Hey folks,

I’m brownbear. Today I’ll be reviewing Axiom, a mod for StarCraft II intended to lower the game’s barrier to entry and provide a gateway to competitive play.

Background

Released in October of 2017, Axiom was designed by TotalBiscuit and built by a team of developers over the course of several months. It was teased prior to release and had a ton of hype going into launch: Polt played it on his stream, the official StarCraft account retweeted it, and the Reddit thread announcing its release received over 400 upvotes.

Axiom makes several bold claims, like the notion that it’s fun to play with someone much better or much worse than you. It further argues that everyone should be able to enjoy competitive StarCraft. I decided to dig further and spent the last two months playing dozens of games of Axiom – all three supported maps, all three races, a bunch of different styles, with both friends and strangers, and on both EU and NA. I engaged with the Axiom developers on TeamLiquid and Discord, talked to my Axiom opponents in-game and discussed the mod’s design ideas with friends. Basically, I tried as hard as I could to give it a fair and neutral shake.

Theme I: Trading one complexity for another

Axiom makes many changes to the base StarCraft II experience, so I’ve organized its feature set into several themes. The first theme refers to Axiom features that attempt to streamline gameplay and reduce complexity, but which ultimately trade one complexity for another. From my perspective these changes make the gameplay experience substantially worse – in the best case, they’re net-zero.

Let’s start with the new economy. In traditional StarCraft II, players pay for everything up-front – before you can start construction of a command center, you need to pay four hundred minerals. Axiom instead uses a pay-as-you-go system in which payment for items occurs as they’re being constructed. This shifts focus to revenue- and cost-per-second – the net total is displayed under a player’s minerals and vespene gas.

The rationale for this change is the avoidance of “dead money”. In traditional StarCraft II, when a player queues 5 marines at a barracks, they pay two hundred and fifty minerals. However, only one marine is produced at a time. This means there are two hundred minerals worth of underutilized production. With Axiom, you’re only ever spending money on items that are actively being produced. In addition, players can begin construction of any building or unit at anytime, theoretically allowing them to focus on what they should build rather than the nuts and bolts of how much things cost.

The downside of the pay-as-you-go-economy is that it introduces a new complexity: overbuilding. In Axiom, if you commit to spending more per-second than you have – and you don’t have a mineral and gas bank – all of your production deadlocks and nothing gets produced. This requires you to either wait for everything to finish (which can take ages) or cancel items until your net-income is greater than zero. Naturally, this never happens in the base game, where everything is paid for upfront.

There are a couple of problems with this. First, it sounds obvious that your costs shouldn’t exceed your revenue, but it’s surprisingly tricky to get right. If you’re in the middle of a fight, frantically queue up a bunch of what you need, and suddenly discover that your production is deadlocked, it can feel really frustrating. Axiom encourages players to focus on what they’re constructing rather than the nuts and bolts of how much things cost, meaning you’re incentivized to make this kind of mistake.

Second, cancellation is jarring because it mixes together different macromanagement design patterns. Axiom’s Universal Command Bar allows players to construct anything from anywhere, meaning they don’t have to go home to macro. But cancellation means going back to in-production items and canceling them manually, the same process as the base game. It’s really confusing to need to know two ways of doing things instead of one.

The overbuilding problem might be a relatively minor issue, but it’s worth noting that the dead money benefit is also small. Only higher-league players play the game with enough precision to necessitate that level of economic efficiency. In other words, Axiom trades a minor advantage for a minor disadvantage. But the disadvantage – overbuilding – affects everyone, not just higher-level players. In fact, lower-league players have much less understanding of how much things cost, making it easier for them to overbuild.

This is a theme I see throughout Axiom, which I’ll label professional perspective bias. Axiom frequently looks at StarCraft from the angle of top-tier competitive play and then designs accordingly. I don’t find this surprising given that high-level players were part of the playtesting team. The problem here is that complexity is relative to your goal – nuances like dead money simply don’t matter for most players, whereas accidentally building more than you can afford affects everyone.

As I mentioned earlier, the Universal Command Bar suffers from a similar issue. It does remove complexity – it allows players to macro from anywhere on the map and construct buildings without worrying about who’s building them. But the new complexity of learning two macromanagement design patterns nullifies these benefits.

The key here is that dumb simplicity beats smart but wrong complexity. It’s great that Axiom tries to provide a smarter user interface, for instance by offering the scan button globally or allowing a player to research upgrades from anywhere. But the interface just isn’t smart enough, meaning that certain interactions – like training workers to the right location – become more complicated than they were before. The “dumb simplicity” of going back to your base to macro is not only easier, it’s also consistent. Players leverage the same design pattern across the board, whether they’re training units, building production buildings or researching upgrades. This is less confusing than a user interface trying to be smart but ultimately getting things wrong.

