“Persona”: splitting myself into two on Twitter

“Are you writing those tweets?”

This was roughly a question someone asked me the other day about my @sean_htch twitter account (where you may have clicked through to this essay). And it’s a good question – if you have met me in person, you’ll know that the stuff I tweet on ‘main’ (@sean_htch) feels a bit distant from talking to me in person. The part of me represented there, perhaps, feels

  • Obsessed with Zelda, especially Link’s Awakening? Does Sean continuously play Link’s Awakening?
  • Wow, this person really loves the PSX, N64, etc
  • Can’t stop talking about their upcoming game Anodyne 2.
  • They are always positive!

Now, of course, that’s very much an exaggeration of some facets of my personality. I enjoy Zelda games as well as old 3D games, but they’re just one of many interests. They just happen to be the best way to immediately convey some of the appeal of Anodyne 2/Anodyne. I do care about Anodyne 2 a lot, as it’s my job, but I care about other things as well. I experience a wide range of emotions. (That was a weird sentence to write.)

If you’ve followed @sean_htch for over a year, you may have noticed that tweets are far less frequent, but when I do tweet, there’s a lot of engagement. That’s all intentional! Here’s why.

This is a graph of my follower count from Feb 2016 to March 2019. The graph continues to the left. In fact, if you follow it to Feb. 2014, followers are at 1700, meaning, more or less, follower count was flat or gently sloping for many years.

I worked on the game Even the Ocean for 3.5 years from 2013 to 2016, so you can see how that corresponds with a nearly flat line on the graph! In the middle of 2018, I decided “something needs to change” about how I approach Twitter. It’s a useful tool for promoting my work and I felt I wasn’t using it optimally. So I changed things and, happily, have had some positive results!

Splitting the Self

Ah… dramatic subheading…

Last year, I asked myself: do my followers, likely who know me through my games, need to hear every little opinion or life thing about me, like that the grocery store near me hasn’t restocked on cookie dough in a month, and that the price of a chicken rice plate at my favorite fast casual restaurant went up $0.50 and it’s ridiculous to charge more money for rice than for a pita wrap?

Well, the answer is no. It didn’t feel healthy to have thousands of people hearing my frivolous tweets. It’s good to have tweets less people see, it’s also good to have parts of you that only offline friends see.

Also, my follower count was more or less flatlined for years, and if I want to stay in this business, I need to stop letting my personal life weigh down my professional work. So as a compromise between locked accounts and mains, I made a ‘sub’. (I know a few other game developers who have done this.)

It’s at @han_tani, and it’s just me using Twitter as a non-professional, i.e. how I’d use twitter if I wasn’t working in games.

Splitting myself into two lets me be more focused.

On my main, I can make decisions that are good for my career as a game designer – I can hyperfocus on making tweets that I know fans of my work will enjoy seeing and sharing. I can tweet important news about my work like console ports. I can share posts like this with a larger audience.

And, I can still do this while having the pleasure of complaining about Tide Pods smelling too strongly on my sub account.

Be a Nintendo

Around August or September last year I started working with The Indie Bros. for assistance with promoting and outreach with Anodyne 2. One useful advice I got was on “being a Nintendo” on Twitter.

Essentially, a “Nintendo” will keep their tweets hyperfocused on getting fans interested in their current work. If there are ever memes or jokes, they always relate to the game. No threads, no arguments, etc. Hearing this sentence made it clear the idea of keeping my main’s presence “all signal, no noise”. Now, I don’t tweet anything unless it contributes to the goal of

  • Promoting Anodyne 2 or something I’m working on
  • Promoting something that helps my company’s stability (console ports, etc)

I will occasionally retweet friends’ work, but I do it less than I wish I could! (FWIW, I do retweet lots of stuff on my sub).

Another counterpoint I always was thinking was “what about using my platform for promoting good things (social justice, etc!)” and that is a fair point! But there are various people with accounts that do a great job at this, and it’s their account’s focus. I feel if I want to do concrete justice actions, I’m better off doing something local – donating to local orgs, or joining a political community, etc.

That is to say, I feel that using Twitter for a career means you need to have a single focus – something the account is known for. And it’s hard for that to be “being yourself,” as everyone is multifaceted.

Being Yourself

As an indie dev, you have a unique advantage, and that is, you can still keep your account as being ‘you’ (in the sense that my main Twitter is Sean, not my company Analgesic Productions). I still think that accounts that are a single person are more attractive than a company, which feels impersonal. (this is basically a fact, given how many fast food twitters now act like they’re a person).

Of course, running a main that’s ‘yourself’, there’s a temptation to tweet more ‘normal’ stuff about your life – but again, that’s what the sub account is for! If you’re a smaller indie, don’t fall into the temptation of trying to show your cool game *and* your cool lunch, no matter how delicious.

The 10-year question

Something I like to think about with curating my online self is – will this curated self successfully age? What should my strategy with Twitter be *now* so that I can still get engagement in 10 years? I asked myself this because near the end of Even the Ocean’s development, I was barely getting much interest in the game itself.

