Late game in strategy games

[Axioms]

When we look at something like GalCiv we are brought back to the idea that the game simply needs to acknowledge the actual point of victory. Some players are able to just decide that they have won and move on but many other players seem to really require an official win screen. Move the win screen condition to the same point that pro-active players usually assign themselves the winner at.

I think the catch is that to move the win screen to that point would mean that the player would start to lose arbitrarily on these grounds. If the player occupies 30% of the map, he's probably into his victory lap. If the AI occupies 30% of the map, it's nowhere close to over, even in a full observer match, as the capacity of the AI to shoot itself in the foot far exceeds the player's at that point.

It also doesn't apply in a game where there are multiple players: One player occupying even a plurality of the map is nowhere close to the end. So you can't draw the win condition TOO generously or the game will be declared won prematurely, but if you make it more stringent, the game could be won and dragging by then.

Lol imagine having a wincon for the AI. This is only for the smoothskin.

As far as multiplayer who cares, MP is trash.

 

[Axioms]

Your empire can't fall unless you play like an idiot.

This. Games do very little to no simulation of various complexities of governance. Large states are notoriously difficult to manage. The larger a state grows, the lesser the extent of administrative control of it's base components by the top leadership. Also. different conflicting interests start to arise and become ever more difficult to get under control with the state growing in size. That is also something games don't even attempt to portrait or if they do, it's a lame slider that gives you time to the next rebellion.

Additionally, AI is often extremely passive. Player clears out his neighbouring rivals, consolidates his domain, takes control of his side of the continent/galaxy/whatever, nobody even thinks to interfere with him and when he's done, borders of AI opponents on the other side have barely moved. No hegemon has emerged, no one came out on top and all that's left is the clean up.

I made a strong effort to simulate this stuff in my game. But it is ridiculously complex to do. It is a very niche genre as far as audience size. Even if you had a perfect UI and so forth most players wouldn't want it.

 

[laclongquan]

Method to deal with with late game boringness:

Add a new, final, boss faction.

Take SMAC for example. Let's assume we are modding a gameplay mod.

IN late game it's usually the time where we are mopping up the remnants of other factions. There's other strong ones, but they just cant compare to our R&D capacity plus our veteran army.

Introduce the two Manifold factions at fixed date that you predict they cant complete the map yet.
+ They will attack to try and wreck up some landmark areas in form of perfect columns (10 very strong units each one). Say, Pholus Ridge Uranium Flats Monsoon Jungle, etc... In case of some areas that get ruined or terraformed into anihilation earlier (like Moonsoon jungle get lowered into the sea for whatever reason)
+ They will raise terrain of a random area 1000m into the air and put down a capital city there (like, driving a meteor onto there to create a landing)
+ The more production your PC faction have plus research done, the stronger their initial force is.

 

[Matalarata]

Eh, a good AI won't sell as much as flashy graphix or marketable stunts. Unfortunately this kind of mentality infected the 4x/Strategy genre before an adequate model was adopted at large. I feel this is the main issue. Granted, coding AI is harder than producing pleasant graphics but I feel we missed a huge opportunity for all these years.
A staggering amount of measurable data (played games) has already been lost without being analyzed. Imagine if an early model was adopted were, like chess, the games of the top performing players were analyzed to introduce new algorithms. Nowadays an AI opponent would be much more flexible and unpredictable. Less cheats for an AI means you're both playing on the same level, the playing field is much wider. We regularly nerdgasm when a game succesfully models things like disruption of supply lines or other indirect forms of attack but said systems are usually asymmetric, simulated.
By evening the playing field you'd allow a number of emergent strategies to be born just by opportunity alone. Instead the general idea is to give an AI flat and broad cheats, with no basis in reality. Small factions fielding (relatively) huge armies due to background income, foreign armies/fleets traversing the observable universe 'coz one AI declared war to someone on the other side of the map and suply is a joke, you get the gist. Those are all effects of said school of thought.

