The other night, after 5 years of putting it off, I finally completed Mass Effect. I see now why it’s so well-regarded.

Briefly, here’s what I think was awesome:

  • Elegant and succinct core plot. Mass Effect’s core plot was actually rather short; about 8 hours of play by my count. This kept it sharp and elegant. It employed some classic tropes, making it accessible, but did so with nuance and context so they seemed new and vibrant. It captured grand scope while remaining anchored in the motivations of individual characters. I’ll refrain from any more detail in case there’s anyone else left who’s not played it.
  • Sharply defined and well-rounded characters. Each of your team member’s motivations and background were unique, and the game wasted no opportunity to expand on these through background dialog, interjections, and excellent character driven side quests. Most other NPCs, too, were detailed in a believable way that made it easy to empathize with them and thus care about the eventual fate of them and their world. This also made side quests feel far less arbitrary and repetitive, despite the limited range of quest activities (kill this, talk to this guy, get this thing) and the frequent re-use of interior maps.
  • Excellent integration of PC customization. So many games let you customize your character’s appearance, background, and other attributes, then limit the impact of those choices to purely mechanical or cosmetic effects. There’s nothing wrong with either; character optimization is, after all, a valid element of play and a lot of fun. Furthermore, mechanical changes implemented well can afford different styles of play that lend a lot of depth – Deus Ex is one of the better examples of this. Mass Effect, however, does the best job I’ve seen in a CRPG of adapting the world and the plot according to those customizations by changing NPC interactions, providing different side quests, and, moreover, by reminding you of your choices at relevant points throughout the game instead of letting them fade away into irrelevancy.
  • Strong setting and backstory.The setting, as developed in in-game dialog and the game’s codex of background information, was coherent, internally consistent, detailed, and mostly plausible given current scientific understanding. Even the three notable exceptions of mass effect (necessary for FTL travel and comms, artificial gravity, and other space operatic tropes), biotic abilities (basically magic), and tech abilities (basically technology that’s indistinguishable from magic) were well fleshed out and justified. Two design details I really liked:
    1. Frequent opportunities to learn more about the world by interacting with computers and other objects, thus drawing you into the setting by reinforcing the feeling that you’re not just playing a game, you’re exploring a world.
    2. The detail, plausibility and novelty of many of the planet descriptions, and the look and feel of the planets you could land on really helped to establish a sense of place for the galaxy as a whole. Driving around lava plain looking for mercenaries is cool, but driving around an alien world with highly active geology caused by tidal effects from a nearby gas giant is way cooler. Maybe I’m just a nerd.
  • High production values. This one’s straight forward – if you design your world with care, but present it poorly, few players will ever take the time to appreciate it. If, however, you employ attractive and above-all consistent aesthetics throughout, in audio, graphics, and UI, it becomes much easier to appreciate the experience as a seamless whole and thus become immersed in it. Mass Effect does this well.

Shepard as I saw her

Nothing is ever perfect, though. I have the following critiques:

