Well...
1.004 was a disaster, wasn't it? Sorry about that.
1.005 to the rescue.
Meanwhile, if you look at the 1.005 notes, you'll notice that there's still an unresolved edge case; a remainder, if you will. Various players have pointed out a few issues at an advanced level of play, and I've been mulling these myself, and am going to make an outline here of what I think the salient issues are.
Potential Problem SpotsTo some extent, the crazy nature of this game is going to guarantee that some strange edge cases come up where the game is too easy. The most specific sources of this (but this is by no means comprehensive):
1. Mythological creatures are much stronger than human ones. Thus without an incentive to build human units, the tendency to just use the former arises.
a. However, the whole point of the mythologicals is that they are stronger than humans. Nerfing them would make the game less fun and less balanced for the regular gameplay style. Plopping down a minotaur to alter the tide of battle means nothing if the minotaur is not extra strong. And that's rather core to the game, eh?
2. In various ways, there are all sorts of clever ways you can create choke points. Whether it's hiding towns behind other towns, or making thin mountain passes, or smiting tiles, or... whatever... these have a lot of potential for abuse. But they also have a lot of potential for fun, and feature prominently in many strategies.
a. You know, even for me, I've remarked many times before that I am a natural turtle when it comes to strategy games. I want a front, and I want some protected behind-the-lines areas. With the rules as of 1.005, you can do a lot of that sort of thing, but not to an extreme degree... most of the time.
3. The god powers are insane, and are meant to be insane. Some of them, like the Skadi ability that makes all those mountains, implicitly cause all sorts of choke points and even complete blockages. I wouldn't want to nerf that into oblivion, because where's the fun in that? This is a rather subversive game when it comes to the genre at hand, and shaving off all the craziness just pushes it back into the land of ordinary.
4. The problems arise when it becomes trivial to solve all your problems with such tactics. Booooring.
Lack Of Consequence Is The Real ProblemIn AI War, the game would be trivial if you could just take every planet and the AI did not get stronger in response to that. If that were the way the game played out, it would just be one long, uninteresting grind. Instead there's all this grand-strategy goodness with hard choices to make as you try to achieve your goals without pissing off the AI too much.
Skyward Collapse is a bit of a different beast, in the main. It encourages you to deal with a swirling mass of chaos, but your only real goal is to tame that. That is super fun (and funny) when you are learning or when you have a challenging scenario, or for the first X amount of time in a particular savegame. But once you have it tamed, then it's just basically "click through for a while until the end, or possibly go for some higher score goals." The fact that you can tame the chaos at all is something I'm not thrilled with, and the changes I've made thus far (such as the 1.004 ones) were basically aimed at making it un-tameable.
The problem with that thinking, of course, is that taming chaos is
fun. And it's satisfying when you do. And it's impossible to stop. There is always someone too clever for the system, who will find a way around it. The better solution is to make them not
want to find a way around it. Either by providing rewards for not doing so, or by providing some sort of counter-balance to the mix.
But that has to be done with care. Once you've tamed the chaos, and, you feel like you've won, it's not fun to keep playing a savegame past that point. Okay. So that causes the endgame of some savegames to be very boring... while others are an exciting and nail-biting experience until the end. How to solve the former without making the latter impossible?
Endgame Reward StructuresI'm not sure how possible it really is to provide rewards that are compelling in these circumstances, unless the potential for further hilarity and "interesting things that crop up only near the end that you don't want to miss" is there. In other words, if you missed some things that only happen in the last x% of the game by stopping early, then that would be interesting. If there was the "coming of Ragnarok" at the end of the game, and you had to basically survive Armageddon right at the end, that would be interesting. Then if there's some point where you've tamed the chaos, it's not boring: it's an opportunity to improve your position to deal with Ragnarok. And if you smash up the Armageddon stage of the game, well -- that's both short and satisfying, so hooray.
But! Those end-game reward structures are also hard to balance, and I think some folks would still find ways to subvert it. So you might even call it impossible to balance. It might still make the game more interesting in general, and thus might be a worthy addition: and it might solve the late-game doldrums in games where the score cap has already been reached. But it might also make the game unpleasantly hard for when you're already struggling to reach your score cap. Those games that are currently exciting right up to the end might become super frustrating instead.
Lobby Options Are Not A Great Way To Solve That Last IssueWe could always add more options to the lobby to allow for customization... and to an extent that is okay, but boy that gets overwhelming fast. I mean, look at AI War's lobby. I don't want to do that approach in this game, as this game simply isn't that hardcore. And I also don't want whatever changes that are implemented to be something that have to be optionally turned on: sometimes a cheese strategy or un-fun doldrums happen to occur right in your very first game, and you're unlikely to have enabled any advanced options then. So then you just conclude the game isn't as fun as you'd hoped, and you move on. Despite the game actually offering something much more fun that you would enjoy, if only you played some more with different options. Having to dig that much isn't a good thing; the good stuff should be right there from your first game.
