quite well, I might add – a diplomatic system known in political science (particularly the ‘neo-realist’ school) as ‘interstate anarchy.’ This is a system in which there are no real binding constraints against the ability of states to wage war (like international law or a ‘concert’ of great powers). Under these conditions, which are by far the most common in human history, states (and also non-state polities, as we’ll see) seek to maximize their security, but everything they do to render themselves more secure – conquering neighbors to get resources, more intensively militarizing their population and so on – renders their neighbors less secure (something called the security dilemma – when any act to increase your security decreases the security of your neighbors). The result is a devil-take-the-hindmost race in militarization and aggression called the Red Queen Effect (everyone’s running, no one gets ahead), a condition in which every state must become the wolf at the door to avoid becoming the prey.

Imperator borrows quite a lot of EUIV‘s simulation of interstate anarchy, but while Imperator has something of an unfair reputation as being just EUIV dressed in a toga (or a chlamys), it doesn’t simply adopt the EUIV system ‘as is’ but makes some seemingly minor but significant changes to it, similar (but in the opposite direction) to the way that Victoria II and III but impose more limits to the anarchy of their diplomatic systems, Imperator deliberately removes some of the last ‘brakes’ to its anarchic system, but more sharply limiting the kinds of diplomatic arrangements and coalitions that can form and generally weakening the external influence of aggressive expansion.

As in EUIV, while a player can play ‘tall’ and focus on improving their starting territory, the far faster route to obtaining the military resources necessary to be safe against predatory neighbors is conquest. ‘Buildings,’ as in EUIV, can offer percentage-based bonuses to production, but the basic resource of economic development in Imperator is pops, discrete units of population which produce resources based on their type. We’ll get more into this system later in this series, but what matters here is that conquest is by far the fastest way to get new pops (a striking contrast, for instance, to Victoria III, where immigration can, in fact, match conquest as a source of new population). Population grows on the order of between 0.1% to 0.5% per month, which is to say each territory (the smallest territorial unit) might take anywhere from 16 to 80 or so years to generate a new pop. By contrast, successful warfare can often move dozens of pops into the player’s core territory via enslavement (we’ll come back to this) and conquest can bring hundreds of pops into the state, providing manpower, levies and gold. The basic equation driving the player and the AI towards conquest remains, simply organized around ‘pops’ instead of ‘development.’1

But as noted, the brakes on this process have been removed or at least substantially weakened. Some of them more direct systems are gone entirely: EUIV‘s system of overextension for ‘uncored’ territory is entirely gone, as is the administrative point cost for incorporating new territory. A rapidly expanding empire may find itself struggling with provincial loyalty as newly incorporated territory filled with pops that aren’t of accepted cultures (we’ll come back to this too) tend to be quite unhappy, but compared to the intensity of penalties for exceeding 100% overextension in EUIV, these difficulties are fairly mild and of course, do not afflict already consolidated territories.

Instead the main systems for restraining rapid expansion remain in the peace system: limits to the warscore value of peace demands and aggressive expansion. In the first case, even with a total victory, it may not be possible to simply annex an entire large rival, because you cannot demand more than 100 ‘war score’ value of concessions in a single peace deal. This is a limit, of course, also in EUIV, however the pricing of territory is much lower in Imperator: a 100 war-score peace might include annexing all of southern Italy, for instance. Indeed, very few starting countries have more than 100 war score of territory at the beginning: most middle and minor powers can be annexed in one go.

Aggressive expansion is similarly weakened. While it does provide the same relations penalty as in other Paradox games, what robs it of its power – so evident in early and mid-game EUIV – is the lack of diplomatic tools to punish a player. In EUIV, a country with high aggressive expansion can generate large containment coalitions, while also poisoning relations with other great powers that the country needs alliances with or to at least remain neutral and not allied to that country’s targets.

Almost all of these options to contain a rising great power are limited in Imperator. The key system here is the state ‘rank’ system, which defines a country’s diplomatic options based on the amount of territory they hold. While the game features defensive leagues – where every state will fight together collectively in defensive wars (much like the defensive half of the function of coalitions in EUIV – they are restricted to countries at the ‘city-state’ or ‘local power’ rank, which is to say countries with fewer than 25 territories. Any country larger than that, which is to say basically any even quite small ‘middle power’ cannot form a collective defense league, but may only form bilateral treaties. The great danger, of course, in EUIV in having a large coalition formed against you is that it could include opposing middle and great powers, meaning that you might have to face your regional rival at the head of a large coalition; in Imperator, this danger is quite intentionally removed.

The rank system further reshaped diplomatic options at the higher end as well. At 25 territories and above, the state gets the ability to guarantee lesser powers, essentially creating a one-sided defensive alliance with them, creating a tool to block the expansion of other powers, however this absorbs one ‘diplomatic relation’ – a limited pool of important diplomatic agreements like alliances a state can have. This is a very limited resource also used by alliances and certain kinds of vassal-state arrangements and states generally have quite few of these (typically around 2 to 7, with more becoming available as the game progresses with investment in the ‘oratory’ technology tree). Since any ‘middle’ power is probably also using some of their relations for client states of various kinds or alliances, the result is they rarely have enough to block every avenue of an opposing major power’s expansion. More critically this limit prevents the formation of large, interwoven alliance blocs through bilateral alliances.

Finally, the largest states – ‘great powers’ of 500 or more territories cannot form alliances at all. Whereas in EUIV, running high on aggressive expansion absolutely could lead to the formation of a bloc of great powers to form, Imperator denies the great powers the tools – defensive alliances and eventually any kind of alliance – in order to work together to contain a rising power.