Other questions: how can I frame my tweets so that they’re both appealing but also there’s obviously enough of a weird/experimental edge so that when I make a less commercial game (think All Our Asias), I can still get people interested? Those are all things worth thinking about if you are going to use Twitter for the purposes of a career. I’d like for Marina Kittaka and I to be able to stay in games and work together!

Well, those questions are too complicated for me to fully finish by my 11 AM deadline, so I’ll stop here. I also want to talk about composing tweets that people like to interact with, visual hooks, various strategies with tweets, and how tweeting is essentially grinding out visibility and increasing the chances of ‘lucky tweets’ or journalists/video makers seeing your game, etc, but that’ll be another day.



For the better part of a year I’ve been developing “Anodyne 2: Return to Dust” with Marina Kittaka.

5 or so years ago, I did an interview about Anodyne 1. I said there wouldn’t be a sequel. Well, now it’s 2018 and, last I checked, I’m sure making a game called “Anodyne 2”. Did something change?

Back then, I was against making a traditional sequel, where we would do mostly the same thing but with new levels. That’s why, instead of just being “Anodyne 1 but different levels,” Anodyne 2 is more the next installment in the “Anodyne Franchise”, like Final Fantasy or Zelda games. Kind of like Nier: Automata, Anodyne 2 is a standalone game, and differs in some ways from the original, but has its commonalities, some narrative continuity, and can be understood more deeply if you’re familiar with the original.

Why didn’t we do a traditional sequel, like Pokemon Gold, Banjo-Tooie, Spyro 2, Dark Souls 2, or most other corporate game sequels?

In this life, we only live so many years. There are certain skills – like making 3D games – that I want to hone and learn, and if a game is entirely just Anodyne 1 again but some additional content and new dungeons, I don’t think that’s the best way to spend my time.

Plus, an “Anodyne 1-2” would be weird. There’s not a great way to create a sequel that incorporates Young. For the most part, Anodyne 1’s story was one and done. Of course, maybe in 5 years I’ll be making Anodyne 1-2 and eating my words. Time changes odd things.

Eh, also, trying to replicate the experience of Anodyne would just lead to it being overshadowed. If you really want Anodyne again… I understand where you’re coming from, but your dream of having a new experience that makes you feel exactly what Anodyne did, is, sadly, impossible. Even if we made the best game ever that was really similar to Anodyne, it would be overshadowed. Anodyne 2 will be a good, memorable time, but it’ll be different.

Some things in life just happen once. In transience is beauty… something, something. Different flavors are good. Spice of life. Etc.

I can’t just keep making more of the same – this world can’t keep doing more of the same.

I think a ‘sequel’ or series installment succeeds when it reflects upon the components that made its predecessor good, and then responds to that material in an interesting way. Nier is a great response to Ocarina of Time. Anodyne is a great response to Link’s Awakening. Likewise, Anodyne 2 is a response to Anodyne. We learn from it, tweak some mechanics, add some new gameplay, remove some gameplay, and incorporate the current narrative ideas and themes and stories that we currently really care about.

As another similarity, the high-level game structure of Anodyne 2 has similarities to Anodyne, but 3D gameplay replaces some of the 2D areas.

The reason Anodyne 2 isn’t just a new IP is because well, the 2D levels play like Anodyne, NPCs are designed with Anodyne’s style in mind, you won’t be able to predict where you’ll go next, etc. The plot is overall clearer but it’s very much still a surreal, dreamy fantasy. Cards and Dust make a return but with different uses. There’s shared elements, just like in a Final Fantasy or Zelda installment. So, it’s called Anodyne 2.


I should mention, we were considering calling it “Anodyne: Return to Dust” or “Anodynia” or something like that. Perhaps one of those choices would show more integrity as to what “Anodyne 2” really is?

But, you have to also look at it from the perspective of us not being Square Enix or Nintendo: it’s going to be far, far more confusing if we don’t put the 2 in there. As an indie, someone might perceive “Anodyne: Return to Dust” as a DLC package! If we use “Anodynia”, that won’t get eyes as fast as “Anodyne 2”. With the ‘2’, it’s obvious that it’s

  • Related to Anodyne
  • A separate game

The ability to call something “Series Name: New Subtitle” and become popular relies on being a series entrenched into culture. We don’t have that clout. So yes, to an extent it is a branding decision, but I hope that makes sense given our hope to continue past Analgesic Productions’ 7th birthday.

Even with this disclaimer, I can predict the exact wording of some negative reviews due to us ‘veering too much off course!’. Well, if that future reviewer is out there, well, I hope you like being screenshotted and used as a joke 3 years from now.

Anyways, I’ll end with this:

We’re the people who made Anodyne 1, Even the Ocean, All Our Asias. We’re dedicated to making excellent work.

Would you really expect us to just make the same damn thing a second time? I hope not!

