As those of you who read my “regular” updates on the forum have no doubt noticed, I’ve gotten a good deal more “home-work” these days than just Imperium.

For this reason, I have spent some of the time during this summer holidays not only working on Imperium, but also speculating as to the possible futures of this “little” hobby project of mine. Basically looking at what is done, what remains to be done, and thinking back over the experiences of the past couple of years.

The status: In theory, almost every part of the game has by now been coded up at least once. The only place where there is really missing some game code is with the interface to the Roman political system, where I am still not 100% satisfied with how the game works. Some of the game mechanics could also still stand being polished (many parts have “only” been iterated on once or twice – as compared to the five or six times I have worked with various parts of the operational game engine) and the GUI needs to be made more user-friendly (in some parts, it is catastrophical). The final part of the game that really needs work still is the AI, where I am definitely not happy with what I see. It is not bad as such (ancient warfare is simple enough that even a simple AI – find enemy army, hit it with a big stick – would work fairly well) but this is the era of Hannibal and Scipio. A little bit of finesse would be nice.

Given money to live on, I could probably get the game in a publishable state within 3-4 months (I, however, build weather databases for a living rather than games and am actually quite happy doing so). At hobby pace, that amounts to a year or probably more of development still. Not much, when looking back, but it’s still a minimum estimate. Given the vicissitudes of life, it could easily take double or triple that amount of time. At which point, there would still a fair amount of work developing art work, game maps, sfx, beta-testing, etc., still left to be done.

I am generally not that good at taking the lessons from my professional life into my hobby ( after all, a hobby is play… ), but one mantra of professional (non-games) software development is “release early and often”. Interestingly, it is one that I heard Soren Johnson echoing in the podcast Three Moves Ahead recently, and it is one that I am now strongly considering applying to the Imperium project, in the interest of moving the project along.

Those of you who have followed this for a while may recall my discussing how to compartmentalize Imperium into modules in the past. The current thought that I have is to carry this one step further and place Imperium 1.0 on the fast-track to getting published as soon as I feel that the parts of the current game engine currently done are sufficiently polished up to stand on its own as a game.

My current idea would be a game called something along the lines of “Imperium: Campaigns of the Ancient World”. The operational engine in Imperium has seen a lot of refinement over the years, along with the battle and siege engines. The graphical UI and the artificial intelligence for these bits definitely need work to be of sufficient quality, but this is a matter of polish, rather than any major work. The main component missing for Imperium to already be such a game is implementing a victory point mechanism/win-lose mechanism, and perhaps a simple scenario-event system (to allow introduction of new leaders, reinforcements, etc).

Is there a market for such a game? Obviously, some of you who are interested in Imperium may not be interested in this idea. This would not be a grand-strategy game, obviously, but rather a game to allow the player to fight the various campaigns of the ancient world on the pure strategic layer (think something along the lines of Imperium Romanum). Although the Imperium game engine is fairly simple, I think it should be strong enough to carry stand-alone scenarios of this kind. And there is more than enough material: potentially a huge number of scenarios could be developed, containing some of the most fascinating characters of the ancient world. Even within the limited time period envisaged in Imperium, we have three Punic Wars, several conflicts in Spain, a couple of Syrian (Ptolemies vs Seleucid) wars, not to mention the Roman civil wars. Much of this is virgin territory at this level of computer strategy games – limited only by the difficulty of collecting historically accurate information for OOBs and such.

The benefits for myself are obvious: I will be able to get a game of this kind finished and ready much earlier than would be possible for the complete package. There should be less bugs due to the smaller scope of the product, hopefully allowing for a more productive beta-test period (allowing us to focus on the gameplay and AI issues). I would also be able to get quality feedback on whether the ideas I have for Imperium work or not. There is no better test for a concept than to have several hundred enthusiasts picking it over for faults to complain about.

