[ Content | Sidebar ]

yakuza 0

May 23rd, 2018

The only Yakuza game I’d played before Yakuza 0 was Yakuza 2. And it had been eight years since I’d played Yakuza 2, so I didn’t remember it particularly well; just a vague memory was that it had some similarity with Shenmue but that I didn’t like it nearly as much.

Anyways, I started Yakuza 0; and yeah, it felt reminiscent of Shenmue but not as much to my taste, harder-edged in particular. But then a funny thing happened: I made it past the intro and was able to wander around town a bit, and the streets started to feel familiar. And, as the eight-year-old memories came back: this was the same part of Tokyo that I’d seen in Yakuza 2, and then a bit further on, the same part of Osaka.

This is something that I have a lot of respect for: my memory is that, while Yakuza 2 didn’t have areas that I liked as much as the ones in Shenmue, they were still treated with care, significantly more so than in most games. But returning to those exact same areas over multiple games (I now assume over the entire series?) is a whole new level of conviction: the series isn’t just following people’s lives, it’s following the lives of people who are embedded in a community, and treating that community, including the physical space, as worthy of ongoing care.

I’m sure there are other series that do the same thing, but I’m having a surprisingly hard time thinking of good examples. Different Zelda games, for example, hit the same notes, but different games in the series take place in what are effectively different worlds; the Mass Effect games return to the Citadel but you see different parts of the Citadel in different games, and most of your time is spent outside of the Citadel in environments you haven’t seen before; each Dragon Age game explores a different portion of the map.


This idea that the Yakuza games are about lives embedded in a community isn’t, of course, limited to the persistence of the physical space. It’s there in the main plot and the way it plays out, with Kiryu and Nishikiyama having family ties to each other and to Kazuma senior, with the Kazuma family being part of the Dojima family and the Tojo Clan, and with the other intertwined relationships that appear in the game and series. It’s there in the side missions and the minigames: these show the ongoing strands of life in Kamurocho and Sotenburi in ways that whichever protagonist you’re playing gets to touch on but that aren’t actively about him. And it’s there in the non-plot personal development threads, Majima’s cabaret club and Kiryu’s real estate business in particular, showing how the two of them might live and grow their lives if they weren’t embroiled in a huge internal gang war.


And, well, that’s a lot for one game to show. When I was playing the game, I felt like it was too much, and there’s no question that the game would be improved along some metrics if it tried to show less, or at least worked harder to have it all fit together.

Concretely, when I started playing Yakuza 0, I was happy to alternate between spending a fair amount of time hanging out and doing random stuff for an hour or two, and then spending the next hour or so moving ahead to the next chapter. And I assumed I’d be able to continue doing that.

But then the plot started to get urgent, with Majima having deadlines (with death backing up those deadlines!) a day away, so how would it make sense to put in multiple shifts in the cabaret club while all of that’s going on? And then all of a sudden I realized I was almost done with the plot and hadn’t even made it through a fifth of the cabaret club; so my choices were to either stop all plot activities and spend the next 5–10 hours of play time only doing cabaret club managing, or to miss out on large chunks of the cabaret club thread; those weren’t great choices! (And, of course, similarly for Kiryu; for the record, I did all of the cabaret club but very little of the real estate business.)


Looking back over the game, though: it’s certainly a game where the plot is important, but the plot is important in service to a larger vision. Maybe I should think about the game through a sort of Cubist lens: the game and series are depicting Kiryu’s life, Majima’s life, and the communities that those lives are embedded in, and it’s looking at all of that through various perspectives. The plot is one of those perspectives, but it’s only one of those perspectives; and the game simply isn’t trying to tie together all of those perspectives into a transparently coherent whole, it’s instead showing different perspectives and points of contact between those different perspectives without worrying too hard about global coherence.

So, in particular, it’s totally fine to be hard to, say, align Majima’s timelines of saving Makoto and running a cabaret club, or for Kiryu to pop over and do some pocket racing while a massive group of Yakuza are trying to hunt for him. Each of those activities is providing a real perspective on the broader context that the game is representing; and it’s totally fine that, if you focus on one of those perspectives, the other perspective starts to get fuzzy, doesn’t fit well.


I’m significantly more curious about the series now than I was after playing Yakuza 2. And hopefully I’ve learned something about how to appreciate the series that will serve me well when I next return to it…


April 10th, 2018

We discussed Celeste in a recent VGHVI Symposium; I was happy to have an excuse to give the game a try, since I’d been hearing good things about it.

And I’m glad I spent time with Celeste: good platforming, and the story had things to say that I’m not used to encountering in video games. Or maybe things to show: internal struggles, and relations between those struggles, other people, and the world.

Having said that, I stopped about two-thirds of the way through Celeste. The platforming was hard, but I basically enjoyed it in the isolated rooms; when there were enemies chasing me, though, I didn’t have time to think and plot my strategy, and when there were relatively open sections, the fact that my brain was in “focus narrowly on platforming puzzles” mode made it hard for me to enjoy the exploration.


Celeste does have an assist mode, and I considered finishing the game that way. But I decided instead to take a different cue from Celeste, learning from its hotel level.

In that level, Celeste passes through a hotel that’s a mess and that’s run by somebody who is in denial that the hotel is shutting down, and who instead insists on seeing Celeste as a hotel guest instead of as a traveler asking for directions. Another NPC (an occasional traveling companion) moves on past the hotel and encourages Celeste to do the same; Celeste however decides to play along with the hotel keeper, acting like a guest and even cleaning up the hotel.

And the result is that the hotel keeper ignores the help that Celeste is giving, and acts as if she’s betraying him when things turn bad. Ultimately, what’s going on is that the hotel keeper has his own issues: he’s dealing with his grief from the hotel failing, Celeste really can’t help with that, and it’s not her responsibility to help with that even if she could.


Which, I realized, is basically how I felt about the game: it has its own model for what it wants to be, and while I’m glad I dipped into that model, it’s not something I wanted to continue to immerse myself in. So I took the lesson of the hotel level as my cue, and moved on instead of continuing to struggle or instead of trying to adapt to the game’s model.

But I am glad of the hours that I spent with it.

vghvi discord

April 1st, 2018

I’ve been hearing mention of Discord for a little while, but it seemed like it was focused on chat in support of PC games, and I don’t play games on PC. (And, for that matter, I also don’t generally play the sort of games where chat while playing would be useful.) But then we used Discord to help coordinate the Rocksmith Backstage Pass stuff, so I signed up; and I’d been noticing the Discord link on Kittens Game, so I joined that Discord as well, and I got some useful tips and a better feel for the game. And then, after connecting my Patreon account, I got added to a few more.

I’m curious about Discord for another reason, though. Over the last year or so, I’ve felt increasingly uncomfortable with participating on Twitter or with having a Facebook account; I haven’t left either site, but doing that is definitely a possibility. But I do have connections with people on both places that I’ll miss if I leave; so it seems worthwhile exploring other possibilities for spaces to chat with people. Slack and Mastodon had been candidates, but Discord seems like a possibility too?

So I’ve taken one step in that direction, creating a VGHVI Discord channel. If you’re not familiar with the VGHVI, it’s the Video Games and Human Values Initiative; it was formed with grand potential plans, but these days it’s just six of us who get together once a month to play Minecraft (on a world that we’ve been working on persistently for over seven years now) and once a month to talk about a game in depth.


Now I have a channel for the six of us to maintain contact, even if I leave other social media platforms. (At least potentially, we’ll see how the experiment goes.) But I’d be quite happy if the VGHVI Discord expanded beyond us, to include more people from the gaming circles that I’ve spent time in over the last decade. So if you’re somebody I know from blogging or Twitter or something and want to talk about video games, or about human values, or something, let me know and I’ll invite you! The sort of game chatter that we’ve done over the last few years is generally some combination of discussing relationships between games and other aspects of human experience or else trying to understand the design of games; we’ll see how the chatter in the Discord develops.


March 20th, 2018

I wish I hadn’t taken almost two months to get around to writing about Oxenfree, because I’ve forgotten most thoughts I had about the game when I played it. I remember going into the game knowing basically nothing about it other than that some of the Spawn on Me folks really liked it; I was surprised to discover that it it was a 2D take on Life Is Strange mechanics. Not the time travel mechanic, but the general feel of controlling a person wandering around, interacting with a small group of other people, making choices that affect how those people feel about you, with something supernatural going on as a vehicle to give more weight to the experience.

Which, honestly, I would love to be a trend (and, for all I know, is in fact a trend): I’m all for there to be more games exploring interpersonal interactions instead of shooting, and doing that in 2D feels like it wouldn’t hurt the core of such games while significantly lowering barriers to entry / costs?

The flip side, though, is that, while I was happy to have played Oxenfree, it (clearly!) didn’t grab me. Not sure how much of that was the story it was telling, how much was the mechanics, how much was my tastes, how much was my mood at the time.


Or maybe I’m wrong in saying that such games wouldn’t lose anything being transported to 2D: maybe it doesn’t make sense to talk about the core of such games as divorced from presentation aspects? So maybe in fact there’s something important about being embodied in a way that a 3D game having you wander around a school (and other similarly contained areas) can emphasize but that a 2D game having you wander all over a mountainside has a harder time pulling off? Or maybe there’s something to the cinematic high production values approach and how I respond to them?

Typing that out, either of those hypotheses actually makes total sense. Still, I’d like to try out more games like this, whether in 2D or 3D. (And, to be sure, it’s not like those two games are the only examples, e.g. The Walking Dead also fits into the box I’m trying to describe.) And I certainly wouldn’t discourage other people from playing it: I’m happy to have played it, after all!

rocksmith jobs to be done

March 7th, 2018

It’s been a while since I’ve written about Rocksmith, but I’m still happily putting time in on the guitar every weekend; so I figured it was time for another post about the game. And, this, time, I wanted to talk about what sorts of things I might like out of the next evolution of Rocksmith.

Which, in turn, raises the question: why exactly do I play Rocksmith, what am I trying to get out of it? Or, if I take a Jobs to Be Done lens, what am I hiring the game to do? Because ideas for the game don’t make sense in a vacuum: there should always be some underlying goal that I have in mind when making any suggestion.