  • Failure to capture or emulate scale.  Space… is big. Really big. You just won’t believe how vastly hugely mindbogglingly big it is.It’s hard to believe, then, that all of galactic society is governed by a council of three. Or that the capital of the galaxy has exactly one trader, one med clinic, one banker, one dance club, and one seedy bar of note. Or that ambassadors would be satisfied with one room embassies. Or that there’s only 60 star systems in the galaxy. Or that all of the interesting features of a planet can be found in square patches 5 km each side. This is a classic problem for game designers – reality is big and complex, but your game has a limited budget, so you have to take shortcuts. There are strategies, however:
    • Allude to greater things. Players don’t need to see or interact with all of the parts of some system, but the presence of those parts should be acknowledged. Instead of forming a council of three, Mass Effect’s council members could have been presented as their spokes-aliens. Depict mass deliberations in the background of a scene, and have them occasionally consult off screen colleagues, and you’ve got something more believable. Add a sub quest that deals with alternate factions within the council, and the rest of the council feels real. Don’t dwell on it, but allude to its complexity. Mass Effect’s elevator newscasts are an excellent example of just this – you get the feeling that there’s a real news-hungry populace out there, eager to know what’s happening. But you never see them.
    • Don’t be limited by travel distance. Sure, travel helps players establish a graphical sense of place, and driving and navigation tasks make for fun game-play when done right. But in most games, travel is a means, not an end. Instead of trying to fit all meaningful resources into a single traversible environment, planets could have been represented as a globe or world map with places of interest marked. Planetary scale would seem more realistic and planets could be re-visited without repeating travel segments, and driving could be reserved for when it serves a dramatic or atmospheric purpose; approaching key installations, for example, as in the core plot. Weirdly, they got this right for planetary systems, but not for planets themselves. The same approach would have worked well for the citadel, too, and even reflects human behavior – most of us have a small number of favored stores & restaurants, ignoring the rest except for special occasions.  The same might apply to traders, med clinics, and other services.
    • Procedural generation. The most direct way to create a feeling of vastness is to actually provide it. Elite 2, for example, included a mind-bogglingly vast array of stars (100,000+) to visit in a game only 400 kB in size. Sure, 99.999% of these worlds were irrelevant, but they were there, and you could visit them, if you wanted. This made the universe seem real and huge, dispelling the feeling of game world as sandbox that so often becomes evident when you look back from the end and see that you’ve explored everything, done everything. Obviously, procedural generation needs to be supplemented with actual design, but this could be limited to customizing 50 or so important star systems. The key challenge here is keeping the UI uncluttered and keeping the player from getting distracted.
  • Uninterruptible animations. Games are interactive experiences, and interactivity is dependent on a sense of control. Unskippable cut-scenes and animations may be warranted in certain rare situations, but surely elevator animations are not. The only value I got out of elevators was from the newscasts in the Citadel, and even these only occupied 50% of the animation time. I really don’t understand why designers persist with this particular class of mistake.
  • Quest interiors. Mass Effect’s many secondary assignments involve visiting facilities on a variety of star systems throughout the galaxy. Disappointingly, only four fairly basic designs were used to represent these: the cargo ship, the above ground building, the below ground building, and the mine / cave. Slight variation was created by rearranging furniture and cargo, but the simplicity of these interiors really undermined the otherwise excellent senses of place created throughout the game. This was particular jarring in comparison to the detailed and diverse environments used for primary missions. Templates are a great tool for building large numbers of environments, but rather than whole area templates, they probably should have used room templates (a la Oblivion), as it’s much easier to explain regularities at that level (standard habitat construction components, for example).
  • Collection Quests. I understand why designers include collection quests, but I still don’t understand why they do them so badly. Their strength is that they appeal to our intrinsic motivation to complete easily measured tasks, and in doing so cause players to explore the world much more thoroughly. Their weakness, however, is that they can flatten the delight of exploration into an exhaustive grind. Instead of telling players ‘you must walk all over every planet to complete this quest’, games should tell players ‘keep your eyes open for hints about where to find these things’, then give them reason to trust that they’ll have ample clues to find them all, with perhaps some gradient of difficulty for the players who like searching for all of the eggs. It’s far more interesting to chase down rumours about the location of hidden artifacts than it is to drive back and forth looking for blips on your radar.
  • Inventory management. As previously mentioned, character optimization is a core aspect of RPGs, and that includes inventory optimization. On this front, however, Mass Effect fails:
    • Aside from a few special upgrades, all items in Mass Effect are labeled Name – Model, with name affecting the look of the item, and model the power. So, a suit of Onyx-I armor from the beginning of the game looks identical to a suit of Onyx-IX, but is much weaker. Upgrades, then, are a colorless matter of incrementally improving certain attributes, to the point that I found it more interesting to find a different but worse suit of armor than an purely better one. Furthermore, several makes and models were strictly better than others, meaning that whole sections of the inventory tree could be safely ignored. As a result, inventory optimization shifted from an exploration of cool things I could trick out my character with to a dry task of making sure all of the numbers were optimal. Borderlands provides a far better example of inventory optimization in an action RPG; another relevant consideration is the idea of ‘incomparables’, as explained in this episode of Extra Credits.
    • Inventory management is an element of gameplay best thought of as a task. You have specific goals in mind, specific choices to make, and can ignore much of the rest of the game while doing them. Like most tasks, then, the UI for managing inventory should be focused on helping you complete that task by providing relevant information when it’s needed and limiting the time delay and effort associated with each action. Mass Effect’s base inventory UI isn’t bad, but it’s not great – it only lists 5 items at once, for example. Where they really drop the ball, however, is in shopping for the rest of your team. You’re only able to modify team inventory when that team member’s with you, or when you’re on the ship (when that team member is not with you). Problem is, when you’re talking to a merchant, you can only compare their wares against the equipment of team members who are with you. So, you’re either reduced to taking your team to the merchant in small groups or writing down the key stats of their gear and managing it by going to their lockers in the ship. Managing this sort of information by hand is not fun and is precisely the sort of thing a good UI is meant to obviate.

All up, though, Mass Effect was excellent, and good motivation for me to move on to the next one.

Posted in Games, Reviews | No Comments »

Since arriving at UW almost four years ago, I’ve been involved in student politics. I’ve not talked about it here as most of it is rather prosaic and, for those not at UW, largely irrelevant. Every now and then, though, something comes up which might be of wider interest..

In early May, GPSS, the Graduate and Professional Student Senate, ran its second Science and Policy Summit, an event for academics, policy-makers, and the general public that looks at the interactions between scientific research and policy development. This year, we ran two panels, one on the impact of bioinformatics on preventive medicine, and the other on the role of science in public political discussion, focusing on the US Presidential debates. As well as the panels, we also ran a series of short talks, inspired by the TED model.