8 Ways To Not Spatially Organize Your Game (Banjo-Kazooie: Part 2)

In light of Super Mario Odyssey’s release, I realized there’s a timely angle to look at BK with. Rather than summarize the levels like with Part 1, I’ll go over some of the major causes of getting lost/confused in BK. By no means do I suggest these are bad things to do, but they may be counterproductive to your goals in a 3D game, especially if you are doing something with ‘box-garden’ design like BK (or Mario 64, etc.)

Maybe this will make your Mario Odyssey playthrough more interesting, given that it is a semi-box-garden style design? Let’s see… at number one…

1. Mismatch of Interior Size to Exterior Visual Size

See TVTropes’ “Bigger on the Inside” page.

At least BK is consistent with this! Nearly every interior space in BK is much bigger than the exterior would suggest. This is something you might *Want* to do in a 3D game. Say you have a central castle in a level that you also need to navigate the exterior grounds of. It feels better for that castle to not be huge and boring to walk around. But if the castle’s interior is proportionately small, then it might not be as exciting or interesting, unless you are working in the style of Dark Souls or similar ‘realistic’ architecture games.

The “bigger on the inside” trope, can often make areas easier to navigate (take small pokemon starting towns). But, when a particular structure that is “Bigger on the Inside” is in 3D and has too many entrances, it can become confusing.

Example: In level 7, Mad Monster Mansion,  the exterior  is realistically sized, to make it easier to navigate around and see, but then the inside is far larger, to make it more fun to move around in.

One problem here is you can enter the mansion through a front door or a window to the left. Though these two entrances are only a few feet apart, they lead to entirely different spaces – respectively, a gigantic dining room, and some smaller  alcove. To make matters worse, the alcove should technically be inside the dining room.

This makes it harder to make a mental map of the overall level. We don’t just have to remember that these two spaces connect. We also have to remember the additional information that the two spaces are related in an non-physically-realistic manner. This mostly matters in box-garden-style games, where you progressively consume and make a map of the level. In a more linear game, it may not matter as much. I do the “larger on the inside” thing a few times in All Our Asias. The difference with AOA, is that remembering the specific connections between spaces is not as important, thus it’s less likely to result in being lost. Though there still may be a sense of disorientation – which may be good!

2. poorly-signed, screen-transition-connected subareas

Poorly-signed means an entrance to an area is not very distinct – it’s hard to identify where the entrance is going.

A screen-transition-connected area is when you go from one subarea of a level to another and the screen fades to black, or there’s  a loading screen, etc. E.g.: Mansion to mansion exterior in the above example.

Most games need screen transitions. It is worth mentioning that the more ‘screen fades’ you have in a 3D level, the harder it will be to remember how to navigate it. How many times have you left some new building in real life, and not been sure which way to turn? The same thing happens in games. A screen transition breaks the sense of spatial continuity and causes disorientation. This becomes more of an issue the larger an environment grows.

Generally a 3D game might fix the disorientation problem with a minimap or some kind of map – something to help you get your bearings when entering a new space. But… over-reliance on a map will reduce how well a player picks up on landmarks and really learns how to navigate a space. Or the game can make the inside/outside relatively clear and with few entrances. E.g., the sandcastle puzzle in Treasure Trove Cove is a very iconic building, and so it’s easy to know where it goes from looking at it. Again, in a more linear game, this is not as big of a problem, because you can generally emphasize that the place you are *leaving* is not as important as where you’re going. But in something with a box-garden design you need to be more careful.


This issue compounds with the interior/exterior size mismatch problem.

The worst textbook example of this, which BK does not have, is straight up teleporting you to another place without any sense of logical spatial connection. Think teleport mazes in puzzle or RPG games.

BK has this issue in a few places.

  • Mad Monster Mansion: the mansion has many window entrances, that are of course hard to keep track of because they look the same.
  • Rusty Bucket Bay: The level’s outer rim with storehouses/etc have many entrances that are hard to keep track of because of similar appearance. The main boat has many portholes you can enter with similar problems.
  • Gobi’s Valley (Level 6, which I’ll come back to soon), which features pyramids and sphinxes with entrances. What’s on the outside doesn’t always match the inside… and so it’s hard to remember where a particular challenge is. Still, there are relatively few rooms in Gobi’s Valley so it doesn’t become a huge issue.

BK does get it right, though. E.g.:

  • Mumbo’s Mountain has a single anthill. Easy to know where that goes.
  • The big turtle and crocodile in Bubblegloop Swamp. Go to a simple place.
  • Christmas Tree and single ice cave in Freezeezy Peak.

3. Swimming

Swimming in BK is a nightmare, probably its biggest mistake. There’s a short air meter, even *enemies*, sometimes a shark that chases you, and in a later level the water drains your air meter twice as fast. Not to mention, swimming is slowly and unwieldy, and you are expected to collect things underwater.