As a result, for larger powers, even just larger regional powers (much less great powers), the primary detriment of aggressive expansion is internal rather than external. Each point of aggressive expansion lowers the loyalty of subject states and contributes to a ticking, monthly reduction in stability. Stability (which runs from 0 to 100) is a crucial statistic because when it is below 50, each point reduces pop happiness globally, which in turn can create mass unrest across a large empire. However, even here aggressive expansion’s impact is muted: retaining high aggressive expansion is generally unsustainable long-term, but aggressive expansion decays relatively quickly and importantly decays as a percentage of its current value (0.2% per month base), while other technologies and bonuses can reduce ‘aggressive expansion impact’ lowering its maluses.

Thus the main impact, in theory, is that aggressive expansion lowers stability which lowers happiness which produces unrest which lowers province loyalty which produces regional revolts. But here again, mechanics matter: all of those effects accrue slowly. 50 Aggressive expansion (a high figure) lowers stability by 0.375 per month, weighted against other factors. As that slowly lowers stability below 50, it will drive pop happiness down, producing territorial unrest. This in turn will slowly chip away at province loyalty (a 0-100 scale) at a fraction of a point per month. On the one hand, this can put the player in a situation where it is only after they have spiked aggressive expansion that they realize they have created a province loyalty time bomb. On the other hand, the player has a lot of tools to apply to this problem, including provincial governors applying repression (which lowers unrest, slowing the downward tick of loyalty) to spending political power to drive up stability, thus slowing this process down long enough to let aggressive expansion decay to manageable levels.

The result is that whereas in EUIV, ‘spiking’ to high levels of aggressive expansion could produce immediately dangerous coalitions that might even act aggressively to tear away decades of expansion in containment wars, in Imperator even relatively poor aggressive expansion management generally produces only periods where the player must turn inward and restabilize, possibly dealing with smaller provincial rebellions in the process. As long as the player doesn’t maintain high levels of aggressive expansion in the long term (thus allowing the penalties to accrue), the damage of ‘spiking’ AE is minimal. The result, as you can see from the screenshots I’ve been supplying here, is that a player state can expand very rapidly and aggressively with minimal external pushback: all of the really meaningful dangers of rapid expansion are internal, which we’ll get to in the third part of this series.

Ancient Anarchy

In short, Imperator takes a system in EUIV that already encouraged aggressive territorial conquest in its simulation of anarchy and removed many of the checks on that process, creating a more intense competition, but also a competition that is far more likely to produce a winner. It takes a relatively talented player, in EUIV, to achieve what I’ve taken to calling ‘hegemonic breakout’ – the point at which a single power ‘breaks’ an anarchic or balance-of-power system – in most regions of the game. By contrast, Imperator, having weakened both its diplomatic tools and aggressive expansion, is built to encourage powers to achieve at least regional ‘hegemonic breakout.’

And to a degree that makes an immediate amount of sense, given the different periods of the games. EUIV covers a period in European history – the geographic area of its greatest focus – where anarchy resolved into a balance of power system which was challenged by, but eventually survived the Napoleonic Wars, thus continuing into the chronological range of Victoria III (and indeed, through that into the early years of Hearts of Iron). By contrast, Imperator covers a period where perhaps the single most famous fact about it is Rome’s achievement of clear hegemonic breakout in the Mediterranean. In that sense, the system is designed to more easily produce one thing everyone knows happened in this period.

The more interesting question is how it produces that result and how well or poorly this maps on to our understanding of the actual historical processes at work. And we can split this into essentially two parts: the diplomatic angle (does Imperator model ancient anarchic interstate systems well) and then the military angle (does Imperator model the reasons for Rome’s success well).

On the diplomatic side, yes, Imperator in its deviations from the EUIV model captures something real about the ancient world, which is, quite simply, the relatively primitive and disconnected state of diplomacy in it. Compared to diplomacy in the early modern period, ancient diplomatic relations both struggled with remarkably more limited information and a shockingly blunt diplomatic language oriented around compellence rather than persuasion.

The key scholarship on this point for the Hellenistic period that Imperator is focused on is a pair of books by Arthur Eckstein, Mediterranean Anarchy, Interstate War and the Rise of Rome (2006) and Rome Enters the East: from Anarchy to Hierarchy in the Hellenistic Mediterranean (2008). Eckstein himself is writing in response to a yet older strain of scholarship that goes back to Ernst Badian2 and W.V. Harris3 which looked at Roman political and diplomatic institutions in isolation and concluded that the reason for Rome’s rise to power was that the Romans were, uniquely more willing to wage aggressive, expansionist war.4 What Eckstein did, using neo-realist theory about interstate anarchy as a lens, was show that sure the Romans were staggeringly aggressive and bellicose, but so was everyone else.

A big part of that, which we need not get into here, is that the internal politics of nearly all ancient states pushed them towards a more aggressive, violent diplomatic stance, as a product of both cultural values (which saw war as normal and a key process in the development of turning boys into men) and political structures (in which military glory was necessary for royal legitimacy or political advancement).

In this sense, it is perfectly reasonable for Imperator‘s diplomatic system to be modeled off of EUIV‘s, because the both simulate periods of quite intense interstate anarchy.

