My post will have three main sections (Introduction, Sources, and Suggestions) and a summary. Thank you in advance for your constructive criticism.
For many reasons, I believe one of the largest challenges to a good (epic-length) TBS such as Elemental will be reducing or eliminating Late Game Tedium (in following: LGT). Late Game Tedium, I propose, may be described as a quality of a game by which a special kind of boredom sets in, one engendered by a combination of two (primary) sources in the complex context of late-game TBS balance. What results is a gap: There is a gap between achieving a satisfying game closure by fulfilling victory conditions and strategically challenging options still open to the AI. You end up having to slog your way through countless turns before winning, and during these countless turns, your advantage is not significantly challenged.
I think the two (primary) sources and one minor (secondary) source can be named in an oversimplified way as followed:
1. Lack of Challenge.
2. Lack of Choices.
(3. To a lesser degree: Micromanagement.)
I will hope to address both of these main sources in the following, and ask for your input as well. First, however, let me try to get a handle on the context of what makes LGT particularly pernicious and particularly special – why it is more than merely two (albeit grievous) lackings I mention here. I think it is this: LGT comes about often (particularly in epic-length TBSs) because a player has achieved a dominant position without achieving a victory condition, and must spend many turns in “mopping up”, during which the number of strategic challenges put open her and number of strategically relevant or polyvalent choices she must make are limited (often due to constraints her opponents face).
In many contexts, LGT is not terribly important. In a two-person MP match, the weaker player might concede, which is not optimally satisfying, but nevertheless often unproblematic. In larger games, this can cause unwanted grief. Wanting to play a game to the end, even if outmatched in wits and power, should be a sign of a good game. In some TBS games, where the scope of the game is smaller, LGT is not an issue because the investment in the games themselves are so small, e.g. Star Chamber: the game is over in 20 minutes anyhow, so there is no problem in slogging it out.
Yet for four reasons, I think this is particularly important to Elemental. Firstly, Elemental claims to be (primarily) a first-class SP game. Remember Brad’s post (http://forums.elementalgame.com/331788/page/2/) from December 2008? “Are you human, are you sure?” In SP matches, it is only you, the player, who can concede – which is not terrible if you are losing badly. But to stop the game because your position is advantageous and you cannot see your opponents posing a challenge to you, yet still have much LGT to wade through to achieve a victory condition, no matter how snazzy the spell effects are going to be, would be greatly suboptimal. Secondly, Elemental will claim to have strong RPG elements and hence a compelling narrative structure to its gameplay arc. The feeling of closure will be very important to playing a game of Elemental; the game should end with the curtain falling, not the main character falling asleep on stage. Thirdly, Elemental wants to have an epic scope: This is running a huge empire, a civilization, over generations; the late game will be large, and our investment in achieving the late game will be large. We want to identify with our empire, and we want the closure in obtaining a hard-fought victory condition.
Finally, I think this topic is particularly important because Elemental will also go the (happily lauded) path of various victory conditions, including the less explicitly dominating ones. In other words, it will include (implicitly) “peaceable” conditions, that have often been associated with “builder”-type victory conditions in other games such as CivIV. While CivIV is an excellent locus classicus in many ways, I think that no TBS game (of at least modest quality) suffers from LGT so much as CivIV when playing as a builder. How often have you been temped to start wars just for something to do? Since your empire cannot build anything but military units after a certain stage anyway, my late game as a CivIV diplomat / builder / scientist was spent hitting return and smoking cigarettes – which is why I uninstalled it. I do not want Elemental to suffer even 10% of the LGT as CivIV did.
SOURCES of LGT
1. Lack of Challenge.
One primary source of LGT comes from the increasingly asymmetrical power structure in late game SP TBSs. By this I mean the common symptom among SP TBS games by which the AI is very challenging at the beginning and increasingly less so as the game progresses.
Take Civilization IV as an example; even after my second game of CivIV, I had to create a mod which attempted to greatly increase the AI advantages (“cheats” and boni the AI got) more and more as the game progressed. This is because the suboptimal AI’s early military onslaughts in CivIV are, when at its “hardest” level, very difficult to survive in the early game, but once you do survive, they pose almost no challenge whatsoever. I had the following dilemma: I either have had to increase the difficulty of the CivIV AI so high that surviving the early game was almost impossible for me (especially as a more builder-type person), or turn it back a notch, so that surviving to the mid-game brought me 185 subsequent turns of boredom.
I notice this in almost every SP TBS game I have ever played, even those with remarkably good AI, albeit to a lesser extent, such as GC2. In other words: I was rarely surprised with a novel, strong challenge in late game. In GC2, once my advantage became clear to me, after surviving the exceptionally challenging early game and the interesting middle game, I became bored in the end game, often merely hitting “return” over and over again, waiting for closure.