Why is it disorienting? Well, you are moving in all three dimensions, freely. Water is usually harder to see in, and because you don’t have gravity stabilizing you, and the controls are so bad, it’s super easy to get turned around or confused as to where you are going. 3rd person cameras can be a little tricky in the water. This makes levels like Clanker’s Cavern or parts of Rusty Bucket Bay much harder than they need to be.  Being in water is sort of like throwing static all over a minimap or just negating all the progress you made in trying to remember where you’ve gone in a space. Add in the air meter, and you now have pressure that makes it harder to think clearly.

I’m iffy about water. I would just leave it out of my games, or make it more of some kind of jetpack like control, where the Y-axis movement is not totally free – some way of making it less disorienting for the player. At the least, keeping your ‘water entry areas’ relatively distinct could be a good move (rather than say, a giant lake you could enter from any point on the shore.)

4. Poorly signed / laid out tunnels

It’s hard to keep your sense of direction in tunnels due to the visual uniformity. It’s also difficult to use landmarks to reference where you have come from and where you want to go. All you see is tunnel.

This is made worse by slightly curving tunnels: you can’t try to remember right angles or ‘easy’ directional changes, you can only follow the tunnel and hope you find an exit. Good way to make you feel lost, though! Mixed with a map, this can actually be okay (think the maze-like tunnel dungeons of SMT Nocturne.)

You can make tunnels harder to navigate with by making their entrances and exits discreet. Put lots of tunnels next to each other to make things even worse.

And to top it off, you could even put tunnels in the water!

Example: Navigating the water tunnels in the final leg of the hub world. This has problems: not only are you using the most disorienting mode of movement (Swimming), but you are connecting places with tunnels.

(Left: an underwater tunnel in the hub world. Right: a tunnel entrance, then the tunnel.)

Or, the tunnels in the water in Clanker’s Cavern. The entrances’ similar coloration makes it easy to forget which tunnel you’ve visited.

Screen Shot 2017-10-26 at 11.25.36 AM.png

5. Lack of Color Variation in Level

Gobi’s Valley: everything is Yellow.


While this at-a-distance picture looks fine, you’re much closer to things in the game’s 3rd person camera, and so even the structural distinctions might get lost amongst the overall Yellow. It’s good that the borders of Gobi’s Valley are brown, at least. Anyways, it follows that if everything is a similar color, it’s harder to tell parts of the level apart.


Compare Mumbo’s Mountain where paths lead you around and there’s clearer color distinctions with the grass, hills, water, central mountain.

6. Lines of sight to other parts of the level broken by walls

When you can’t see to the rest of a level, you can easily feel lost or boxed in… in some ways this is a less intense ‘tunnel effect’. You aren’t sure where you are going or how you are oriented. Mad Monster Mansion has this problem: the hedge fences are tall, uniformly dark, so it’s really hard to tell what side of the mansion you’re on sometimes.  Click Clock Wood can be tough to remember where you are (at first) because your line of sight to the other side of the level is cut off by the central giant tree.

Compare Rusty Bucket Bay, which has a central boat in a pool, then a perimeter area. You can at least see what side of the boat you are on, so it’s easy to tell where you are overall in the level (though not necessarily easy to get to somewhere due to the water – RBB has its issues.)

BK pretty consistently increases the ‘broken lines of sight’ as the levels get bigger, partially because of more landmarks / more open space. In some ways, BK’s difficulty comes from having to remember more and more of a 3D space in your head. I think spatial organization topic is why Mumbo’s Mountain (Level 1) feels so strong – it’s really easy to get a sense of the level’s scale at a glance.

Level 4, Bubblegloop swamp (mentioned later), breaks lines of sight – but because it organizes its space in a hub-spoke form, it’s not too bad to navigate. I kind of like the hub-spoke structure: put challenges or singular things at the end of a spoke, so people know how to navigate back to the hub to find the next thing they didn’t do or see.

7. Lack of Landmarks

I wanted to throw this one in here, BK actually does this really well. But  if you are designing something in an ‘box-garden’ fashion like BK, you should make sure the sections of the level have some distinct feature to them – a landmark – that can be used for players to navigate with and judge their sense of place. One issue I’ve seen with first-time Unity students is making the terrain tool and creating high, bumpy and indistinct mountains. A good way to make landmarks is to do concept sketches of a level and include distinct sections/buildings/geological forms.

Landmark examples…

  • The termite hill in Mumbo’s Mountain.
  • The central mountain of Treasure Trove Cove.
  • Clanker, in clanker’s cavern.
  • Freezeezy Peak’s Snowman.
  • Click Clock Wood’s mumbo house.

I really love the structure of Bubblegloop Swamp – while I think its color choices can be a bit uniform and confusing, it has a central location with different branches, and each branch has clear landmarks to let you know what the branch is. My one complaint is the maze in the bottom left, which is hard to see from the center, and thus, hard to remember that there’s a Mumbo hut there (a place you need to go to transform into an alligator.)