But Eckstein also touches on the nature of ancient diplomacy generally. We have, over the centuries, developed layers of diplomatic niceties and coded language, designed to present a country’s interests and goals in the best possible light and to enable as much cooperation as possible. By contrast, ancient states tend to be shockingly blunt. Ironically in this the Romans were much better than normal, doing things like engaging in polite euphemism concerning the real status of subject communities whereas Greek diplomats tend to be surprisingly explicit. The Athenians in the fifth century, for instance, take to referring to their allies-turned-subjects in the Delian League (also known as the Athenian Empire) as the πόλεις ὅσων Ἀθηναῖοι κρατοῦσιν, “those poleis which the Athenians rule.” Polly Low in an article that engages in something of a defense of such undiplomatic diplomatic language nevertheless refers to this as the “language of kratos [strength, power]” and it is ubiquitous in ancient diplomacy, with powerful states bluntly threatening smaller states (or equals) to try to force submission.5

I should note, that blunt language also tends not to be used in private settings. Often these are speeches addressing things like open sessions of the Senate or public assemblies of all citizens, or addressing a king in front of his court and retainers. In short, this bluntness gets used in settings where upsetting the audience can create political pressure on key leaders to act in ways they know are bad strategically. But that is because in many cases, the diplomatic channels to deliver these messages directly to decision-makers instead of explosively in public didn’t exist.

And that gets us neatly to the information problem. Ancient states often had very limited information about the political situations in other states, making it very difficult to coordinate. We are used to diplomatic systems where most countries maintain a permanent diplomatic presence in most other countries – embassies and consulates – which both create an open, active channel for communication but also for information. That embassy staff can keep track of local politics, read the local newspaper and also perhaps engage in a bit of spying and communicate that information back home too, giving government decision-makers some insight into how other countries might respond to certain actions, how tense relations are and so on. But the first permanent embassies of this sort began to be established in the late Middle Ages and really only spread in the 1400s and 1500s.

Instead, the Roman Senate was often reliant on sending out commissioners (legati) irregularly to investigate certain issues or on waiting for complaints from other regions to arrive, if Rome had no permanent military presence there. We see this process play out repeatedly, particularly through Livy, in the period from 218 to 168: the Senate is often reliant on having Roman-aligned, independent powers (like Saguntum in Spain, Pergamum in Asia Minor or Rhodes in the Aegean) tell it that Roman interests in an area are threatened and try to push the Roman Republic to act. Having no permanent staff in the region feeding it information, the Senate is often reliant on these ambassadors – not permanently stationed, these are ambassadors sent on one-off missions to Rome – address the Senate, occasionally opposed by competing ambassadors from the other side. Naturally the quality of information that comes through these very motivated interlocutors tends to be quite poor and the Greek states are able more than once to effectively engineer Roman intervention against the Antigonids and Seleucids on that basis.

And that goes a long way to explaining why alliances and coalitions to thwart the rise of Rome were so infrequent and unsuccessful. Generally speaking, we can point to three examples of such coalitions, but all of them suffered from fierce coordination problems.

The first of these was the Third Samnite War (298 to 290) which began as a war between Rome and the Samnites (the peoples of the southern Apennines) but expanded when many of the Etruscan communities and eventually the Senones, a Gallic people. This was still essentially a local coalition between northern Italian polities (the Etruscans and Senones) and southern Italian ones (the Samnites) on the sort of scale that Imperator could model with a defensive league, although this was a war that expanded after it had begun, rather than beginning with all participants.6 That said, even at this scale, the coalition suffered severe coordination problems: the Romans win in part because not all of the coalition against them was present at the decision Battle of Sentinum (295).

The second coalition effort, if we are to understand it as such, was the Pyrrhic War, in which the Greek states of Southern Italy (notably Tarentum) invited in Pyrrhus, king of Epirus, in a bid to peel back Roman power. The Samnites eventually join in this effort, but the Etruscans and Gauls do not and once again coordination problems plague the effort. Still, this too is something that could probably be modeled either by an alliance or a defensive league in game; Pyrrhus was an effective military commander with a large army, but Epirus was not a major power – as demonstrated by Pyrrhus’ repeated inability to force his way as a serious player in the struggle for power in Macedon, itself the weakest of the Mediterranean great powers, by far, in this period.

The last effort to string together an anti-Roman coalition is certainly the Second Punic War, which actually manages to get two great powers – Carthage and the Antigonids – along with quite a lot of Gauls and of course some of Rome’s socii in revolt. Once again, however, the efforts were hardly coordinated. The Antigonids didn’t so much join a coalition as they waged an independent, parallel war with Rome, opportunistically trying to take advantage of Rome’s distraction (something the AI is plenty able to do in game). Notably, Philip V of Macedon enters the war late and leaves early when it becomes clear he isn’t going to be able to accomplish much, since the Romans controlled the Adriatic.

Compared to the sprawling and complex European coalitions (and the equally sprawling wars they waged) in the Early Modern period, all of these coalitions were indeed small and poorly coordinated. Absent diplomatic institutions like permanent embassies, combined with a generally lower degree of cultural connectedness7 made forming vast large coalitions much harder, in turn weakening the ability of a ‘balance of power’ system to prevent hegemonic breakout by ‘ganging up’ on the winning party.

As, indeed, famously happened: Rome was able to deal with its rivals one at a time as it expanded, rather than all at once. So while this system doesn’t feel particularly bespoke to Imperator, the adaptations here do seem to express something relatively accurate about antiquity: that the interstate system was anarchic, and while powers did engage in balancing behaviors, these were limited by the fairly crude diplomatic tools available at the time. As a result, the system was unable to prevent the emergence of a single dominating great power: Rome.

Now that gets to the question of why a single dominant hegemonic power emerged in the Mediterranean, but not why it was Rome, so how does Imperator treat that?

Why Rome?

Now normally I would lead in with Imperator‘s game mechanics, but I think it is going to be easier to follow if we lead with the history and then check the mechanics against it, since Imperator doesn’t so much have a single system to represent Rome’s military machine as it has a collection of bonuses and tools.