Though there’s no one answer for what I’m trying to get out of Rocksmith: I’m getting different things out of it at different times, and I’ve gotten different things out of it over the years as well. Which, to be sure, is a strength of the game: it can build on the same underlying capabilities to serve players in different ways, or to serve the same player in different ways at different times. (That latter point is one difference between Jobs to Be Done and the “Persona” concept that I see in UX design: focusing on the end goals rather than on the people. Not that Personas are bad, just slightly different.) But this lens is, I think, still useful as a way to focus on the effects of different design choices.


At any rate, here’s my list of jobs:

  • Improving your skill at a technical level: going from zero guitar experience to being able to play basic songs to tackling harder and harder stuff.
  • Enjoying the experience of moving through music: playing through a song that you like to listen to or that like how it feels on your fingers, then moving onto the next song, potentially going through dozens of songs in a single session.
  • Honing your performance: figuring out what it means to play a single song well, not just on a “did you get the notes right?” level but at a deeper musicality level.
  • Improvising: being able to jam with a band without specific notes to play, just a chord progression or a basic riff or something to build on.
  • Understanding the bones of the music: appreciating music theory and the theoretical underpinnings of composition.

I’m going to talk about each of those in the context of the game, how Rocksmith currently supports those modes of learning and how it might support them still better.


Improving your skill at a technical level

I didn’t start playing Rocksmith from zero knowledge: I had a tiny bit of guitar experience from messing around one summer in college a couple of decades back, and I played Rock Band 3’s Pro Guitar mode for a year or so before I started in on Rocksmith. (In fact, the Squier Rock Band 3 controller was the guitar that I used when first playing Rocksmith; Rocksmith quickly made me realize the problems with that, so I bought a much better guitar a few months later.) Rock Band 3 had a lot of problems when it comes to learning guitar, but it got me past the initial hand pain of barre chords, taught me how to move around a fretboard, and gave me a baseline skill level of pick usage (including alternating strumming).

Having said that, I am 100% positive that Rocksmith would have been a much better place to start. It would have taught me all of the above; I might have taken a little longer to focus on barre chords, but I’m not sure that that’s true or that that would have been a bad thing. (Rock Band 3’s controller also had some limitations that had serious conceptual musical downsides, but I don’t want to go into details here.) And Rocksmith comes with a nice set of lessons to introduce you to guitar techniques.

What really makes Rocksmith special in this regard, though, is the way it adapts songs to your level, and does that on a per-phrase granularity, with (typically) a dozen or more difficulty levels available for every single phrase. And, to be clear: this isn’t just a way in which Rocksmith is better than Rock Band 3 was, it’s a way in which Rocksmith is better than human teachers or printed music.

Because one key to learning of any sort is to constantly be going up against challenges that are within reach but beyond your comfort zone. (“Deliberate Practice”.) And arguably Rocksmith’s single most important didactic design decision is to do exactly that: it is constantly asking you to stretch your limits on every part of every song that you can’t yet reliably play all the notes on, but it stretches you in a way where the next goal always remains within reach.

A second key design decision in this regard is Riff Repeater mode: once you get to where the next step on a given phrase really is a significant challenge, then you’ll need to put in some effort to work on it; so Riff Repeater lets you play that phrase over and over again, with various knobs (speed, in particular) to help you work your fingers until they’re doing the right thing. (And to help your brain understand exactly where and how your fingers aren’t doing the right thing!)


You can make a huge amount of progress with the tools that Rocksmith currently has. Having said that, if I wanted Rocksmith to do better in this regard, it seems like there are two potential areas for improvement: detecting techniques more accurately and more broadly, and having a richer holistic understanding of the player’s skill level.

Rocksmith gives notation for a fairly wide range of techniques, but it doesn’t actually grade you on most of them, it restricts its grading to playing the right notes at the right time. Which, in general, is the right choice: most of the time, I don’t actually want the game to give me a grade as to how good my palm mutes are.

I do want it to grade me accurately on the notes, though. Which it does a pretty good job of, and it errs on the side of accepting what you play, which is also the right choice: it’s very frustrating to play the right thing and have the game tell you you’ve done something wrong.

But occasionally the game tells me I’ve done something wrong when I don’t realize what I’ve done wrong. A lot of the time, I can eventually figure it out: e.g. I have a habit of not always strumming through three-string power chords, and sometimes the game is correctly pointing that out! (Not that there’s a huge musical difference between two-string and three-string power chords, but I would like to have the pick control to reliably choose the one or the other.) Similarly, I occasionally get the rhythm a bit wrong on a section without realizing it (especially when a song has similar sections with varying rhythms). So, in both of those situations, I appreciate the game telling me I’ve made a mistake; I just wish that it could tell me what mistake I’ve made. (At least if I’m playing in Riff Repeater mode.)

But whenever I encounter those situations, I also don’t always trust the game. This comes up most frequently in bends: Rocksmith won’t reliably detect bends unless you pause at both the start and end of the bend to give it a clear signal on the pitch of the note. Which isn’t right musically; and it means that, if it says I’ve gotten a bend wrong, I don’t know if I was being sloppy about the ending pitch, if I was being sloppy about the start and started bending before I’d strummed at all, or if I was playing in a reasonable way that was just too fast for the game to detect.


The lessons are good to start with, but they have their problems as well. Sometimes playing through the lesson is frustrating, because it doesn’t want me to get past a section until I’ve 100%’d it; if I’m having trouble with that section because it’s too hard for me, that’s fine, but if I feel like I’m doing what it asks and it’s still showing failed notes, then that’s annoying, and it feels like it happens in the lessons more than it happens in regular playing. (I’m not sure if Rocksmith tries to detect techniques in the lessons that it doesn’t try to detect when playing songs? It feels that way…)

More broadly, the game doesn’t feel like it has any real concept of what techniques I’m good at, what I’m bad at, and how to help me with the latter. I’ve been playing the game for years, and just this morning, when I was playing through a song, the game had, as one of its recommended items, to play through the Chords 101 lesson; trust me, Rocksmith, I’m not going to get anything out of going through Chords 101 again.

The other tool the game has for technique improvement is the Guitarcade games. Probably I should spend more time on those; but what I find happening is that it asks me to do something faster and faster, I start failing, and (on many of the games) it feels like my failure has as much to do with slow detection (e.g. taking a while to decide that, yes, I have strummed the chord it asks for) or failed detection (harmonics, bends) than it has to do with my failures. Which is frustrating, and I react to that by staying away instead of trying to pay attention to the part of the grading the game is doing that really does reflect on me.


Ultimately, if I look at Rocksmith through this lens, I’d like the game to better understand what I’m good at technically and what I could use work on and to have it more frequently give useful, actionable advice on the latter.

The flip side, though: a lot of this is on me, and the tools are there for me to take control. Sometimes I don’t know what I need to get better at or how to do it, but a lot of the time, I know what I need to work on, I just need to put in the time; and Riff Repeater helps me do so. Or at least it helps me do so if I can map my technical difficulties to specific sections of songs; but usually I can, because if a technique doesn’t show up in a section of a song, then I’m not likely to care much about it!


Enjoying the experience of moving through music

This is the aspect of Rocksmith that I find most seductive: just diving into songs. Picking a song, or having the game pick it for me; playing through it once, maybe playing it a second time if I enjoyed it and I either want the experience of going through the song again the same way or the experience of trying a slightly more complex version now that I’ve gotten the phrases leveled up a bit.

I can easily spend a couple hours solid just doing this. And it’s clearly something that the devs care about: Rocksmith 2014 was better at this experience than the original Rocksmith, and the Remastered patch added on still more improvements.


What more would I want when approaching the game in this mood? Honestly, the basic flow is really good: I’ve got various ways to pick the collection of songs that I want to play, and then the game is happy to either let me go through the selected songs manually or to serve them up to me itself. In the latter mode, it can sometimes be a little hard to pause things if a song particularly catches my interest, but that’s only a very minor issue.

So, really, it feels like the major remaining issue is the number of songs that are available at all: the number of songs I have available in my actual library, and the number of songs that are potentially available but that I may or may not have purchased.

The first step there would be to remove that distinction: switch to a Spotify model where everything’s available as long as you pay a subscription fee. This is clearly the way the music industry is going; I imagine Rocksmith will switch to this at some point, I just hope all the licensors are on board.

But, as vast as the Rocksmith library is, it is nonetheless even more vastly smaller than the total library of guitar music! There are certainly incremental changes that I’d like (e.g. add in some complete albums), but, ultimately, that’s a problem that will need algorithmic solutions: growing the library by depending on manual notetracking is not going to scale.


I assume that, 20 years from now, there actually will be something algorithmic that can do a credible job of acting like Rocksmith for my entire music library, or for the entire world of streaming music. (People are already giving this a try; at some point I need to give Capo a spin to figure out how good it is…)

I don’t expect that product to be an evolved version of Rocksmith, though, because it would require a complete business model shift: this looks like a classic low-end disruption scenario to me, and that almost never turns out well for the incumbent. But that future is also far enough out that I don’t see any point in worrying about it now, anyways: if the Rocksmith folks wanted to worry about it, I’d tell them to try to make money by satisfying their existing customers, see if they can use Jobs to Be Done theory or something to set up a lateral move, and in the mean time explore automated technology somewhat themselves, as an internal tool to speed up the process of note tracking.


Hmm, that last sentence makes me wonder: maybe I’m wrong and automated song detection is a sustaining innovation instead of a disruptive innovation? I was going to say that I doubt it, because of the business model shift, but maybe I’m wrong about that, too: Rocksmith should be able to successfully shift gracefully from a “buy individual songs” model to a “pay a monthly fee to stream from a fixed library” model, I think. And shifting from the latter to a “pay a monthly fee to stream from almost anything” model, with most songs handled algorithmically but some popular ones handled with human intervention, might actually work just fine?

