Autoqueue's Tremendous Tradeoffs for Age of Mythology

Looking at the pros and cons

Age of Mythology: The Titans came out 21 years ago, and its latest remake is set to hit the streets in just a few months. AoM brings back a lot of good memories for me, and I’m pretty excited for Retold. I’d honestly love to ladder a bit ahead of release, but I can’t seem to find anyone queueing for EE, and I can’t be bothered to dig out my old discs for Voobly.

Retold’s upcoming release got me thinking about autoqueue. Autoqueue is a feature in Titans that enables players to automatically re-queue units at production structures, both economic and military. It’s convenient, offering “perfect macro”* at the click of a button. Not surprisingly, that kind of thing wasn’t popular among the competitive crowd, and Titans struggled to hold on to top competitive players from vanilla; many of them returned to Age of Empires II.

I think it’s worth thinking about the pros and cons of autoqueue. It’s funny, because I think within many real-time strategy communities, you see routine calls for more simplification and more streamlining of macromanagement. Yet, very few mainstream games have implemented a feature like autoqueue as comprehensively as Titans did. I’m really surprised by this, even if my take on the feature is not all that positive.

Pacing Problems

I talk a lot about pacing on this Substack, because I think it’s an underappreciated aspect of the RTS experience. What a game feels like to play matters a lot, and how the gameplay is paced is an important contributor to that.

Part of macromanagement is periodically building new stuff; it’s an active process, not something that just happens on its own. You’re queueing up workers, queueing up military units, hotkeying the stuff that was already built, going back, checking out what’s up. It’s all part and parcel of the tactile experience of playing these games.

Tactile! Real-time strategy games are unique because you’re there, in the action, building stuff, moving stuff, doing stuff. And yeah, I understand that the more you play, the more this becomes a subconscious thing and you’re not necessarily mentally present all the time. But you’re still there, physically, and the stuff that you’re doing is still affecting you, even if you don’t necessarily stop and consciously think to yourself, gee, I should build some marines, shouldn’t I?

(It’s kinda like playing a shooting game - yeah, you probably internalize a lot of the mechanics after awhile, but that doesn’t make them any less enjoyable.)

I think autoqueue presents a problem because there’s no other game mechanic that steps in and replaces this tactile process. And I honestly think that’s intentional - I think the designers made an assumption that the machinery of macromanagement was somehow dull or boring or unnecessary. This theme shows up repeatedly in interviews with folks in Ensemble; like this one, prior to the release of Age of Empires II:

Archangel: What will you do, if anything, to give [Age of Empires II] more substance and length for two not equally matched players? Will strategic implementations be more of an issue, against fast moving of the mouse and remembering keyboard shortcuts, in determining the winner of a game?

Bruce Shelley: We hope that several new features such as permanent farms and formations will allow players to get better at strategy and not just at using the interface.

Or this one, several years later, prior to an expansion pack of Age of Empires III:

Age of Empires III: The Asian Dynasties is heading to store shelves shortly. What do you think of Age's latest incarnation and how different or similar it is to the original Age title?

BSC: I like TAD a lot. I think it is a very good add-on for Age of Empires III. I like it that India has finally got into one of our games as a civilization. I think that TAD is a nice blend of new and old, with the wonders back, for example. The biggest changes to me over the entire series are changes in graphics (nice 2D to stunning 3D), the gradual reduction in micromanagement, the increasing sophistication in the single player campaigns, and the continual innovation in gameplay features.

(emphasis mine)

I didn’t develop these games, so I can’t speak to what actually happened. But in reading these and other interviews, I discern an intentional design shift away from mechanical controls (“the interface”) and toward - as Mr. Shelley describes it - “continual innovation in gameplay features”.

And that’s all well and good; I don’t want to downplay the meaningful progress on the user interface. I fully agree, for instance, with that Age of Empires II quote - it definitely feels more comfortable to play than Age of Empires I, and that’s just comparing vanilla to vanilla. The Definitive Edition brings an order of magnitude more convenience, with better camera controls, superior pathing, and so on.

But I don’t think enough credit was given to the fact that part of the reason people play real-time strategy games is because they like real-time strategy mechanics. On a purely physical level, it’s fun to build stuff. It’s fun to select the buildings and press the buttons and watch what happens. It’s part of the tactile experience; it’s what you do, in the course of the game.

And that’s what I’m trying to drive at - periodically building and managing new stuff is part of the gameplay loop of playing real-time strategy games. You queue up some stuff, you go off and do other stuff, and you come back to look at the cool stuff you’ve done - and you repeat that, many times over the course of the game. The physicality of is meaningful, even if it becomes automated with time. And in its absence, the peaks and valleys of gameplay don’t translate as well into a fun end-to-end experience, at least from my perspective.

Competitive Conundrum

Autoqueue enables a higher baseline macromanagement capability in competitive play. In all RTS games, average competitive players - who are, themselves, much better at the game than average campaign players - struggle to maintain constant production. Just building workers and army units continuously is sufficient to be an above average player - even now, when these games have gotten more professionalized and competitive.

I understand the idea that this is a little bonkers. A caricatured way of looking at it is that the majority of ladder climbing is just building more stuff. I don’t think that’s a fair characterization, but I can see how people get there. And honestly, it is true, in the narrow sense that if you set out to get better as efficiently as possible, you start by continuously building workers and army units.