The standard scholarly explanation for why Rome, of all of the Mediterranean powers, emerged victorious of the anarchy of the third and second centuries is very old indeed, dating back to the historian Polybius, writing himself in the middle of the second century during the final phases of that hegemonic breakout. Polybius in his histories contended that the Romans had a number of advantages – they were more adaptable, had a better military system, a better organized state – but the advantage modern historians have most focused on is manpower. The Romans had more soldiers.

Polybius illustrates this in an astounding passage in book 2 (2.24 to be precise), where he relates the result of a comprehensive Roman census in 225. It was an unusual accounting – the Roman census usually only concerned citizens, but in 225, expecting a major war with Gallic peoples to the north, the Romans ordered a general accounting of all men liable for military service in Italy, both Roman and ‘allied’ socii, which is to say subject communities. Polybius reports the totals: 699,200 men liable for service as infantry (42% citizens) and 69,100 liable for service in the cavalry (38% citizens), a total of 768,300 men. There are, I should note, a few quirks in Polybius’ total and subsequent scholars have made some corrections and emendations; P.A. Brunt read it as 550,000 iuniores truly available for conscription against 785,000 total adult males, while L. de Ligt reads it as just 526,000 iuniores against c. 750,000 total adult males.8 For our purposes, the differences don’t matter. Those numbers are enormous, and yet because we know the Romans did, in fact, keep records to this effect for use in the dilectus, they’re also pretty reliable (and consistent with other Roman census figures, which we know were the product of a hand count).

Simply put, no other ancient state could come anywhere remotely close to matching those sorts of numbers and the military strength they implied.

The most recent and rigorous treatment of this is M.J. Taylor, Soldiers & Silver (2020), which presents I think the most careful and responsible examination of what we might call the ‘manpower thesis’ out there, though here is where I have to admit that the pure manpower thesis is one of the things I aim to knock down in my own book project, by arguing for the importance of more resources than just manpower in producing this result. Nevertheless, Taylor estimates, quite conservatively, that the maximum actual Roman deployment – because no state can put 100% of its adult males in the field at once – was around 185,000 during the Second Punic War (212 and 211, to be precise). Carthage, the next best, peaks at 165,000 in the field in 215, but drops rapidly from that figure, unable to sustain it against casualties and cost. Behind them were the next two great powers, the Ptolemies and the Seleucids, which Taylor puts around 80,000 a piece, the former in 217, the latter in 190; in truth I might nudge these figures upwards to account for garrisons and such, but not very far. Certainly no higher than around 100,000. Finally, there are the Antigonids, the smallest and weakest great power in this period – mightily fallen from grace since the Battle of Ipsus (301) where they could deploy some 70,000 men. By the mid-third century, the Antigonids control only Macedon proper and Greece, their peak deployment comes in 171 at just about 50,000 men.

In short, then, the Romans could effectively bury any possible opponent under a tide of Romans, though in the event they rarely needed to: against generals not named ‘Hannibal’ or ‘Pyrrhus’ the Romans had no problem routinely winning while outnumbered. Consequently, Rome tended to engage in military operations on multiple fronts simultaneously, spreading its huge military deployments over multiple theaters and was more than able to absorb setbacks on any one of them. One of my own interventions on this debate actually intensifies this disparity: I can show that not only do the Romans have a lot of troops, their troops are, man for man, more expensive in total resources (though not in cost to the state – the Romans shift most of the costs on to the farmer-citizen-soldiers themselves) than those of their rivals. So the raw deployment totals tallied by Taylor actually understate the scale of Roman advantage, though they are directionally correct – Rome is deploying substantially larger military resources than Carthage, which in turn nearly doubles the deployments of the Seleucids and Ptolemies, who in turn almost double what the Antigonids can field (though the ability of the Antigonids to still compete seriously in that system should warn us that materiel isn’t the only thing that matters!).

How was that possible!?

The short answer is: buy my book (when it is done and then comes out), because it is literally a 150,000 word answer to this question (just wait for, “Section 2, Chapter 6, Part 2c: Warlords and Youths” and suchlike).9 The much longer answer is…my book (when it is done and then comes out).

But the medium length answer is this: Rome’s decision to adopt an expanding, incorporative citizenship – in contrast to the closed ethnic hierarchies of its peer competitors – gave it a huge core of citizen-soldiers who could be called on to serve at their own expense. These fellows weren’t professionals, but rather conscripts. However, Rome’s habit of always being at war meant that these soldiers would all serve quite a few years (around seven) meaning that any given legion was a mix of experienced veterans and only a few raw recruits (quickly trained by the veterans), giving these armies a degree of training and discipline unusual among citizen-militia armies.

That alone would have made Rome a viable but not overwhelming competitor. But the real key is the way the Romans structured their control of Italy. Instead of the common way empires were structured – tributary empires where the imperial core taxed its subjects to pay for an army to extort more taxes and so on – the Romans instead made conquered Italian peoples ‘allies’ or socii in Latin. These allies supplied troops for Rome’s armies and – because they were products of the same violent cauldron of Italian conflict – they fought just as hard and the same way the Romans did. These fellows provided a little more than half of all Roman armies, massively magnifying the military power Rome could bring to bear. And they tended to stick with Rome, even when the going was tough, because the ‘deal’ the Romans offered them was a good one that the Romans took seriously.

