I’m sure most of you remember the novel-sized retrospective I did on the original Mass Effect trilogy that ran from July 2015 to June 2016. At the end of the series we were still looking forward to Mass Effect Andromeda and wondering how that would turn out. Since then the game has been released, received mixed reviews, became an industry-wide joke due to bugs and glitches, was patched up by the developers, and then faded from public memory. It began as a hot mess and ended as a disappointing footnote in the history of BioWare. Now I’m finally getting around to playing it, over two years after my original Mass Effect series ended.
Frustratingly, this game is neither as good nor as terrible as I’d hoped. This is not a return to the energetic worldbuilding of Mass Effect 1, but neither is it an affront to reason like the final act of Mass Effect 3At least, not the main story.. There are a few good ideas here, mixed in with the bog-standard gameplay, open-world busywork, and cringy dialog.
While Mass Effect 3 was at times frustrating and irritating, Andromeda‘s great sin is that it’s merely disappointing. There’s not much to get worked up about here. Mass Effect 3 had the problem where its story unraveled right at the moment where it should have started wrapping up, and it was the conclusion to a story we’d been following for three games over the course of five years. This game tells a stand-alone story. The stakes were lower this time around, and so were our expectations.
If you’re a longstanding fan of BioWare then one question probably hung in your mind as you played through Andromeda: What happened to this game? Janky animations, un-BioWare gameplay, linear dialogs, Ubisoft-style open-world stuff, pervasive bugs, oddly scripted cutscenes, and a host of other problems all work to pull your mind away from the game and turn it towards thinking about its development.
The previous games came from BioWare’s Edmonton studio, while this game was created by the new studio based in Montreal. I deliberately avoided looking at the credits during the writing of this seriesNot totally successful, mind you. Partway through writing this series I stumbled on a few names I recognized. Still, I did what I could to keep “the writer” as abstract as possible., so I didn’t know which (if any) writing staff or creative leads this game might share with the previous entries. As in my previous write-ups, we’re going to pretend that this story is the work of some singular unnamed writer. I find this makes it easier for the analysis to feel detached instead of coming off like an angry hit-piece.
Obviously I don’t have any special insider knowledge about the development of this game. I’ve never met anyone from BioWare and I’ve never personally worked on a AAA game. I’m stuck on the outside looking in, just like everyone else. Most of the behind-the-scenes information I have comes from the Kotaku article The Story Behind Mass Effect: Andromeda’s Troubled Five-Year Development. Sure, there are other sources of information out there, but as far as I can tell they all lead back to the original research done by Kotaku.
While I don’t have any insider information, I do feel like this series would be incomplete if I didn’t offer my own take on how things went wrong. As far as I can tell from reading articles and playing the game myself, the problems with Andromeda can be traced to three key decisions: Changing game engines, trying to make an open-world game, and ramping up the scope of a Mass Effect title to include much more content. All three of these are bad decisions, but they’re bad decisions that all exacerbate each other.
The first bad decision is that they decided to…
1. Switch to the Frostbite Engine
As the Kotaku article explains, the previous entries in the series used the Unreal engine. But for Andromeda, they switched to using EA’s in-house engine called Frostbite.
The thing about changing engines is that it requires you to retrain your entire creative staff. If you’ve got three dozen people who are all experts in using Unreal Engine to pump out game content, then switching to a new engine will effectively turn them into a bunch of rookies. They may have to change tools, which is hard on both productivity and morale. If you’re fluent in one 3D modeling program and you’re forced to adopt another, then you’re going to spend weeks struggling with what used to be simple tasks. The team will certainly need to adopt a new workflow. Sure, your artists can get the basics down in a few weeks, but they probably won’t be fluent in the new tools until they’ve shipped a game with them. (This game came from a new studio, so it’s possible they hired Frostbite experts. Still, they no doubt began with a seed team from Edmonton or Austin.)
The new engine will also oblige you to restart your entire asset library from scratch. Let’s say you’ve just spent the better part of a decade building up a vast collection of talking animations, ship models, weapon models, character models for several different species, models for “furniture”Computer consoles, crates, chairs, doors, and everything you use to fill in the environments.,and modular interior room sectionsSounds and texture maps should translate just fine. It’s the 3D models that are a problem.. If you were sticking with the same engine you could draw from that collection to fill in your gameworld, and maybe just touch up the oldest ones to make them appropriately “next gen”. Even if you end up remaking a lot of them over the course of development, it’s really nice to have the old content to serve as a placeholder during production or a fallback option if the schedule gets tight. But if you have to wipe the slate clean then you end up starting with a blank canvas.
