[The Leaders of Civilization](/blog/syph/civLeaders)
The final major piece that I want to talk about are the leaders of Civilization. At first glance, they seem sort of like the only way that the game could have been done, but there are a couple of things implicit to this implementation that are worth considering.
The Starting Point
The most important decision is to have all of the players start in roughly the same position (difficulty bonuses notwithstanding). This adds a sense of fairness to the game and so adds to the sense of achievement when you do well.
It’s also a nod to the board game roots of Civilization and makes multiplayer a much easier proposition.
The Fallen Empires and Upliftable Species of Stellaris make for an interesting alternative. Players that enter the game at a substantially higher or lower level than the other players certainly texture the experience. However, with both of these cases, the difference is so large as to substantially constrain the actions possible with those players.
The asymmetry of different leaders having different special abilities is key to forcing some variety in player strategies. As is, games often feel the same in terms of what you build and what you research. Through leaders that nudge players toward certain strategies, the game naturally changes from playthrough to playthrough. As Genghis Khan, I built more horses than I ever had before and the combination of missionary and conquistador for Phillip 2 made for a unique game.
Additionally, it helps substantially with the fiction. Leaders based on real world people have expectations that come with them. Given how different Cleopatra and Teddy Roosevelt were in real life, players expect them to be different in the game as well.
I think that a fundamental fact of present-day game design is that you cannot get AI to do a good job at mimicking real people for non-trivial games. For Civilization, it’s clear that the AI often just doesn’t understand the situation that it is currently in and doesn’t know how to extricate themselves from tough spots. There are certainly a number of changes that can be made to significantly improve the intelligence of the computer-controlled players. However, the larger issue is that they are simply asked for too much.
The first thing that the AI has to achieve is fun for the player to play against. Challenge is an important part of this for a lot of players, but the two attributes are not interchangeable here. There are also players who like beating up on weak AI or just don’t want to have to deal with the AI as they build the empire that they want to see.
The other important thing for the AI to achieve is transparency. For a game like Civilization, a highly random AI just doesn’t feel fun as it invalidates the long-term planning core to the game. The agendas of the leaders of Civ6 do something toward this, but they are still likely to do things that feel out of character, like declare wars of opportunism.
These wars are actually a bigger part of the game than they should be, so I’m going to go a little bit deeper into them here. Players actually are also often willing to declare war on a militarily weak neighbor that they’ve had hundreds of years of friendliness with. When the player does it though, the chain of logic that leads to the action is understood. As the player doesn’t have the same level of insight into the mind of the AI as they do into their own mind, it just seems capricious when the AI declares a surprise war. It also just doesn’t feel like the leaders that you play against should be as focused on victory as you are.
This may seem small, but I think that it was actually a really important part of the Alpha Centauri experience. Many of the technologies had flavor text with quotes from the leaders of the game. This does a lot to add personality to the game by giving the leaders more space for expression. It also makes the game feel more interconnected and so more real. The depth of the immersion of Alpha Centauri is unparalleled and the key to why it is my favorite game of all time.