The downside, of course, are that such an idea may fall flat on its face. I am not worried so much about the game flopping (I do think the game system is good – obviously – and if it isn’t, the sooner it flops the better so I can move on ), but there simply may not be a market for this kind of game. I have discussed the idea with one potential publisher for Imperium, though, and they have given me a tentative thumbs up. So this post is primarily to give you a chance to let me know what you think (all three of you who read this blog).

Following this plan does not mean that I’ll drop the other elements of Imperium. Once I have the operational module done, I would be able to continue polishing the political modules and the plan would be tosubsequently release the political modules + the grand campaign as an expansion pack to the core operational strategy game.

The pitched battle system in Imperium has gone through many iterations; I can count four or five, just of the top of my head. And we’re talking of fully-fledged iterations; with the engine implemented and playable. Some of these have been good, others, less so. The key problem that has led to so much experimentation has been the need to find a battle resolution system that is both interesting and AI-friendly. In short, if you are facing AI-Hannibal, even with Scipio Africanus, you should be worried even knowing that you can influence the course of events…

The current – and final – version of the battle system is an auto-resolution system where the player’s general, army composition, and luck hopefully blend to give a satisfactory battle result. Several of the iterations I worked through were much more interactive, but they all invariably resulted in long playing times (a problem, given that 250 years will see a lot of battles) or impossible AI (which undermines the strategic game). The current system gives quick results, yet hopefully still provides enough interesting decisions to enhance the game narrative.

Pitched battles are resolved in three distinct phases: skirmish, clash, and pursuit.

In the skirmish phase, all the units assigned to skirmish as well as skirmish detachments are engaged in combat (a “skirmish battle” as mentioned in the previous post, is just this phase of the battle). An army will want to have at least one regiment of skirmishers committed here, as the main battle line troops may suffer casualties from missiles if the army’s skirmishers are completely overrun. The result of the skirmish may also impact the combat value of the units in the ensuing battle directly; units that are shot at and harassed while deploying into battle formation being unlikely to fight at their best in a pitched battle.
The clash of forces is naturally the high point of the battle. The regiments of each side are lined up against each other with the frontage potentially being constricted by the type of terrain present. The player has some limited control over how forces are lined up (strong right flank, left flank, center or perhaps a weak center) and allocates units between flanks, battle line and skirmish line. Units that have skirmished will participate in the battle, though their earlier exertions will mean that they fight at lowered strength even if they have not suffered casaulties. Needless to say, infantry units will tend to be more useful in the centre of the battle line than on the flanks.

Combat is then resolved with forces clashing “row by row”, starting with both flanks. If a force succeeds in routing its opponents, it may swing inward and take other enemy units in the flanks and rear. This mechanism makes it possible for strong cavalry units to be decisive if they are not neutralized (or at least blunted) by an opposing cavalry force. The battle is decided, however, by which side wins more of the clashes on the battlefield; it is thus, in fact, possible for a heavily packed infantry army to win even if it loses on the flanks (essentially, what the Romans tried – and failed – to do at Cannae).

Casualties are relatively limited in the clash of forces (typically 5 to 20%); the real problem occurs in the pursuit phase. As the word implies, this is where the winning side tries to inflict additional casualties on the losing side. The system is very simple, with cavalry and light infantry being more deadly than the heavy infantry. Regiments that have been outflanked or surrounded during the battle will suffer much higher casualties. A powerful cavalry force can wipe out its opponents in this phase.

So far, the results of the battles I have seen have been very satisfactory. The battle resolution system fulfills a number of key requirements that I aimed to see in the game. Heavy infantry are the king of the battlefield, but the cavalry have the possibility to be decisive. Missile troops (i.e. skirmishers) are neither simple support troops or paper in the cavalry-infantry-missile RPS trinity, but have an individual and important role on the battlefield. If the system works as intended, players will be encouraged to build armies similar to those that campaigned historically by the game mechanisms, rather than just for the fun of it.

Whether they do so, of course, remains to be seen. For now, however, this concludes my discussion of the combat mechanics. Hopefully, these posts have provided you some insights into the ideas and thoughts that have gone into the battle aspects of the game.