Consequently, Rome could ask its citizens, and the citizens of its ‘allies,’ to provide their own equipment, serve at relatively low pay (the socii pay their own troops; Rome merely needs to call them up and feed them) and fight hard, because for both groups military service was how they served their own communities, and how they proved their worth and status in those communities. In short, Rome took the intensive, powerful recruitment you get in something like a small polis and ‘franchised’ it first over an enormous citizen body and then an even larger collection of socii, resulting in the Roman Republic being able to call on a huge proportion of the resources – manpower, but also money, metal, food, non-military labor and so on – of Italy. The Roman resource pool was not itself larger, but the Romans pulled much more out of it.

So how well does Imperator simulate that?

Patchwork Italy

This runs quite quickly against some fairly fundamental game design concerns, particularly the need for player choice and balance and also the need to create states and polities which are broadly legible to the player (both when being played by them, but also when played against). For the first, while Imperator starts in 304, before the Roman military machine is fully operational (which we can only begin to see clearly, I’d argue, in the Pyrrhic Wars, 280-275), having a situation where a competent Rome player can amass an essentially unstoppable military force just a few decades into the game and proceed to effortlessly steamroll every opponent would hardly be good balance. It’s accurate, but it would hardly be fun. Consequently, we might expect Rome’s advantages to be trimmed back at least a bit to give the other states a decent shot at prevailing, whereas historically, only Carthage realistically came particularly close to holding up against the Romans (we really need to do the Punic Wars on here at some point; perhaps next year).

The deeper problem, however, is that Imperator struggles to simulate the structures which enabled Rome (and to a substantial but lesser extent, Carthage) to develop so much military capacity.

Imperator features three sorts of armies: levies (raised directly from controlled regions), legions (standing professional forces which use up the levy-capacity of a region, usually the capital) and mercenary armies, which are hired with gold. Each region has a certain amount of levy or legion units (‘cohorts,’ a term which is a bit jumping the gun for a lot of this game) determined by the number of non-slave integrated-culture pops in the region. Casualties in levy or legion cohorts are in turn replaced by a second resource, manpower, which is primarily generated by the ‘freeman’ pops (representing the free peasantry; we’ll come back to pop-types) but citizens and tribesmen also produce small amounts as well. Note that manpower and levy-capacity are separate, though incorporated-culture pops produce both.

Meanwhile, Imperator has a variety of different ‘subject nation’ statuses but the one that concerns us here is the ‘feudatory,’ which is used at the beginning of the game to represent at least some of Rome’s nascent socii-system, as well as Carthage’s relationship with some of it’s subject states in North Africa. And I appreciate the effort here at the outset to try to model the sort of composite states Rome and Carthage were, but that effort stumbles over substantial problems, both in its initial structure, in the mechanical rules that determine it, and in how basically any game is going to actually progress.

The first problem is simply the starting map position. Rome begins with four feudatories, all very small: Fretania, Pelignia, Nuceria and Marsia (the use of the toponym rather than demonyms here is striking and we’ll come back to it; it ought to be the Fretani, Paeligni, and Marsi (and Nuceria, which is fine)). But a huge chunk of Campania is simply represented as core Roman territory and this is much too early for that. Capua, for instance, represented here as part of the Roman Republic, should certainly be a feudatory as it was a community of socii (famously defecting to Hannibal after Cannae!). Critically, Campania was a mess of different smaller communities (of which Capua was the most important), which actually matters, because some of them defect to Hannibal and some of them don’t.10 Large parts of Campania were only brought into the core Roman territory (the ager Romanus) as punishment for those defections, while parts of Campania remained socii all the way to the Social War.

Consequently, the game essentially flips the proportions that Rome is going to have with its socii throughout its history: the socii from an early point modestly outnumbered Roman citizens, but in Imperator, even at the very beginning of the game, the Roman Republic is far bigger (399 pops) than all of its feudatories combined (23, 52, 22, and 30 pops; 127 combined). And that really matters because it changes what was historically a mid-sized Roman Republic leading a larger ‘alliance’ (really, subject) network to a larger Republic that is strong on its own, which happens to have a few, largely negligible ‘hangers on’ vassal states.

And that isn’t likely to change either: as Rome expands, either by player or by AI, it is going to tend to absorb those feudatories and – unlike the actual Roman Republic – not create any new ones, meaning that often by the early or mid-third century, Rome will have direct control over the whole of the Italian peninsula with no socii at all, a thing that did not happen historically until 87 BC.

The reason the player will do it is that whereas historically the socii system was crucial to Rome’s ability to mobilize military force in Italy, far more effective than direct imperial rule might have been, in the game feudatories are generally inferior to direct control. The way feudatories work is that they join all of their overlord’s wars with their own armies and contribute 15% of their manpower gain to the overlord. In a war then, feudatories make it slightly easier to replenish damaged armies and contribute their own, independent tiny armies (which have a nasty habit of just getting annihilated, ‘stack wiped’ in the parlance, because they’re too small). But that’s now how Rome’s socii system worked! The socii didn’t supply small, independent armies, instead their military system was wholly subsumed by the Roman one – there were no independent socii armies.

More broadly, the player is going to want to move out of this system for obvious reasons: why accept 15% of a feudatory’s manpower gain when you could control that territory directly and get 100% of that gain? Especially because feudatories require being in the same ‘culture group’ as the overlord, which reduces happiness penalties for not being the primary state culture (meaning it is less punishing to absorb them). Moreover, while feudatories contribute some amount of manpower, they do not contribute levy-capacity and generally speaking it is levy-capacity, which governs how many troops a state can field at once, more than manpower, that is important to maximize. While feudatories do have their own armies that fight the same war as you, unity of effort is important and it is simply more valuable to have that levy capacity under your direct control.

And I think the developers seem to have known this, because the AI makes no effort to generate a Roman-style system of socii either, instead directly conquering Italy as it can, and incorporating the feudatories back into the main Roman state as quickly as possible. The socii system thus becomes an odd quirk of Rome’s starting position, quickly abandoned by both the player and the AI, rather than the key to Rome’s military success.