8. Excessive + confusing vertical layering

This relates a little to how swimming can get confusing, with having too much freedom of movement in 3D space. When a level starts to build tons of vertical layers, it gets harder to remember where you are – you aren’t just thinking about your flat, X-Z plane position, but also the vertical Y position.

In BK, most clear is the sense of place you lose when climbing up the giant tree in Click Clock Wood (see 2:35 on here). While you can sort of remember where you generally are on the spiral path, it’s hard to remember what is all the way below you.

Treasure Trove Cove initially feels daunting because you have both the ground-level paths and the rocky-raised-structure paths.

This is also why a lot of early 2D platformers can be confusing – too much vertical layering, and the camera is very close to the player so it’s hard to see what’s around you.  (Watch Toy Story 2 for a while.) Maybe this is also why Symphony of the Night and Super Metroid feature minimaps. This is partially why Even the Ocean featured mostly linear levels.

It’s not like you can’t vertically layer spaces or have vertical variation – but it might be good to keep the vertically layered space linear (like the timed challenge in Bubblegloop Swamp – see 1:34).

Now compare that to say, the scaffolding of an apartment building construction site, with intersecting scaffolding – that could get confusing, fast, depending on the level design!

Screen Shot 2017-10-27 at 3.50.21 PM.png

That’s all for Banjo-Kazooie – it’s a nice game to check out if you haven’t, at least for the first 2 levels !

If you’d like to stay up to date, please subscribe to my blog, or follow me on Twitter!

I’m Sean Han Tani and I worked on Anodyne and Even the Ocean. I write a lot of music and also teach music and game design at SAIC in Chicago. Now, I’m working on a 3D game coming out this year, All Our Asias.

You can check out more of my writing here.

Walk Geometry and Perfect

Even the Ocean took a Long Time to make. Preliminary development was from March to July 2013, really kicking in around August of that year, not ending until November 2016. A lot happened in between, and I wanted to talk about a few art projects that popped up for me.

Walk Geometry / 散歩幾何学 (February 2015 to August 2015, roughly)

More info

Walk Geometry was a project inspired by how much I enjoyed walking and progressing through Dark Souls 1. It consisted of a series of vlog-style videos recorded with my phone from a first-person perspective, narrated by me and sometimes other people I was with. It is structurally similar to the one-take vlog – the videos record my entire progression and activities within a space, sometimes pausing for inappropriate social contexts to record in. However, the narration of the video focuses on how the spaces made me feel, sort of like as if I was ‘playing’ these spaces like the popular  walk-and-look-primacy games (also known as ‘walking simulators’). This style of walking is a little different from just being observant, as you feel a pressure to think of ‘interesting things’ to say about what you are experiencing, sort of like creating a Let’s Play (a video where you record playing a video game and talk about it.)

Though I still enjoy recording live video in this format from time to time, the bulk of it happened from feb to aug of 2015. For example, shopping at a trader joes, walking down an apartment staircase, going to my apartment’s roof, driving to and shopping at a super target(!), etc. There have been other videos, like “standing on a street corner” and so on.

It’s not a particularly ‘brave’ sort of video making (though it might have looked weird at certain moments…), but looking back at this playlist (and the other videos on my YouTube which fall under the category), it’s interesting how easily we can categorize these pretty menial things – walking to a store, walking to the train station, taking a bus, and how we usually consider them as being under the general umbrella of ‘mediocre experiences’, yet really are something unique each time, evidenced by the title.

Back to the Dark Souls reference, what I really liked Dark Souls is how the levels were just small enough, and you moved just slow enough and with repetition (due to difficulty being killed by enemies) that you intimately learned the ‘geometry’, or shape and layout, of the levels, such as castles, woods, swamps, caves. The combat required you to be extra-aware of the ground and obstacles, too.

Anyways, I encourage anyone interested to try out this style of video-making, if possible! Clearly there are dimensions of one’s appearance and language and context that come into play, possibly eliminating one’s ability to safely video-make. If you can’t do it in real life, doing it in the context of your favorite game may reveal aspects about the game’s design you hadn’t seen before, as I did with the short-lived Walk Souls series.

Walk Souls (Feb-Apr 2015)

Around the same time as the above project I did a project called “Walk Souls”, which as you might imagine, was a Let’s Play series of Dark Souls 2 and Bloodborne (Both unfinished) which focused more on observing how the experience of being in the game’s spaces changes over time.

So a normal video would consist of : first clearing away all the enemies beforehand, then recording walking around and discussing the area. In retrospect, enemies are an important part of the Dark Souls landscape, so it’s debatable whether or not I should have eliminated them before recording. Observations usually consisted of how environmental objects or shapes changed how tense or open or familiar a place felt, how an area’s location in juxtaposition to other areas can change one’s perception of the place.

These two projects had some value in helping with the 2nd half of Joni Kittaka and I’s Nov-2016 game Even the Ocean‘s development – at that point, mostly thinking about how to lay out levels and construct Power Plants. However, I didn’t really get the chance to synthesize these ideas into new games until 2016.