I guess the problem is competing against something like Capo that’s a one-time fee piggybacking on existing payments to a traditional music-listening streaming service, but that doesn’t sound so intractable. And maybe the existing business model would make the Rocksmith developers reluctant to really focus on getting good algorithmic note tracking. Not sure…


Honing your performance

Rocksmith does encourage you to learn all the notes in a song. It nudges you towards 100% in various ways: gradually leveling up sections of a song, changing the color of a section when the game is willing to show you all of the notes, a Score Attack mode which then pushes you towards going on longer and longer streaks without missing a single note, and Riff Repeater mode to support you in learning individual sections that are giving you trouble.

And, once you’ve shown you can play all of the notes in a section, Rocksmith then asks you to memorize the song, by switching into Master mode. Its implementation of Master mode in Learn a Song mode is a little iffy, honestly: sometimes it throws me into Master mode inappropriately (especially when there are multiple similar sections in a song), and sometimes it obstinately refuses to go into Master mode. But it’s still pretty useful as a tool, and more so as an aspiration: the point is to really learn the song. And Master mode is there in Score Attack as well, to test you more comprehensively: would you be comfortable playing the whole song in a band, without the notes in front of you?

Knowing (and being able to play) the notes is just the table stakes for being able to perform a song, though: there’s a big difference between being able to fumble your way through a song more or less correctly compared to getting something deeper out of the music. Here, Rocksmith’s tools are fewer, but it has some: the fact that it shows you accents and vibrato is a reminder that not all notes are created equal, that you should think about which ones are more important and how they work together.


Helping you develop the musicality of a piece is, of course, an extremely difficult task for software to undertake. Software can only grade what it can measure; and different interpretations of the same piece can be equally valid, and I wouldn’t want Rocksmith to judge me for every time I different from the interpretation of the recording I was following. I won’t swear that there’s not more that Rocksmith could (perhaps optionally) do — e.g. sometimes my chords are muddy, sometimes I play notes in a way that makes it clear that I’m just barely managing to get by, and perhaps Rocksmith could point that out? (Which, I think, it already tries to do with its suggestions of which sections to work on in Riff Repeater, but that could be more actionable.) But maybe this is an area where I don’t want Rocksmith to take the lead.

But I do want Rocksmith to support me when I’m thinking about how I want to perform a piece. Which, of course, I try to do when I’m playing, but it’s hard, especially if I don’t have the piece ingrained in my fingers. When I’m working on a piece on the piano, I’ll stop, think, experiment with fingerings and phrasings for a measure or a few notes, write down stuff on the score; if Rocksmith could let me do that, then that could be nice, but I don’t have any super concrete ideas how to do that beyond what’s already there in Riff Repeater.

Rocksmith does have a significant advantage over sheet music over a piano, though: it should be able to record what you’re doing easily enough. And, in fact, this is the one thing from the original Rocksmith that I miss in Rocksmith 2014: when you finished a piece in Rocksmith, it would play it back for you, letting you hear your performance, showing you the notes (even if you’d been in Master mode), and showing you which notes you missed.

And it was really eye-opening to hear just how bad I sounded on sections where I could just barely 100% them: I realized how much work I had to do beyond just improving my completion percentage. Even on sections where my fingers were more confident, though, listening to myself helped point out sections where I stumbled a bit or where I just might want to think more about how to approach a phrase; and it was extremely useful as a followup when I was playing in Master mode, as a learning tool for the sections that I’d thought I had memorized but, in retrospect, hadn’t. (Admittedly, Rocksmith 2014 has much better Master mode tools than the original Rocksmith, but even so, I think seeing the notes and seeing/hearing my mistakes after the fact would be useful.)


So that’s my main request for a future Rocksmith in terms of improving my musicality: some sort of Review mode. It would play back your performance; it would show you all of the notes in the song, even if you’d played in Master mode; it would show you your mistakes; it would let you drop into Riff Repeater in the middle of a piece to work on a section that doesn’t sound good.

If the Rocksmith developers had more ideas about how the game could give you suggestions about musicality, Review mode could be a place for that, too: it would be a context where you can think and talk about your performance after the fact, which removes issues of grading and real-time feedback. So if they have ideas about ways the game could give feedback that aren’t reliable enough for real-time scoring, and that do a better job of matching the player’s performance than the current rudimentary “here’s a lesson you should try” capability, then this could be a place for that. But that would all be a bonus: just being able to listen to your performance in a way that ties into existing Rocksmith capabilities would be enough.



When I play guitar with my coworkers, we spend a lot of time doing jam sessions; it’s fun, I wish I were better at it! And of course Rocksmith 2014 took direct aim at this with Session Mode.

I don’t have a lot to say here, unfortunately: I’ve spent several hours with Session Mode, I’ve gone through all of the Session Mode missions, and I’m quite glad I did, but ultimately I only have so much time that I spend with Rocksmith each week, and I’ve been spending that time playing songs instead of improvising. Also, I don’t have much domain knowledge here; I don’t know that I’m significantly worse at improvising than my coworkers, but it’s not something that I’d feel comfortable doing in a more formal situation, either.

So, while I genuinely have a lot of respect for Session Mode, I don’t feel like I can say anything particularly productive about its strengths and weaknesses. I’m leaving in this section because I think that “be able to play in a jam session without being embarrassed” is an important Job to Be Done for Rocksmith; I’m glad the Rocksmith developers have been thinking about it, I hope they continue to do so. (And if I had more time to practice guitar, I’d definitely carve out time to spend in Session Mode!)


Understanding the bones of the music

Basically: what can Rocksmith teach you if you want to write your own songs, or just understand what the composers of the songs in the game might have been thinking?

This has significant overlap with the previous job, of course; maybe I should have combined them into a single entry. And Session Mode does have some tools that attack this directly: it focuses on both scales and chord progressions.

When playing through songs themselves, Rocksmith doesn’t help as much with the underlying musical structure; it gives you the names of chords, but that’s not really what’s important, what’s important is there those chords are in relationship to the underlying key.

Though actually two decisions that Rocksmith made in Learn a Song mode do help with this in a roundabout way: it lets you play whatever you want during blank spaces in the song (and doesn’t actively penalize you even if you’re experimenting when it tells you to play specific notes), and, when you’re playing a new song, you don’t generally get to see all the notes. The latter decision actively poses the question of what notes would make sense in this context; and the former lets you experiment with how to answer that question.


I’m vaguely curious what it would be like to play Rocksmith with it labeling chords with I / ii / IV / V instead of, say, E / F#m / A / B: that feels like it would help me train my ear to recognize the different chord progressions, which would in turn help me understand them better. But in general I don’t have a lot to say about this job, and of all the jobs I’ve listed here, this one feels like the most of a stretch; Rocksmith made the right choice to focus on improvisation over composition / music theory, I think.


If I had boil this down, what I most want is: 1) let me review my performance after playing; 2) do a better job of quickly and accurately detecting what notes I’m playing; and 3) keep the business model up to date so it continues to be worth supporting Rocksmith. And, of course, the developers probably have completely different ideas, either to help with the above jobs or for completely different jobs; Rocksmith 2014 certainly brought a lot of improvements that I wouldn’t have known to ask for. As long as I can keep on playing guitar, and can keep on getting better, I’ll be happy…

twist steps and refactoring

February 18th, 2018

One theme that constantly comes up in Tai Chi classes is keeping control over where your center of gravity. At any point in the form, you should know whether your weight should be completely over your right foot, completely over your left foot, somewhat on the right side, somewhat on the left side, etc. You can see an example of that in this stepping technique drill video by my teacher: he shifts back and forth from leg to leg, and pauses with all of his weight on his stationary leg while taking a step with the other leg. (Later on in the video, he even raises his moving leg to emphasize the single-leg position.)

That’s an isolated exercise, but there are several places in the form where you’re doing a similar sort of steps. And, a few months, ago, when my teacher was giving me pointers after watching me go through the form, he noted on one of those sections (Twist Steps) that I wasn’t shifting my weight properly, and suggested that, before moving my foot forward, I should sink into my weighted leg, to help ensure that I was positioning my weight properly.

I worked on that over the next week or two, and it felt like my technique got significantly better, and not just on that one part of the form: I could feel my body clicking into place on transitions in a way it hadn’t before. I wasn’t sloppily combining weight shifts with other movements (or at least I was less sloppy!): at each point, I knew where I was and what I was doing.


When I’m programming, I do test-driven development. Which means that I know what I’m doing at each point. Sometimes I’m writing a failing test, sometimes I’m getting that test to pass, sometimes I’m refactoring to improve the quality of my code. (And sometimes I’m thinking about what to do next!)

This clarity of state reminds me of practicing my form: I have some specific behavior change that I’m working on at any point (similar to knowing the next move in the form); I know whether I’m adding a red test or making it green (similar to knowing whether I’m moving my feet and what my weight is doing); and I pause periodically to refcator (similar to the weight sinking mentioned above, and to the exhaling and relaxation that I do in pauses in the form).


Of course, the Tai Chi form is an artificial practice: the moves have applications, but they’re quite different from how you act in other contexts. For example, when walking down the street, I don’t lift my left leg, move it over to my right leg to make sure I’m balanced on my right leg, then stick my left leg back out and put it on the ground without placing significant weight on it, and finally shift to my left leg while maintaining contact on the ground with both legs. Instead, I just move my left leg straight forward in a way where I’m somewhat off-balance for most of the movement. (Though, arguably, I’m less off-balance while walking normally now than I was a year ago: if I’m thinking about it, walking really does feel different now than it once did!) So why don’t I do the same thing while programming: treat test-driven development as a valuable exercise but not my normal mode of behavior?

When doing Twist Steps, my teacher says to think about the move as if you’re walking on thin ice: don’t put your weight down until you’re sure the ice will hold and that you won’t slip and fall. The thing is, while programming, you’re walking on ice a lot of the time: there’s almost always something that can go wrong if you’re not careful. If your code base is decent quality, most of the time it’ll be okay for a while if you march boldly ahead, but every so often, surprises will pop up, and you’ll fall flat on your ass in the good case, sink into icy water in the bad case.

So spend your time feeling out your territory while programming: that’s the red-green part of the TDD cycle, small steps guided by tests are what help you avoid slipping too often. But in programming, unlike when walking in the physical world, you’re directly affecting the territory that you’re navigating: you can make it more or less slippery as you walk around. So refactor constantly: that improves the quality of your surface.