Simply Letting The Game End After Score Gates Are ReachedThis really isn't great either, for a lot of reasons. Sure it lets you skip the doldrums in savegames where you hit them, and doesn't impact the games that are already nail-biting to the end. So from that vantage it's an elegant and simple solution. But I think it's not a wise move in the end, for these reasons:
1. The score gating provides some secondary goals, but it's all too easy to subvert a score goal with specific tactics. So while the score gating is interesting to a point, I am not sure how central I see it as being to the game. You can also turn it off, and then the doldrums are right back in those games. Some people don't WANT score gating, so what do we do to help them?
2. Overall the score gates are something that you can set differently from the other difficulties, right? So you can set a super low score gate, and super high other difficulties, and basically win during the age of man in a super boring way. The game is over hardly before it's begun.
3. In general the goal of the game has always been to last a certain number of turns, and changing that is... pretty drastic. It's likely to have a lot of secondary effects as unpleasant as the things reverted in 1.005. I'm not sure that something that fundamental should be changed; at least by the current evidence I'm not convinced.
4. Not having the game last a certain amount of time, but having the ability to end early, also impacts our ability to later do other victory conditions. Aka, things like peaceful victories or other challenging things. Kind of like building a spaceship to Alpha Centauri in Civilization, for instance. In those cases I would think that they actually WOULD potentially allow you to win the game early, but only under certain really difficult conditions in the third age. Those sorts of secondary win conditions can have all sorts of very specific rules; but what we're talking about here are the
general rules that would affect everything, even secondary win conditions, and I think the ability to end early in the general rules would be very constraining.
Reactive BalanceDon't you just
hate rubber-band AI in Mario Kart and certain other games? That's really frustrating, because it's completely arbitrary. You start doing well, so mysterious your opponents also start doing even better. Bleh.
But with AI War, the "AI Progress" is more a manifestation of
Newton's Third Law of Motion. That's interesting, because it's neither arbitrary nor unpredictable. You know that when you push X hard against the AI, the AI is going to push back Y hard (in some unpredictable manner, but still you know generally the scale of what will happen). This is where the strategy comes in, because you're deciding where and how to push, and how hard. You try to make your pushes as efficient as possible, and this is fun and interesting.
To me, that's a funky mix of both the carrot and the stick. The important thing is that it's fun and interesting, though. It also solves the issue that most strategy games -- including Skyward Collapse, Civilization (all of them), Age of Empires (all of them) and so forth -- all have. Namely, that past a certain point your victory is assured and you're just going through the motions. You've passed the tipping point, balance is in your favor, and you can't lose. That's why
quitting early is sportsmanlike in Starcraft 2. Why drag out something pointlessly?
With AI War, thanks to AIP you don't run into that because the final showdown is epic and your risk of loss is often extremely high right at the end. Running into that is at worst a very edge case, anyhow (and sometimes when you think you've run into that, that's actually a bit of hubris and right when you lose in a surprising fashion because you missed something the AI was doing). All of that is thanks to the reactive balance nature of the game, and that's something I'm really fond of; saying "gg" and quitting early is a bit of an anti-climax in a lot of ways, unless it's really right near the end (in cases like Chess, where again it's sportsmanlike to bow out before the actual checkmate once you are seeing there is no possibility of escape; but that's waaaay in the endgame, not just a "marshaling of forces" in the midgame).
More to the point, there isn't an opponent in Skyward Collapse, and detecting valid "gg" situations is notoriously hard unless we go back to the score gating thing above (which again I don't like). So that's where reactive balance comes into play, potentially.
Napkin Sketch Of Reactive Balance For Skyward CollapseThis is just something brief I've been thinking about this morning, so it's not polished or ready for trials. But basically, what if the game were detecting certain circumstances, and this added to some form of AIP-like meter? Chaos, perhaps? Possible candidates for chaos-increasing circumstance might be:
- Way more mythologicals out on the board than humans.
- Not enough human units being destroyed in a certain number of turns.
- Not enough red or blue buildings being destroyed in a certain number of turns.
- Too high of resource stockpiles.
- Unused action points going above 9 and thus being wasted.
- Excessive smiting.
- Towns that go too long without seeing much enemy presence.
Those are just napkin-sketch ideas, so please don't get all in a tizzy about any specific one you might not like. I'm not thrilled with all of those either, but they're somewhat illustrative.
As Chaos goes up, presumably some new Bad Things start to befall you, thus making the game suddenly harder. This sounds like rubber-band AI, but there again it's really not because Chaos is easily avoidable... and also survivable. Specifically:
- You can easily play a brinkmanship sort of game where you go for high Chaos and an unorthodox strategy, and then it's exciting to survive the Chaos effects in an atypical way. It's like some of the people who play AI War with super-high AIP. Fun! Glad to see variety in strategies.
- If you don't want the Bad Things to happen, it's not that hard to keep Chaos down. You mostly just play the game like most people already do, and it doesn't really even come into effect.