But, experientially, I don’t think that’s how most people approach competitive play. Sure, people grind and practice and whatnot, but ultimately I think most people spend most of their time queueing up and playing games. Your average ladder player isn’t sitting around doing mechanical drills for the majority of their playtime, or studying the meta and trying to theorycraft like it’s a college course. They’re just, you know, playing the game, or watching someone play the game, or chatting casually about playing the game on a message board.

And I think that’s worth remembering because within the narrow context of “people who just play the game a bunch”, getting better at the mechanical side is straight-forward, and getting better at the strategy side is not. Actually, I would say the strategy part of real-time strategy is really hard. And it’s not something you necessarily get better at quickly, or even at all, as time goes on. Take this breakdown, for example, of how Japanese men-at-arms into archers can play out in Age of Empires II, that I wrote about way back when in my Age 4 launch critique:

I won’t quote the five paragraphs from the original script (how the hell do I expect anyone to sit through my videos) but hopefully the picture captures the gist - developing the tech tree is civ-, map-, and situation-dependent. There are many civs that can open men-at-arms into archers, but they all play differently. And it’s incredibly hard to remember it all, let alone the particulars of your opponent’s civ, too.

It can thus be quite frustrating to grind out strategy improvements because a) it’s hard and b) you can still lose even if you get it right, because execution is still a thing, even in an autoqueue world.

The nice thing about mechanics is that they just naturally improve as you play more. You get a continual sense that you’re better, somehow, than you were 3 or 4 months ago. And in this day and age of everyone-recording-everything, you can go back and see how much better you’ve gotten. It’s fun, and easy, and low-stress to get better at the day-to-day mechanics of the game.

(A random aside, but it’s remarkable to me that we continued to see plain old mechanical efficiency improvements in pro-level StarCraft II a decade after release. You can go back and look at supply counts at various convergent points of different types of builds and it’s amazing how it’s climbed higher and higher on average. The skill ceiling on basic execution and build order efficiency is crazy.)

So, yeah, sure, ultimately the best players at the game are really good at using “the interface”, scare quotes and all. But that’s partly in the sense that people who do something a lot are going to get good at going through its motions. That doesn’t necessarily mean that’s the only thing the game is about, or that the challenge in executing those mechanics perfectly is somehow the totality of what the gameplay experience ends up feeling like. Actually, I would argue that by enabling players to improve via mechanical skill, you actually create the time and space they need to digest the strategy side of the game.

Autoqueue Alternatives

I mentioned earlier that I wanted to discuss the pros and cons of autoqueue; astute readers may be left wondering if I misunderstand the meaning of the word pros.

I don’t think autoqueue is a bad idea, and I certainly don’t think Ensemble had bad intentions with its inclusion. Rather, I think a feature like this has tremendous impact on the gameplay, to the point that the whole game needs to be designed around its inclusion. You can’t take the gameplay of a standard RTS and throw an autoqueue wrench in there; it just doesn’t fit.

I play the hell out of Factorio, for instance, and that game is nothing but autoqueue - it’s an entire game about automation. And I love it. I’m not some addict to pressing buttons, and I don’t think most other competitive players are, either. But in the context of competitive RTS, automation changes the gameplay so significantly that the rest of the game has to be proactively adjusted in response.

One simple idea I’ve tossed around over the years is simply nerfing the mechanic itself. What if autoqueue had a time buffer in between production cycles, or built stuff slightly slower than manual production? I think either of these ideas could work, but they strike me as annoying more than anything else. It’s almost like the game is goading you into interpreting it as a button pressing exercise; you’d need to playtest it carefully and monitor user sentiment.

I think StarCraft II’s solution is elegant in that it pairs numerous interface improvements with genuine new gameplay ideas, like a focus on mobility, multi-pronged attacks, and asymmetric mechanics. I actually think this is the “real” solution, as far as what Ensemble was looking for - a game whose interface truly doesn’t get in the way of a casual player.

But just to lean into it a bit more - if you’re really intent on automating production, I think the foundational role of macromanagement ought to be reconsidered. Maybe production, conceptually, no longer makes sense - design the game to be purely about unit choice, for instance (ala Direct Strike). Or, maybe production is just significantly downplayed relative to micromanagement - dumb down the units, harden up the rock-paper-scissors, and dial up the twitchiness.

Or, maybe something else. I had a few other ideas in this space, but I need to think them through a bit more.

Final Thoughts

I’m still really looking forward to Retold, and I’m actually not too concerned either way as to whether it includes autoqueue, despite any misgivings I may have on the mechanic. Personally, I rarely used it back when I played competitively; mostly, I just really want to ladder some Age of Mythology, and I’m OK if my opponents are using autoqueue to macro better.

I’ll add this, too: Ensemble’s commitment to gameplay innovations was really admirable. They took risks and exercised genuine creative vision with each game they put out. Not every idea was successful, but I think their fearless approach to creativity is the reason they put out such legendary games in the first place.

I wish every developer took swings as big as this one!

Until next time,
brownbear

If you’d like, you can follow me on Twitter, Facebook, and Instagram, and check out my YouTube and Twitch channels.

* “Perfect macro” is an exaggeration because macro involves unit selection (not just unit production) and, in particular, optimal composition selection (which may require stopping production at some points).