And, if you’re doing that, then yes, you can walk faster! Which can tempt you to stop doing TDD; but, the better the quality of the code, the easier it is to add that next test that makes your movement safe. And, as with the form, if you practice mindfully over and over again, you have the potential of moving more quickly while maintaining discipline: you can stop and think if you want, but you can also move quickly without losing clarity.

super mario odyssey

February 8th, 2018

The obvious points of comparison for Super Mario Odyssey are the latest Zelda game and Super Mario 64. Breath of the Wild in particular pairs with Super Mario Odyssey as a statement game: they’re the most recent entries in Nintendo’s two most important franchises, working together to launch a hugely important console for Nintendo. And they share something new in their design: just as Breath of the Wild sprinkles shrines and Korok seeds all over its world, so too does Super Mario Odyssey manage to hide a remarkable number of moons within each of its levels.

That comparison, though, points out why, for me, Odyssey is no Breath of the Wild. Odyssey is a very good game, to be sure; but what I found most magical about Breath of the Wild is how natural the world felt. Not only did every hill and every tree feel like it was in the right place, but every time you came across a Korok, a shrine, a village, it seemed like that was exactly where it belonged.

In Super Mario Odyssey, however, the levels just don’t fit together the same way: they’re segmented instead of flowing, and the result is that each level felt to me like an agglomeration of puzzles and locations rather than a whole. I was going to say that they feel designed rather than organic, and that’s true, but that doesn’t quite get at the issue: overtly designed structures can have their own beauty as well, where every portion is there for a clear reason. But Odyssey doesn’t do that, either, or at least doesn’t do that across the scope of each of its levels: individual puzzles are well designed, but the fabric between them isn’t, or at least isn’t to the extent that Breath of the Wild managed.


Breath of the Wild is, of course, a very high bar, and for most games and most series, it would be a completely unfair comparison. But with a Mario game, it’s not an unfair comparison, and in fact the game showed why not at two separate times.

One is after the end of the main narrative arc of the game: you return to the Mushroom Kingdom setting, with a level based on Super Mario 64. And as soon as I landed there, I felt at home, at home the same way I felt in Breath of the Wild. You can make a case that that’s nostalgia, but I don’t think that’s all of what’s going on there, or even most of what’s going on. It’s a castle and a castle grounds, and the grounds are designed in a way that feels right to me in a way that most of the other levels didn’t, with hills and trees and paths and water just where they belonged. In most of the other levels, the locations of moons made sense because that’s where a Mario game would put a moon, but there was a tension between the parts of the levels that were there for platforming and the parts of the levels that were there as connective tissue; in the Mushroom Kingdom level, though, I didn’t get that separation.

The other reason why I felt at home in the Mushroom Kingdom level was the music, and the sound design more generally. Again, I’m sure that some of that is nostalgia, but I also again don’t think that that’s everything that was going on: the music in Super Mario 64 was a lot better than the music in Super Mario Odyssey. (Admittedly, this is an area where Breath of the Wild doesn’t do so well, either: Ocarina’s music was much better.) And a lot better not just in an abstract sense of preferring the earlier game’s tunes, but in a situated sense of being the right sounds at the right time.


There was actually one other level in Super Mario Odyssey that felt as good to me as the Mushroom Kingdom: New Donk City is a ridiculous name, but it’s a great level. It felt a little corridory to me when I was going through it initially, but once I made it past the boss fight and the sun came out, I really liked the level: full of people who had a reason to be there, full of buildings that made sense in that context, but the buildings also all served a second purpose as platforming navigational challenges and a third purpose as suitable settings for hiding isolated puzzles.

And, while I wasn’t actively impressed by most of the music in Super Mario Odyssey, I loved the song that caps off New Donk City. And I loved the set piece that that song fits into: organic puzzles are great, but artfully designed set pieces that combine theatre and gameplay into a virtuostic package can be special as well.


Don’t get me wrong: I thought Super Mario Odyssey was a very good game. I was happy to play it, I stuck around for quite a while in each level the first time after making it past that level’s boss fight and I spent a fair amount of time with the game past the end of the main story. It’s just a very good game that actively invites comparison to both Breath of the Wild and Super Mario 64; at its best, it even manages to come off evenly in those comparisons, but most of the time, I wanted a little more…

rethinking my twitter usage

January 29th, 2018

Three months ago, I removed Tweetbot from my phone: I was spending too much time on Twitter, and getting too caught up in the hour-by-hour drama of politics. And that was definitely the right choice: I spent less time distracted and less time worrying about stuff that I can’t do anything about on the timescale when those worries appear.

I didn’t stop reading Twitter, though: I read it on my iPad at home instead. And what that revealed was that Twitter wasn’t causing problems just by being a distraction: the volume of time I spend reading Twitter instead of doing other things is higher than I would like.

Thinking about that more: if I run a thought experiment of what I would spend my time on if I magically had much more non-work time, I absolutely would not answer “follow more people on Twitter”. So that’s another signal that, yeah, I should reduce my Twitter usage.


The next question is: how? The obvious thing to do is to follow fewer people, until the amount of new tweets each day matches the time I want to spend on Twitter. A second option would be to stop being a Twitter completist: put in a time budget and accept that sticking within that means that I’ll miss stuff. And, of course, there’s always the option of quitting Twitter entirely; certainly the company’s behavior makes that attractive.

The problem with the first option is that I feel like I’m getting something valuable out of everybody that I currently follow on Twitter, I’ve pruned my follow list pretty well over the last year or two. The problem with the second option is that I have friends whom I follow where I actually don’t want to miss stuff they post. And the problem with the third option is the combination of the problems of the other two options.

That second option, though, gives a potential way forward: most of my timeline consists of people who aren’t my friends and who post in fairly high volumes. I’m following them for a reason, they do post interesting stuff, but, honestly, I’ll be fine if I don’t read everything they post or even most of what they post.

So I think what I want to do is divide people into two classes; find a way to be completist for one class; and only browse the other class.


The next question is: how? And I could use some advice here: this feels like it must be a standard problem, but I don’t really know how to solve it. Ideas that I’ve had so far:

1) Create two separate accounts, following the different groups of people on different accounts.

2) Only follow my friends on my main account, but create a private list for people I want to browse who aren’t my friends.

3) See if enough of my friends are on Mastodon (or micro.blog, but that seems less likely) that I can make do with that.

4) Write a client of my own that can do this!

Also, whatever solution I end up with, I have a fairly strong preference for being able to carry it out on my iPad, though I can imagine using a computer to browse non-friends.


Thoughts? Any ideas that I’m missing? Does Tweetdeck have some magic solution to all of this? Here’s how I see the above:

1) This seems straightforward enough; though one downside that I see is that I can see myself accidentally replying from the wrong account, and in general account confusion isn’t great.

2) It feels weird not showing as following people whom I actually am paying some amount of atention to? And I’ve never used lists before, so I don’t really know how that works.

3) This is pretty tempting, actually: I do want to move off of Twitter, after all. I feel like other options aren’t well enough developed for this to work yet, but who knows.

4) This idea is, of course, ridiculous. The only serious argument for it, though, comes from the thought experiment mentioned earlier: one of the things that I would spend more time on if I had more time is some sort of personal programming project, and learning about iOS programming would qualify. Also, I wouldn’t have to write a full-fledged client: it just has to be enough for a subset of personal use, so rough edges are fine, and it doesn’t have to support reading direct messages or being notified about replies, I can use other apps for that. And it also has the side benefit that it would give me an excuse to pay money for an iOS developer account, which means that I could keep my custom Tokyo Mirage Sessions iMessage sticker pack permanently installed on my phone instead of having its certificate expire every week!


Writing that out, I think I’m going to take option 2, but I’ll actually think somewhat seriously about option 4? And if anybody has any pointers about this, I’m all ears: it feels like what I want to do is standard enough that somebody must have written up the best way to accomplish this?


January 25th, 2018

Gorogoa felt right to me from the very beginning, from this screen that begins the game:



The picture in the window is lovely, drawn with care; but look at the texture and slightly off-white color of the paper that makes up the boundary of the screen, and look at the untextured white of the window’s boundary strip. This is a game that feels alive, a world where I want to spend time.

Watching the initial scene through that window (and clicking around in it, serving both as a touch of interactivity in what would otherwise be a cut scene and as a gentle tutorial teaching me about changing perspective) did nothing to disabuse me of that notion. And then the one window became four tiles (one filled and three empty), and I got a new interactivity mechanism: it showed a view through a window, and I could move the outside of the window to a second tile.


So that’s the puzzle mechanic: different takes on changing perspective. You can zoom in, you can zoom out, you can pan side to side; you can decompose and recompose, removing outer layers or superimposing them. And you can combine by side-to-side adjacency as well. It’s a mechanic that I haven’t seen before; it’s a mechanic that I enjoy.

Gorogoa’s use of that mechanic turned out to be a good match for me difficulty-wise, too: there were one or two places where it took me a while to figure out how to progess, but I made it past those, and felt like I learned something about the game’s internal logic in the progress. I like Monument Valley quite a bit, but that game decided to go fairly far in the direction of having puzzle solutions be clear; Gorogoa stepped back from that noticeably, but I never ended up needing to head to gamefaqs. It admittedly doesn’t have the sort of puzzle design that supports deep reasoning, admittedly — ultimately, you’re mostly presented with a relatively limited set of potential actions (places to tap, mostly) at any given stage, and trying to figure out what combination of scaling and adjacency will let you progress — but there’s enough there to give you something to think about, and a bit of lateral thinking as well.