There were about 10 talks, 10 minutes each, delivered by a mix of graduate students, post-docs, and faculty, and of a very high quality. Here’s a few topics to whet your interest:

  • Couch safety, various failed attempts at regulation, and the fate of your cat
  • The ups and downs of developing tourism as a means of restoring communities and ecosystems in SE Asia
  • A passionate argument for the necessity of cosmology research
  • The woeful state of healthcare in the US (not health funding, but healthcare itself), including some rather damning statistics and factoids, presented as humorously as possible

We recorded all of the talks, and they’ve now been posted to YouTube. They’re all rather interesting, and worth at least a look, even if the film quality isn’t all I’d hoped when I filmed them.

Posted in Conference Notes, Science & Technology | No Comments »

Spacey fun at the symphony last night, with Legeti’s Atmosphères and Strauss’s Also Sprach Zarathustra, both used in 2001: A Space Odyssey, and Holst’s Planets.

Atmospheres was subtle, relaxing, and almost entirely without melody – a piece for meditating to, if stark and slightly scary. I really needed that, as Thursday was just one of those days – at one point, I’d even been the target of a precision bombing run by a passing bird. Two things about this piece. Firstly, most recordings don’t do it justice, because it’s all about orchestral color, and that often just doesn’t come through. Secondly, it’s very quiet in places, which offers a great opportunity to really appreciate the inability of some people to simply sit still and be silent.

The opening of “Also Sprach ..”, unsurprisingly, did its best to shake my innards to jelly with awesome – the sheer weight of the music is that much more spectacular when you’re up close with full orchestra and pipe organ. Like most, I wasn’t too familiar with the rest of the piece, but I’m now inspired to fix that.

Sadly, no Blue Danube.

The Planets, being one of my favorite pieces, were up for more criticism than the rest. They played it alongside a sequence of photos and reconstructions from various space probes, many of which were awesome. Unfortunately, while most of the performance was serviceable, there were a number of missed notes in the lead brass portions in several movements. I also think the conductor stuck too rigidly to the meter in places, and as a result, certain sections came off a little mechanical and slightly too fast. That said, my reference performance was conducted by Karajan, and he’s famous for interpreting the meter and tone of a piece in his own way.

Posted in Art & Photos, Music | No Comments »

Saw a cool presentation today about an implant-based therapy for epilepsy at the Center for Sensorimotor Neural Engineering by Jaideep Mavoori of NeuroVista, a company here in Seattle.

The idea is that if someone is warned of an imminent seizure in advance, say 20 minutes out, they can remove themselves from unsafe or embarrassing situations, take other precautions (lying down, perhaps), or take fast acting drugs that might stop it from happening. This is a big deal, as it helps the 30% or so of those suffering from the disease for whom conventional drug treatment doesn’t work. It may also allow some of those receiving drugs to come off them, reduce their dose, or shift to less effective drugs with fewer side effects. It might also help reduce the number of deaths from epilepsy-related accidents – 50,000 annually in the US, apparently.

The technology’s actually fairly simple. There’s three main parts, none of which appears to be particularly magical.

  • an array of electrodes implanted on the surface of the brain, beneath the skull, but outside the dura mater, and thus not in contact with or penetrating the brain itself
  • a set of signal processing and machine learning algorithms that classify brain patterns into risk levels based on training data previously collected from that individual and from the public in general
  • A mobile device or phone app that warns the patient of periods of increased risk

It’s in trials at the moment in Australia, and is apparently performing well, with no known associated adverse events.

Apparently, this all started out with Jaideep’s PhD research into the sensorimotor function of moths – he basically designed chips and implants small enough to put in a moth, then studied the different nerve signals associated with its wings as it flew. They also did trials of the epilepsy detection technology on dogs, as they also suffer from epilepsy. Unfortunately, I was unable to find copies of the cute pictures he showed in the talk.

If you’re interested in hearing more, there’s a 5 minute video article from ABC News in Australia talking about it. It’s formatted a bit weird, so you might need to download it and switch to the second audio track in VLC.

Edit: Apparently the electrodes are implanted beneath the dura mater, but outside the arachnoid mater. So, between the second and third membranes that encase the brain.

Posted in Medicine & the Human Body, Science & Technology | No Comments »

    Reviews, rants, reflections, arguments, scrawls, ideas, refutations, pontifications, rhetoric, records, accounts, journals, scraps, plans, authentic articles of thought.

    No artificial ingredients.
    May contain pretentiousness.
    May reflect personal bias.

    • Hey baby! Do you want to taste the sting? #PostsThatNeedContext 2011-12-29
    • Today's new word: apophatic - adj, beliefs that god can only be known in terms of what it is not. Opposite, cataphatic 2010-01-24
    • Naptime over. Now becoming fully cognizant of all of the little things I need to catch up on. Foo! 2010-01-14
    • Anyone got suggestions on Twitter clients for Windows. I'm using twhirl - got anything better? 2010-01-14
    • More updates...