Flash of Steel header image 2

Developer Interview: Michael Akinde

November 21st, 2005 by Troy Goodfellow · 3 Comments · Ancients, Imperium, Interview

In the first of what I hope will be many interviews with independent strategy game developers, Michael Akinde agreed to answer some questions about Imperium: Rise of Rome, game design and the place of history in the game business.

This interview was originally going to be the first thing published on my new domain, but my intended host had some issues that need to be straightened out. Here it is in any case.

Akinde is a regular reader of Portico and occasional commenter. Hopefully he can answer any further questions any of you might have about Imperium.

(My questions are in bold, his answers follow.)

What gaming experiences have been the biggest influences on Imperium?

A tough question to start out on. I have played a lot of games, and try to learn something of both the good and the bad from each of them. But if I were to pick just one gaming experience, it would probably be playing the boardgame “Republic of Rome” in two-player mode. It’s interesting, because although you’re playing against “dumb” solitaire rules controlling the other factions, it still manages to produce a very tense and exciting game. The idea for doing Imperium originates in games of the Avalon Hill classic.

What about the ancient world makes it an attractive game setting for you?

I don’t know if I consider it an attractive game setting as such. I enjoy ancient history, and that more than anything else drove my decision to do a Roman-style game. Another reason, perhaps, is my opinion that there is a lack of games providing a reasonable model of ancient warfare.

For example, consider “tech trees” in the ancient world. The pace of modern development makes such a game mechanism easy for us to accept, but it makes no sense in the ancient world. To put the issue in perspective, one need only observe that a Roman legionary of the Punic Wars time-travelling into the future would have been able to fight at no disadvantage on the German frontier four hundred years later. Despite this, almost every strategy game of the ancient world contains some form of tech trees.

I realized at some point that if I was ever going to have a chance of playing a strategy game of the ancient world the way I think it should be,I’d have to build it myself.

What has been the hardest part of the game design so far?

Striking the balance between complexity and simplicity. The AI has been a huge concern for me right from the beginning of the project, and it has been a factor in all of the major design decisions. Achieving a competent, non-cheating AI is really only possible if the game mechanics are simple.

Making it historically sensible and compelling to play at the same time, however, is no easy task.

Have you had to cut anything that you wish you had kept?

Not really. I have had to scale back on the diversity of government types in the version of the game I plan to release, but assuming the game has even a little success, I would hope to be able to support the game with some fairly frequent updates adding the extra governments. It’s an approach that has been carried out pretty successfully with Stardock’s “Galactic Civilizations”, and has the added benefit of inhibiting piracy (who wants to hunt around for a new “cracked” version every other month?).

If there is anything I regret, it is probably that I haven’t cut more. The project started out too big to begin with, and the long development time is a direct result of that. Developing and publishing the game in smaller increments right from the start would probably have been a better idea.

Is the decision to make the game turn-based rooted in personal preference, or are there practical reasons for going that way?

It is purely practical. Making a game real-time increases the demands on the graphical interface (lots more animation) as well as the complexity of the AI. Then you have the AI competing for CPU cycles with the graphics, and that invariably means that certain kinds of strategic analysis become impossible. Another is that real-time game mechanics opens up the window to all kinds of “gamey” strategies that an AI can hardly hope to emulate – for instance, consider the many subtle ways in which the precise timing of moves can be used to frustrate the AI in Europa Universalis. Real-time has its advantages, but compensating for its disadvantages would have added even more development time to what is already a pretty huge project..

One of the big draws for me is your attempt to make character and personality of generals and kings shape the game. How will this work?

One could define history as an account of events caused by generals and kings, and this is one of the basic ideas around which the game mechanics revolve. History doesn’t just happen – it happens as a result of the actions of people.