Consider a bug I frequently experienced in Axiom, wherein a newly built command center would not automatically produce workers. To resolve this, I manually constructed workers from the training tab – but this would produce workers everywhere, not just at my new base. As a result, I’d have way too many workers at my main, but not enough at my other bases, forcing me to manually move workers. This is new complexity. Zerg faces a more severe version of this problem with its queens, which also get built everywhere. You can work around this problem by selecting a hatchery and building a queen locally, but not only does this contradict Axiom’s design goal – because it forces you to go back home and macro – it’s really confusing to have two different buttons to build queens.

Global Rally Points are another good example of this “smart but wrong complexity” idea. They allow players to globally rally units through pre-designated groupings. But the groupings don’t make sense – there’s no reason for Starport units to rally separately from Barracks units because bio pairs with medivacs. Zerg is worse, splitting ranged and melee units into separate rally points. This is not how most of Zerg’s army compositions work.

The base StarCraft II experience, from the perspective of a new player, is much better – manage your army as one big blob with the select-all army hotkey, and maybe use rally points if you know what you’re doing. It’s dumber for sure, but it’s also easier and more correct. It doesn’t “intelligently” lead you down the wrong track.

Theme II: Design misunderstandings

The first theme included features which trade one complexity for another, affecting the gameplay in a net-zero or even arguably net-negative way. The second theme covers features which seem to misunderstand the design underlying StarCraft II, leading to “fixes” that are not good.

Let’s start with Smart Creep. Here’s how Axiom defines it:

Spreading creep is essential to Zerg victory but can be overwhelming to manage as a less experienced or slower player. Axiom features a smart creep system, in which creep tumors will spread automatically in a logical way across the map, after a player has manually placed their first using a queen.

Most Zerg units move faster on creep and therefore perform better in fights. This is especially true for melee units like zerglings or banelings that need to physically connect with opposing units. Zerg is therefore not balanced around always fighting on creep. Off of creep, they would be too weak and have no ability to push into their opponents. On-creep, they would be too strong and too difficult to break. Neither of these are desirable outcomes.

Instead, Zerg is balanced to work both on- and off-creep. The developers built-in a trade-off: creep makes Zerg units more effective, but spreading it requires time and attention. In your average competitive game, you have roughly the same “amount” of attention as your opponent, meaning how you choose to allocate it defines your particular playstyle. If you spend your time and attention spreading creep, you’re not spending it doing something else.

(You could apply similar logic to injects, which Axiom also automates.)

Spreading creep automatically removes a mechanical burden while retaining the benefits. This doesn’t make any sense without an appropriate counter-balance. Furthermore, even at my level (Master’s 3), it’s common for Zerg players to spread very little creep, causing me to question the original assertion that “spreading creep is essential to Zerg victory”. This seems like another example of professional perspective bias (where spreading creep is essential), mixed with a lack of knowledge as to how StarCraft II functions.

The redesigned supply system also speaks to a misunderstanding as to why supply exists. While I appreciate that Axiom tries to force players to build more production in order to get more supply, the complete revamp of the supply system is the final nail in the coffin for the game’s pacing.

When most people think about pacing, they imagine story-driven games: the ups and downs of the narrative arc that define a player’s journey. But really, every game benefits from good pacing. Pacing is just the flow, the feel of playing a game on a moment-to-moment basis.

Many of Axiom’s features interfere with StarCraft II’s built-in pacing. First, Axiom eliminates the macro cycle concept by allowing players to train units from anywhere in any quantity instead of rhythmically going home to macro. Next, it eliminates the need to regularly build supply structures. Finally, it encourages players to focus on what they’re building rather than the nuts-and-bolts of how much things cost, meaning there’s no logical flow to when players begin constructing buildings.

In-aggregate, all of these features add up to an experience that seems to vault back and forth between frantic activity and dead silence. I’ve heard the same criticism of the base StarCraft II experience, and to some extent I agree with it – many interactions in the game have a certain “suddenness” to them. Rather than correct this problem, Axiom exacerbates it by removing the hum-drum work of going home to macro.

Real time strategy games have featured supply structures for twenty years, and it’s not an accident. It’s part of how games are paced. Titles that removed this system (like Rise of Nations) tended to focus on other things, like a much more complicated economy – in other words, they tried to come up with interesting things to do other build than supply depots. Axiom doesn’t do that, leaving a hole in how its games flow.