The previous two posts have explained a bit how the units and the leaders function; all of this comes together when two armies meet on the battle map. The result is that the player is taken to the battle resolution window, where the player has the opportunity to set up the army and select what “encounter strategy” to follow. Depending on which side manages to grab the tactical initiative, this will result in one of four situations.

Perhaps the most common situation when two sides meet is that one of them withdraws from battle. In this period, it was almost impossible to force a battle on a reluctant enemy, and the same will be the case in Imperium. Withdrawal comes with a price, though; whenever a General withdraws from battle, it will affect the “imperial morale” of the empire and the prestige of the General. The price will depend on the size of the forces and the relative strengths of the armies involved. The people are obviously not going to be as concerned when a force of 4000 men refuse to fight with 40,000 men, as they would if the opposite was the case. This is one more mechanism which I expect will make skilled enemy commanders like Hannibal a nasty proposition in the game. Imagine the 20+ Roman units that you have finally collected down in Apulia suddenly running into Hannibal with his 16 or so units. You might not fancy your chances too much, knowing that a large proportion of your troops are still recruits, that Hannibal is the best Commander of the period and leads a veteran army… nevertheless, you will have to make a choice: do you run away, further depleting Roman morale and the prestige of your General? Or do you face down the Carthaginians on the banks of the Aufidius?

(NB: Imperial morale is an important aspect of the imperial system; somewhat similar to “stability” in the EU games. Time and victories raise morale; withdrawals and defeats lower it. Low imperial morale may result in the empire’s forces becoming demoralized, lowered force quality, and revolts in the cities of the empire. Imperial morale at its lowest will literally force a dynasty to its knees. Prestige is the reputation of a character in the game; a vital aspect of the game).

Ambushes are a special ploy; possible only in very particular situations; namely when one side significantly outscouts the other. Rough terrain (such as will be found abundantly in Spain) will improve the chances. Given the opportunity, the potential ambusher will then have to decide whether to ambush and what proportion of his force to dedicate to the ambush. An ambush is not automatically successful (it depends on the relative Guile skills of the two commanders), and a failed ambush could have serious consequences for any subsequent battle (especially if the ambusher has overcommitted forces to the ambush). This will hopefully mean that ambushes will tend to be small-scale, with the large-scale full army ambushes only attempted in especially favorable or desperate circumstances.

Defenders in appropriate terrain will be able to attempt to maneuver, to defend a river crossing, block a mountain pass, or to seize the high ground. In these circumstances, the attacker will have to make the choice of whether to attack at a disadvantage, or withdraw from the confrontation. The mechanism allows for Thermopylae-like battles, where a smaller force may be able to beat back a much larger force.

If there is no ambush or withdrawal, then a battle occurs. Battles can be either pitched affairs where both sides line up on a selected battlefield to fight it out, or it may be limited to skirmishing. Regiments can be assigned the skirmish role (manipular legions will automatically send forward their skirmish detachments); if the battle is limited to skirmishing, only these units will be involved in the fighting. In a pitched battle, all units (who have not been placed in “reserve”; i.e., out of the fighting) would be involved. Skirmishes can end up quite bloody, considering the limited number of units involved, but have the advantage for the losing side of being of less consequence than an open battle. This makes, of course, skirmishing a favored tactic of the weaker side in any conflict. Similarly, states with lots of light cavalry (that usually have high skirmish factors and quite low combat values) will also favor dealing death with a thousand cuts.

In the next and – I think – final post on this subject, I’ll briefly step through the combat mechanics that the game will  use to determine the winner and loser of a battle.

Over on Flash of Steel, Troy Goodfellow has raised some interesting questions in a response to the previous post. I shall have to pass over most of the discussion for another time, as it would involve a rather more detailed discussion of the game’s character model than would be useful at this time. It does however remind me of one relevant aspect of the character model which I left out of the previous discussion.

Like many other strategy games that model characters, the persons in Imperium also have character traits. I didn’t touch upon them much in the previous discussion, as I am still considering balance issues. I find the “this trait adds 1% to you gold” and “+1 to your command skill” type of traits extremely uninteresting; but at the same time, one has to be careful that the character traits do not overshadow the character’s other attributes.

