On Autoqueue For Macromanagement - And How to Evaluate Accessibility Features For RTS
I think this is a tricky subject where my views have evolved with time
Age of Empires IV launched on Xbox home consoles a few weeks back. As part of supporting a gamepad control scheme, the developers added a number of accessibility features that aren’t available on the PC version, such as automatic queueing of workers at town centers and automated economic balancing based on a player-defined resource split (e.g. 50% food/25% wood/25% gold).
Naturally, people started to discuss whether the same accessibility features should be ported over to PC. The response was… negative. And to me that’s not super surprising, because folks in the core RTS audience sometimes react negatively to new accessibility features in existing games. It’s a funny sort of dynamic where players tend to interpret existing accessibility features as good old fashioned good-game-design, but see new accessibility features as “dumbing the game down”.
I get it; I used to be broadly opposed to most new accessibility features, too. And I think that was a fair perspective with merits; but I realized over time that it was also a little narrow-minded. This week I want to write a little bit about how my views have evolved and talk about the framework I use to think about accessibility features nowadays, which I like to think is a bit more neutral and fair and (hopefully) aligned with better player experiences.
How Players Spend Time
I think it’s reasonable to frame real-time strategy gameplay as a prioritization exercise: out of a million different things a player could be doing, like managing their economy or controlling their army or building new units or researching upgrades or what have you, they select what deserves their attention at the present moment. And the reason this prioritization is important is that real-time strategy games run, well, in real-time, meaning that wheels are always turning in the background regardless of whether you’re consciously aware of it.
To use a simple example, if you’re busy dealing with an opponent’s incoming attack, that’s occupying a chunk of your time and attention that can’t be devoted to, say, rebalancing your economy. At least, not at that particular moment. And while your attention is preoccupied, your (unbalanced) economy will continue gathering at its current distribution of workers, perhaps creating a lopsided surplus in your resource reserves.
(Now of course you can caveat this by observing that RTS games have multiple input devices - a keyboard and mouse - and that the keyboard specifically enables players to do a lot “in the background” while their camera is focused on something else. And I think this is a super interesting segway into the difference between how players segment their time and how they segment their attention. But here I’m just concentrating on the idea that players can only do so many things at once and that they are continually tasked with prioritizing their actions. I kind of speak about time and attention interchangeably to keep it simple, even though technically I shouldn’t.)
Building on this, if you string together a bunch of these present moments across a chunk of time - say, a minute or two - you end up with a distribution of how players spend time. Maybe it looks something like this:
Now this is pretty contrived, so let me try to use a real-world example. Let’s suppose you’re looking at a span of gameplay in which a relatively novice player wants to take a big, 100-unit army and attack their opponent’s base with it, all at once. What might that chart look like in StarCraft: Brood War?
That’s kind of it, right? You can only select a handful of units at a time and those units require a lot of clicking to get from point A to point B. For a novice player, they’re probably devoting all of their time and attention just to completing these tasks, because they’re mechanically challenging.
Now, what does it look like in StarCraft II?
StarCraft II has unlimited unit selection, good pathing, and strong unit AI. It’s really easy to move your entire army from point A to point B, meaning you can spend a lot less time and attention on it. Personally, I would opt for PB&J in this situation.
Of course, I’m meme’ing. And, importantly, the story doesn’t end here; I know from experience that this is where a good number of folks rush to the comments to proclaim that actually, there’s way more to army control in StarCraft II than just a-moving. Well… yeah, obviously. Thanks for your contribution. The point here is that given an equal amount of time and attention, the player gets more mileage out of the controls in StarCraft II than they do in Brood War. And that frees up their time and attention to do other things; and in the case of a novice player, I genuinely think that can mean kicking back and enjoying a snack while their army melts everything, because hey, even I do that during the campaigns sometimes.
Intentional Design
A point I’ve tried to make over the years is that freeing up players’ time and attention via accessibility features impacts competitive and casual players in different ways. Casual players are just trying to play and enjoy the game, beat the mission they’re on, stuff like that. If you make the game easier, that doesn’t mean they’ll fill the newly freed up time with more tasks. They might sit around and think more; they might spend more time marveling at the pew pew of a game’s graphics; they might literally eat a sandwich. Better controls empower them to do what they’d like, and “gameplay-as-a-prioritization-exercise” may not even apply as a useful mental model.
