In terms of the difficulty, it's funny because you guys were all losing an unexpectedly large amount right at first. But clearly once you graduated past a certain amount, certain strategies are dominating. This is one of those give-and-take things with balance of any game, really. It's a razor-thin line between making something impossibly complex and something that is too laid back, sometimes. It might not feel like it, but small tweaks can cause huge havoc as you may have noticed.
Woes make the game more interesting and varied, but I agree that I don't think they should be the sole source of difficulty itself. Balancing the factions should be hard in and of itself, and the woes are there to add variety and extra challenge. Really the same is true of the bandits, to some extent. If your main source of threat is the bandits, we're not doing something right.
In terms of not building siege workshops at all, that kind of was a lightbulb moment for me. The health of buildings is fine, incidentally; they should be really not at much risk from minor guys, and they are not. It's those siege weapons that take them out super fast, and not having those on the board is a big problem and thus causes the issue.
The core thing that is off here is that we removed the scores and score gating -- with good reason, but still. In our original designs, you had to have a certain amount of carnage in order to win the game. If you didn't, then you'd lose when the turn timer ran out. Hence you were encouraged to use all the crazy stuff in order to maximize carnage, siege weapons included.
Setting a secondary goal like that, especially something score-like, just doesn't work though. It feels too complex and arbitrary, and is just about impossible to balance because there are always exploits. What is really needed is a more in-game reason for the carnage where "otherwise worse things happen" sort of situation. Woes answer that to some extent, but they don't address siege weapons or blockading off certain towns, etc. Their random nature even means that the more of them we add, the more wildly the core difficulty will fluctuate. Some games would be super boring once you pass a certain skill level, and that's deadly to the long-term interest of the game. So while I think more woes are a great idea in general, they don't help (and in fact to some extent hinder) here.
The problem is the motivation to fight. Your guys are bloodthirsty, and you have to build some of them to beat off bandits. Then further carnage happens based on them rising up and fighting each other. That part works, but since bandits don't have towns the siege units are not involved (except when bandits use them). So the one most obvious thing is to start having some bandit towns pop up, which was something I'd already planned to do through woes. But I think that is too core of a thing to be in woes; I think that needs to be alongside woes, really; the more subsystems that are interacting under the hood, the more interesting things emerge (to a point, anyhow; too much and it gets overwhelmingly complex).
This brings my thinking back to AI War, actually. Late in the alpha for that game (completely private), it was still just being too easy much of the time. Players could just build up a critical mass of planets, and then the AI wouldn't stand a chance no matter what it did; unless I turned the AI difficulty knob up so high that the players stood no chance. Either it was pretty much impossible to lose or to win, once you passed a certain early point in the game. The solution that I came up with to that is well-known by now: AI Progress. That made you weigh each decision, and governed how you took planets. That in turn made the AI balance-able, and kept the depth there.
With SC, you're not taking territory or whatever, so a similar mechanic doesn't make sense. Well, not TOO similar anyhow. Remember score? In the original designs of SC, this was intended to be almost like AIP except in reverse. You had to shoot yourself in the foot enough to win, basically. I thought that was the most interesting thing, because other guys shooting you doesn't feel very god-game-like. The woes and the bandits are interesting, but I don't think they can serve the role of replacing what the score was trying to do, in other words. The problem with score was that it was bloody complex from both the developer and player standpoint, and felt super arbitrary.
Okay, so we need a mechanic that has fewer moving parts, and thus which can both be balanced by us and understood by players. AIP in AI War has a wonderful simplicity: mostly that increases by taking planets and that's it (in the original 1.0 version that literally WAS it). And there's a few other things that contribute, but mainly that's it. Simple, easy to understand, and bloody hard to manage at a high difficulty level. That's what SC has.
That leads me to the idea that I've had today, based off of Mick's (rather distressing, I must confess) notes: Unrest. These people are bloodthirsty, right? We've established that. They want to fight, and will do so as soon as they exist. The problem is, right now there's not much incentive for you to create them except to fight off bandits, and then you just balance the remainder. Crime was once upon a time supposed to solve that issue, and I guess it kind of did, but it was too opaque and not that fun. Also it was something that could be gamed in various ways.
So, Unrest. This would be a global number in your HUD at the top, like AIP is in AI War. So right away, we get away from that hidden-ness of Crime, which I like. The core idea is simple: a) Unrest goes up when red and blue don't hurt each others' towns enough; and b) bad things happen when unrest gets too high, and you get into a death spiral basically.
That's the core design, stripped of any details. I think that's solid, and is what the game needs. There are a variety of ways that could be implemented, but I think I have a pretty solid idea for the first go-round.
1. Regarding part (a) above, the "goes up when red and blue don't hurt each other's towns enough" bit:
- This is a great place to encourage siege weapon use.
