It’s What You Do With It That Counts: Factual accuracy and mechanical accuracy in Crusader Kings II
Guest post from Robert Houghton. Robert mainly researches urban and episcopal history in Italy in the eleventh and twelfth centuries, and also has a strong interest in the portrayal of the Middle Ages in the media. He currently works at the University of St Andrews and as a research consultant for Paradox Interactive.
A few weeks ago, Paul posted about the mechanics of historical computer games and how it’s more important to have accurate mechanics than it is to have accurate facts. I’d like to elaborate on this with the example of Paradox Interactive’s Crusader Kings II, a game which I have had the pleasure of working on since January 2013.
Crusader Kings II has sold over a million copies since its launch in February 2012. The game puts the player at the head of a medieval dynasty at any point between 15 September 1066 (just before the Norman Conquest) and 29 May 1453 (the fall of Constantinople) and, like other historical grand strategy games, gives them extensive controls over many social, economic, political and military decisions. When the selected character dies, the player takes control of his heir. If you run out of heirs, it’s game over. Expansions have extended the game geographically into Africa and India and chronologically back to the age of the Vikings and, in a forthcoming expansion, Charlemagne.
The most obvious thing that sets Crusader Kings apart from other games of its genre is its sheer scale. The game allows you to play as any head of an aristocratic family: from the mighty emperors of Germany (Lord over several dozen dukes and counts), to the humble jarl of Shetland (vassal of the Lord of the Isles who in turn was vassal to the King of Norway). As any day from the period can be selected as the start date, this requires a massive database of historical figures.
The ongoing development of this database is powered in part by a sprawling international team of volunteer beta-testers and research consultants. This group comprises interested academics, part-time historians, students and gamers who support the research behind the game. Data for this database are compiled from research and class notes, consultation of books and websites, and, occasionally, original research into historical sources. This is used to create characters, define local, regional, national and imperial borders, and construct family trees and networks of vassalage. The end result is an elaborate social/political/economic database.
But this is a game, not an academic research project. Crusader Kings is not meant to be a wholly accurate record of the medieval world. Certainly, accuracy is attempted wherever possible and where it does not interfere with game design and balance. But, ultimately total accuracy is often simply impossible. There is insufficient evidence. Rulers of kingdoms and empires are relatively well known for most of the period, but there are still some significant gaps— particularly in regions and periods with few surviving written sources. As you go further down the social pyramid, sources become increasingly tenuous and fragmented. Solid knowledge becomes informed guess, which becomes something that “sounds about right”.
But unlike academic histories, the game must present absolutes. Precise dates are necessary for the database files. There is no way to represent conflicting theories about the year of death of a Viking warlord, for example, or the specific familial relationship between Ethiopian kings.
This all sounds very much like a focus on the accuracy of the game’s historical facts rather than the accuracy of its game mechanics. However, for all the effort poured into the historical database at the heart of Crusader Kings, this can only serve as the starting point for each game played.
The second the game begins, divergences from history begin to occur. Crusader Kings is, indeed has to be, inaccurate. This is where the game mechanics kick in, and where, in my opinion, the game really shines.
The traditional game mechanics of grand strategy games: constructing buildings, researching technologies and raising armies, have been augmented with extensive role-playing elements. The player has control of the actions of his chosen character and can influence, but not dictate, how this character, his (or, occasionally, her) court and relationships with the wider world develop. Although abstractions have to made for the sake of playability, the role-playing elements make this feel like a game built to represent the middle ages from the ground up, not a generic grand strategy game with a medieval coat of paint.
For example, your character may throw a feast for his vassals. Rather than simply providing a flat relationship bonus, this triggers a series of complex events and decisions. Will the player spend lavishly on food or try to cut costs? Will he hire the travelling band of musicians or the fire-eaters? This will effect what happens at the feast: spending too little on food may lead to attendees complaining about the quality of the fare. Musicians may appeal to some characters, while others would prefer the fire-eaters. Beyond this, there are chances for all attendees to develop friendships and rivalries or to seduce other characters. This can all have immediate and long term consequences for the game. It is entirely possible to spend more time on these role-playing elements than in the running of the kingdom.
Crusader Kings avoids most of the common gameplay trends present in other historical grand strategy games. Conquest is not the sole purpose of the game: acquiring “prestige points” (through acquisition of titles and power) is the only stated goal. Beyond this, players are left to pick their own challenges. History is not determined solely by the heads of civilizations: admittedly, it still rests in the hands of ‘great men’ but does at least supply a much broader range of great men.
Furthermore, cultures change and adapt to the world around them: the norse will eventually start to identify as Danes, Swedes and Norwegians while those who have settled elsewhere may become Normans for example. Rulers come and go with their own strengths and weaknesses: the player doesn’t simply control one undying totemic figure [unlike in Civilization -Ed.]. Crusader Kings builds on a solid historical database to create a working model of medieval society.
Realistic or Accurate?
Crucially, the game is, by necessity, inaccurate. But, it is not wholly unrealistic. This distinction is important. Harold crushing William at Hastings would be inaccurate. Harold, with his army of elephant riding mercenaries from Bengal, conquering the Russian steppes would be unrealistic. [Though awesome! –Ed.]
The former reflects good mechanical accuracy while the latter does not. Happily, no matter the start point, Harold’s victory at Hastings is much more likely than his pachyderm divisions wintering in Moscow. The latter should only be possible through human intervention: such an unusual achievement would not be sought by the AI (or most players) but is nevertheless possible for the player, albeit requiring a reasonable amount of effort, quite possibly through the abuse/manipulation of game mechanics.
I’d like to conclude with two thoughts:
1) The ability to resort to fiction to fill in gaps in the historical database (albeit in a realistic manner) allows the creation of a more complete world than is possible solely through the use of the sources, which is absolutely necessary for the game to work. While the game’s database couldn’t hope to pass muster as an academic work, it allows the creation of a fully functioning, if simplified, model of the medieval world in a manner which can’t be produced by relying solely on the sources. This is immensely useful for teaching purposes and could be of auxiliary use in suggesting new avenues of research.
The models and mechanics used to create these games represent simplified theories about the underlying structures of medieval society. As such, they can potentially be used to support theories about how the medieval world worked. Of course, the games are not meant as simulations and can never replace academic study. That said, they could be useful tools to augment these studies.
2) The same issues which influence factual accuracy also influence mechanical accuracy in all historical grand strategy games. In fact, these issues can be even more pronounced. While dates, names and cultures may be debated, rival theories about how society functions are much more complex and fiercely disputed by academics. The game designers have to pick one of these theories to follow, rejecting all others. The need to create artificial precision of facts and mechanics means that the games give no indication of academic disputes over these facts and mechanics. These disputes and uncertainties are a central aspect of the field and the inability of these games to present these debates is perhaps their greatest weakness as learning tools.© Copyright 2014 Robert Houghton, All rights Reserved. Written For: The Public Medievalist