Why must 4x Endgames take so long?

I'm playing Rome: Total War again. It's always been my favorite Total War game, and I'm back into it because a) it's a game on my shiny new Macbook Pro and b) it's awesome.

Anyway I'm playing as the Red Romans, and I've basically got the game won. I control all of Western Europe and Britain; I've got a giant army in Italy just waiting for the Civil War to start. I've got endgame tech units. Nobody can come close to me; all of the other Roman factions combined aren't nearly as mighty as I am. And yet, I'm probably only halfway through this game!

The winning condition is holding 50+ provinces and Rome. I've got 30. Which means I need to conquer twenty more. Which, at this point, is asinine. It's busywork. Nobody can stand against my legions; nobody has been able to stop them in a long time. And yet the game's going to make me fight thirty more sieges and Lord knows how many more field battles before it lets me win.

Why do 4x games do this? From what I recall the same thing happened in Civ IV an V: I'd definitely have the game won, but I'd have another four or five hours of grinding out the victory. Civ at least had the option of factions surprising me by hitting other victory conditions (stupid diplomacy...) Rome doesn't really have any of that; while each faction theoretically has a game-winning condition they're all so close to "conquer everything" that it's obvious nobody is going to achieve it except me.

This is why I always play use the "short campaign" option in Total War. But, from a philosophical perspective, I agree with you.

Yeah not choosing a short campaign was a mistake. :(

I'm also always playing on a tiny map in Civ5 so the entire game takes only two hours.

Yeah. It's annoying; for the last while I've been inflicting typically 8:1 casualties on my enemies in battle. There's literally no way they can stop me at that rate. It's not that I'm a good player either; it's just that my units are teched up to hell and the game tends to reward a fairly conservative approach.

If you know you've won, why keep playing? Yeah, you miss out on the end-game victory screen, but that's not really worth another 10 hours or whatever. I don't think I've ever completed a game of Europa Universalis 3, but it's one of my favorite games ever.

It's missing adjustment to game realities in computer games.

The board game Diplomacy has very rigid combat rules, which generally mean that once you control more than half of the resources you cannot be beaten. Having one less supply centre (17 instead of 18) means that there is still a good chance for you to be defeated and completely eliminated if the other half of the map is unified against you.
In MP games with less rigid systems you need even more of an advantage as even controlling a majority of all resources (which can be economy, military, or otherwise flavoured) might not grant you a superior strategic position. For instance Greed Corp, a board game-ish computer game where two players working in tandem have a very good chance beating the leading player, even if he has a big edge.

Many strategy games always have a potential for inferior parties to create a cooperation power multiplier which makes it possible to dethrone a leading player. Lots of (good) boardgames have that as their core principle with player interaction/cooperation trumping ingame advantages.

The problem is this doesn't work with AI's. Having a player control 33% of the universe/world in a game against AIs means he will win. Game designers just fail/refuse to acknowledge it.

It's not always just a matter of becoming invincible, I think some games just don't scale well past some point.

There comes a point when short of spending hours on each turn, you're no longer capable of understanding how systems are interacting, even though your empire is still running, mostly on its own. I've seen this happen in Shogun 2, when you start idly building stuff in provinces just because you have the money but it's no longer a choice with effects you can observe.

I think it's a good choice to impose time limits on the player, otherwise there's too much temptation to farm resources and the game just drags on. This happened to me on my last XCOM campaign. There's a point where the aliens just stop being a real threat and the game gives you as much time as you want to build a squad of gods.

Both in this XCOM campaign and in my latest Shogun 2 FotS campaign I had no guarantee of victory, in the case of XCOM because the last mission is still hard.

Strategy games in general could use more restraint, they are often at their best in the early to middle game when there's an actual challenge and then just become busywork with huge unit stacks. Even in an RTS like Dawn of War 2, which is otherwise very elegant, the endgame gets confusing and unmanageable.

I really enjoyed this aspect of Atom Zombie Smasher. The points on the victory track go just quick enough so that you're not bored while still having a sense of progression and a full campaign arc.

In the Total War games (bearing in mind I haven't played Rome) it's the fault of having simplistic victory conditions. "Hold X amount of territories" means just that, regardless of relative strengths. While it would be difficult to fine-tune I do think the game calculating the likelihood of your enemies turning the tide (assuming a minimal degree of human competence) would be possible, and if that's very unlikely allow you to force a defeat.