- It starts to become more of a thing late in the game as the doldrums would hit if you've hit your score gate; although, the risk of tedium there might go up if you're trying to avoid chaos and thus just going through lots of busywork to avoid it. That would just drag out the doldrums even more, in less of a fun way, so that's one hole poked in this idea right there.
Secondary ObjectivesThe "Edicts" idea was suggested by players, partly implemented, and then cut because both score and Woes handled the situation in a more graceful and less-complicated manner in the main. I think that was the right call, especially given the former design of edicts. However... some sort of secondary goals I think would be a good thing, potentially. It makes it a lot harder to balance the chaos, and when combined with Woes and the score gating could make for very tense moments.
Previously you could select your edicts from the game start menu, and that was not great. I'd rather do something more unpredictable, like a fourth difficulty setting (sigh, but unavoidable perhaps) for Edicts. Basically that works just like the Woes one.
Then at the start of each round, a new Edict is rolled randomly. You have to achieve your existing goals,
as well as the current edict. If you don't manage the edict or the score gate by the end of the round, you lose. This would need to have an Off option like the score gating -- some people simply aren't going to like this, and that's fine -- but once it was refined enough for general use would probably default to Easy or similar.
What the heck would good edicts be? That's hard to say off the top of my head. I think that a lot of our prior ones that we had designed were bad. Things like certain kill counts, building a university, or whatever. Those are all things that might happen in the course of normal gameplay, and thus which you might just happen to cause unintentionally.
The worst thing is if you reach age 3 and immediately you both are assigned and complete a new edict. Here's your edict -- and you've already won it! To some extent that means that the edicts would need to be something that only counts activities while the edict is active: so if the goal is "kill 200 human units," then it doesn't matter how many already died before the edict came down -- it means 200
more. But even that can be kind of boring, because if you have a huge engine of war around, then just killing 200 human units is trivial. Or if you are already struggling, then killing 200 human units might be impossible. Blargh! You can see why we scrapped edicts.
I'm not sure how to reconcile those sorts of issues yet -- the risk of making the game unpleasantly complex or too impossible to win is very real with this sort of thing. But the idea really does smell like it has some promise, to me.
Added Complexity Of The Good KindOne of the things I like about reading
Boatmurdered is how many things are going on at once in Dwarf Fortress. You're trying to mine metals, keep dwarves happy, ward off sieges, keep nobles happy, keep your trading partners happy, deal with animals messing with you, deal with monsters leaping out of the abyss, and many other things.
It's more of a sandboxy game in a lot of senses, though, because you're trying to build up a fortress in the style of your liking. There are a lot of moving parts, and you're trying to make something that is creatively unique and yours. There's not that much granularity in Skyward Collapse for you to do that. Games like Pharaoh or Age of Empires or whatnot are more the level of granularity of Skyward Collapse.
Part of that comes from the huge focus on the military. Having civilians wandering around on the map was decidedly boring during the alpha, so we cut the chapmen and the traders. However, one thing that I think would be really interesting for the game is having more civilians buildings that don't contribute to the military efforts at all. Making those necessary in some fashion (such as dealing with the demands of the traders and the nobles and so forth in DF) strikes me as a potentially very fun thing.
The key thing is that with more complexity of that sort, if it's done in a fun fashion, it becomes increasingly hard to balance the chaos. Past a certain point it becomes impossible, I would think. Whether or not the game is still any fun (or even understandable) at that theoretical point is a matter for debate, of course. But I think that there is a lot of latitude between here and there, and again this is something that could be optional or only for harder difficulties (or indeed a fourth difficulty option).
Rather than edicts, perhaps there is simply a "civilian difficulty level" or similar, and depending on how you set that is how demanding your civilian populace is. Exactly what that entails is still something that would have to be designed out, but this strikes me as perhaps the most promising first avenue of all for both how to handle some of the cheese cases and how to reduce or eliminate the end-game doldrums (although the ragnarok thing for the end of the game is perhaps the best thing of all for the doldrums, honestly; but this and that are not mutually-exclusive).
The VerdictOverall the impressions of players seems to be that the game is really fun and interesting as it is. I rocked the boat too much recently while trying to solve some problems for edge cases where there were players who were less into the game in general. I want to avoid doing that again.
THAT said, I also do want to support an ongoing, interesting, advanced level of play. If you've "got it figured out" after a few weeks of play... well, you've gotten your money's worth anyhow, but I'm still not thrilled about that. I'd rather that it was still a lot of fun and equally challenging (assuming you properly graduate yourself to higher difficulties) even after months or years of play. We have AI War players with 600+ hours of playtime on the game on their Steam profile, and with the combination of free updates and expansions here, I'd like to see this game have a similar level of longevity. That desire is part of why I had such an overreaction to the cheese tactics that came up, because I feel like that sort of thing really undermines the longevity of the game.
But ultimately, whatever is done needs to have more thought and discussion around it before something is done. So... what do you think about all this? What ideas do you have? Josh and I will still be thinking about it and talking about it ourselves of course, but many heads are better than two.
Cheers!