Since the chance of a revolution in the gaming AI field is slim at best, I'll also vouch one of the most elegant (imho) solution for a game doomed by a sub-par AI: environment. By using the map itself as a "neutral player" even resource starved devs can do much to keep a game entertaining. The old Fall From Heaven mod for Civ IV, for example, uses the armageddon counter to change the playing field as you go, with the world being slowly plunged into hell. Good factions need to go out of their way to crusade against those bringing evil into the world, else your late-game empire will crumble from the inside, as your pastures turn into hellscape and grain becomes razorweed. Evil factions are usually a bit better equipped to tackle the changes but still need to deal with the same dangers and the addition of a pretty strong Good faction which will only spawn if the armageddon (and the Infernal faction) are themselves rolling.
More recently, Endless Legend and the winter mechanic come to mind, although that's far more manageable for an experienced player. It still adds an element of quasi-randomness you need to plan ahead to and play around.

Now, obviously this is far from being a true solution to the underlying problem, player and AI play different games with different rules. I think it's still something that should be considered. Especially nowadays, with so many small, talented dev teams which are unable to seriously tackle the AI development issue. A dynamic, reactive world can prolong a 4x appeal and semse of "freshness", if you get what I mean!

 

[3 others]

Missed this topic back when it was fresh, but here are notes from a few mods that have made creative attempts to tackle late game bloat.

Crusader Kings 2 Historical Immersion Project modpack (or one of the included mods to be precise) has a mechanic called Imperial Decay. It is a negative modifier for Empire-level polities which affects pretty much everything from vassal rebelliousness and tax revenue to foreign relations. Imperial Decay is raised significantly during each ruler death and lowering it is extremely costly (loads of money, rarely available decisions, putting down rebellions). After a few generations - with corresponding increase in Imperial Decay - each succession is a potential crisis as the most ambitious vassals start making demands or putting forth claimants to the throne.

I was pleasantly surprised to actually have an internally fractional Empire in my hands and over the centuries having to make the occasional concession to the Council privileges and even actually losing one succession war, relegated to the status of a mere King, and seeing another dynasty rise to the Imperial throne. With all this infighting, any outwards expansion was more of an afterthought for the rest of the game, save a few oppourtunistic claims from rival nations in midst of their own Time of Troubles. All in all, it felt like a reasonable representation of an Imperial lifecycle and it worked really well with the vassal mechanic in Crusader Kings.

I'm sure that Imperial Decay can be gamed like any other mechanic but it is a good attempt at applying an internal handbrake to the common strategy game snowball effect where success breeds more succcess. I seem to recall from forums chatter that Imperial Decay is a pretty contentious mechanic as 'punishing' the player for being successful is generally discouraged in game design, but once you're making mods only/mostly powergamers are bound to be interested in, you're free to disregard that rule.

Civ4 mod Fall from Heaven and its dynamic terrain was mentioned in the previous post. Even standard Civ4 integrated some challenge game types that I believe originated in Civ3 community, such as One City Challenge and Always War modes, but Fall from Heaven unsurprisingly goes even further than that. It has a pretty unique game mode called High-to-Low where once you have the highest score in the game, you get demoted to the civilization with the lowest score and have to start playing as them. Twice. On your 3rd civ, you can finally win the game, although I guess you could also win a cultural victory without having the score lead at any point. I've found that the game mode provides a more evenly spaced out 'challenge' throughout the game. Usually in singleplayer, you reach a tipping point after the tight early-midgame, after which the rest of game becomes more-or-less obvious mop-up but in High-to-Low, you will encounter two serious-but-entertaining difficulty spikes over the course of the game.

The downside of the system is that very prone to randomness. You might be relegated to a hopeless Civ one turn away from defeat, or a low-score sleeping giant (a.k.a. Doviello) with the largest standing army in the game. It's also a pretty arbitrary way to play a game unless you enjoy LARPing as a PwC/McKinsey bloodsucker (which I do, thank you very much).

Come to think of it, the closest comparable is playing Football Manager as a journeyman manager, taking less prestigious jobs wherever you can and hustling your new teams to relative success. It's not a direct comparison regarding late game bloat as the game world in Football Manager is 'persistent' in that the global football ecosystem existed before your character began his career and it will continue existing with little structural changes after you retire. A game of Civilization has a beginning and an end from settler-and-scout to a global empire, which means that playing Civ will be fundamentally different at different stages of the game.

 

[what am i doing]

I'd lean towards the optimal solution being a hard cap on directly owned & managed territory (edit: not strictly hard as in "completely locked", but hard as in "extremely high penalties as you exceed it"), with further expansion having to come from vassals/substates, and a further limit in turn on the number of vassals/substates that can be had, requiring them to be consolidated into larger entities that can pose more of a threat if they rebel (or incorporating mechanics that let many smaller vassals generally rebel at the same time and in the same area).

