February 2018 Update
In this update we'll be focusing on two things: the basic design thinking behind the diplomacy system, along with showing off some new elements of the game from a recent playtest of mine in screenshot-form.
The post started to get a bit long, so I've decided to save the specific details as to how diplomacy will work (e.g. Relationship Levels, Global Reputation, Leader Personality Traits and Interaction Types) for the next update. We'll start off though with a high-level look as to the challenge of diplomacy in a complex strategy game.
What "is" Diplomacy?
Diplomacy is one of the biggest challenges in the strategy space, in large part because it's trying to simulate something that's hard to wrap your head around even in the real world.
There are some basic tenets that people agree on when it comes to good military strategy: divide and conquer, pay attention to supply, hold the high ground, etc. But what does "good diplomacy" look like? Sometimes negotiating averts a major war, while other times it simply brings "peace in our time". What looks like prudent flexibility to one can be seen by another as an unforgivable betrayal.
So, yeah, a tough thing to model!
Of course there are elements we can try and incorporate such as personality, trading, making promises, punishing liars and traitors, etc. but it's much harder to simulate all of this than a simple resource-based economy (and even those are hard to get right). Is there room to make something really nuanced and revolutionary here? I think so, but probably not as just one feature among many in an already-complex game.
A few weeks ago I asked on Twitter what people thought made for a good diplomacy system, and I received a lot of good answers. There was certainly some common ground, but the biggest lesson I took was that there was no general consensus - I think mainly due to the challenge I spoke of above.
Another challenge is that diplomacy is meant to simulate the nuance of human interaction. We're not necessarily trying to represent systems here, but more, well, feelings. Alas, this isn't really something that game AI is up for at this point in time, in large part because it's AI, and there's nothing you can do as a developer to convince players otherwise. Regardless of individual moves, it simply feels different playing against a computer. That is starting to change with the kind of work Google's DeepMind has done with AlphaGo, but that is the crown jewel of a multi-million dollar research studio on a game whose rules have been fixed in place. So we've got a long way to go before the 4X genre will be revolutionized in this way.
The AtG Diplomacy Design Pillars
So what are we doing in AtG then?
The focus is to come up with an approach that does something interesting and new while most especially making sure to avoid pitfalls of past games, and with that goal we've settled on three fundamental pillars.
Distinct, Predictable Personalities
"Oh wait, I know this guy… Awww man."
The biggest problem with most diplomacy systems is that they're too random. While there are probably always well-intentioned rules under the hood which enables AI players to reevaluate their situation and change their minds when it makes sense to do so, in reality this usually ends up turning into, "AI declares war, then asks for peace 10 turns later, then declares war again 10 turns after that".
We'll specifically be avoiding this pitfall in AtG in a couple ways. First is through a focus on Personality Traits. If Attila with his 'Aggressive' Trait finds you nearby then you can be pretty sure war is coming soon (unless, of course, you bow before him and give into his mostly-symbolic demands that you know will soon be on their way). Sometimes war will be a good idea for Attila. Sometimes it won't be. But most important of all is the fact that he wants it. Not every Leader will be this extreme, of course, but it's important to know what you're getting.
Tough, But Clear Choices
"Ugh, I was trying to be friends with both of these guys…"
A common problem I see with diplomatic systems in other strategy games is a focus on minutiae, particularly on the trade front. Having a really complex trade system seems like a neat idea, but it usually ends up turning into a game of, "always trade X for Y, then try to exploit the AI out of all their money". In AtG trade will not be a focus - in fact, it won't even be present at all. Instead, the focus will be on the relationships between leaders.
One leader might demand that you choose between him and another leader. In line with the first design pillar though this should always be somewhat predictable - if you try and be friends with a leader with the 'Jealous' Trait you know that also trying to be friends with someone else will trigger him to challenge your loyalty.
Most of the time it's going to be impossible to make everyone happy and keep all of your stuff and your pride - but that's part of the fun of figuring out how to best adapt to and "solve" diplomacy.
A Few Basic, But Powerful Player Knobs
"I'm going to tell this guy to pound sand!"
The final pillar is based around the concept of player agency, and ties somewhat into pillar #2. Players should still be able to steer things diplomatically, even if a lot of the game will be responding and adapting to the other tribal leaders.
Sometimes you just want to vent frustration at someone, and in AtG a lot of the time you'll be able to get your way. A leader with the 'Meek' Trait might always give in to the first Demand For Tribute, making the strategy here more about optimizing what to ask for and when. Demanding something from 'Proud' Attila might be guaranteed to fail every time, and draw his wrath - but in return your Global Reputation might receive a large boost, allowing you to build a friendship with another leader.
In the next update I'll go into more specifics as to how diplomacy will work in AtG (the 7 Relationship Levels, how Global Reputation works, the list of possible Interaction Types, etc.), but for now we'll wrap things up here. But before we go we'll go over some new screenshots and show off some of the new recent additions to the game.
The first couple images here show off the new tutorial system.
It's mostly made up of basic popups triggered by particular events (e.g. if you're running out of food), but the cool part of the system is that most of it is optional. This supplements the fancy tooltips-in-tooltips system we started working on early in the project, and together should provide a much smoother on-ramp into the 4X experience than any previous title.
The system is also cool in that the tutorial messages themselves can be nested multiple levels deep.
We've also made sure everything is accessible in one place, just in case you want to go back to something later, or maybe turn off the tutorial system entirely and explore the in-game help at your own pace.
You can access this screen either by pressing the '?' button in the upper-left or by pressing the ? key. Not particularly fancy, but it gets the job done!
Speaking of ways to make the game easier to play, I may I've touched on the 'Notes' system in the past, but I can't help but show it off here now, as it was a really helpful feature in my latest playtest.
Right-clicking on any Clan Card brings up a screen which allows you to attach a colored note to the bottom.
This feature is rather handy, as it makes it easy to keep track of who you want to do what, something that's pretty important in a game in large part about managing Clans! It's especially useful when you have to stop playing for the night and would otherwise have no clue what you were up to the next time you pick things back up.
It's also possible to write notes on the map itself in order to keep track of spatial information, e.g. where to construct that Logging Camp or cut a path through the forest in order to make it easier to get around.
The last screenshot I'll include shows off the new 'Declare Kingdom' button you might have noticed in one of the previous screenshots. It doesn't show much in and of itself, but I promise the button does work! Just need a bit more Parchment...
These screenshots are actually from a pretty interesting playtest that I've been writing up notes for. I'll probably compile them into a future update post giving a more in-depth look at how the game plays out.
This was a pretty tough and interesting game where I found myself in the far north without many Resource Deposits but plenty of Forests to harvest Timber from. It also brought up some interesting design questions (How accessible should Resources be? How much variance between starting locations should there be?), so it would be a fun game to explore in more depth.
That's it for now. Thanks again for reading, and we'll be back with another update soon!