Competitive players experience accessibility features differently. For them, the goal is to maximize the value of the controls and saturate their time and attention with as many value-adding actions as possible. Accessibility features that simplify the gameplay do not actually make the competitive experience easier; they just make it different, because the freed up time shifts to a different set of actions. Time and attention simply get divided up in a new way.
And this is a crucial observation. Different is not better; it is just different. The question is whether the skill ceiling design following an accessibility improvement is superior to the skill ceiling design before it.
Let’s finish the StarCraft II example. The game adds innumerable niceties in its game engine compared to the RTS games that came before it. This is great, but from a competitive perspective, it meant that the developers had to plumb other venues to keep the skill ceiling high. What they eventually arrived at in Legacy are design features like an emphasis on multi-pronged attacks that are easier to execute than to defend, the addition of game-ending mechanics like disruptor shots that demanded ungodly army attention from players, the use of design patches to constantly mix up the meta to prevent it from solidifying, and so forth.
This is all well and good, and as I’ve pointed out, over and over, it was phenomenally successful. But from the point of view of competitive players, StarCraft II saturates their time and attention just as much as any previous real-time strategy game did. It just saturates it in a different way.
(No sandwiches to be found here.)
I adopt a similar perspective when evaluating the addition of autoqueue to Age of Mythology way back in 2003. Offered as part of the Titans expansion pack, autoqueue enabled automatic production of units at any building. It was certainly convenient, but it wasn’t paired with any new changes to the skill ceiling design, in a game with a large macromanagement focus. The result was a game that genuinely did feel “dumbed down”, with the unfortunate side-effect of balance problems taking on a larger role in the game’s meta. It was substantially more difficult to outplay your opponent when they had a mechanic in their backpocket that guaranteed near perfect macromanagement. And I think the inclusion of autoqueue, more than any other game design decision, is what caused the exodus of many top players from Age of Mythology after The Titans expansion was released.
And so it was from this perspective that I became generally skeptical of new accessibility features, because it seemed so challenging to shoehorn them into existing games without unintentionally worsening the skill expression implementation. Worse still is that developers often tacitly assume that accessibility is an unalloyed good for casual players; and as I’ve written before, that’s not a safe assumption to make.
But Wait, There’s More
Anyway, I feel that everything I’ve said up until now is reasonable. I don’t begrudge anyone whose views on RTS accessibility end here. But for me personally, two things shifted my views in a new direction.
One is the longevity of games in the competitive RTS space. StarCraft II and Age of Empires II have been out for 13 and 24 years, respectively, and yet they continue to be market leaders (again, specifically in the competitive space). It feels to me like it’s mildly unreasonable to say that fundamentally modifying a game’s skill expression design is off-limits, when people might play that game for more than ten years. It changes the calculus of what risks are worth taking: even a year of worsened gameplay as the developers figure out a new mechanic might be worth a longer-term transition to a significantly better competitive experience.
Two is becoming a new fan of Formula 1, and observing how that sport periodically modifies its regulations to keep competition fresh and displace top-performing teams. That of course comes with natural ups and downs (not all of which I fully understand as a new fan), but it seems to have been beneficial to the sport’s long-term health. It shows that well-managed intentional changes can succeed even in environments with lots of moving parts (no pun intended) and already-baked-in physical infrastructure.
Basically, I came to the conclusion that long-running live-service games require a different perspective from games that might only be on the market for a few years. It’s worth taking chances to re-balance, re-design, and generally re-fresh the overall skill expression design for the benefit of a game’s long-term health, because short-term churn is worth years of improved gameplay. And that opened my eyes to the notion that even features like autoqueue - features that in the past arguably hurt a game’s competitive scene - might have their place, so long as they come paired with an overall plan for a game’s skill expression design.
Nowadays, I like to think my perspective is a bit more open-minded. My ask is simple (I hope): if you’re proposing a new accessibility feature, then how do you intend players to spend their newfound time? What’s the before and after time and attention split, at least at a high level? If you can offer good answers to those questions, then maybe you’re onto something.
Let me try to illustrate with two examples, starting with autoqueue for workers. The challenge with stuff like this is that producing workers is often part of a real-time strategy game’s pacing design; it’s part of the rhythm of the gameplay loop. It’s a tactile part of the building experience, which differentiates real-time strategy games from more detached city builders or management simulations.
I think if you want to bring in autoqueue for workers, that’s reasonable so long as you account for the game’s pacing and replace manual worker production with a more compelling mechanic. I don’t really know what that mechanic would be in the context of a game like Age of Empires IV, so I generally lean toward keeping this particular manual action in-place; but my point is that if this is something you’re passionate about removing, this is the approach I think would be most successful.