Worse, the team was switching to a less mature game engine with fewer features. The Unreal Engine is about twenty years old and is a generalized set of tools to support many different types of games. Frostbite is ten years old and is mostly focused on linear first-person shootersAlthough the FIFA and Need for Speed teams were able to adapt it to their needs.. There are assumptions built into the engine that simply don’t allow for non-linear exploration-based content. You’re supposed to run through a level, shoot all the dudes, and then leave for the next level. The engine didn’t originally have allowances for the idea that the player might have a branching dialog with someone in zone ATalk to the quest giver on a space station. that will cause changes to the content in zone BOnce you have the quest, the bad guys spawn at their designated location on the planet. that can be reactive and persistentOnce I shoot the bad guys, they need to stay dead even if I fly to another planet. and which can make further changes to the contents of zone AOnce the dudes are dead, I can come back to the quest giver for the reward and wrap-up conversation.. All of that needed to be added.
This engine switch is a heavy burden to place on a team. All by itself, it would make for a challenge. But on top of that the team decided to…
2. Create a Procedural Open World
If not for the Kotaku article, I wouldn’t have believed this. You certainly can’t tell from playing the game. Apparently the team wanted to run with the idea of being a space pioneer and give you massive procedurally-made worlds to explore.
While this does indeed sound like a great idea for a videogame, it sounds like a horrible idea for a Mass Effect game, a worse idea for a team dealing with sweeping changes to their development tools, and worse still for a game built using the Frostbite engine.
Mass Effect is a game about telling science fiction stories using cutscenes and branching dialog trees. In a story-rich game like Mass Effect, the writer needs some control over the order in which you encounter content and the pacing of that delivery. I won’t say it’s impossible to do that in the context of a procedural world, but it’s never been done before in the AAA world and there’s no obvious way to reconcile these two vastly different ideas.
This idea ran against the genre of the series and the new engine. It’s an idea that leaves a lot of question marks in the design and requires a lot of R&D to sort out. It’s not something you should be messing with when you’re also changing engines and trying to rebuild your entire content library and trying to break in a new team.
This feature was obviously abandoned for a more traditional design, but it still ate up a lot of time and development resources. This is bad because they also decided to…
3. Ramp up the Scope
This is a big game. There’s a lot of content here. Sure, a lot of it doesn’t fit together and none of it is polished, but someone on the team felt like Mass Effect needed to “go big”. Possibly different creative people had different ideas about what content the game should include and what it should be about, and rather than settle on a singular vision they just decided to take two and a half conflicting game designs and blend them all together. Or perhaps the creative leadership just had runaway ambitions. Either way, we wound up with this oversized, unfocused mess.
- In terms of story, the main plot has you facing off against the bad guy for control of the space-MacGuffin. This is a bit like the core of Mass Effect 2, where you have a chain of fixed missions that are more or less divorced from the side content.
- The main plot is about the MacGuffin, but most of the gameplay is focused on making five different planets habitable and sorting out the local political conflicts to make the places safe for colonists. Each planet has its own questline to sort the place out. Back in my Mass Effect 1 write-up I divided the game’s story into “episodes”. Each of these five planets feels like at least an episode worth of content.
- As with previous titles, each of your six squaddies has a story you can go through to unlock their full potential. Unlike previous titles, some of these stories can span multiple missions and run in tandem with the main plot.
- On top of the main plot and the planet-fixing stories, you have the missions where you rescue the arks. These missions exist in a sort of neutral zone between the other content. One of them acts as a stop on the main story, and another is the personal mission for one of your squad mates, while the third is completely standalone. You can sort of claim these parts have already been accounted for above. But these missions have more in common with each other than they do with anything else in the game, to the point where they sort of feel like yet another independent questline.
- There are more romance options than ever before. You can romance five of your six squad mates, plus a couple of additional crew members, plus a few random people sprinkled around the gameworld.
- There’s this whole side-plot where the player unlocks Alec Ryder’s memories to learn the backstory of Alec, the Andromeda Initiative, and AI buddy SAM. It’s a very simple story (no branching) but it’s really heavy on content. Every memory is a full cutscene that runs for several minutes.