It's also the fault of allowing the AI to generate severely underpowered armies quickly, which then attack the super armies you have traipsing across their land. Playing whack-a-mole with tiny armies as targets is just boring. If the AI was very competent this might not be such a problem, because they could fight a sort of rearguard action, doing enough damage to slow or cripple your advance. That would have the air of plausibility to it. But past a certain strength they're just going to be stomped and the battles only serve to waste time. This might be better represented by having the AI build up as large an army/armies as they can based on their current resources and then have those try to halt your advance. If you smash them in the ensuing battle(s), and you're close enough to their territories to give them little time to rebuild, they should recognise hope is lost and give in.

GoG's release this week of the original Imperialism prompted me this morning to listen to an old episode of Three Moves Ahead in which they do an Imperialism retrospective. And one of the things they praised it for was avoiding the endgame slog. They attributed that to Imperialism's UI, which they felt was streamlined in such a way that the endgame could be played as quickly and efficiently as the beginning. Not having ever played the game, I can't speak to the truth of that. But as a hater of the Civ games endgame slog, it got me tempted.

1. 4x game designers are lazy.
2. The vast majority of game purchasers never get to the end game.
3. The kind of 4x game players who do get to the endgame love tedious micromanagement.

While this may be generally true, I'm the sort of bloody-minded self-hating idiot who suffers to the end for the sake of it. There are probably other miserable souls like me.

There should be a self help group for people like us. Endgame Sloggers anonymous or something.

As for why this keeps happening, I think it's due to games not building in systems to limit the amount of doodads you have to deal with at any given point. Take xcom, for example: there was a conscious design decision to limit squad size because larger squads at endgame were really too cumbersome. I'm not saying that 4x games should just arbitrarily limit someone to x number of units or cities, but as the game goes on and you have more cities, units or whatever, there should be ways to group those interactive objects up so that you only have to deal with a smaller number of them at a time.

I keep hoping someone will make a game which increases in scope as time goes on. You start off micromanaging a single city, with handful of units. That progresses to a handful of cities and a regiment or three. After that, you start dealing with provinces and armies, and on and on.

Not all love is the same!

I don't think it is generally true.

I don't know if you can do that without making the game feel too disjointed. When the scale of your decisions increases by an order of magnitude, does anything you did before really matter? It's great that you were able to turn your tiny village into the leader of a big city state, but none of that is really relevant anymore. There is stuff like Crusader Kings 2 and Romance of the 3 Kingdoms 8 (?) where you can start as the lowest member of a court and try to gradually work your way up to King. You're always using the same types of units, though - just the numbers are going to get bigger.

I think adjustable victory conditions are a great way to handle the endgame slog. Set your own sweet spot.

I agree that it's not a trivial problem to deal with. Simple slapping on a higher level of city abstraction to a game like civ4 would be a terrible idea, but at the same time, the games themselves change their scale as they progress. Losing a single warrior in the early part of civ4 can be crippling, while losing nearly an entire stack in the latter part of civ4 can be considered acceptable. Providing ways to deal with that reality need not be any more disjointed than how these games already feel.

I agree this is a huge problem with 4X games, and indeed, with strategy games in general. I don't think it's monocausal - this really is a case of "all of the above". Bad AI that can't compete with humans past the early game; designs that equate power with quantity of resources (more cities and more armies, usually); a general tendency for designers to distinguish their game by adding more, and more complex, systems; "kill 'em all" or "occupy too much of the map" victory conditions... they really are all to blame, and I'm sure that's just the tip of the iceberg.

Leaving aside RTS for one moment, I do want to give props to the two 4X games that solved this problem -- Armageddon Empires and base Shogun II. AE is built on scarcity: before a match, you add units to your deck as you would cards in a CCG and by and large, those are all the units you'll get. So even if I'm sitting on a mountain of resources that would do Scrooge McDuck proud, I can still only deploy, say, the 3 Vengeance mechs I bought during the deck-building phase. AE also gives players who have run away with the game a handy "I win" button -- if you have a Scroogian pile of resources, you can bypass the endgame grind by building an H-bomb for a virtually guaranteed knockout blow.