In this manner the player would have their imperial heartland, which they administer directly, and then the greater empire, which is usually not managed as well and can sometimes break away or collapse (but with the player having cheap or free imperial restoration CBs). So the player can map paint and have a sprawling empire, but without having to micromanage most of it, and with the potential for large portions of it to be rapidly lost or regained, getting more severe as the empire grows. At the same time the core country can be much more stable, secure, and prosperous if the player manages it well, so ideally having their core country be safe would alleviate player irritation over more remote parts of their empire rebelling or having incompetent governors.

 

[Riel]

Other than improving AI, diminishing returns is key. EUII, was pretty good at having an interesting end game, not to say you couldn't, once mastered, overcome the diminishing returns mechanics, but it certainly held its challenge a lot longer than any other long spanning strategy game. A shsame Paradox has shafted most of those mechanics or nerfed them so much that they are no longer an issue.

 

[Gilius Thunderhead]

Because you are playing them right.

In a zero-sum game success breeds success, and you can't avoid it without blatant cheating on behalf of AI or frustrating restrictions ("punishment for success") that kill any joy from progressing in the game.

This is the crux of the issue. Also, what The Brazilian Slaughter said regarding easy to manage empires that ignore entropy. This comes with a problem of definition: is the strategy game in question a simulation or a sort of electronic "board game" in which you play quick matches against adversaries?

If you want to play a quick match, then you need clear victory conditions and a consensual, that is, 'narrow', set of rules and procedures. In a match of chess, a superior player will win every time and his victory can be dissected by retracing his moves, while the losing player can similarly observe and learn from his own mistakes in an objective, impersonal way, as if they represented some universal chess player. This is allowed by the logic of the game, which is highly deductive, that is to say, universal. Every "moment" of the game can be isolated and studied as a set piece by the detached observer. And when one talks about deduction, one speaks about being absolutely right or wrong, which extends to the victory conditions, which are naturally zero-sum. This is intrinsic to the game. Any attempt to add granularity by means of simulationist, realistic or roleplaying additions would tarnish the elegance of the game considerably.

However, the domain of strategy proper is much larger than solving logical puzzles or making genius moves five steps ahead, based on assumptions. We don't need to get too involved in definitions. We can look at world politics in a conceptual framework and ask ourselves what would traditional concepts such as "winning" and "losing" mean in this context. No one ever though that winning meant that you had to eliminate every opposition, for example. In Rome TW you could paint the entire map red, while historically this was obviously impossible (some games try harder than others at hiding their "board game" nature). Why this was, and continues to be impossible is an open question, but certainly the reasons are varied, and some of them may be inextricably linked to human nature.

Let's be clear, though. If you're going to simulate imperial expansion, it really wouldn't make sense to depict a single "entropic" variable that would scale in relation to your territory. Sometimes the opposite, in which larger entities (even "too large) are more stable, is true--imagine the chaos that would erupt if big historical territories like China, Russia, etc. had to deal with breakaway "provinces". The reality of the matter is always layered and often paradoxical.

In other words, there's no way to simulate such complex realities in a consensual, easily definable way. It soon turns into a "project" where people use computers to paint abstractions and "fill in the blanks", much like our universally praised climate models. Every setting requires different rules and variables to depict whatever process it is imagining, and each game must possess its own metrics for success, if it needs those at all. I really liked playing Victoria 2 for this reason, because it allows you to roleplay(that's the keyword here) basically any scenario that is historically plausible.

I mentioned Victoria 2 for its openness, but it's obviously a very limited game as far as strategy is concerned. How to make such a game challenging, apart from generic improvements to the AI, which most games would benefit from? This is entire point of the thread, which only now I'm coming to.

Obviously, you CAN add more granularity. What happens if you add a province to your empire? Does it make the game easier or harder? If the game becomes easier, is it plausible that it should be so? Apply the same reasoning to every aspect of the game. Is it plausible that you should have political stability for 50 years, that no one tries to kill the king, that no general rebels, that you never have a bad harvest, that peoples's loyalty doesn't change, that people will live in the same place and have the same jobs, etc.