More importantly, though, the game grabs my soul. Looking at the basic visual design of the game (the aforementioned textured paper outlines) makes me feel better, looking at the drawings in the game makes me feel better, seeing the hints of the world that the game presents makes me feel better, dancing between scenes and making connections makes me feel better. That’s a rare experience, even outside of games; I should seek it out more.

move your dna

January 14th, 2018

Move Your DNA is the latest book I’ve read in order to try to understand how to move / position my body in a healthy manner, and it’s quite interesting in a way that, I think relates to Kegan’s stages of understanding. Because the main point of the book are that your body, even when it is doing things you don’t like (e.g. causing pain because of bulging lower back disks!), is responding as best it can to the circumstances you put it in. And, as a corollary, doing any one thing (taking an action, being in a position) over and over again, almost no matter what that one thing is, isn’t going to maximize health, since your body will be over-adapted to the forces that that context creates. (Again, my lower back problems, which I suspect were a result of my body’s attempt to deal with the amount of sitting that I did and the particular posture I used while sitting.)

And that’s all well and good, and it certainly makes sense. But it’s a sort of making sense that can easily lead in a Stage 3 direction: without the guiderails of Stage 4 systematicity, it’s too easy to interpret whatever sort of common sense actions you’re engaging in as an appropriate amount and type of shaking things up, without really understanding what’s going on. And, of course, it’s an answer that raises questions about the wisdom of following Stage 4 answers: if you’re not careful, those can end up warping your body in different ways!


For example: how much does Move Your DNA suggest that I should worry about the Gokhale method being bad because it instills new habits which could cause their own damage if followed regularly? I’m honestly not sure what the answer is to that; right now, though, I’m not too worried. Bowman isn’t agnostic about different positions, either, there are some that she likes more than others; and it feels to me like Gokhale’s idea of a desired neutral body position (which I have found very useful) matches reasonably well with Bowman’s starting points?

I’m less sure about some of Gokhale’s other recommendations, the ones that are designed to stretch your lower back in particular. They feel like they replace one set of external forces that our anatomy isn’t well adapted to with another set of forces that doesn’t match anything that would have been regularly encountered in the evolutionary environment, and that seems potentially dangerous if taken to an extreme. But the flip side is that, done appropriately, they feel to me like therapeutic exercises rather than something you’re supposed to stay in regularly for long stretches of time, and in that context they’re probably fine / good?

And certainly some of what the Gokhale method has revealed to me matches some of Bowman’s major points. E.g. Gokhale has some techniques to alter your shoulder positioning; they have definitely had an impact on my shoulder positioning, in a way that feels helpful to me, but more than that they have revealed to me how much I (and other people I see as I look around) have been trained to hunch and to position shoulders and rotate arms towards the front of their possible range of motion, which does feel off. And it’s actually still the case that, now that I’m aware of my shoulder positioning, my two shoulders don’t feel symmetrical; I assume that that’s a consequence of differential arm use. (Though who knows: I spend an awful lot of time typing, and there my positioning is symmetrical! At least I think it is…)


The other body positioning and exercise regimen that I have to compare Bowman against is, of course, Tai Chi. Bowman is a proponent of exercise, but she also cautions repeatedly that repetitive exercise (e.g. lots of running; especially on treadmills, but also on flat roads, though she likes it quite a bit more if you run on varied terrain) doesn’t bring the benefits that you’d hope. It conditions your body to expect the repeated context that that exercise provides, and while having a second repeated context added in isn’t a bad idea, it still leaves out a whole range of potential forces that your body could be adapted to.

I’d like to think that Tai Chi provides a large enough range of movements and forces on your body that it’s less susceptible to Bowman’s critique than some other forms of exercise, but who knows. The (Chen Lao Jia) first form has 75 parts to it; there’s a lot of repetition in there, but still, it’s asking you to a reasonable range of different things with your body? And the weapon sets add still more forces (e.g. the Dao with its forcing you to respond to the inertia that comes from its weight), and I’m finally experienced enough to be able to start learning the second form, which adds still more into the mix.

Though the flip side is that Tai Chi does have its repetitive practices; and, in fact, it sometimes takes those practices to extremes. (Which is the Chi part of the name: it’s a character 極 that means extreme!). For example, standing meditation, where you stand in place for, say, 20 minutes at a time, is something that my teacher recommends.

Arguably, though, Bowman’s book sheds an interesting light on standing meditation: being in one position for 20 minutes is going to have an effect on your body, but if you’re spending that 20 minutes focused on something else (typing away, say), then you’re not going to be paying attention to those physical affects or trying to shape them in a way that’s helpful instead of harmful. Whereas, with standing meditation, your physical positioning and understanding your body’s reaction to that is part of the goal, as is tweaking your positioning as your understanding develops so that you can guide your body into a positioning that helps you.


Hard to say; I don’t have nearly a deep enough understanding of the consequences of what Bowman is saying to be able to use it to seriously analyze other activities at a more-than-superficial level. And, while I’m curious about the book, it’s rich enough that really diving into it would take quite a bit of time; and I’m choosing to only budget so much time to that sort of physical experimentation; right now, Tai Chi is eating up (and, actually, slightly expanding) that budget.

But I do think that Move Your DNA is potentially a pretty interesting / important book? And it’s definitely good to get some encouragement to try opportunistically moving in different ways: I wrote a bit of this post while squatting instead of sitting in a chair, and I’m straying off of sidewalks a bit more. And one concrete effect that the book has had is that it’s had me notice my feet, and the signals coming through the soles of my feet more. I’m sure it’s really a culmination of the Gokhale book and Tai Chi, but when Move Your DNA started talking about feet as highly flexible and able to sense quite a lot if you let them, I realized that, actually, if I paid attention to my feet, I really could feel quite a lot through them and have them in a range of positions even inside my shoes.

Which wasn’t so much the case a few years back when I was experimenting with lightweight shoes: back then, I had a hard time really feeling like my feet were responding to the environment, and if I went too far in the lightweight direction, I felt like I was impacting my body in unfortunate ways. Maybe it’s time to give that a try again? Bowman actually has written a couple of books on that subject, too…


January 7th, 2018

Undertale is charming in a way that I found surprisingly refreshing: it’s telling a story, the story leads off with a major conflict but almost immediately shrinks to a personal level, and then the personal level gets a lightweight touch. And that doesn’t combine to make the game feel slight: it makes the game feel human.

And that’s all pretty remarkable, but I don’t have much to say about that aspect of the game. So, instead, I’ll talk about another design choice Undertale made: your ability to make it past enemies by empathizing instead of fighting.


I’d heard about that going into the game; I’d also heard that a pacifist route was quite difficult, and so for a first playthrough you probably shouldn’t go full-out pacifist? But it wasn’t that bad at the start: confusing, yes, because I got out of a battle but I didn’t get any experience so I couldn’t tell at first if I’d instead triggered some sort of escape mechanic?

Eventually, though, I realize that there were changes in the UI (yellow text, basically) that signaled that I’d triggered the ability to resolve battles peacefully; so yes, I was resolving conflicts, I wasn’t just avoiding them. I did wonder about the consequences of not gaining experience, but I figured I’d worry about that more when the time came.

Then I hit a boss battle, and Undertale seemed pretty insistent that talking wouldn’t work to get past that conflict; after trying a few times, I concluded that the game was telling the truth about that fight, so I switched over to fighting. Which wasn’t too bad; and, after that, I jumped from level 1 straight to level 5 (and from 20 to 36 hit points).


That was pretty clearly the end of a chapter of the game, and it seemed like a surprisingly large jump in my level, more than would be justified by a single fight. So at this point my mental model was that the game levels you up somewhat at the end of chapters even if you’re taking a pacifist route. Which makes sense as a design choice; hopefully the game won’t descend into the BioShock approach of making the two paths numerically identical, but being able to continue in the game as enemies get leveled up seems like a good idea.

My mental model was not, in fact, correct, however: there is actually a hidden pacifist approach to that boss. (Which annoys me a bit — I don’t really like that sort of active misdirection or needing to resort to GameFAQs — but only a bit.) But it took me quite a while to realize that I’d misunderstood what happened: my one instance of leveling up was after a special event, so I wasn’t sure what qualified as a chapter ending boss fight in my hypothetical model. Eventually, though, I’d made it to enough sorts of areas and fought enough unusually strong enemies that it was clear that I wasn’t going to be leveling up every so often, and a look at GameFAQs confirmed that I’d missed a pacifist approach to that boss.

Which made me a bit sad, if only because that one boss was a quite nice person, and probably the nicest person that you actually fight against, so I wish I hadn’t killed them! The thing is, though, having 36 hit points instead of 20 hit points was really useful: it let me make it past strings of normal enemies even as I approached the end of the game, and while, as the game went on, I did have to retry boss fights and use items during boss fights to boost my health, the boss fights remained doable.

So, ultimately, the game probably would have been much less pleasant if I had gone completely pacifist; and, if I hadn’t, my choice would have been to occasionally slaughter non-boss enemies even though that probably wasn’t necessary (which wouldn’t have felt right) or to fight one of the other bosses (which might still have been difficult, and which might have been less fun given that I actually enjoyed some of the boss fight interactions). Given all that, I probably accidentally mostly made the right choice (which, of course, the game strongly nudged me to do); I just wish that first boss hadn’t been such a nice person!


Ultimately, though, it turned out that 36 hit points wasn’t quite enough: I made it up to Asgore, one of the final bosses of the game, I was already on the edge at the end of my abilities, I was a little low on health items, and there wasn’t a shop nearby to allow me to easily replenish. I took a couple of swings at him, but I decided that I wouldn’t enjoy the combination of backtracking to get items and repeatedly fighting Asgore to improve my skills that would be required to make it past him. (In retrospect, I guess the third option would have been to level up by fighting some wandering monsters; for whatever reason, that didn’t even come to mind at the time.)

So I decided to give up then and stop playing. I almost always complete games that I start, so in general I would expect this to leave a bad taste in my mind. But, with Undertale, somehow that actually felt satisfying. Basically, I was playing as a kid going through a world of monsters: I was trying to talk my way out of trouble, and I’d made friends along the way, but ultimately, there were a lot of scary folks around. And also, the monsters had been treated quite badly by humans, and hence had reasons to want to attack those humans. So having the game come down to a final battle between a king who wants to do best for his people but who doesn’t feel good about killing a human kid and a human kid who’s gotten surprisingly far but is really over his head, and then having the king kill the kid, seemed to hit the right note of melancholic response to a situation that’s bad for everybody.