If the right balance is found, however, the traits of a person would be one of the systems used to distinguish between commanders. A Hannibal, for instance, might possess the “Ambusher” trait that would make it easier for him to get chances to ambush opponents. Fabius Maximus Cunctator might be served by a “Cautious” trait, which would provide the reverse ability. A Julius Caesar might be distinguished by “Celeritas” trait, providing him with an initiative bonus and reflecting his famed speed of operations.

The subject of characters is one I shall return to, as they are central to the entire gameplay mechanic in Imperium. The next couple of posts, though, will return focus to the combat system.

The effect of commanders on battles in the ancient world is the subject of many books and varied opinions. In Imperium, I have preferred to go with the idea that the overall commander significantly affected the outcome of the battles. Thus, who commands the battle should have an effect on the in-game results, above and beyond the actions of the players. The player should notice, in short, the difference between a Varro and a Hannibal.

Persons in Imperium are equipped with five skills and three of them are directly relevant to the battle system. These skills are Command, Combat, and Guile.

When we consider the track record of the more capable (ancient) generals; it is easy to observe that one of the key abilities is their ability to throw the enemy off balance and dictate the pace of events. This is one of the way the Command skill is used in the battle resolution system; when forces close for battle, the relative Command skills of the generals involved are used to determine the course of events. Command skill also directly influences the in-battle combat value of units, reflecting the greater efficiency of competently commanded units.

The Combat skill is used to represent both the hand-to-hand fighting skill of the General, as well as his “Presence” on the field of battle. Classical warfare of this period had a tension between the ideal of the “Heroic Warrior” and the more responsible image – espoused by Polybius and Onasander – of the general as a “Battle Manager”. This tension is built into the Imperium battle system by the choice of where you deploy your General. Generals can be ordered to “Stay Back” (representing the battle manager), or take the front by being attached to a specific Regiment. Generals attached to a regiment boost the regiment’s combat value significantly (although modified by the Combat skill); however, they are also at a far greater risk of being killed or injured during the battle with potentially devastating consequences for the army. Generals who stay behind the battle lines can also be killed or injured in the course of the battle, but the risk is lower, and the immediate consequences for the army are not as serious.

The plan is that some cultures (for example Gallic culture and some variants of Hellenistic culture) may have a Heroic ethos, meaning that their armies will actually fight worse if the General does not fight in the front line (reflecting the cultural pressure). Whether fighting in the front line or lurking at the back like a smart Roman, the Combat skill will be important for the General, as it determines the likelihood of the general surviving episodes where his life is put into danger… such as of course in battle.

Guile is, of course, the classic ambush skill; it is used anytime base cunning is required. It is used to determine the success of an ambush. Thus a Roman general with a low guile rating and an army poor at scouting marching into a region occuppied by Hannibal with excellent Numidian and Spanish cavalry (superior scouting units) and a master of deception risks learning the same lesson about the art of ambushes in Imperium that Flaminius historically learned at Lake Trasimene.

Skills in Imperium range from 0-10; with most persons ranging in skill from 3-7. A historical Hannibal at his peak might rate a 10 in Command, a 6 or 7 in Combat (a 10 would be reserved for someone like Alexander the Great who led a charmed life on the battlefield), and a 10 in Guile. A Julius Caesar might be 8 or 9 in Command, 8 or 9 in Combat (Caesar effectively inspired his troops from the front line in times of crisis), and 8 or 9 in Guile. Persons also have traits, some of which are naturally going to be military in application.

I’m hoping that the above model will permit for a reasonable simulation of the key characteristics of ancient generals. It does achieve at least some of my objectives for the system. Skilled commanders will definitely have a significant impact on battles, making it possible for them to win battles even when outnumbered. It will also be possible for a general with brawn but little brains to win victories, simply through the effect of sheer personal combat prowess. Oh yeah… and facing Hannibal will be nasty.