In place of that actual system what Rome gets in Imperator is a series of direct and indirect bonuses. The direct bonus is that Rome gets a +2.5% flat bonus to levy size for ‘Roman Heritage’ (which makes it all the more important to absorb those feudatories so that the bonus applies to them too, as they get different ‘heritage’ (state-specific) bonuses). Indirectly, Rome’s ‘Roman military traditions’ (a military tech tree; each culture has specific ‘military traditions’ they can unlock) includes another +2.5% levy size multiplier, along with +5% manpower recovery, and substantial bonuses to infantry stats, along with access to a set of ‘military reform laws’ all of which (save for the ‘Marian Reforms’) come with substantial bonuses to levy size (5%, 7.5% and 2.5% respectively). Thus, unable to model the system by which Rome raised large armies, the game instead gives Rome a set of bonuses to ensure that it simply gets larger armies, regardless.

On the one hand, this is a theory of history, an awareness of what I’ve termed above the ‘manpower thesis.’ On the other hand, it is unfortunately a fairly unsophisticated version of that theory of history, understanding that Rome had ‘allies’ and that Rome had an advantage in manpower, but not quite able to put into mechanics how those two things relate. I don’t think this is necessarily because the developers didn’t know, but rather because simulating complex, composite states like the Roman Republic in a way that is fun to play and clearly legible is really hard.

Marrying Your Manpower

Carthage suffers the same problem, but processed through an uncritical acceptance of Polybius’ narratives. To be fair, Polybius’ position, that Carthage’s armies were made up of ‘mercenaries’ is still repeated uncritically in most textbooks, even as it has long been challenged in the specialist scholarship, so I am not surprised that this was the model that Paradox went with. But as with Rome, Carthage is a much more complex creature than the game lets on in important ways.

The tricky thing here really is simply the word ‘mercenary.’ The word Polybius is using when he describes Carthaginian armies as ‘mercenary’ is misthophoroi (μισθοφόροι), literally ‘wage-bearers.’ In English, the word mercenary implies a foreign soldier, serving purely for pay in a conflict to which they have no real part, but misthophoroi doesn’t carry all of those meanings. Instead, a misthophoros is simply a soldier whose compensation is in the form of a wage rather than rewarded with land settlement or a duty of citizenship; we might actually better translate misthophoros in some cases as ‘professional’ rather than ‘mercenary.’ Even then, I will note, Polybius is irregular in his use of the term (particularly when he turns it to the Hellenistic kingdoms), often classifying soldiers as misthophoroi or not when it suits his rhetorical purposes. He isn’t making things up, mind you, but Polybius will sometimes emphasize that wage-earning and in other cases de-emphasize it, as suits his purposes. He is trying to draw a contrast between Rome’s citizen soldiers and ‘inferior’ mercenary troops, but in doing so obscures that in many ways the Roman and Carthaginian systems are more similar than he’d like to admit.

Imperator, however, takes him at his word. Dotted around the map, throughout the game are mercenary companies which, for a large sum of ‘gold’ (I really wish the currency in Imperator was, as would be more correct, talents of silver) can be hired. These companies have the cultural unit mix of the culture that dominates the area they are hired from, and you can hire mercenaries from another country’s region easily enough if it is in your diplomatic range. Just as Rome gets bonuses to levy size, Carthage gets penalties to levy size, but bonuses to mercenaries, able to hire both more and (through their military traditions) more cheaply. Combined with Carthage’s strong economy (and a 10% export value bonus), the Carthaginian player is encouraged to rely on these mercenaries, recruited from wherever.

And you can see how that seems like a reasonable and straight-forward extrapolation of what Polybius says about Carthage’s armies…but that’s not how Carthage’s armies worked.

On the one hand, it is the case that Carthaginian citizen soldiers are exceedingly rare. Apart from serving as generals, Carthaginian citizens didn’t fight much and the Carthaginians do not seem to have put much value in one’s ability to do so (which doesn’t meant Carthage didn’t value war or was peaceful; it was aggressive and bellicose, like everyone else). But Carthage didn’t just hire mercenaries from wherever; instead Carthage recruited troops from the territories it controlled, directly or indirectly, and then paid them. That is a system which, at a good deal of abstraction, describes the military systems of most large ancient states (except Rome, which didn’t pay the socii).

Taylor, in Soldiers and Silver (2020), does a good job of taking an inventory of the various sources of Carthaginian manpower, particularly in the Second Punic War, where we can see it most clearly and identifies fundamentally four core sources. First off, Carthage employs a lot of troops recruited from the regions of Africa it controls, sometimes described as ‘Africans’ and sometimes as “Libyans” in our sources, often as many as 40-50,000 of them (it’s unclear if there is always a meaningful distinction between ‘Africans’ and “Libyans” as described by our sources, by the by). These fellows often serve as the heavy infantry backbone of Carthaginian armies, though some ‘Libyans’ may have been light troops.11

The next key source by distance (but not size) were the Numidians. There were two key culturally Numidian kingdoms, Massaesylia, and Massylia, and at any given time at least one of them was effectively a vassal-ally of Carthage. Carthaginian generals – the title was rabbim or rab mahanet – who served for long periods, have a notable tendency to marry into the royal Numidian families, especially strange because the impression we get is that the rest of the Carthaginian elite did not marry out. But, thinking back to how mobilization works in non-state or proto-state societies, you can immediately see that those generals are using a standard tool of non-state aristocratic power building: those marriages give them the ability to call those ‘Big Men’ with their retinues to war, giving them access to Numidian cavalry for their armies. And Numidian cavalry was some of the best in the ancient Mediterranean, so that was a formidable advantage to have. While Carthage recruits tens of thousands of heavy infantry from its settled, urban African holdings, Numidian cavalry was a scarcer resource, a few thousand here, a few thousand there, perhaps never more than 10,000 or so total, but very valuable.