That’s how my playthrough of Undertale went, and it left me with a huge amount of respect for the game. Because the almost-universal default for RPGs (and, indeed, for a wide range of games) is for you to play as a mass murder; usually as a psychopath, occasionally as a person who expresses regret but then returns to slaughter. And I don’t feel at all comfortable with that these days.

One possibility for dealing with this is to allow a “good” route, and to provide mechanisms where the good route isn’t that hard, or, in the worst cases (BioShock again, though I wouldn’t call that game’s good route actually good in any serious moral sense), isn’t any harder at all than the bad route. But a choice like that is, from my point of view, a moral abdication: it’s letting you feel good about yourself without seriously grappling with the nature of the mass-murder route. And, of course, not letting you have that choice is a different sort of moral abdication: only allowing the possibility of choosing horrifically immoral behavior or not playing the game at all is not successfully confronting evil either.

Undertale, in contrast, confronts you with a choice but doesn’t sugar-coat it: you can be a decent person, and in fact there are a lot of other people in the game who are also decent people and who explicitly decide not to attack you even when told to do so. But, if you do that, the game is going to be quite a lot tougher; you can choose what sort of person you want to be, but that choice is going to have consequences, and you won’t like them.


I’m not going to present Undertale as the ultimate state of what I think moral choice in games should be like. For one thing, the RPG legacy of constant violent encounters is still there; for another thing, in the real world, decent behavior generally does actually have an impact, it’s just that the impact is unpredictable and plays out in a way that is often functionally quite different from the impact of bad behavior.

Though of course, in the real world as in Undertale, lots of people don’t benefit at all from good behavior in the sense that RPGs like to measure, with some clear external number (e.g. your bank account balance!) going up. So actually maybe Undertale is doing better than I gave it credit for in the previous paragraph: maybe it’s doing a good job of modeling the experience of structural oppression, where you’re constantly under attack, where you might be able to make it out of some individual instances of that attack with your soul intact by dancing around cleverly enough that you don’t get hurt too much, but where, ultimately, for almost everybody, the sum total of those assaults will wear you down too much for you to win. (Or at least Undertale’s pacifist route is behaving that way; the route embracing violence becomes much less realistic in that reading, because that route should also turn out difficult.)


Interesting game. (And charming, too, even though I didn’t talk about that much!) I’m very glad it’s out there, I’m very glad I played it, I hope it will spark thoughts in other game developers.

console upgrade cycles

December 29th, 2017

When listening to video game podcasts talking about the Xbox One X, they were generally disappointed because they didn’t see where the console fit in: there aren’t any games exclusive to the Xbox One X, and people who really want the best graphics will play games on PC, so who is going to buy it? And maybe they’re right from a business analysis point of view (but maybe not, I’m not at all convinced of the business acumen of the enthusiast press), but it felt weird to me to listen to that, because what those podcasters saw as problems felt to me like benefits.

Basically, what I want is to have a video game console family that takes its cue from the iPad / iPhone. So I don’t want big discontinuities: I want to be able to keep on playing my favorite games (or, for that matter, to be able to try older games for the first time) instead of, in the good case, having to rebuy games to play them and, in the bad case, not having access to certain games at all unless I plug in old hardware and hope that it still works. And I want to be able to do that without static hardware: if I’m replacing my hardware (which does wear out, after all!), I want to be able to get something with technology from the last year or two, instead of potentially being stuck with five-year-old chips. (Or, alternatively, having the choice to be able to buy two-year-old technology at a discount also sounds good; it’s not a choice that I personally normally make, but it makes sense in a lot of contexts.)

I could get that continuity with a PC, but that misses another aspect of the iPad / iPhone: I like being able to buy a game and have it just work, and to be confident that it’s not going to interfere with the running of the system as a whole or leave strange tendrils around if I uninsall it. And I also prefer the way consoles fit into my life: PCs feel much more isolating in that regard.


I’m not completely against discontinuities in console generations: if there’s a physical constraint that you want to change, then embracing that change is probably more likely to be successful than bolting it on. But there’s only one manufacturer who has the vision to (sometimes!) pull that off. So sure, Nintendo can keep on doing its own thing; but what I want out of Microsoft and Sony is incremental improvements based on PC hardware, based on the controller scheme that’s been working fine for the last two decades, and with an operating system that’s evolving to meet current needs for game distribution and online interaction. Or, in other words: more consoles like the Playstation Pro and the Xbox One S / X.

I am a little curious how this will play out from a business point of view, especially if Microsoft sticks with an incremental approach while Sony decides to jump to a Playstation 5 that doesn’t allow two-way compatibility across generations for most games. I’d like to hope that Microsoft will start quietly gaining ground from people who are buying a new console when a prior one breaks: they’ll spend more time having better options across multiple price/performance preferences.

Of course, what really matters to me is not having to rebuy a thousand bucks worth of Rocksmith songs and then having to level them up again…

nuclear war

December 28th, 2017

After Trump won the election, I was worried that the United States might actually slide into fascism. It’s a little over a year later, and, honestly, I’m still worried about that, but now I have a new worry: that we might have a nuclear war, with both North Korea and the United States launching successful attacks, probably with the United States firing first. And what especially worries me here is that I can’t actually see a good reason to think this won’t happen, and in fact it feels like it could happen literally any day now.

When worrying about fascism, I could at least be optimistic that political institutions would successfully stand against it, and that, as a last bulwark, protests could be successful. I’ve seen both bad and good signs in both of those potential protections against fascism, but at least there’s a case there.

With nuclear war, though: what exactly is the case that Trump won’t launch a missile against North Korea? That he’ll be horrified by the potential consequences? Trump has consistently demonstrated an almost total lack of that sort of reflection. That he’ll want to think hard about consequences of his decisions? Trump has shown a staggering lack of interest in getting real information, preferring instead to be led around by blowhards on TV. That he won’t respond emotionally to events? Over and over again, Trump has shown that he can be baited by a tweet. That his advisors will convince him otherwise? I’m sure that some of his advisors think that a nuclear first strike would be a horrific idea, but I’m equally sure that some of them think it’s a great idea. That somebody in the military chain of command would refuse the order? I see no reason to believe that. That the Constitution says that Congress has the power to declare war, not the President, so surely he wouldn’t be able to? Okay, I’m just including that one as a joke, and I can’t blame that one on Trump or the Republican party: we have decades of active bipartisan undermining of that portion of the Constitution.

There is, I guess, one potential source of optimism: Trump talks up stuff (whether past actions or future actions) an awful lot more than he actually does stuff. And, actually, now that I type that out, that really is a reason to be optimistic. Hmm…


If we launch a nuclear missile and North Korea fires back, what next? On a purely selfish level: I can imagine North Korea deciding that a West Coast US target would make sense, in which case San Francisco is a logical target. I assume I and my work are far enough away from San Francisco to survive that; I’ll just hope that North Korea doesn’t decide that Silicon Valley isn’t a better target. I don’t think there will be too many nuclear exchanges past that: I doubt North Korea has all that many bombs and missiles, so I guess it turns into a conventional war after that? (Which, sadly, is something we’ve seen a lot of.)

I was going to say that I’m also not worried about other countries joining in the exchange; and I do think that that’s true. (Though other countries might be targets of North Korean attacks: people in Seoul probably have the most cause for worry, and Japan as well. Or rather, the most cause for worry outside of North Korea itself.) But I don’t really see China being complacent about the United States taking military action on their border, against a state that China has supported in the past; I have no idea exactly how that would play out in practice, though.


Setting aside the affects on me locally, what about on the nation as a whole? Probably Washington D.C. is a more likely target; I don’t know if we’re sure that North Korean ICBMs can reach that far, but it sounds like they can? And if that happens, I have no idea what the consequences will be for our system of governance, but I can’t imagine they would be good. Even if North Korea chooses another target, though, we’ll have millions (hundreds of thousands if I’m optimistic) of deaths in the country; we’ve had a decade and a half of overblown reaction to a couple of thousand people dying in an attack, this would be much worse.

And, of course, we have a long-standing tradition of using wars and other forms of demonization as excuse to ram through favored political platforms, especially repressive ones; this would absolutely be no exception.

I’d like to think that Trump getting us into war would make more people realize that he was an awful choice as president; but I think more people would respond by saying that a war isn’t time to argue about the president. And I also think that the media and government would work together to downplay protests, even if they turn out to be massive.


That’s domestically; what about internationally? We’ve already lost a lot of international respect over the last year; but I’d have to think that a nuclear first strike would cause a lot more countries to treat us as actively dangerous. Which, I’m sure, third world countries already do: we’ve spent the past century and more going around the world overthrowing governments and installing puppet regimes. But if more powerful countries start seeing the US as a source of instability rather than stability, that feels like a state change? And if it leads to serious sanctions (which, I think, I would hope it would?), that would ratchet up the tension.


I dunno. I guess, typing this out, I’m actually a little more optimistic, because I really do believe that Trump is almost all bluster, that he himself doesn’t do much. Not that he didn’t, say, support the tax bill or trying to get rid of Obamacare, but Congress was driving both of those; he probably had more ties to the Muslim Ban, but even there, I suspect it was advisors driving matters in a way that they wouldn’t be able to drive a nuclear first strike so easily. So I guess this is the time to be grateful that we have a president who far prefers to be at his golf course instead of doing any actual work…

nier: automata

December 25th, 2017

There was a part a few hours into Nier: Automata where I got actively angry at the game. I was going through the desert, the robots were telling us not to hurt them, and 9S was telling me to ignore what the robots were saying: they were just repeating words without any meaning behind them.

To which my reaction was: seriously? I get it, we’re going to come to the realization that we’re actually the bad guys here, but can’t you be a little less heavy-handed about it? And maybe, you know, don’t force us to engage in mass murder while we’re coming to that realization?

Fortunately, the next two segments of the game helped me get past that point: the amusement park showed robots that weren’t attacking me, and while 9S was dubious about that, he was willing to accept that maybe slaughter wasn’t necessarily required. And after that comes Pascal’s village: robots who are explicitly and thoughtfully actively seeking out peace. So yeah, 9S’s behavior in the desert was heavy-handed, but the game at least made it past that fairly quickly.