Supply structures also play a key role in providing incremental value. If a strategy had to kill an opponent in order to be viable, most strategies would become unviable. Supply structures help close this gap by allowing players to find value even if their attacks aren’t game-ending. A roach timing might not kill a Terran player or force out enough units to justify its own cost, but it can kill a couple hundred minerals’ worth of depots and force a supply block. I don’t think it’s an accident that each race’s supply structures play a critical role that force them out to exposed locations – or, in the case of Zerg, all around the map.

Before I go further, let me make my clear that I don’t believe supply structures are a must-have in order for a real time strategy game to work. Rather, I’m pointing out that they do play a critical role in StarCraft. Whittling that role down without an adequate replacement ultimately removes compelling gameplay from the experience.

Finally, I’d like to discuss the free workers. Axiom justifies this by noting it tones down the impact of economic harassment:

The destruction of workers will still massively impact a players income, but losing a mineral line is no longer the death sentence it so often once was.

First, a big impact of losing workers is their production time, not just their cost. Second, StarCraft II’s emphasis on economic harassment stems from many factors: the three-base economic soft cap, the importance of cost-efficiency, the relative lack of incentive to engage an opponent directly, etc. I’ve written about this at length, as have numerous other authors. Axiom doesn’t address any of these issues.

Most importantly, in combination with the pay-as-you-go economy, free workers incentivize players to expand across the map as quickly as possible. Managing this feels much more difficult than the 1 or 2-base economies you see in lower-leagues. I don’t think it’s an accident the co-op developers limited themselves to 2-bases per mission (or, in some cases, 1).

If you skip the fast expansions and follow Axiom’s advice to “see a strategy in a competitive match from a pro-gamer [and try it out yourself]”, you’ll discover that the new economy is in some ways more vulnerable to economic harassment because losing mineral income disrupts anything that’s currently in-production. What was once affordable suddenly deadlocks, delaying your response and forcing out cancellation operations.

Theme III: Things that don’t make sense

Let’s start with Axiom’s commitment to provide the same maps as the pros:

Axiom will use a regularly updated, smaller pool of 3 maps. Every season we will choose the 3 most balanced maps in the season’s pool for our rotation. Players can choose any of these maps to create a lobby in. If you see a strategy in a competitive match from a pro-gamer, you can try it yourself in Axiom and see how it works out.

Committing to periodic development work fixed to an unreliable dependency’s schedule doesn’t make sense. This is not sustainable in a free project that does not generate revenue – priorities change, people find other interesting things to do, etc. It would have been better to select timeless, classic maps like King Sejong Station or Overgrowth.

You can see this for yourself – 2017 Season 4 is now a month-and-a-half old, but Axiom does not yet support any of its new maps.

Axiom also remaps all of the game’s hotkeys. This is a bad idea:

• Folks that do use hotkeys will be frustrated that all of the game’s hotkeys have been remapped. They are negatively affected by this change.
• Folks that do not use hotkeys – ostensibly the game’s target audience – will not notice that all of the hotkeys have been remapped. They are not affected by this change. You could argue that Axiom trains them to use hotkeys, except that doesn’t make sense because Axiom’s hotkey setup does not appear in any other game mode.

There are only downsides to this change. This was the #1 complaint I heard when I asked people for their feedback on Axiom.

Finally, Axiom seeks to “[take] away barriers such as high actions-per-minute and complex UI.” However, Axiom’s UI is substantially more complicated than StarCraft II’s. It’s visually overwhelming and contains loads of additional buttons.

I also didn’t find the production tabs metaphor to be intuitive: even after dozens of games I still found it jarring to look down, not see the thing I want, and after a second or two realize I needed to switch to a different tab.

Everything Else

There are a handful of other features that I’ll discuss briefly:

• Production Tab – cool idea, and partially bandaids the cancelation problem mentioned previously
• Automated Control Grouping – I was confused by this because I thought control groups were too complicated for Axiom’s target audience. In any case, this is offset by a clunky delay in control groups appearing on the UI after they are manually set, making Axiom feel unresponsive.
• Double HP units – while this introduces balance problems, I don’t think that matters for the target audience of Axiom. If people like it, then keep it. At least it’s an across-the-board change rather than something like automated creep and injects, which only affect Zerg.

Overall

Axiom misses the mark. Its features frequently trade one complexity for an arguably worse alternative, misunderstand StarCraft II’s underlying game design, or otherwise don’t make sense. I would discourage any new player from playing this mod, and instead redirect them to the Tutorial, Co-op, or Campaign.

For example, I got a non-gamer friend of mine into StarCraft through custom games and co-op. Despite little prior experience with RTS and relatively low APM, my friend now enjoys playing StarCraft as well as watching StarCraft esports.

Axiom is still in development, meaning some of these problems are fixable. I’d encourage the developers to more actively engage with the Axiom community: the last message in Discord was made on October 28 while the last new post on Reddit was written more more than a month ago.