Meanwhile, the short campaign in Shogun II is paced so that, when you hit the point of diminishing returns (around 17 or 18 provinces -- the "short" victory condition is 25), the game transitions to an endgame showdown against almost all the AI factions. Not only does this mean the late game is as tense and exciting as the early game -- you are in very real danger of wiping at the end if you're careless -- but it gives each campaign.... almost a story arc; it makes the whole game feel like an exercise in preparing your realm, armies, and alliances for that final battle. (Unfortunately, from what I've heard, the long campaign still suffers from the same problem described in the OP, but the short campaign is an absolute masterpiece of strategy game design.)

God don't even get me started. Rome doesn't provide a way to flag cities that aren't building anything (at least not that I can see). The only way for me to determine which cities have free build queues is to click through all thirty (and counting) cities on Every. Last. Fucking. Turn.

The part that burns me is they're sometimes tremendously ahistorical. Medieval is probably the best example of that.

Rome does that as well, with the Civil War. When you hit a certain point the other three Roman factions (along with the Senate) turn against you. The problem is generally that they're very lax in recognizing the threat; by the time they turn against you it's too late. In the game I'm in I'm still buddy buddy with my Roman friends, even though I control all of Western Europe. The time to have turned on me was when I was consolidating in Iberia; now I'm just taking random territories and waiting for the Senate to start the civil war, so I can steamroll with my (already prepared) armies over their lands.

Yeah - the ironic thing is that the Shogun II "realm divide" is clearly inspired by crossing the Rubicon, but I could never get that far in R:TW for exactly the reason you named in the OP!

It's even worse than that. In many TW games, it's "hold x territories on such-and-such a date". It doesn't matter if you're holding those territories 50 years earlier, you still have to next turn-next turn-next turn until you get to the chosen date. I've literally had to next-turn my way through over 100 turns just to get to the victory condition date. It's been a big flaw in many games in that series. I don't necessarily agree that Shogun II fixed that - they just traded one problem (getting to the date) for another (where everyone turns against you regardless of your history with them).

The OP does make a good point - most TBS games really hit a wall during the endgame. Even the best TBS game series of all time - Civilization - has always had endgame issues. It's a tough balance. You want players to be able to get that final victory, but not too easily and not too quickly. So balancing out that with overwhelming victory becomes a challenge - and most designers opt for the "safer" route rather than open up some quick-and-cheap early victory path. If one path to victory can happen substantially earlier than the rest (let's say you could trigger a diplomatic victory with the Apostolic Palace), then everyone just goes for that path and the strategy of the game is found lacking. So to balancing things out, designers generally make players play it out. Not to mention that you have more units/buildings/options available at the end of the game - something you've been building towards - so each turn takes correspondingly longer.

In fact, I can't think of any TBS game that doesn't bog down somewhat at the end of the game. I'm sure there has been one, I'm just not recalling it.

Civ at least has the option of multiple victory conditions. I did a LP on qt3 a couple of years ago where I was merrily rolling over the entire world and then wtf the Koreans won a diplomatic victory out of nowhere. They were a tiny pissant nation but the fact that I was totally neglecting the UN subgame gave them the win. I think I ended up backing up a ways in my savegames, boomed to nuclear weapons, and then nuked the shit out of them. Problem solved.

Rome's problem is that - especially in the long game - there's really only one victory condition, and it's conquer the world. Right now there's only one viable world conquerer, and that's me. I'm going to win this game, it's just going to take me forever to do it! The short victory conditions are generally more local (instead of ocnquering the world you conquer your little corner), which at least turns into potentially interesting races. I had a game a while back where I was playing the short campaign, I was about three territories from a win but my armies were strung out and I was faced with the need to consolidate when I got the message that Egypt was about to win. I ended up having to throw some inadequate armies at enemy cities and it turned into a pretty desperate grab, so that worked out well.

So maybe the problem (with Rome at least) is the sheer grandiosity of the victory condition. Satisfying as it is to see the entire world map turn red, it just takes forever!

Having just concluded a siege: one thing that keeps me playing is the opportunity to snipe the enemy king with onagers throwing fireballs. That's always hilariously satisfying.