Aside: Photography

These things about Dark Souls and Walking did get me interested in photography, and I wrote a few essays about photography and games at the end of 2015/start of 2016, as well as reading about photographers and making a bigger effort on phone-photography. It’s a latent interest that has made playing games and walking and generally observing feel more valuable. Anyways, these ideas led into Perfect.

Perfect (Feb-July 2016)

Inspired by a short story of the same name (and similar themes), Perfect was a HaxeFlixel “Software/Game” I worked on in between Even the Ocean and other stuff in the first half of 2016. It’s an adventure game about the reaction of a community to the aftermath of a disaster in an future Taiwanese immigrant community in the US, set in a sci-fi setting (a gigantic supermarket where people now inhabit). Some of the themes were about the future of labor automation (such as how it may not lead us to leisure, but lots of people basically feeding machine learning algorithms – or manually filtering bad content from Youtube/Facebook.)

The intention was to use a lot of these ideas about spaces and games, photography – and create a “Music Album Game” which consisted of a bunch of screens with wireframe art, connected by ‘links’ (underlined in blue). More details. The method of exploration was inspired by early point and click adventures and how they always give a single fixed camera angle of a space, with no camera panning. By using this and lots of abstract shapes to construct the spaces, I could try and conjure a ‘decaying remembrance’ aesthetic for the game, as Perfect was a story of someone in 2100 or so, ‘experiencing’ the recorded memory of a journalist from ~2070, as the journalist went around interviewing people.

The graphics were very simple, done by rendering wireframe meshes using scaled 1×1 pixel sprites. You could easily do something similar in Unity! The “mountain” effects were done by using hidden shapes to determine ‘forces’ on the sprites.

Of interest to me was the website-like interface. Rather than moving in real time, you only get ‘photographs’ of an area, and move from shot to shot. At times you were supposed to find ‘interviews’ of people, and listen to the game’s music while reading the interviews. Music-wise, it was an experimental way to try to foreground a game’s music and writing rather than its ‘gameplay’, and I thought of this project as a ‘software’ more than a game. But it looks like a game to me, now!

Instead of a game driving the music… I was trying to get the music to define spaces within a game world, and help define what narratives would appear. Example, a song called ‘club’ would take place in a ‘club’ and you would interview a club-goer about what he thought about the disaster. But that sort of just feels like an (interesting) game to me. Either way I was interested in different was to approach music-making for games, in terms of the planning stages influencing what the music would be.

I had split the game into three sets of themes (kind of like how we divided Even the Ocean) – Physical, Digital and Algorithm. As you progressed through the game you would reach more Digital and then Algorithm areas. Physical were things like: Balcony, Exterior, Desert, Club, Lake. Digital, like Shopping Carts, Desktop. Algorithm – Random, Vector, ACK, Governance, etc.

In practice, the themes and structure of the game were too nebulous for me to really grasp in a consistent way, and I found it hard to find time to work in between Even the Ocean and travel, so I abandoned the project around July last year. Moreover I was iffy about structuring a game around a terrorist event for a country (Taiwan) which historically has been very well off in the USA due to immigration policies for its country.

In retrospect, I should have laid off the race themes and focused on the labor automation ideas, rather than trying to difficultly address race ideas in a future setting, which is prone to Many Problems. However there is much sci-fi and research being done on automation already, so I’m not sure how much I could have contributed.

However, I did get a lot of neat music out of working on the project! And the wireframe style is still a neat idea I now have as a ‘tool’ in creating game spaces.

Either way, a lot of teaching happened (I teach at SAIC now), and releasing Even the Ocean happened. I’m quite excited to have found enough free time in the past months to really start working on a game project I (naively?) believe I will be able to finish.

Next time(s)…

  • A discussion of my current (codename) game ‘Project Shiho’ and its relation to the projects mentioned in this essay
  • A discussion of “over the clover” by Hiroshi Yoshimura and my remix of it
  • Something about SMT Nocturne?
  • Thinking about live music, or game music..
  • Something else?

4 Good Songs of 2016

This is a pretty incomplete list… I will keep better track in 2017. Anyways. Here are four songs I really liked. These songs all do great jobs of evoking interesting digital spaces.

Bryant Canelo – all this water you cannot drink


The song develops in the way you might peel back the world surrounding us to uncover, slowly, more about it. I would call it ‘realistic’ in the sense it sort of jumps out at you and asks you to recontextualize your surroundings.

The skittering trill throughout the whole song always has an unsteady feeling, both from the mechanical nature of trills in general, as well as its pitch changing (or a filter envelope moving around, not sure). Further contributing to this are occasional one-off percussion giving the impression of ‘did I hear that or not?, ‘dissonant’ midi-guitar-like plucks, and further percussive and choral layers.



ella guro / Liz Ryerson – (untitled sitting in the rain track)


I heard this song without knowing it also was in Robert Yang’s game, No Stars, Only Constellations. I’m particularly fond of the gently delayed, overlapping guitar sounds, especially around 1:19 when a single sequence of notes is played in different rhythms. I enjoy the layering of vinyl-ish static sounds in the song as well to give an impression of ‘recollecting’ a memory.