The problem with granularity and complexity, in general, is that the AI needs to become exponentially more competent. You can program a computer to win at chess, but you can't make it understand the Middle East, even if it's Fantasy Middle East (I should patent this idea before some 4X company gets to it).

My other, more economical idea, is to have the AI behave sort of like the aliens in the They Live movie. What do I mean by this is that the AI should be very good at doing AI things and recognizing AI behavior in others, while being extremely sensitive to things that fall outside what it deems to be "normal" behavior. If the player suddenly starts behaving like a human being who is aware of their existence, they should become dispositionally hostile towards him. You could invert the analogy, to humans who find out about aliens, whatever. The point is, the AI should be more aggressively conformist and punish the player who is too smart for his own good. This is not different from any political system, where upstart powers (even very small) who grow too quickly will arouse the suspicion and, eventually, hostility from the status quo powers. What you have now, instead, is the pseudo-Hobbesian system of every AI for itself, which is highly exploitable by the player. The player should, instead, be more encouraged to hide his own power or to postpone his victories until he secured wider support.

 

[mondblut]

The point is, the AI should be more aggressively conformist and punish the player who is too smart for his own good. This is not different from any political system, where upstart powers (even very small) who grow too quickly will arouse the suspicion and, eventually, hostility from the status quo powers. What you have now, instead, is the pseudo-Hobbesian system of every AI for itself, which is highly exploitable by the player.

You have just invented Paradox' "badboy" cooldown, the device patented to suck all fun from their games.

The player should, instead, be more encouraged to hide his own power or to postpone his victories until he secured wider support.

And what is the player supposed to do while postponing his victories? Watching the paint on the map dry?

 

[Gilius Thunderhead]

You have just invented Paradox' "badboy" cooldown, the device patented to suck all fun from their games.

My objective was precisely avoiding such artificial cooldowns and mechanics like "Infamy". That's not the AI who is punishing you, it's the game giving you a yellow card for some supposed infraction you're committing. Once the cooldown wears off, you can resume your behavior without consequence. IRL this would have the opposite effect (imagine Germany biting off a piece of France every 5 years, or Brazil quietly munching on its SA neighbours like a fat spider while they quietly wait their turn), in that everyone would be driven to paranoid hysteria and gang up on the infractor.

What I propose (and I realize it's not easy, although less difficult than developing genius AI that can beat the player 1 vs. 1), is to make the AI as paranoid as humans naturally are. If the player behaves in a way that is ostensibly more competent than the AI, the AI should feel threatened by it, even if no "fouls" are committed, the same way you'd be apprehensive if you found out the guy sitting next to you in an MMA champion. Obviously, if competence and aggression are combined, the paranoia should be much greater.

And what is the player supposed to do while postponing his victories? Watching the paint on the map dry?

If all the game has to offer is map painting, then it should be treated as a board game and my above points don't apply.

 

[mondblut]

My objective was precisely avoiding such artificial cooldowns and mechanics like "Infamy". That's not the AI who is punishing you, it's the game giving you a yellow card for some supposed infraction you're committing. Once the cooldown wears off, you can resume your behavior without consequence. IRL this would have the opposite effect (imagine Germany biting off a piece of France every 5 years, or Brazil quietly munching on its SA neighbours like a fat spider while they quietly wait their turn), in that everyone would be driven to paranoid hysteria and gang up on the infractor.

So, instead of becoming fun once in five game years, it is supposed to stop being fun five years in till the end of the game? Sounds exciting.

If the player behaves in a way that is ostensibly more competent than the AI, the AI should feel threatened by it, even if no "fouls" are committed, the same way you'd be apprehensive if you found out the guy sitting next to you in an MMA champion.

Yet most inferior countries tend to suck up to USA instead of teaming up against it.

If all the game has to offer is map painting

Always has been.

 

[deama]

I wonder when they'll finally start to implement that alphastar machine learning AI. Sure it's not as good as some pros, but I bet it'll give most vets a hard time.

 

[Norfleet]

Every time a thread on this topic comes up I have always said I would like to see an expanding map that adds new and bigger foes (or friends) as the game advances. Preferably that you can interact with some while they are still off map entities.

I commented on this idea before, yes. It's most easily done in an environment where the map is functionally infinite, like spess or something, so that you can have distant things that don't actually have to exist for the player to be running a race against from the very beginning, that you can then fully manifest once the player gains the power and reach to observe it, like off-the-visible-map entities held in a state of quantum indeterminacy until the player can observe them to collapse the wave function.