- On top of all that is a huge collection of random sidequests. Even after getting all five planets to 100% habitability and completing the game, I still had dozens of open quests and countless more quest markers on offer. I had emails offering me quests, people on the NexusThe central hub space station. offering quests, people on planets offering me quests, and even people on my ship offering me quests. None of them are particularly deep or engaging and they offer very little worldbuilding. Most of them follow the form of “click on three things” or “shoot three groups of dudes” or “find and scan three things”. It’s shallow content, but it still has voiced dialog and brief cutscenes associated with it. Which means these quests were eating into the time budget of all the other content.
That’s What Happened
On top of these three bad decisions was the unavoidable fact that this team needed to do some sort of soft reboot / reset to the Mass Effect universe, because making a same-setting sequel wasn’t possible and the public didn’t want a prequelAgain, according to the Kotaku article.. Any one of these decisions would have created challenges for the team, but when you combine them you end up with an insurmountable mess. The team was in a situation where making content was going to be harder than ever while also embracing a design that required more content than ever before while also messing around with experimental designs that required lots of prototyping while also breaking in a development studio that was at least partly new to the franchise.
I know it’s popular to dump on EA, but I don’t think they can claim more than a third of the blame for this. I’m sure the move to the Frostbite engine was their idea. Frostbite is EA’s in-house engine, and this move was probably done so they wouldn’t have to pay the licensing fees for the Unreal Engine. It’s entirely possible the accounting nitwit who imposed this decision on BioWare didn’t even understand why using Frostbite for a BioWare game would be difficult. You can imagine the technologically illiterate thinking behind this: Both games are about shooting dudes! How hard could it be?
But that decision wasn’t enough to kill Mass Effect Andromeda on its own. If the team had stuck to the tried-and-true design and limited the scope of the project, they would have enjoyed much better odds at making it out of development with a polished game.
But Shamus, isn’t it unfair to fault the developers for trying new things? They’re just trying to push the medium forward!
If this was a new franchise, then I’d agree with you. My problem is that they were “trying new things” that ran counter to the kind of game they were supposed to be making. If they had a design idea ahead of time that would be one thing, but from reading the Kotaku article it sounds like they began making the open-world stuff first, without any clue as to how they could reconcile that with the linear story they wanted to tell. “We’ll figure it out later” is a pretty dangerous attitude, particularly when you’re also trying to change to a game engine with no support for “open world procedural” content.
A Lack of Polish
What we ended up with is a scattershot game that’s being pulled in six different directions at the same time, built atop wobbly technologyYes, the Frostbite engine is pretty solid. But the Frostbite+RPG engine was still coming together at the very end of the project., where all of the content suffers from a glaring lack of polish. It’s a sad end to a long development cycle and possibly even the end of the franchise.
Over the next 22 entries, I’m going to go over the major story arcs of the game and point out the failings. I’m not going to do a deep dive on the entire game mission-by-mission the way I did for the previous entries. This is an enormous game and covering it in detail would take forever. Worse, most of my analysis would boil down to “This mission is poorly paced, poorly animated, poorly scripted, and contains too much repetitive busywork.” These problems are pervasive, and I don’t feel the need to document it every time they crop up.
 At least, not the main story.
 Not totally successful, mind you. Partway through writing this series I stumbled on a few names I recognized. Still, I did what I could to keep “the writer” as abstract as possible.
 Computer consoles, crates, chairs, doors, and everything you use to fill in the environments.
 Sounds and texture maps should translate just fine. It’s the 3D models that are a problem.
 Although the FIFA and Need for Speed teams were able to adapt it to their needs.
 Talk to the quest giver on a space station.
 Once you have the quest, the bad guys spawn at their designated location on the planet.
 Once I shoot the bad guys, they need to stay dead even if I fly to another planet.
 Once the dudes are dead, I can come back to the quest giver for the reward and wrap-up conversation.
 The central hub space station.
 Again, according to the Kotaku article.
 Yes, the Frostbite engine is pretty solid. But the Frostbite+RPG engine was still coming together at the very end of the project.
Starcraft: Bot Fight
Let's do some scripting to make the Starcraft AI fight itself, and see how smart it is. Or isn't.
id Software Coding Style
When the source code for Doom 3 was released, we got a look at some of the style conventions used by the developers. Here I analyze this style and explain what it all means.
What did web browsers look like 20 years ago, and what kind of crazy features did they have?
Deus Ex and The Treachery of Labels
Deus Ex Mankind Divided was a clumsy, tone-deaf allegory that thought it was clever, and it managed to annoy people of all political stripes.
Fixing Match 3
For one of the most popular casual games in existence, Match 3 is actually really broken. Until one developer fixed it.