A more interesting example is the skill expression design in Age of Empires II - specifically, its complex and challenging Dark Age. I mean, I do in good faith understand the critiques that the game starts off too slow, but I also quietly wonder whether the people saying this stuff have ever really tried to pull off a perfect Dark Age. It’s not easy! Actually, it’s pretty frantic. And over the years one of the things I’ve struggled with is the fact that villagers in Age 2 often don’t behave the way you’d expect. They require manual handholding to work consistently, and if you’re trying to pull off one of the tighter meta builds (e.g. 18-pop 1-range archers), you need to babysit them. A few examples:
Lumberjacks getting stuck
Lumberjacks walking all the way around the lumber camp to get to a new tree (also causes the wood line to become unbalanced)
Villagers killing extra sheep unnecessarily
Food villagers occasionally needing to be garrisoned/ungarrisoned because they chose a strange way to path after a manual food drop off
Boar villager pathing in a strange way and needing to house block the boar
… etc
Some of these are bugs, some of these are bad AI, some of these are just players chasing efficiency. But personally, I find them tedious, and I don’t know if I would call them a compelling part of the game’s skill expression.
Now, I think you could justifiably fix most or all of these issues outright with no other changes, and the game would be fine. I think many people underrate how much power this would take away from the very top players who’ve mastered this stuff, but that’s sort of an irrelevant aside. Point is, nothing about Age 2 would break if villagers ceased to be complete putzes. (At least, I don’t think.)
But just for the sake of completing my thought on skill expression, I think the Dark Age would be more interesting if this tedium were replaced by a new form of skill expression. And again - just for the sake of argument - I’ll suggest an idea: adding a second scout.
MegaRandom frequently features a second scout in its map generations, and I personally love it. For novice players, they could just auto-scout - no problem at all. But for competitive players, there’s so much to think about. The devs could make laming more defensible, and then it could become a legit back-and-forth of each player fighting for their opponent’s boars. You’d be able to scout more consistently with sufficient multi-tasking ability. There’d be more interesting pressure attacks you could do with the Feudal Age transition with two strong scouts, potentially preventing full wall-ins more easily. The game - in general, and with the right tuning of early game laming - would become faster-paced and more aggressive, and yet simultaneously play out more consistently and with less luck - less weighed down by chance lack of scouting, missing of early game sheep or being lamed.
I mean, I don’t know if it’s a good idea. Certainly it would require a lot of re-balancing and re-designing of the Dark Age to arrive at a happy state. And even then, it’s probably still a bad idea! But it sure sounds better than fixing stuck villagers on wood, doesn’t it? And it’s the sort of thing that would take a lot of time and resources to get right, but I’d argue the end result has a good shot at being meaningfully better than the current Dark Age.
Putting this idea aside, the broader point here is that I think there’s potential for Age 2’s Dark Age to be significantly better. I haven’t played the game seriously in a long enough time to offer a fresh opinion on whether that’s where the skill expression design could benefit from the most love. But I think it can be noticeably improved, and we shouldn’t assume it’s in some sort of perfect final state.
Final Thoughts
I think skill expression design benefits from being both wholistic and intentional. I get that I probably underrate the value-adding potential of organic development in this area - too many years of watching good competitive games descend into stale or unenjoyable metas - but I also think there’s real merit to a steady hand from a committed developer continuously pushing a game in a better direction.
I spoke earlier about the longevity of games like Age 2 and StarCraft II. I don’t think that’s an accident - both titles benefitted from numerous years of development support. For instance, the skill expression of today’s pro-level Age 2 - the actual stuff that people are good at, not just how good they are - would be unrecognizable to someone in 2000 or 2001. And that’s in large part thanks to slowly upleveling the title’s gameplay design across updates and DLCs.
I think we should feel comfortable applying the same lessons when looking forward, too. I get the concerns around new accessibility features - the dumbing down of gameplay, the risk of ruining the competitive experience. But I think these risks are worth taking. It’s worth investing in making skill expression better in existing games on the market, even if that means making big changes or redesigns. Ultimately, we’re not designing for next week or next month; we’re designing for the quality of gameplay years from now. And that ought to change our calculus of how we evaluate risks of major changes.
Maybe - just maybe - it could even mean bringing back autoqueue.
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.