- This is a great place to encourage NOT having towns be absolutely turtled and too well-protected of towns. I.e., having all towns engaged in warfare as the Crime mechanic was trying to do.
- This also is a great place to be outwards-looking-in, in terms of looking at damage a town takes, rather than focusing on military units it produces, which was another problem with Crime (far back towns could produce military units fine, but still be perfectly safe themselves, which was of little help.
So, the design this puts in mind for me is:
- For every turn that passes for a town, its personal Unrest counter goes up by a certain amount.
--This breakdown can easily be seen in the overall single unrest number when you hover over it.
-- This increase would vary by difficulty.
- For every tile you smite, Unrest goes up by a certain amount in the nearest town.
-- So no just faffing around smiting tiles, either.
-- And if you're trying to temporarily blockade off a town that is normally open, you also pay for that. It's possible to do still, but there's now a downside.
- For every building that is destroyed in a town, the Unrest goes down by a certain amount.
-- Note that it's hard to destroy buildings without siege weapons, so they would have to be on the board.
-- Note that you can't always control what your dudes do, so sometimes you'll get unintended consequences of them destroying more than you meant them to do.
- The Unrest growth rate in towns should be something that increases sub-exponentially but certainly more than linearly.
-- That way destroying buildings frequently in a town can keep unrest almost at zero, and that's great.
-- But by the same token, just destroying a building very periodically, or hiding one town and letting its unrest grow while you keep all the other towns near-zero isn't possible.
-- In other words, ignoring any one town is a runaway disaster, because of that growth style. You can't leave one town with high unrest and the rest with low and have it balance out.
-- In general things become more difficult for the player to balance, because no town by definition now has to be on the front lines, and has to fight.
- Thematically speaking, the reason unrest is caused in a town when they don't get their stuff destroyed actually is backwards from that. The other faction starts resenting that undamaged town more and more, is what is thematically going on. It's not that the people get mad if their personal stuff doesn't get blown up or something.
- Having the crazier god powers also give a reduction in unrest also is something that we can put in as kind of a "this is my panic button" sort of fashion.
-- Paired with woes and the god power cooldowns, I can't see any of the god powers going unused in advance play with this.
-- This also gets back to the original idea for god powers, which were that they would be a super huge source of score, and thus you'd be inflicting these on yourself to win, and then playing cleanup.
2. Now, about part (b) of the above, the "bad things happen if unrest is high" bit.
- This is a great place to have a second source of chaos other than woes, and of a different flavor than woes.
- It strikes me that something relatively simple here would make the most sense, because woes are already the main source of personality between games and I don't think we need to try and do that with two separate mechanics. That would just be too much.
- As a core tenet here, the basic idea is that the game needs to become unwinnable at a high difficulty level if unrest gets away from you.
- Unlike AI War, it needs to not lead to stalemate situations, but instead should actively kill you. Because of the nature of this game, that's actually almost implicit in any potential design here, so that's good.
So, the design this puts in mind for me is:
- Bandit towns become a thing, but only based on unrest.
-- The higher your unrest goes, the more bandit buildings get placed.
-- It might not even need to be whole towns, but just isolated barracks or whatever that wreak havoc on you. Or at any rate a cluster of military buildings and possibly some towers; nothing else is needed.
-- This is yet another reinforcer for "you absolutely cannot do without siege weapons in the hands of red and blue," which is good.
- For every bandit building on the board, I think that unrest should be reduced by a certain amount. But not actually on a per-town basis. Oh, no. This would simply be a negative unrest counter from the bandit buildings.
-- So in other words, the positive unrest from your individual towns would still be just as high, and growing at just an awful of a rate if it's too high.
-- And also, if you kill a bandit building then your unrest shoots back up again.
-- Thematically speaking, the bandit buildings "dampen" unrest for a temporary time because people are so distracted by them. But as soon as the bandits are gone, the old anger just comes right back.
Bear in mind that I'd make this relatively tame on the lower difficulty levels. And ramp up steadily to where Expert is absolutely all about managing unrest, really. While juggling woes and regular bandits as a side thing. But the core thing is still managing that conflict between red and blue, and making sure that neither side gets just too boiling with anger. Otherwise some of them "go rogue" and a bandit town pops up. There would literally be a popup to go along with this, informing you of the even, why it happened, and some thematic flavor bits, etc.
The core idea behind this game -- at an advanced level -- was always that you'd have to self-inflict pain in order to win. That was something that the score was designed to do, but that's been kind of lost in the shuffle as the design solidified and improved in pretty much all areas except this one loose end. I think unrest will bring back that element in a big way, so I'm pretty excited about it. I'm going to try to get that in there today; thoughts are naturally welcome!