Liz’s music is important, especially in the context of games, where it is actively contributing and help push existing boundaries on what music for games (and music in general) can be.

emamouse – STP


See this song in a different context, with emamouse’s visual work: (6:20) http://psalm.us/mousewasher.html

emamouse puts out a lot of music, I was glad to be able to meet her at a show in Tokyo last August. A lot of emamouse’s songs have a lot of motion going on a lot – I think everywhere in this song, some synth, somewhere, maybe more subdued at times, is always doing some sort of scale-like run up and down. It’s something I’d like to do more of but is pretty hard to pull off! It’s interesting to zoom in on parts of the song and uncover new layers. Passively listening sort of gives you this kind of murky and dark flavor for the song, but like exploring an area in a game, you understand more of it and see new things over time.

That, and the hook at the start is really catchy, using a vocal sample which is characteristic to some of her other work.

Gazelle Twin – Outer Body

Buy gazelle twin’s music here

I heard this some point last year – and it’s really catchy. It’s has this repeating percussive bit reminiscient of times in AAA games where you’re approaching or waiting for action to happen. To that end it would be neat to see Gazelle’s music within the context of a game!

The vocal samples partway through the song add another interesting layer to the mix. The “ambient” aspects of music are fun when the artist drops in different parts, creating a panorama of sorts of a musical space. In this song, bass clips pop in, reversed percussion samples, processed vocals, etc. Looking forward to more music by Gazelle Twin.


SAIC’s Experimental Game Lab, Fall 2016

This past fall,  I taught undergrads and a grad student at the School of the Art Institute of Chicago (SAIC) as the lecturer for the Experimental Game Lab class. There was a lot of luck in being able to get this opportunity (thanks William Chyr). I am teaching there for the near future and am in talks for developing a game music composition and critique class for the fall.

The Class

It was a studio-style class, meeting at nights twice a week for three hours. I could see this being nice during the summer, but teaching 6-9 PM is kind of drab once day light savings time kicks in.

I decided to make the first half a discussion/lecture, then the second half ‘studio’ time, or time for students to work on their projects/troubleshoot.

The class structure and content were up to me to design.  I decided it would be a mix of introductory game design with Twine/Unity and critique and critical play of games (to try and increase the variety of games students were exposed to.)

One issue with the class was that it took place on Tuesdays and Thursdays – so I had to figure out what games I wanted to assign by Thursday to have them ready for the next week. In practice I was really busy with releasing Even the Ocean (please buy it) till November so I ended up having to assign 3-4 games over the weekend and a few games on Monday/Tuesday. As you might expect most students had trouble playing the games assigned on Monday/Tuesday for the Thursday class.

Part of why this was the case was that students seemed overworked – most taking 4-5 classes a semester, requiring many hours per week in-class and with projects to work on. I wasn’t really able to assign as much work as I’d like to, but perhaps this was also to blame on assigning too many game projects (more on that later.)

Luckily for next semester it’s a once-a-week 6 hour class, so it’ll be easier to structure – just assign once a week and discuss all of them for half of the next class.

Overall I managed to assign around 50 games, which was about my goal. Including student critiques that brought the total number of games discussed to around 100.


For each game I assigned, I gave some readings (often reviews or other criticisms), and then I wrote up some questions, which were usually short essay questions but I only required a few sentences. I tried to focus on questions requiring students to deconstruct how a game worked, or how effective a design decision was, or occasionally asking to think about the game’s presence (or lack of) sociopolitical themes.

Some students consistently gave good responses, others half-assed them. If anything I need to be clearer on the expectation and then enforce that in grading.

Question examples…

Yume Nikki: Did the presence of the ‘effects’ change how you perceived and made your way through the game’s world? How would the game have changed if there were no effects to find?

Lacuna III (An visually ‘spooky’ game): The NPC dialogue in Lacuna III is an interesting choice – although some NPCs may speak cryptically, a few speak in friendly, almost cute ways. What’s the effect of this on the player’s experience and perception of the game?

LSD Dream Emulator: In the dream-like games we’ve played so far, you transition from one place to another through doors, waking up, dying, or portals. LSD Dream Emulator adds a new way to the mix. What is the overall effect of colliding with any object functioning as a trigger to the next dream-world?

Now that I’ve generated around 50 games and written questions for them, that should buy me some extra time this semester to polish up questions, readings, and find new games to replace my weaker choices. I would like to include more short writings of my own on the games – I managed to do this for a few games, but for most I only had time to write summaries. I think it’s important to introduce why I’m picking a particular work. The past few weeks I’ve played ~80 or so games so I’ve found a handful I think are worth assigning.