The next key chunk – and by 218, by far the biggest – were Spaniards. Carthage had maintained trade contacts and a mix of both small trade posts and colonial settlements in Spain from a relatively early date, but it is after the end of the First Punic War (264-241) and the subsequent Mercenary War (241-238; a revolt of Carthage’s African mercenaries along with the subject communities from which they were recruited, which Carthage defeats) that a family of Carthaginian generals, the Barcids (first Hamilcar and then his sons Hannibal, Hasdrubal and Mago) begins a large-scale conquest of Spain. That said, when the Second Punic War starts in 218, Carthaginian control – very real and very violently enforced – on the Mediterranean coast of Spain is still quite new. The Barcids rule through a mix of diplomacy with ‘tribal’ leaders – including some marriages to Spanish tribal princesses – and force. They rely, in particular, on the support of key Iberian warlords called in our Latin sources reguli (‘little kings’ but more clearly ‘warlord, chieftain’) with whom they have personal alliances. In short then, this seems like a mix of colonization, but mostly an extension of the Numidian system to embrace a whole new set of Iberian peoples.

Those relationships in turn provide Carthage access to enormous numbers of Iberian troops over the years, making up around half of all of Carthage’s soldiers, providing both Iberian infantry (‘mediums’ who served as line infantry) and cavalry in quantity. Taking a snapshot in 215 (the year when Carthage has the most troops, almost 165,000, under arms) there’s about as much Iberian cavalry in Carthage’s army as Numidians, and about as much Iberian infantry as there are Africans and Libyans combined. Moreover, as the war wears on, while Carthage’s supply of African troops seems to decline, Carthage regularly raises huge fresh armies in Spain (until it is lost to them in 206), so Carthaginian reliance on Spanish manpower grows over time. For the curious, this is almost entirely Iberian manpower – the Celtiberians, though bellicose and famously willing to serve as mercenaries are outside of the zone of Carthaginian control and so figure only infrequently into their armies.

Finally, when Hannibal crosses the Alps and approaches Italy, he finds a lot of Gauls in the Alps and Cisalpine (Italy-facing) Gaul who had bones to pick with the Romans and were thus willing to side with Hannibal. Something on the order of 25,000 of these fellows work their way through his army over the course of his campaign, but we don’t see larger numbers of Gauls generally in other Carthaginian armies. Hannibal has them because he marched through their homeland and they share an enemy, though we ought to imagine their recruitment too probably looks like the Numidian and Iberian case (minus the marriages): alliances with local tribal leaders giving Carthage access to the troops.

Now it is the case that Carthage is clearly paying these fellows, but as you can see here, they’re not quite mercenaries in our sense of the word. All of these soldiers are being recruited from various peoples we might correctly describe as subjects of Carthage. The key difference between this system and the Roman system is that whereas Rome offers its subjects protection and immunity from tribute and then gets their soldiers ‘for free,’ Carthage imposes tribute and uses that money to pay for their troops. In that sense, they are misthophoroi – wage-earning troops – but hardly external foreign mercenaries. Polybius wants to make the point that Rome’s citizen-soldier model is more effective (and he’s right), but that isn’t the same as supposing that Carthage orders its armies on Amazon. Notably absent in nearly all Carthaginian armies are the Greek, Thracian and Galatian mercenaries which were everywhere in the Eastern Mediterranean. Carthage surely could have hired them, but didn’t, instead opting to recruit from its own domains.

We cannot know – no source tells us – but to judge by the sums of money Carthage is spending, one reason is that their sources of manpower in the West seem to have been a lot cheaper. We hear of one planned cash infusion of 1,000 talents (6 million drachmae) and another infusion where the number is lost in the text (Livy 23.13, 23.32), while recent estimates of Carthaginian revenues tend to be between 15-20 million drachmae.12 That’s a lot, but much, much less than the c. 50m drachmae we suppose the Seleucids pulled in annually, or the 75m the Ptolemies did.13 So with less than half of the revenue, Carthage is putting out twice the soldiers, suggesting they’re paying a fair bit less for these fellows. What is striking is that it isn’t clear that Carthage sacrificed much quality for this: their armies had plenty of war elephants and on average don’t seem to have been much more cheaply equipped than their Hellenistic peers.14 Sure, Carthage is employing lots of cheap light troops (Iberians, especially), but so are the Hellenistic states.

In short, then, one thing that is really important to know about Carthage’s armies is they’re not just mercenaries hired from wherever: if they had been, Carthage could never have matched, much less doubled the deployments of the much wealthier (in state revenues) Hellenistic states. Instead, Carthage is using its territorial control and aristocratic alliances, it sure seems (we can’t know for certain) to pay something like a ‘discount rate’ for its vast armies. Alas for Carthage, that was still too high as compared to the ::checks notes:: nothing that Rome paid for the socii.

The Trouble With Complex States

So on the one hand Imperator wants to express some true historical things about Rome and Carthage: that the former relied on citizen-soldiers (both Roman and then the citizens of the socii) while the latter relied on hiring wage-paid troops from its imperial possessions and vassal polities.

