Dynasties and Characters
The aspect that perhaps required the most work in Kingdoms was evolving the rather simple system of Empires’ generals into a much more developed one. The task initially seemed monumental because it’s no secret, and it’s even obvious, that players need to identify a game’s signature by comparing it to another. So, when you see a medieval strategy game announced, you wonder how it compares to Paradox’s Crusader Kings III, which is the current benchmark for many. Of course, it would be presumptuous to believe that Kingdoms could equal CK3 in every way, and that wasn’t the goal anyway, as what we wanted with this new game was primarily to create a military and economic strategy game, so we had to strive for an achievable goal.

Ebb and flow, one of the goals of the game. The changes in Byzantine dynasties are clearly identifiable here.We thus asked ourselves what was imperative and what was more about simulating the life of a ruler and their court (which is the core gameplay of CK3, and so not something we wanted to tackle). The choices are quite pronounced in Kingdoms, as everything revolves around the sovereign and their kingdom, and this is where we stop, by design. The court is defined in relation to them, and even though there are nobles who are not of their lineage, other characters can be entirely removed when the sovereign changes. There is thus no proper historical record of all the dynasties of the nations on the map; this is intentional, partly to maintain the focus on the main experience of Kingdoms, namely military campaigns, economic developments, and of course the rise and fall of nations, but also because trying to compete with CK3 on what it does best would make little sense.

Hats off to the artist who made 900 portraits by hand. No generative AI was used!To summarize, we aim to offer players a different experience, not a copy of another game. This is something to consider if you are designing your game: what are its identity(ies), its strengths (marketing department would name that Unique Selling Points or USPs!)? What does it want to achieve, what are you willing to put in the background to ensure the player’s experience is what you want? A game is not a hodgepodge of ideas…


Religions
Religious conflict, in the context of a game only (sic!), is good! It creates what we call fault lines and is thus a rather effective engine for generating tensions. In parallel, if you impose some disadvantages for attacking your co-religionists while offering bonuses for waging war on nations of another faith, you effectively recreate what happened historically. In Kingdoms, there’s no need to do anything for Christians in Spain and Muslim Taifas to go to war (though we added an event that unites them with the arrival of the Almoravids). The progressive deterioration of relations (which can be avoided if you have a treaty, as happened historically between some nations) will naturally trigger the algorithm giving claims of one nation to another, which will also cause a quicker deterioration of relations, leading to war!

This is a crucial point for good historical design: always try to ground your design elements to reality. When you start imagining mechanisms that are too convoluted and have little to do with what happened historically, you can be sure you will lose some players, and ultimately the situation you achieve will be implausible. Conversely, designing rules inspired by history that are logically linked will lead to a game that is essentially historical or plausibly historical. The beauty of these logical sequences is twofold. Firstly, these sequences are quite resilient against divergences (while not being totally constrained, it is possible for a player, with directed effort, to counteract them), so even if something does not happen historically, the system naturally returns to a historical state via the pressure exerted by different mechanisms supporting each other.

A Crusader detachment made up of 2 Danish Huscarls, Polish Armoured Nobles, and 2 Spearmen from Saxony.Secondly, a logical design inspired by history that is a series of small interdependent rules has the advantage of being easily memorable for the player. It’s a bit of “go with the flow,” an easy mnemonic for them. Just think that religious tensions degrade relations, that you claim the territories of nations you covet or dislike, and there you go, you know the rule without learning it (and you anticipate what will happen, which provides a geopolitical context) because everything seems natural: I play a Christian from Spain, I know I must prepare for a series of conflicts with my southern neighbors. And I also naturally understand that my Christian neighbors will also go to war, which will bring us closer together. From there, the player chooses their path(s), including some possible devious ones! (for example, being the best friend of a Christian neighbor, then completely absorbing them). The possibilities are numerous and replayability is improved.


An Engaging Experience
The last point I wanted to address was the computer opponent that players face and what is ultimately its role, as the adversary. Even though things may change in a few years with the emergence of generalist AIs, I realized over 20 years of developing strategy games that it is vain (or even vainglorious) to believe that a so-called AI (which isn’t an AI, of course, it’s a linguistic shortcut) could rival a player, even an average one. The game’s AI, an opponent of a few thousand lines of code, does not compete with a human brain of 100 billion neurons developed iteratively over tens of thousands of years, that’s obvious. As a result, one must then ask whether the AI can be anything other than a roadblock or a foil. But consequently, is the player’s experience of the game engaging and interesting? Not necessarily, if you quickly find yourself in a mop-up phase or a “push mindlessly my armies,” ad repetitam (and often ad nauseam).

The Papacy is AI-only by design. Achieving historical limitations and behavior for a playable Papacy would require extreme difficulty and months of work.A beginning of a solution can be found in the approach taken in pen & paper role-playing games (and I was a game master for many years in several campaigns!). The goal when you are a game master is not to make the players (your players) live a miserable experience that ends in mud, blood, and oblivion. With a clever mix of tension, twists, and also benevolence, you guide them towards epic adventures that end in a grand apotheosis. This is indeed the approach used by Rimworld and to a lesser extent AI Wars, both very good games with different profiles but sharing the concept of storytelling.

That’s why AGEOD’s games also take, modestly and in a measured way (I want to emphasize this!) this approach to AI difficulty levels. Some players struggle with the concept that not all game rules are symmetrical between human and computer opponents, and that even the default level, the balanced one (named so for a reason), makes adjustments, either for the computer or the player. The goal is to have a strategy game that offers challenge and tension, maintaining this for as long as possible. This doesn’t mean everything is leveled and neutralized to achieve this trajectory, but let’s say there are some rubber banding mechanisms to dynamically balance difficulty and obstacles, and there are rule adjustments.

Having the AI comment on your kingdoms is also part of the storytelling experience.But again, the goal is not to make the player’s life miserable, to punish them; that’s where the benevolent aspect, derived from role-playing experience, is a plus. When you are human, you don’t want your troops to be automatically controlled by your suzerain, and this is not the case in Kingdoms, and it is indeed a ‘human’ privilege that the computer opponent does not have. But similarly, the computer player will be less likely to face a coup from a disloyal peer because it does not optimize as well as a player, taking into account dozens of constraints, and the distribution of its peers. It’s also in some way pragmatic. You don’t need more vanity than necessary when developing a complex game. Do you really want to spend two more weeks of development (which will cost you the non-development of many other features or UI improvements, everything has a cost in time) just to boast that your AI is almost as capable as an average player (and no better) at assigning its governors, taking into account loyalty, weighted by the bonuses they provide, contextually in each situation? What does this ultimately change to the player’s experience if the AI still suffers civil wars? (and it will, because it does not manage loyalty rules as finely as the player, regarding who can be army commander or peer of the realm, but it’s a bit less tied to each peer’s loyalty). Not much, AI nations suffer civil wars, and this remains generally linked to loyalty rules, period. Equity and balance remain.I am well aware that this is also a divisive approach, but it is the result of many years of experience and hundreds of feedback and game analysis. For better or worse, it is AGEOD’s approach… While awaiting the supposed advent of super-game opponents driven by AI. Not sure players will appreciate it either, because it comes back to what makes a good game experience. Not necessarily a game where the ruthless, logical opponent crushes you, but a game where there is constant tension, twists, and perhaps, ideally, a great comeback, giving you memorable moments. Your best games are those that unfolded like a good movie tells you a story that keeps you glued to the cinema seat at the end… not a factual documentary, interesting, but a bit too conventional.

Until next time,