 

[Galdred]

Let's be clear, though. If you're going to simulate imperial expansion, it really wouldn't make sense to depict a single "entropic" variable that would scale in relation to your territory. Sometimes the opposite, in which larger entities (even "too large) are more stable, is true--imagine the chaos that would erupt if big historical territories like China, Russia, etc. had to deal with breakaway "provinces". The reality of the matter is always layered and often paradoxical.

Actually, China has had a lot of civil wars, and one of the famous quote of the Romance of the 3 kindgoms is:

Luo Guanzhong — 'The empire long united must divide, long divided must unite; this is how it has always been.'

So yes, dealing with breakaway provinces and warlords has always been a concern of the central power in China, and all have not been successful.
For Russia, it is a bit different, but the country population is mostly concentrated in a few key regions.

Obviously, you CAN add more granularity. What happens if you add a province to your empire? Does it make the game easier or harder? If the game becomes easier, is it plausible that it should be so? Apply the same reasoning to every aspect of the game. Is it plausible that you should have political stability for 50 years, that no one tries to kill the king, that no general rebels, that you never have a bad harvest, that peoples's loyalty doesn't change, that people will live in the same place and have the same jobs, etc.

The problem with granularity and complexity, in general, is that the AI needs to become exponentially more competent. You can program a computer to win at chess, but you can't make it understand the Middle East, even if it's Fantasy Middle East (I should patent this idea before some 4X company gets to it)..

That is why the "entropic" variable (or set of variables) approach is not too bad: it makes it way more readable for the AI.

The point is, the AI should be more aggressively conformist and punish the player who is too smart for his own good. This is not different from any political system, where upstart powers (even very small) who grow too quickly will arouse the suspicion and, eventually, hostility from the status quo powers. What you have now, instead, is the pseudo-Hobbesian system of every AI for itself, which is highly exploitable by the player. The player should, instead, be more encouraged to hide his own power or to postpone his victories until he secured wider support.

Have you been in a board game group where the players tacityl decide before hand to murder the best player early on?
It can get quite boring for him. That is one of the main issues of "winning in 4X":
Of course, human players in a boardgame would do just that (and 4/5th of the time, overcommit to remove the most obvious threats and still grand an easy victory for the next in line, but I digress). Players usually expect the AI to LARP in these situations, and not play the metagame to win.

 

[Norfleet]

Large states are notoriously difficult to manage. The larger a state grows, the lesser the extent of administrative control of it's base components by the top leadership. Also. different conflicting interests start to arise and become ever more difficult to get under control with the state growing in size. That is also something games don't even attempt to portrait or if they do, it's a lame slider that gives you time to the next rebellion.

It's actually not that large states are difficult to manage, or that conflicting interests start to arise. It's that once the player solves any given problem, it stays solved because the player solved them, and the player remains, eternal. In real life, the guy who solved the problem dies and the situation becomes unglued because his talentless understudy doesn't know how to handle this anymore. Empires tend to fall over generations rather than overnight. But in the game, the player is essentially an immortal spirit that provides continuity of purpose and memory. Once a player solves whatever obstacle you've placed in his path, it will stay solved, across generations of leadership in the game, or even across games.

Have you been in a board game group where the players tacityl decide before hand to murder the best player early on?
It can get quite boring for him. That is one of the main issues of "winning in 4X":
Of course, human players in a boardgame would do just that (and 4/5th of the time, overcommit to remove the most obvious threats and still grand an easy victory for the next in line, but I digress). Players usually expect the AI to LARP in these situations, and not play the metagame to win.

I've been that player before. It usually resulted in gitting so extremely gud that people give up entirely on winning first place and consider it a win to be second place, instead coming to view my presence in the game as a force of nature or elder god to use in their plays for second place.

This is especially the case if you make a point of making sure that if they DO try ganging up on you, ONE of them is going to DIE, and which one of you wants to be that person? You? You? Yeah, I didn't think so. The winning move of the best player in such a scenario is to trap the others into a Prisoner's Dilemma situation, where the ones that work against you, are most likely to suffer horrible fates and come out personally worse, but the ones that collaborate and betray the others will come out ahead for sure, though not first.