Where Imperator struggles on both points is capturing exactly how these systems worked, and I think the problem, apart from some outdated or insufficiently specialist history at work is that these states are just really complex in ways that are very hard to capture in a game context. Rome, if proceeding historically, by 218 should have dozens of socii or otherwise subject communities; Polybius breaks the socii in his census (Polyb. 2.24) into seven large groups, but each of those was composed itself of many subordinate communities, each notionally with just a bilateral relationship with Rome. For a player running another country, trying to even parse such a vassal swarm on the map would be very hard to make anything other than frustrating.

But the broader trend, which we’re going to see become a theme in this series, is that Imperator struggles to model the Roman Republic in particular because the design precepts of the game – and indeed, the broader genre of Paradox grand strategy games – demands a single set of rules that encompass most, if not all, of the polities on the map. Crusader Kings (both II and III) has traditionally dealt with this problem by picking a core focus kind of government (vassalage-based polities) and leaving the odd exceptions (Republics, the Byzantines) for later expansions, enabling the design to focus on a single broad model of governance.

But that approach simply isn’t available for Imperator, because the spotlight has to be on the Roman Republic, which as a state works very differently than every other Mediterranean polity. It is perhaps most similar to Carthage, but as you can see above, not very similar even then (something that will become even more clear when we get to politics). Indeed – and this too will be a trend – Imperator is best at simulating not the Roman Republic on which it is notionally focused (I mean, the title is the Roman word for ‘victorious general’ and Rome), but the major Hellenistic monarchies of the East.

And, to take a brief stop, the Levies-Legions-Mercenary system actually simulates those armies – Hellenistic armies – quite well. Under the ‘Royal Guard’ military reform (easily obtained for those kingdoms) they end up with an army that looks quite a lot like a Hellenistic army: a professional core raised from integrated culture pops (read: Macedonians) living in the imperial core (Syria, Mesopotamia, Alexandria – where the Greek-speaking colonies are!) reinforced by levies raised from their outer domains, with some Macedonians in them, but also bringing large numbers of local troops fighting in local styles, finally occasionally supplemented with mercenaries hired from the kingdom’s border regions. But it doesn’t fit the armies of the Roman Republic particularly well at all, which is a problem as those armies as the ones that are supposed to be the stars of the show!

I certainly think it is possible to create a military-political system which captures the complex ways that Rome and Carthage raised armies, but such a system would have to be granular and complex in ways that would fit every other power in the Mediterranean very poorly and would probably make playing a lot of them a tremendous chore. Instead, the clear choice is for different states to have lots of bespoke mechanics: one military system for Rome and Carthage, another for Hellenistic monarchies, a third for non-state peoples. But development time and resources are limited. One suspects the plan was always to launch with a single system that didn’t quite fit anyone perfectly, but captured something about everyone and then in DLCs and expansions to flesh out culture-specific systems with different rules, but of course Imperator did not develop a player-base sufficient to justify that development. In part, I suspect, precisely because its single military system ended up feeling a bit flat.

It is a nasty catch-22 inherent in the era, but also part of why I really hope Paradox does, in fact, go back to this era, either with an Imperator II or another, similar project, to iterate on these mechanics. After all, Rome wasn’t built in a day – and that’s where we’ll turn to next: Imperator‘s take on pops, development, ancient economies and urbanism.

1. I get the sense – from public statements and dev diaries, I have no special visbility into Paradox development at present – that a move from development to pops may be in the cards for ‘Project Caesar,’ which would bring that series’ expression of this system into line with Imperator‘s.
2. Roman Imperialism in the Late Republic (1968)
3. War and Imperialism in Republican Rome: 327-70 B.C.. (1979). On the author, please note.
4. On this, see also K. Raaflaub, “Born to be Wolves? Origins of Roman Imperialism.” In Transitions to Empire: Essays in Greco-Roman History, 360-146 B.C., in honor of E. Badian, edited by R. W. Wallace and E. M. Harris (1996).
5. P. Low, “Looking for the Language of Athenian Imperialism” JHS 125 (2005): 93-111.
6. This is something, I note, that Paradox’s design philosophy has changed on. Both EUIII and Victoria II feature a number of ways for wars to expand after they’ve begun, with new participants ‘jumping in’ after hostilities have begun. It’s not hard to see why they’ve shifted to a system where the participants in a war are basically fixed once it has begun: having a ‘surprise’ power jump into a war after it has started is extremely frustrating and unpredictable for players, even if it is a thing that happened a lot, historically.
7. That is, the lack of Latin and later French as a shared diplomatic language (Greek comes close in antiquity, but not quite close enough) combined with the shared cultural touchstones provided by Christianity in Europe.
8. Brunt, Italian Manpower (1971); De Ligt, Peasants, Citizens and Soldiers (2012). Among other technical issues here, Polybius has quite clearly double-counted a few small groups and has not accounted for less than a 100% response rate to the census.
9. Before my publisher reads this and panics, worry not – I plan to have normal human section divisions. But also the third subheading under the second heading of the sixth chapter which is in section 2 is, in fact, titled ‘warlords and youths’ in my current draft.
10. The key communities here are ::deep breath:: Cales, Volturnum, Casilinum, Capua, Atella, Abella, Calatia, Sabata, Suessula, Acerrae, Nola, Nuceria, Surrentum, Pompeii, Herculaneum, Neapolis, Puteoli and Cumae. For more on this region in the Second Punic War, the thing to read is absolutely M. Fronda, Between Rome and Carthage (2010).
11. On the tactical composition of Carthaginian armies, see J.R. Hall, Carthage at War (2023), 28-30. We know a lot less than we’d like.
12. Taylor, (2020) and D. Hoyos, Hannibal’s Dynasty (2003).
13. Figures from Taylor (2020) again.
14. This is actually a question I go at in some detail in my book!