In Imperium, the idea is that you get to control a group of these people (a faction). One of these will be your faction leader – essentially the player’s avatar – and how your faction does depends a lot on his actions, and how the other characters react to them. There is a pretty rich personality model built into the characters, and though I will not have the
ressources to fully explore all the gameplay possibilities in there, I hope it will be possible to provide a pretty unique gameplay experience with it. I would like the player to (at least occasionally) forget the spreadsheet. I will have succeeded if the player upon considering Gaius Servilius Trebonius thinks “Do I dare send that cranky old git to Cilicia?” rather than “He’s a 4-star general, excellent!”.

The other aspect of the character model, is that they are my vehicle for driving the “historical narrative”. I am not a big fan of event-based gameplay, and other than a little generic randomness (bad harvests and the like), there are going to be very few “historical events” in Imperium. Anything worth representing in the game can happen in regular gameplay, provided the right circumstances arise. For example, the great slave revolt of Spartacus is not coded to happen in 73 BCE – but slave revolts can happen anytime the slave population in a province gets high enough (poverty in the countryside helps as well), and there exist slave leaders with sufficiently high charisma and intelligence (to become more than just bandit chiefs). This of course sets the stage nicely for historical characters such as Eunus, Salvius and Spartacus.

Since historical characters tend to be born at historical times (unless one switches off this option), this means that the revolt of Spartacus probably will happen around 73 BCE – except, of course, if the conditions of the time make a major slave revolt unlikely. It also means that you could have some completely unknown (i.e., randomly generated) slave leader spark of the great slave revolts of the game. Essentially, the historical characters in the game are “personality-coded” to push the historical narrative in certain directions. But because they work within the framework of the game (rather than imposing changes from the “outside”, as an event system tends to do), the historical narrative will hopefully maintain cohesiveness.

Has it been hard to get the word out about Imperium?

I don’t think so. I can see that there are a fair amount of people following the development of the game, and considering I haven’t done very much to bring them in (no flashy screenshots, for example), I am pretty content with the attention the game has got.

Where do you draw the line between historical simulation and history game?

I don’t know that I actually draw much of a line anywhere. I guess it depends on how you define the two.

A simulation is by definition an abstraction of a real world situation, and a historical game is the same to me, except that you, as the player, can tweak the variables to see where it goes. Where historical games differ, is what aspects they have chosen to focus on in their simulation, and how conscentiously they deal with those aspects of the “simulation” that they are not focused on. Of course, it tends to be a problem that the abstractions chosen are often pretty poor from a simulation point of view – i.e., they lead to historically implausible outcomes even for the core game focus. But that is an entirely different issue.

Of course there are a (very) few games which are nothing more than a game mechanic dressed up in historical “clothing”; but I tend to not consider those as “history games”..

How have you managed the challenge of independent development?

I just try to do a little every day – even when I don’t have time or enthusiasm for the work. That is really all there is to it.

Is it too early to get a general idea of the release time frame?

I do have some idea of when I would like the game to be done, but even after the game proper is done, there are still matters such as developing the final graphics, testing game balance and AI that need to be finished. Ask me again when the game has been signed with a publisher.


3 Comments so far ↓

  • roboczar

    Vry, very cool. I’ve been looking forward to this title for a while!

    Good job on scoring the interview Troy, I hope there are many more to come.

  • roboczar

    um, also, don’t put too much credence on the sense of security that Stardock’s method seems to demonstrate.

    GalCiv and its subsequent patches/expansions were quickly cracked and distributed by many groups because the installation files were saved locally. You would be far more secure by adopting the services of a firm that handles distribution in a Steam-like fashion. While Half-Life 2 was cracked fairly quickly, it was a very involved process, and Valve has virutal omniscience over the Steam network and can identify clients that don’t ‘play nice’ and subsequently blacklist them. It raises the risk bar for those that would attempt to crack the software.

  • Michael A.

    Thanks for the comments.

    With respect to the “Stardock System”, I have no particular intention of signing with them at the moment. However, their philosophy of releasing new material (rather than just patches), is IMO a nice way to keep a game fresh and slow down casual pirating. New games tend to get cracked immediately – however patches and patch-like expansions don’t have as high a priority and are often at least a couple of weeks behind. The actual technical solution to this is another matter. :-)