2. Lack of Choices.
Elsewhere, in what I hoped to describe as the axiom of TBS fun (http://forums.elementalgame.com/374597 ), I claim that any TBS will be flawed which lacks the lack of a continuous flow of interesting, challenging strategic options at a player’s disposal. Yet this grievous state can obtain in any phase of the game, be it early, middle or late. Any game designer must avoid this systematic shortcoming at all costs, it appears to me.
In the context of LGT, however, at least one additional aspect rears its head: You have already done everything. In your cities, you have already built your barracks and upgraded them, you have already installed your housing upgrades and your temples, every square meter of your land has been covered in railroads; outside your cities, you have explored the world, perhaps you have even discovered the technology or magic for the removal of the fog of war from the entire globe, all of the resources have been discovered or found. At least one of the “Xs” in 4X has been removed.
I will take Civilization IV and GC2 as another example. In CivIV, at a certain stage, there is nothing left to build. Your cities are done; you can only produce military units, research, or money. Your workers can clean up the occasional landfill, but otherwise stand idle. In GC2, most planets were “done” in the middle game; there were no more slots to build on, and no more improvements to be made (as unexciting as that was) to the existing ones.
Now it is time to “exploit” (the last of the four Xs) – but we still need a continuous flow of many meaningful choices to make. Particularly builders will have a difficult time here if the only thing left for them to do is armed conflict; micromanaging Corporations is not fun enough. We need more choices, the stream of which should not stop in late game. I do not want to wait and wait – for the ability to upgrade my lasers just one notch more. In CivIV, I set my queues to make mechanized infantry, jet fighters, money – and smoked cigarettes and hit return. In GC2, I built constructor ships, smoked cigarettes and hit return. It wasn’t long before I uninstalled.
But even if you are not a builder – although this option should be open to you, as Elemental claims it will have multiple discrete VCs – armed conflict will need to have disjunctive but viable options for resolution in late game. Battles and choice in battles still need to be fun late game. Dominions3 did a decent job here, better than GC2 did, even if it failed at making the preparations for battle good (organizing and micromanaging them, getting your troops to the right place at the right time, automating queuing, etc.)
I will not address micromanagement here explicitly, because I think it is self-explanatory that excessive micromanagement becomes exponentially burdensome in late game and thus is to be avoided if LGT is also to be avoided.
So… time for some ideas on what might help.
In the following, I will humbly propose some humble suggestions and welcome your additional feedback and constructive criticism. Thank you in advance.
NB1: Some cases will try to address Source 1 more explicitly, others Source 2, only one for micromanagement, while I have none in general for addressing both simultaneously.
NB2: Please note that in at least two cases, I explicitly suggest that my idea be implemented in the form of an optional toggle switch to be used in game setup. I say this because I can imagine that some people (mostly, I suspect, people who see the SP aspect of the game as secondary to its MP value) would find the suggestion either too difficult or too unconventional.
Suggestion 1: Player-Customizable VCs.
There were at least two posts mentioning player-customizable VCs, here http://forums.elementalgame.com/361897 and here http://forums.elementalgame.com/373483. I think that implementing this should not be too difficult, and well worth the effort. If objectionable to some, one could always imagine that this option could be toggled off. Let us design our own VCs from a customizable and combinable list of drop-down options. Let the VCs, in addition, be variable for each (AI) player – in other words, let Player B be able to win if P, Q, R, and T obtain, but Player C might be able to win if only P, Q and R obtain, while Player A (say, you as the human player) must need fulfil P through Z. This might be as simple as a number of Master Quests that need be completed to as complex as obtaining a Master Quest and simultaneously having an offspring of yours in each house whose population is below X but whose prestige is above Y while also holding 25% of landmasses with forest regions on them and making sure that no dragons currently live in the world, or some combination like that. In addition, the concept of spatially attributable Victory Points (Dominions 3) could be used, e.g. having (randomly generated) tiles of land be valued as a Victory Point, and (one) condition of victory might be owning a certain number of these Points (in conjuction, perhaps, with other conditions).
Suggestion 2: Epic-Length Dynamically Scaled AI Difficulty:
Even after my second game of CivIV, I had to create a mod which attempted to greatly increase the AI advantages (“cheats” and boni the AI got) more and more as the game progressed. This is because the suboptimal AI’s early military onslaughts in CivIV are, when at its “hardest” level, very difficult to survive in the early game, but once you do survive, they pose almost no challenge whatsoever.
I would like to see Elemental’s AI come with a feature out of the box which allowed it to scale. In other words, if toggled on, this scaling AI Difficulty gets few boni in the early game, but increasingly more and more as the game progresses, so that in late game, the AI is capable of still engaging its counterparts on a number of levels.