I myself posted a question on the Axiom Discord a week ago and have yet to receive a reply. You could say this is a result of the holidays, but that’s yet another downside of pushing a community-supported mod over the default experience. Christmas is precisely the time when people open up new games and start playing them.

One thing I’ll point out is that, from my perspective, Axiom does feature a structural flaw that might be difficult to change with a balance patch: it’s 1v1. When I commented on the downsides of this decision, a commenter on TeamLiquid pointed to Hearthstone as a counter-example. This is not a good comparison because Hearthstone and StarCraft are very different games: Hearthstone features a large pay-to-win element, a large RNG element, is turn-based, and contains collecting and deck-building mechanics that are inherently enjoyable on their own.

A more general problem with casual 1v1 in StarCraft is that real time strategy games snowball. Unlike, say, tennis, StarCraft doesn’t contain a reset mechanism – a small advantage in the early game will snowball into a large advantage in the late game. To undo this, you’d need to change some core aspect of the game, like modifying the binary win/loss mechanic to a points-based system, adding a reset mechanic or adding a big RNG element. Axiom doesn’t do any of these things. Based on the large number of lame excuses I heard from players losing to me in Axiom, it doesn’t seem very fun to play with someone much better than you.

Historically, the most popular casual multiplayer modes in real time strategy were team games or free-for-alls, often on custom maps. I don’t understand why Axiom ignored this. The player pool seems to bear out this problem – over the past month my average wait time for an Axiom game was over five minutes, and a third of the time I couldn’t find a game at all. The feedback I consistently received in my Axiom games was this: “really cool, but not for me”. There was always some hypothetical other player that’s intended to enjoy this mod, but it was rarely the person who was actually playing it.

Motivation

If you’ve followed my work for any length of time, you’ll know that I rarely put together critical content. Personally I think it’s better to focus on what’s good, talk about why it’s good, and encourage more of that good stuff rather than tear things down. Praise in public, condemn in private, all that. Even in this article, I’ve tried as hard as I can to be neutral and charitable in my feedback.

I decided to write this piece for two reasons. One, I feel that Axiom exaggerates “problems” within StarCraft II in order to justify its own existence. I find this approach disingenuous given that StarCraft II is arguably the most successful competitive real time strategy game ever made. Here’s the first sentence on Axiom’s website:

Our goal with the Axiom mod was to provide an experience that captured the essence of what makes Starcraft and Real time Strategy so compelling, while taking away barriers such as high actions-per-minute and complex UI.

This is pointless vilification. No one needs high actions per minute to enjoy StarCraft: how do you think it manages two million monthly active users?

Axiom goes further in paragraph three:

Watching Starcraft is one of the best eSports spectator experiences there is and many viewers would love to play like the pros but simply can’t meet the demanding requirements of the base game. Axiom aims to change all that and let you play a version of competitive Starcraft that is easier to get into, yet gives you access to all of the tools and strategies that a pro-gamer would employ in front of crowds of cheering fans in an arena.

It’s not fair to characterize StarCraft as featuring “demanding requirements”, even on the lower-tiers of the competitive ladder – watch some Bronze League Heroes if you don’t believe me. It’s a strawman intended to make Axiom seem friendlier by comparison. Sorry, but I find this distasteful. Promote creative works on their own vision and merits, not by putting other stuff down.

(And this is only Axiom’s official marketing materials – out of courtesy, I’ve excluded the stronger language used by the developers on places like TeamLiquid.)

Axiom itself features a bunch of problems that remain unfixed more than two months after launch. Auto-walls are still really buggy, control group formation feels delayed, workers get overtrained onto one base, the map pool hasn’t been updated for the new ladder, the hotkeys were pointlessly remapped, etc. For a mod that brazenly points out “problems” in the base StarCraft II experience, Axiom doesn’t hold itself to a high quality standard.

Final Thoughts

My second motivation for writing this piece is to emphasize that StarCraft II is a really successful game. It’s a model for how to build a modern-style RTS that appeals to both casual and hardcore gamers alike. I think StarCraft II’s decline from the #1 esport to what it is today has led some observers to conclude that it must have “failed” in some way, but the truth is that it has a really strong legacy. More than seven years after its release, it has over three hundred thousand active ladder players and around two million monthly active users. For a real time strategy game, those are insane numbers.

There’s a ton you can learn from studying StarCraft II’s design – if you’re interested in my personal take, I recommend you check out my in-depth commentary series on the game.

Social Media And Stuff

Thanks for reading! If you enjoyed this piece, I’d love it if you followed me on Twitter and Facebook and checked out my game-design focused YouTube and Twitch channels. All the best and see you next time.