How to build a less shit 4x experience:

Step 1 hire Vic Davis
Step 2 obey him
Step 3 have raunchy sex with a dozen hookers on a pile of money after your game goes triple platinum

For me, the end game comes down to how well I can process all the information I have on hand. It seems like a simple problem that as a game progresses, there is a lot more to take on board. Early stages, it might be easy to know what is happening with a small selection of units, a handful of cities and the particulars with my neighbours. But there will be a point at which I'll suffer mental fatigue because the amount of extra information coming in becomes too much, or I have too much enthusiam for the next turn and break down pressing the enter key rather than micro managing some lesser task. The Total War games are terrible in that regard. All too often, I'll forget that I have some diplomat/princess/assassin stationed off in enemy territory because I'll get caught up in some other conflict halfway across the world. I've only done the short campaign for RTW once, and even then I was happy to have finished it as I felt I was done many turns earlier. I'd reach the point of apathy in some aspects of my empire management as they fell right down to the bottom of my priorities list, whereas early game, I'd be looking into such a problem with a lot more detail and fine tuning if needed.

What made Civ IV a lot easier in the end game was the BUG mod which presented a lot more information in a much clearer way. If I can have information available at my fingertips, and a means to easily compare x vs y, or see what has the most z or who hates these guys the most, and why do those guys hate me, then I feel empowered to do more each turn. Rome Total War etc felt more an exercise in frustration when it came to important information, such as knowing when there were idle units, being able to see what would cause my treasury to crash (or what cities/provinces were underperforming) or even something as simple as knowing which province I had specialised, if I had a few days break from the game.

Despite the many frustrating and poorly thought out parts of Gal Civ 2, I did like the part of its UI that made empire management on the whole much easier. I could set rally points when my ships were produced, or ships of a particular make at the very least. I also had a fairly easy way of seeing how my planets were performing, and it was dead easy to specialise them right from the start by queueing up the improvements needed.

Alpha Centauri definitely becomes a slog / bog in the endgame, but damn if it isn't an enjoyable one.

There's an inscrutable little icon on every city that is building something, so you can use the lack of an icon as a clue.

But, yes. Basically the game becomes micromanagement hell after about 20 or so provinces.

Imperialism 2, while not entirely bogless, handles the "scaling" into the late game far better than most 4X games.

Distant Worlds had the benefit of a large amount of potential automation, but even with all that, it suffered from bogging late in the game. (Though, part of the problem is that the computer is not as smart as you would like it to be, so you end up wanting to handle more things manually).

The simple answer to the OQ is (as Jason also mentioned):

The vast majority of games (including those played by gametesters) never, ever reach the end-game.

Think about it. Many - if not most - 4X games have game lengths from start to finish of 12 hours or more (Civilization 1, for instance, was around 600 turns long - at an average of 2 minute/turn, you'd still end up with 20 hours of gaming). It takes a special caliber of gamer (not necessarily a compliment) to play a game of that sort to completion. And testing the end-game? Forget about it - can't be done effectively. Presumably, developers are happy if any of the testers even reached the endgame a couple of times.

Armageddon Empires is able to handle the endgame well because of two things: limited actions and limited turns. And even then, the game is often over a fair while before you see the victory screen.

That's because 4x games straddle an ugly line between sim and game. If they picked game, they'd be a vic davis style experience. If they picked sim they'd be closer to ck2/vicky2. Instead they're ugly boring frankenbeasts fixated on BIGGAR THY NUMBARS gameplay.

Civilization offers many different map sizes, and most turns only take seconds on the smaller ones. I routinely complete entire Civ5 games in two hours.

That's because Civ5 is a lot more streamlined game experience than any previous Civ game. And - presumably - the short game version was tested a lot more and a lot better than the large map games (at least, that seems pretty likely based on the game mechanics).

I don't like playing on smaller maps or with faster turns - for me Civ is about building a great civilization, not winning a game. And smaller map sizes and all the other stuff makes Civ more like a game to me. Unfortunately that does mean, that the endgame becomes quite a slog - especially if you know exactly what you have to do to make victory a certainty, but actually doing it presents no challenge. But then I'm too much of a gamer to just declare victory - I need the graphs and cheevos before it's real.
Stupid me.

How is Distant Worlds for this? That had quite a few victory conditions, and I recall some of the early complaints were that games ended before they began.

It didn't launch in that state though, but yes, the UI for the late game was much improved upon.

Well, it depends. If you want to follow the full "plotlines" from a one planet tech 1 start, the game is inherently going to be dragged out because the events involved are just set up to take a long time. The victory screen will probably have appeared hours before that. If you simply want to achieve conventional victory, it typically doesn't take that long to do so.