It’s more subtle than that, though: the behavior of the robots in the amusement park and in the village are really quite different. The robots in the amusement park are peaceful (or, when not, are explicitly coded as broken by the other robots), but they’re peaceful in a creepy way. Most of the amusement park robots are repeating the same moves over and over, they’re acting, well, robotic. In contrast, the robots in Pascal’s village feel like people; or at least Pascal does, and the others at least feel like they’re moving in that direction.

So if the robots in the desert are as far away from the amusement park robots as the amusement park robots are from the village robots, then maybe 9S is actually correct: maybe they really are just saying words without having any idea what those words mean? Also, 9S has seen a lot of robots, so maybe he does have reason to believe that meaningless recitation is the norm for them.


Looping back one more time, though: in the real world, wars do indeed happen. And the people on the other side are people like you; people who believe that they’re fighting for the right thing, people who are scared. Or people who aren’t even fighting, but the war has come to where they live, and there they are.

Yet fight them our soldiers do. I can say that 9S’s behavior is ridiculous, but I live in a world (in a year!) where social structures (and not just in the military, not just in the police) evolve to emphasize seeing people not just as others, as different, but as Other.

So 9S’s behavior isn’t the unrealistic behavior: what’s unrealistic is me thinking that I can avoid that, that I can just choose not to take part. (That has certainly been a lesson of 2017: for example, how much the United States has been built on white supremacy, how present that legacy is, and how you can’t simply choose to step away from it even if you benefit from it and would rather not.)

Which, in turn, raises the question of how to model this concept, this mindset? Except that, of course, the answer is: video games model a mindset of othering, a mindset of the soldier following orders, all the time, they do so as their unquestioned default behavior. They tell you you’re the good guy, they tell you who the bad guys are, they present information to justifying that labeling while not actually letting you interrogate it, and then they put a weapon in your hand and tell you to use it. If you do, you’re a hero; if you don’t, you’re worse than not being a hero: you simply can’t play at all. So what made me uncomfortable about Nier wasn’t that it wasn’t letting me behave morally, it’s that it was making that question explicit.


I appreciate how Nier’s looping encourages me to rethink as well.


The robots show gradations of thought and consciousness, reflections of what it means to be human; the androids do as well. The extent to which they’re following orders versus responding to the details of the situation, for example: we see the Resistance members on Earth versus the soldiers based in the station, and even the various androids who have struck out on their own.

But also their range of responses: 2B in particular is not great at responding to social cues at the start of the game. Which, at first, I interpreted as her not having the same range of emotional responses (which then translates into a narrative of her learning to feel); but I’m now much less sure that that’s the best way to interpret her behavior. I think it’s a mixture of her being a soldier combined with her not reacting so fluently to certain interaction modes and cues.


You can imagine a lot of games made out of the aforementioned elements: in particular, you can imagine a game that would use this world to focus on different aspects of what it means to be human, and that either takes that in an anthropological direction or in an optimistic direction.

Nier didn’t make either of those choices, and I kind of wish it had. Instead, it gets more and more nihilistic as you progress from endings B through D; ending E holds out more hope, but ultimately it’s a very grim game. Admittedly, it’s been a very grim year, so maybe it’s the game that we deserve in 2017, but I don’t feel like the thoroughgoing bleakness of Nier’s outlook was teaching me much.

Maybe that’s me, though. There are, after all, bleak structural forces in our world: so perhaps Nier is illuminating them in a useful way? I could see that in the individual tragedies in the game, but I had a harder time seeing that in its structural bleakness.

best practices

December 14th, 2017

A quote from Anil Dash’s article about Fog Creek’s new project management tool, Manuscript:

Be opinionated: Manuscript has a strong point of view about what makes for good software, building in best practices like assigning each case to a single person. That means you spend less time in meetings where people are pointing fingers at each other.

Here is my opinion: if you want to talk about opinionated software (which, by the way, is a concept that I do agree with, even when I disagree with the specific opinions), then own it. Don’t start covering your ass in group legitimacy (before that sentence has even ended!) by saying that your opinions are actually a “best practice”.

Dash does, at least, try to explain why people would feel that individual assignment is a best practice, an opinion worth having. But geeze, that explanation: we’ll avoid finger pointing by making sure that each case has the name of the person you should point a finger at? How does that work exactly?

Don’t get me wrong: he’s got a coherent point of view. As far as I can tell, he believes in primarily optimizing for individual developer productivity. And yeah, if I preferred to work that way, I’d want to assign tasks to individuals, too. But say that, don’t talk in the abstract about best practices.


Though, looking at Manuscript’s feature list, I see no evidence at all that it’s actually opinionated software, so probably the “best practice” empty phrasing is closer to the truth. Take their section on Scrum:

Construct and plan iterations from existing cases. Create milestones for your backlog, sprints, doing and done — or any other structure your team or project needs. Customize each project’s structure to match your ideal Agile or Scrum workflow.

Followed soon by this wishy-washiness:

Estimates can be added to cases from the planning screen, and if you prefer story points, then you can switch between hours and story points too.

And there’s a Kanban section too. So: use Scrum, use Kanban, use your own homegrown process, use hour estimates, use point estimates. Do anything you want, we’ll be happy to support it! (At least as long as what you want to do doesn’t include pair programming, I guess.)

Dash’s article quotes a tweet with the following lament about Jira:

Why are you customizable to a fault, except in the ways I want you to be?

But isn’t that exactly what the above bits from the Manuscript feature list are promising as well?


Ah well; not a big surprise to be disappointed by marketing for enterprise task management software…

layton’s mystery journey and whackamon

December 11th, 2017

Layton’s Mystery Journey was actively disappointing. Following Layton’s daughter was a nice enough change of pace, I suppose, and the series is a good fit for the iPad; but the game didn’t have soul, and the puzzles weren’t enough for me.

For example, you start off by meeting a dog who tries to hire you to figure out why he can talk: but then another, more urgent case comes along, you do that instead, and the question of how the dog can speak never comes up again. And you have some boy who follows you around making puppy-dog eyes: I guess it’s still an improvement on the gender politics of the earlier games in the series, but only barely. The main other person whom you regularly interact with has their personality filled out in very broad, stereotypical strokes; all the other characters have one distinguishing feature and zero depth.

The puzzles are fine, but nothing at all new compared to other games in the series. The visual art isn’t awful, but it isn’t good: the dog always seems like he’s floating off the ground, and characters wave their arms or recoil in shock in ham-fisted ways. And breaking the game up into lots of different cases with only a vague hint at an overall story isn’t particularly effective plot-wise, and makes it harder for you to get to really like the city. So I think I’m done with the series unless something changes.


The other game I played recently on my iPad is WhackaMon. Which I started on the laptop, but it involves fast clicking on different areas of the screen, and doing that with touch is a lot easier. This game the only reason why I’ve ever logged into Facebook Messenger, and I’m certainly not going to continue to do so now that I’m done with the game, but if it’s the only way to play an Eyezmaze game, then I’ll put up with that.

Unfortunately, WhackaMon isn’t one of my favorite Eyezmaze games: too much clicking, not enough thinking, and not quite charming enough. Though there is some thinking involved in the clicking, and there is some charm in the standard Eyezmaze building up of a more and more settled area; it’s too bad that there’s not more thought involved in the building, though. And Facebook Messenger actively gets in your way: I accept some amount of being asked to spam your friends, but being asked to do so immediately after building a new structure is not only probably too often, it actively gets in the way of your enjoying that new structure.

Having said that, I’m glad I played it: I spent a pleasant enough three or so hours tapping on stuff and figuring out systems. And I’m certainly glad that Eyezmaze is continuing to make new games.

post-systematic flexibility

December 10th, 2017

David Chapman has, among other things, been writing about modes of approaching meaning, in a way that’s informed by Robert Kegan’s developmental psychology. He’s written a summary of this recently on one of his blogs, and he discusses it frequently on Meaningness (see e.g. this post and posts it links to), but I thought he had a particularly good discussion of it recently on the Imperfect Buddha podcast. (You can skip to about 22 minutes in if you want to skip over the discussion of the state of Buddhism in the west.)

He focuses on stages 3, 4, and 5 of Kegan’s model. Stage 3 is characterized by a focus on communal values, individual relationships, emotions, and experiences. Stage 4 is systematic: it accomodates complexity in a rigid way, by mapping it to a model. Stage 5 is meta-systematic: if you’re in stage 5, you’re skilled with dealing with interface between systems and reality, and can handle use that to handle vagueness while embracing precision and complexity.


I’m trying to come to grips with whether or not I think this is a helpful model. (And, if so, in what contexts it’s helpful, or how that help manifests.) For now, I’m having a hard time thinking about it in terms of an individual’s development as a whole, but it seems to me like a plausible match to how somebody thinks about specific aspects of their life?

For example, I’m a software developer who has spent some amount of time thinking about and experimenting with agile software development. So it feels to me like I can tell the difference between stage 3 and stage 4 use of agile: stage 3 agile is saying / believing that you do agile because that’s what cultural forces present as normal behavior, while if you’re asked what you do, you have some idea that agile = scrum and it means that you have standup meetings once a day, call each two weeks on the calendar a sprint, and store a backlog in Jira. (And a stage 3 agilist will do all of that while happily continuing to have separate requirements, design, implementation, test, and maintenance phases, and while constantly generating estimates and plans that are far more ambitious than what they actually get done in a sprint.)

Whereas a stage 4 practitioner will say that the phrase “we do agile” doesn’t make sense, because agile isn’t a methodology, it’s too vague for that. But they’ll have a precise idea of what it means to follow, say, Scrum or XP, and they’ll be skilled in following that precise model and helping teams follow that model.

Which, in that light, means that I’m probably not a fully stage 4 practitioner, because I’ve never been on a team that followed Scrum or XP as a whole, or that had a well-considered homegrown system that it actually stuck to. (Which doesn’t mean that I’m in stage 3, either, because I’m generally quite aware when teams aren’t following methodologies, either external ones or ones that they’ve written down for themselves.) But, if you go down from full methodologies to smaller practices, like test-driven development or refactoring, I can make a better case that I’m a pretty solid stage 4 practitioner.

