Jon Shafer has just posted an extensive discussion of his original design for Civilization V and what he changed for his new project, At the Gates. It’s a great read for Civilization fans, and for anyone interested in the design of grand strategy games. Keep in mind that Jon left Firaxis before a multitude of patches and the massive Gods & Kings expansion substantially improved the gameplay, fixing some of the issues he mentions. In the rest of this post I’ll highlight and comment on some particularly interesting passages.
Out of all aspects of Civ 5 that I was involved with, I’m particularly proud of what our team accomplished with the UI. […] My one disappointment with the UI was the general lack of “power features” tailored for hardcore fans. […] One of my early goals was to have an alternate “expert” switch that you could flip, adding a significant quantity of detailed information to the screens and mouseovers.
Completely agree on both accounts. I love the Civ5 interface but some things are needlessly clunky when you already know your way around – why do I have to re-enable the city build queue every time? Patches did add more detailed information, though, and Jon promises that At the Gates will ship with an “expert mode” out of the, er, gate.
Diplomacy & AI
My original goal was for the AI leaders to act human. But humans are ambiguous, moody and sometimes just plain crazy. This can be interesting when you’re dealing with actual, real humans, but I learned the important lesson that when you’re simulating one with a computer there’s no way to make this fun. Any attempt to do so just turns into random, unproductive noise. […]
In Civ 5, you might have been lifelong allies with a leader, but once you enter the late-game he has no qualms backstabbing you in order to win. With this being the case, what’s the point of investing in relationships at all?
That’s an old problem with strategy games that feature any sort of diplomacy. AI players realistically backstabbing the human player might be aggravating enough, but in practice the AI’s evaluation of strategic threats and opportunities is so unreliable that the betrayals are usually random rather than realistic. Civilization fans are well-acquainted with AI players who love to declare suicidal wars, yet refuse to exploit profitable opportunities suggested by the human player. This results in a single-player experience that’s very little different from permanent total war against all AI players.
Jon’s solution for At the Gates is obvious, although perhaps somewhat unsatisfying: step back from attempting to create artificial human players, and turn the AI players into predictable gameplay systems instead.
Like other 4X games, diplomacy in ATG is built around your “relations” metric with other leaders. But compared with Civ 5, what goes into that number and what it does is very clear. For example, if you’re at -5 with a leader, he’ll never trade with you, while at +10 he’ll always agree to help out in a war if requested. Rather than trying to decipher what the RNG-based AI is “thinking,” your objective is instead to find as many ways as you can (afford) to boost that Relations number. Once you’ve done so, a variety of options for how your new friend can assist you become available.
Similarly, the overall AI in At the Gates should be stronger than in Civ5, by way of being both simpler and less random.
Another problem with my AI was the randomness, which is something I’ve already talked about at length. The computer opponents were weighted towards a variety of possibilities, with a healthy serving of RNG (random number generator) on the side. This meant they floated from one “strategy” to another without any real cohesion behind those decisions. This approach is nice in theory, but if you want a strong AI there are times when you need to force it to behave in very specific manner.
Once again, I fully agree with the diagnosis of the problem, at least seen from my outside perspective. It was painful to watch the Civ5 AI randomly shuffle its units back and forth, clearly attempting to do something sensible but never quite succeeding. Much tweaking of probability values in subsequent patches has eventually improved the situation, but it’s obviously better to build an AI that doesn’t need such lengthy trial-and-error tuning in the first place. The possible downside is an AI that’s so predictable as to be boring and easily exploited; we’ll see how At the Gates manages to steer clear off that cliff.
Resources & Happiness
In Civ 5, players ended up with easy access to a bit of every resource and there was almost no reason to trade. In the real world, swapping goods is worthwhile because of the effects of supply and demand. In Civ 5 there was almost no demand since you could be virtually self-sufficient. This will be completely different in ATG, where the threat of critical shortages will always be right around the corner, and bringing in much-needed resources via trade might very well be necessary for survival.
My removal of the health system in Civ 5 also had repercussions elsewhere. This greatly reduced the value of non-strategic resources (like wheat), and in retrospect it’s clear that I didn’t manage to fill that void with something else. ATG has far fewer resource types than Civ 5, but the ones which do exist are all very important. The map is absolutely vital in a 4X game, and that needs to be the case for everything on it as well. If you see something on a tile and think it’s not a big deal, that is a flaw that needs to be fixed.
This is sadly all too true. Luxuries are virtually the only kind of resources I ever trade in Civ5 because they are the only ones whose absence is really painful. Patches and the expansion made other resources somewhat more valuable by increasing their bonuses with certain technologies and buildings, but they are still nowhere as critical as they were in Civ4 or Civ3. It’s good to hear that At the Gates will revert to those earlier designs.
Carrying forward lessons from my experience with global happiness, ATG is much more freeform when it comes to expansion. There are factors in the game which discourage mindless spamming of settlements, but none of them are as heavy-handed as exponential maintenance, corruption or empire-wide unhappiness.
I’m ambivalent here because I’m one of the apparently few Civilization players who don’t want to build the biggest possible empire on the biggest possible map. I prefer managing just a few cities on a small map. Still, as long as that strategy remains viable in At the Gates I’m not complaining.
My removal of the research/commerce/culture sliders also came with positives and negatives. I’ve always found fiddling with sliders in strategy games to be boring busywork, and in that sense I don’t miss them. But the sliders also had a hidden value that I didn’t realize until later – they gave players the ability to shift directions at any time.
I never liked those silly abstract sliders that miraculously transmute science into gold, or vice versa. Allowing the players to change direction is a worthy goal but I hope whatever implementation At the Gates ends up with is better than a return of global sliders.
Policies & Combat System
Jon concludes with a discussion of policies – everyone loves them (me too!) and they’ll return in At the Gates as part of the new Romanization perks – and finally, the combat system. I think his criticism of Civ5’s one-unit-per-tile system is overly harsh. Just as I never liked managing tons of cities, I also never liked how earlier Civilization games made you incessantly spam units. Jon cites lower unit counts and longer build times as necessary disadvantages of 1UPT, but from my perspective they were great improvements! At the Gates returns to unit stacks, but all hope is not lost because it will also have a fairly brutal supply system that should hopefully prevent a return of the dreaded Civilization killer stack.
2013-02-16: Jon Shafer’s follow-up article, Building an Empire Builder, discusses the creation of 4X games in more general terms, especially the importance of random maps and the technical and artistic challenges they pose.