This option could have a drag-down selection list, e.g. How great should the difficulty of the AI scale as the game progresses? “none” (“off”), “slight”, “moderate” “great”, “tremendous”, “huge”.
Suggestion 3: Hidden VCs / Hidden VC information.
This could be a simple toggle: If “on”, then (human) players cannot see how far other (AI) players are in obtaining a VC. If espionage is implemented as a part of the game, and I hope it is, then this option might become part of a difficult espionage quest.
Suggestion 4: Narratively Dynamic AI-controlled End-Game VCs.
This suggestion will not come across as bizarre to people who are familiar with level-type quests in RPGs or RPG-FPS hybrids. Nevertheless, I have not seen it implemented in 4X TBS games, and so suggest it be kept as an optional toggle in game setup, for those for whom it might be problematic.
The scenario is this: You think you really need to complete Goal G to win, and you are well underway in achieving this goal; you have obtained a strategic advantage, yet need another X turns in order to obtain closure. Suddenly, however, you see that something dramatic has come about: You no longer (only) need to obtain Goal G, but also, additionally, P, or P and Q, or P Q and R within a certain number of turns.
You might be familiar with this from “missions” in RPGs or FPSs or hybrids, in which you have finished a particular (sub-)goal, but the level or mission is not over due to a new narrative complication (e.g. you don’t just need to find the Queen, you now need to sneak her out of the palace, and you cannot use your magic X all of a sudden because an evil Y has put a blocker on the palace grounds; and suddenly, you tripped an alarm so that you only have 10 minutes before a magic bomb goes off…).
I suggest that if toggled on, the AI be able to calculate when it is behind, but not behind enough to be obliterated for closure within the next Y turns (Y could be a variable the player could choose upon game setup from a drop-down choice menu), and that this triggers an event which requires an additional goal to be met (or be stopped) in order for the human to win (or not to lose). Not for you? Toggle it off.
Suggestion 5: Bona Fide “Are You Human? Are You Sure?” Mean Bastard AI:
This suggestion comes from a post Brad made in December of 2008: Are you Human? Are you sure? I suggest an optional toggle available at game setup which lets AIs act as if they were hardcore cutthroat minimaxing gamers out to win, or at least stay in the game, at all costs; when enabled, the AIs should be able to recognize when a human player (question: any player?) appears to be coming threateningly close to running away with the game, even if that AI does not have enough espionage available to determine how close to achieving a hidden VC the human player actually is. If the AI gauges a player to be getting close to obtaining a strategic advantage which will let her win in the long run (a long slow 150-turn slog), obtaining a dominating turtle-position, then the AIs should form a schoolyard pact and gang up on the (human) player, much like humans might whisper "Don't trade with Bob! He's getting too powerful! Let's take X and Y away from him and divide it up!" The AIs should be willing to take larger risks, cooperate, and step back from certain short- or middle-term goals they had set; they should do this to stop a (human) player from dominating in the long run -- particularly in the end game. The AIs should enter “bastard” mode and be willing to take risks and sacrifices to harm the human player once her advantage seems to get too large; the AI players, if this mode is toggled on, should be more willing to form pacts against the human player even if diplomatically well-inclined; as long as that AI would not “win” by having its (human) ally (also) win, it should act as a human player would and attempt to prevent the human player from winning. In an extreme form, one could imagine that some of the AIs are “spiteful” and would even be willing to suffer a great (short-term) disadvantage for even the slim chance of being able to stay in the game later, or even gain pleasure in “going out with a bang” and harming a player instead of suffering many turns of attrition. I realize this will have obvious consequences for calculating diplomacy and diplomatic victories, consequences that might be unappetizing in certain contexts; so toggle it off if bothersome.
SUMMARY and FURTHER DISCUSSION
Another element contributing to LGT is also vast micromanagement which often comes with having a large empire, one which by its very nature only comes about after the mid-game in empire-building games such as this. I do not have any (novel) suggestions to combat LGT stemming primarily from this aspect, but I do think that game structures that explicitly encourage turtling and stockpiling are more susceptible to LGT than game structures which do not. This is not to say that obtaining and holding a strategically defensive position should be unviable – that would be madness. Tactical defense is and should be more viable than tactical offense, all things being equal (which, of course, they never are); but not strategically. Yet stockpiling and turtling are not the same as taking a defensive stance. Game structures which reward you for expending or trading resources at your disposal instead of saving them up might be helpful.
In general, I want to be able to have many strategically important polyvalent choices open to me in the late game. I want to be challenged and I want to be able to be surprised in the late game.
While I was enticed with GC2 for the first couple of games, and I still like the idea of playing excellent games such as Dominions3 in SP mode, I no longer do so, because I get bored in late game. Please let Elemental not suffer this fate!