I’d like to more strongly theme the games together. E.g. “newsgames”, “political”, “architecture”, “dreamscape 2d”, etc. I did attempt to do this for some weeks, but other weeks I had to do some ‘easy’ weeks like assigning some 3D games that weren’t Great but that I thought would be easy for students to emulate in their work, or even assigning my own games (which I’m still feeling okay about doing – though it does feel a little weird at times.)


We had four crits over the semester. I think this was too many. The first three required entire classes (of which I only had 28 classes in total), the last final crit took the last two days of class.

The idea was: Twine game, 2D Unity game, 3D unity game, Final project. It was an easy way for me to get a decent syllabus done, but it’s clear in retrospect that expecting four finished projects with the expectation of a crit was too much.

Instead what I’m shifting to for the spring semester will be a few shorter assignments focused on literacy with tools and design exercises, and then just a midterm and final project critique.

In the future I’d like to incorporate more interesting design exercises – perhaps when I get more confident with teaching. Robert Yang, Jake Elliott and Paolo Pedercini’s teaching work have been good reference points in this regard.


I was initially nervous about these at first, but then I realized it was basically a feedback session and it got a lot easier. It helped I had taken a few art classes in undergrad. One problem is that it got a little awkward when it was unstatedly obvious the student wasn’t satisfied with their work or, maybe, knew they turned in something of poor quality. But I never really brought this up and insisted on investigating the work anyways, and that seemed to have been fine and perhaps contribute to class goodwill.

I made sure to take notes for each student… which helped the discussion get going when the student had trouble talking about their work.

Because we are dealing with games and not traditional artworks, students must play the games ahead of time. Which is fine but this ends up having to shave a few days off of each student’s development time for a given project. This should be less of an issue since next semester will only have two crits.

General lessons on teaching

I didn’t anticipate how much of a distraction electronics would be. It feels a bit absurd to do this, but I’m planning on banning all electronics during class discussions, unless a student specifically requests to use something for note taking. Which I would then grant permission for (under the threat of a grade reduction if they take advantage of the policy). I’m not really the most confident when it comes to punishments but it seems like a necessary thing for discussions to be productive. Some students were producing music, chatting, etc.

Additionally, I’m implementing a no late work policy. I had a few students who couldn’t keep up with the work and ended up missing many assignments. I think this will help to prevent work from piling up for those students. In terms of projects the no late work thing sure sounds stricter – but I think it is also necessary. I had students turning in things a few days late or even on crit day and it really fucked up the crits sometimes since only I and a few others would have anything to say about the work.

If anything I want to emphasize that it’s okay to turn in unfinished work/WIPs. But it’s hard, and I assume it’s something students don’t feel as proud of doing.

As for discussions in class, I only did this a few times but taking notes during the discussion is really helpful over a 90-180 minute discussion, in order to reference back things people said. Should have done this more often. Might have been anecdotal, but I felt like discussions were livelier when I sat at the same big table as students, rather than project the game onto the screen during discussion. I’m still for live play, but it seems better to have another student play the game – I think students may see it as easier/more ok to be distracted if I’m up at the front of the class.

Unity (3D)

For 3D, Vanilla Unity is a hard thing to teach to non-coding students. It’s an extremely powerful tool and as  you might expect if you’re not used to coding it can be hard to track down problems.

From the students, I mostly ended up with little/no-interaction first-person games. I tried to teach coding for a few days but this was pretty much a failure. I think teaching coding would be too difficult alongside the other goals of the course – I’d probably have to axe game assignments – so I’m focusing on looking into scripting tools this semester.

This semester, I’m requiring (non-programming background) students to use Playmaker if they want to use Unity 3D, hopefully that will result in better games as it (I hope) helps abstract away some of the coding and get them more into design, rather than worrying about randomly downloaded libraries not having namespace collisions or figuring out C# syntax.

2D Unity

Unity 2D wasn’t much better. For getting student to just make a simple game, there are so many caveats that make this a nightmare… objects disappearing because their Z coordinates change for whatever reason, it’s hard to get pixel-perfect set-up, the sprite animation system in Unity can quickly become tedious.

So I’m using Stencyl instead. Game Maker feels a lot more ideal and documented, but I’ve managed to figure out enough with Stencyl that I feel comfortable teaching it. It also has drag and drop coding, which, while tedious, is less error prone.

And, it works natively on Macs, which is necessary  because my students are only guaranteed to have Macs.


Despite shortcomings with the course the students seemed to all be satisfied. On a few accounts I heard that it was one of their more well-organized classes. Evaluations will come out soon so I can see some more feedback, hopefully. Using different tools should help – I would rather leave the students in a place where they can go further with the tools rather than being stuck at a coding roadblock.

In general…

Teaching is pretty fun, especially when getting a good discussion going or when a student produces interesting work. It’s nice to stay up to date on aspects of youth culture.

I was pretty lenient with grading this time around – not sure how much I should change that.  I think for the most part, students did put in good effort and their projects will be improved by structuring the class slightly differently, and changing up what tools we use.