Just for the sake of convenience, here's a copy of Threetoe's post from the blog in case anyone else wants to get caught up quickly:
How hard a game should Dwarf Fortress be? The more technical among you have already found the text files that govern this. The various triggers that cause everything from dragon attacks to goblin snatchers are relics of a more primitive game where nothing existed outside the patch of mountain you dug into. Now the world is super-big, filled with all sorts of bad stuff waiting to find you. The triggers still exist, simulating the rumors of your wealth that the trade caravans spread to the outside world. However, nowadays, instead of generating an army from nothing, enemies are drawn from existing populations all over the world. The triggers are pretty much the same, but it is possible to do stuff like tick off the elves with a raid in order to get them to attack early. That might give you some control, but you are still on the same clock for everything else.
It's for the sake of the noobies that the triggers exist. No one wants to get overrun while they are still figuring out how to play in year 1. There has to be a default difficulty. We are in the process of logging the invasions versus different scenarios to calibrate this. Should it be based on population? Maybe traded wealth? Maybe you think you should have some time to prepare for the invaders with your hundred dwarves. Or maybe you want the dragon to come right after you buy out your first caravan. We are going to make this all customizable in a settings menu, but what about before you mess with that? How hard should the game be?
Sound off on the regular channels! We are monitoring you.
I kind of feel like he's talking about difficulty sort of in terms of combat/"world threat." In my mind, there's more of a built-in threat slider than a "clock" in the game in terms of how much wealth you generate over time, and how much you buy from caravans, because you have a degree of control over those things. If you're not ready for "big threat," you can take care not to generate too much wealth as you make your preparations for big threat, e.g. by using low-value materials or sticking to utilitarian items or the like, and also not to export too much valuable stuff. This lets you tell the game to some extent how much threat you think would be fun or exciting or whatnot at that time. To me this system is already really really elegant, one of my favorite in any game really, because controlling the amount of wealth you generate and deciding on your approach to the next caravan and so on is itself a kind of fun and interesting little set of games to play I feel like, so it's neat to communicate this larger gesture of "let's dial the threat up" or "not so much threat just yet" or whatnot through the means of that game. It also feels sort of like cool RP to me to do—laying low and building in secret, or being brash and confident and showy, make lots of fancy art and having a rowdy tavern and so on. Through this means, you can think about what sorts of dwarves these might be and why they might have come to this place and so on during the early game, so I feel like the "classic" trigger system is really nice both on a narrative level and on a mechanical level.
Of course, it tells the story of like, "word spreading about your fortress" in a pretty impressionistic way. I think it works really well on that basis, like on the basis of just giving you the visceral feeling of that happening, but I also do totally like the approach of having actual rumors that spread around the world and other simulated settlements and societies that they spread within and so on. That gives you a whole new perspective to look at the idea of "the story of your fortress" from, the sort of "macro high fantasy" angle, legends passed down through the ages and elaborate geopolitics and so on, and it's neat to imagine something on that kind of huge narrative scale playing out on as minute a level as one reindeer vs. beak dog interaction in front of you or whatnot. So I'm totally happy with the current direction.
There's an obvious sort of tension there with the old system of course, just because it's so vague, it just says like "(X) shows up when (a/b/c) numbers align" or whatnot, so it doesn't map very obviously onto a system that's trying to simulate large-scale sociocultural interactions. In that context, your fortress is just one settlement among many basically, and it doesn't really make as much sense to put you in this like super protagonistic role that the old trigger-based system has you in. The difficulty with that of course is that like, if you look at it from the perspective of vaguely medieval history or something, a "highly accurate" simulation might have 80 years of peace in a large city or back-to-back wars sweeping through a tiny town, in either case due to forces well beyond them. If you were just playing on the scale of one settlement in that kind of context, you would only expect to have an impact on large-scale patterns of peace and conflict in the world around you if you were very politically prominent relative to surrounding settlements. You would only expect to reach that kind of position well into the game, when it's less important for you to be able to carefully restrain the amount of threat you're under. So, from the earlier perspective of "the joy of playing to tell the game how much threat you want early on," this tendency in a "highly accurate global politics simulation" feels kind of awkward.
What I think might work in well in this context is to have every settlement follow the same set of triggers in the simulated larger world as in the player's fortress, or at least a similar set of effects. Like, a model of a small town that looks more like, "no one knows they're here yet, so they're safe," and where there is some early stage in a settlement's development where it is "lying low," and in a way that in the player's case they can kind of influence in various ways by adjusting their playstyle. I'm not sure the current set of what stimulates rumors quite does this, because it seems to center mainly on artifacts, which are produced at a relatively constant rate which the player doesn't have much control over. It does seem like it would make a kind of reasonably-fast linear ramp-up in difficulty over time, which would probably be all right, but I think it would be lovely to preserve the ability to "lay low" or "be brash and bring on the big threat" in the early game even within the larger political simulation. In a way I think this goes beyond "mere difficulty"—even for an experienced player, it has to do with the sort of story you feel drawn towards that time around.
To be honest, I kind of feel like it would work fine to keep the existing triggers as they are for the player's fortress more-or-less and just treat the other settlements in terms of the rough wealth they've generated over time and their population size. Then there could be a larger world with complicated large-scale stories playing out and the player fortress would fit elegantly into it in a way that preserves the lovely quality of the early game. I think giving a kind of "manual threat slider" in the settings is great too, as long as the game has any sense in the background of a "sliding threat scale," just because that gives another fun way to play the game, experimenting with the different threat settings or designing little "scenarios" you want to play out. If you're treating the other settlements similarly to the "old school" trigger model, it kind of implies that every town has a "threat level," so artificially lowering or raising any town's threat level (including the player's) doesn't seem too weird still from the perspective of the game design.
I guess, at the same time tough too, I don't necessarily think the triggers are the be-all-end-all of this kind of design for sure or something, I just do think they work quite well, specifically basing the threat level off of generated and exported wealth. That approach is nice in that it's worked into the game everywhere on a variety of levels, how you approach it has to do with your holistic sense of what your fortress is all about that time instead of just like "what the game thinks you should do" or whatever, and it's never exactly the same problem twice. Probably any system with those properties would be at least as good, if it turns out that other things about the larger-scale simulation fit in too awkwardly with the generated-and-exported-wealth-leads-to-threat approach. Wealth is nice because it's a number you can just kind of attach to everything freely and the player will accept it to some extent, because it has to do with how much things are worth in the
dwarves' world, so although we would expect e.g. gold to be valuable it's fine and even desirable to some extent to surprise the player here and there with how valuable something might be or not, so you can base your idea of the worth of everything largely on other concerns then seeming credible to the player. It's nice to have some idea that you can work into the game everywhere where the player won't have too strong a set of fixed ideas about how the thing should be. I'm trying to think of other aspects of the game that are similar, but population and created artifacts and so on, while similarly abstract in scope, are much less under the player's control. There's the biome of course, but even that is a part of your thoughts about how you will generate wealth, because it determines the medium upon which you'll do this.
Those dwarves, so ruthlessly industrial… XD