And if we move outside of software development, I can tell a similar story: e.g. I’m quite sure that my Tai Chi teacher has an excellent systemic understanding of Tai Chi (and hence I also believe that it makes sense to talk about a systemic understanding of Tai Chi), I’m equally sure that I don’t, but I also feel like I’m learning relatively concrete facts and improving in ways that I can point to? So I’m consciously trying to start the journey towards a stage 4 understanding of Tai Chi, I just haven’t gotten very far.


Stage 5 is more of a mystery to me. One of the points of stage 5 is that systems are only models, and hence are always flawed. But the issue there is that there are multiple ways that you can get to a rejection of systems: you can take a stage 3 approach of not really thinking about them seriously; you can take a nihilistic approach (Chapman calls this “stage 4.5” and is pretty worried about it) of correctly understanding that systems are always imperfect models and using that as a reason to reject them; or you can take a stage 5 approach of appreciating the nuances of the boundaries between systems and reality. Which should mean that you can use the power of systems in contexts where they apply well, you can avoid them in contexts where they don’t apply well (or, potentially, switching to a different system that applies better there), and you can tell when you’re near the boundary, using the system to inform your actions but not to rule them, and potentially using your observations to update the system as well.

At least I think that’s what stage 5 means: but it also feels to me like my understanding of all this stuff is probably basically at a stage 3 level? Chapman sounds sensible when he talks about this, it feels to me like he’s getting significant value out of it and believes that it’s tied pretty well to other forms of thought that he finds valuable, but I can’t say that I’ve seriously tried to put the framework to use. So, ultimately, I’m mostly just parroting / cargo culting what he says, which (I think) is stage 3 behavior?


One feeling that I’ve had over the last few years: more and more, when making programming decisions (broad design decisions, narrow decisions about what to type now, decisions about how to segment my work while trying to go from my current state towards a desired future state), my mind is starting to associate weight to those decisions. And here, by “weight”, I mean that my mind literally associates certain decisions with something that feels heavier or more solid, whereas other decisions feel like more of a haze. Hmm, I guess weight alone isn’t actually all that’s going on in my internal perceptual apparatus: e.g. there are some that feel like pebbles, solid and reliable but also like small steps, some that feel like mist, where I don’t perceive any weight but I also don’t understand what’s going on, and some that feel like they’re crumbly terrain, actively and concretely dangerous to proceed along. So maybe it’s more of a combination of weight and texture?

If I wanted to try to tie that into this Chapman / Keegan model, maybe that’s saying something about the boundary between stage 4 and 5? The areas where I have these feelings are situations where I don’t just know how to follow a given system, I have a pretty good idea of what the specific consequences are of doing so or not doing so (or doing so in different ways). So that means that I’m getting a better appreciation of reality pushing back (the “interface between systems and reality” that I mentioned above): when a certain question is answered well within a given system, when I’m pretty sure a given system is accurately warning about something, when I’m on the edge of a system, when I’m pretty sure I should work within a different system, and when I just don’t know?


Hard for me to say: like I said, I don’t understand the theory very well. And, for all I know, I’d get as much from linking my understanding to any other random list, e.g. The Five Levels of Taijiquan. (Different numbers in that book’s levels, though!) And, don’t get me wrong, there are certainly areas where I’m firmly in stage 3: e.g. when reading Twitter I’m just as likely to react to events in a way that ultimately comes down to group membership as anybody else is. But it is nice to start to have a deeper sense of what substantial expertise might feel like…


December 3rd, 2017

I guess I played Paperclips enough that I should write about it here? Or rather I spent enough time watching it in my browser, or I spent enough time being distracted by it, or something.

Paperclips isn’t the first cookie clicker I’ve played, but it’s the one I’ve played the most; I think it’s the only one I’ve made it to the end state of, and certainly the only one I’ve replayed. And the narrative, as slight as it was, was actually a rather good fit to the mechanics.

Mechanics-wise: it’s all about bare numbers, and the game helps you think about them by exposing derived information (rates, in particular). And there’s enough complexity that it’s not obvious what the optimal strategy is at any given point: you basically know what to do, but you have a couple of directions you can go when optimizing, and also you don’t know when the next deflation event (or, more rarely, cataclysmic change, e.g. a new currency introduction) is that will invalidate all of your current calculations. And, if I’d want to think about it more, there would have been more that I could have dug into: e.g. part way through the game you start picking a competitor in a robot prisoner’s dilemma tournament, and I haven’t figured out (either theoretically or empirically) which strategy is the best.


That’s the game play; but, ultimately, much of the time you’re just sitting and waiting for stuff to happen. (Maybe buying more production capacity every once in a while, but not in a way that makes a real difference.) And, most of the time, the game is even happy to play itself on autopilot: continuing to make more of the relevant currencies without needing explicit action.

So you could imagine having it run in a background browser window while you, say, write a blog post or something, checking in once every half an hour. I found that very difficult to do, however: there’s always something just around the corner, some slight reward for spending three minutes watching numbers go up and then clicking as soon as possible.


There is, fortunately, an end state to the game. Which gives you two options; one is to start over, with slight tweaks to the numbers; the other is to contemplate the void. I picked the first option the first three times I finished the game, and I have no complaints about having done so; after that, though, I picked the other option, and I’m glad it existed.

genre insecurities

November 28th, 2017

If you were to ask me for, say, a list of my top five favorite movies, I don’t know exactly what the full list would look like, but most of the time both Spirited Away and Pom Poko would be on there. Which, it turns out, I have somewhat mixed feelings about: even admitting that I don’t have a particularly thorough movie background, is a pair of fantasy anime movies that could reasonably also be labeled as children’s movies a place where I (a 46-year-old man) want to put my stake in the ground? Shouldn’t I prefer movies that are more thoroughly grounded in a range of life experiences?

The above, of course, isn’t any sort of case against holding those movies in very high esteem: as phrased there, it’s completely unsupported genre snobbishness. And I wouldn’t put up with that sort of snobbishness in any other art form: I grew up in a context that, say, valued literary fiction over science fiction or romance, that valued classical music over pop music, that valued a whole load of things over video games (to the extent that video games even existed while I was growing up), and I’m pretty confident in saying that those blanket valuations are ridiculous, that literary fiction and classical music are just different genres. I can still see the effects of that context in my psyche, but I can also consciously set it aside. (And, don’t get me wrong, it’s not like anybody told me not to read science fiction while I was growing up or to not listen to pop music when I went through that phase in high school. And, also, don’t get me wrong: if I were to make a similar list for music, classical music probably would be extremely well represented.)


Setting both anti- and pro-genre snobbishness aside, though: you can learn from any genre, so I’m sure I’ve got gaps in my taste that arise from my genre choices: I did actually read a fair amount of literary fiction in grad school, and it was productively different from what I’d been in the habit of reading. And there are also stereotypes that I see in some of my habitual genres that I’m actively unimpressed with: e.g. the “anointed savior of the world” trope I see in so many games and also in comics (both American and Japanese, both in print and animated forms).

Worry about that latter stereotype is probably what’s really going on in my psyche here: I do enjoy wish fulfillment, but I think it’s healthier for me personally if I don’t spend too much time diving into it. Instead, I’d prefer to have a healthy balance of art that focuses on the small scale, on the details of what exists, and on actual people.

Having said that, too much of a focus on small scale personal concerns can be associated with its own negative stereotypes that I’m equally dubious of: e.g. literary fiction about middle-aged men unhappy with their marriages and instead finding a match with women in their twenties. I don’t have any more respect for that sort of wish fulfillment than I do for RPG “savior of the world” wish fulfillment; but if we can step away from that to something that feels more like real interactions between real people (and, yes, with real problems), then that’s important.

But at any rate you can of course focus on details and on people in any genre. Returning to science fiction, Trouble on Triton puts you in the head of somebody so you can how he interacts with other people, what he wants from those interactions, the pain that he gets from that, the pain that others get from that, and the self- and outwardly-inflicted nature of the problems surrounding him; and the novel’s nature as science fiction lets it generalize those experiences in a way that clarifies by the distance of the setting.


I said above that I’d prefer to have a healthy balance of art that focuses on the small scale, on details, and on actual people; that’s true, but only half true. My relationships with my wife and daughter are both extremely important to me; and if art can shed light on that, that’s great. And work involves people too, of course; and I do care about my friends.

But, granting all of that: I’m not a people person. Also, a lot of the classic literary themes actually aren’t particularly reflective of my life: happy, stable marriages and careers aren’t in general the subject matter of great novels. (Not that our family doesn’t go through rough patches – this last year in particular has been quite a bit rougher for us than I’d like – but still.)

Instead, a lot of what interests me is trying to figure out systems: figuring out what code and computers are telling me, solving puzzles of one form or another in my spare time. Which doesn’t mean that I don’t like small scales and details, because as I get older I find more and more that listening to details is an excellent path into broader concepts. But still: figuring stuff out gets me going, and that’s going to inform my artistic choices. Not necessarily in a direct way, I don’t particularly want to read books featuring programmers, but in a metaphorical way, I want to read books where reading them feels like uncovering and making sense of a conceptual space that’s new to me.


I led off by bringing up Spirited Away and Pom Poko; this focus on systems and details is easier to see in Pom Poko, because it’s a message movie, in multiple ways. It’s about growth and the negative affects growth has on the environment, on animal life in the environment in particular. It’s about the process of change, focusing more on the loss that change entails but still allowing you to see the benefits. So there are conceptual spaces to explore here, and to test your understanding of via exploration of tradeoffs.

And Pom Poko certainly focuses on the details, and on people. (I mean, mostly on tanukis, but still.) How individuals react to change in different ways; how life continues in its pattens despite change. It does this without grandiosity and without catastrophizing at a broad level: ultimately, the tanukis lose their battle, but most of them survive and adapt nonetheless. Though many of them don’t survive: the movie doesn’t catastrophize, but it doesn’t pull its punches.

Spirited Away isn’t the same sort of message movie: it’s about a very capable girl who turns out to be friends with a river god. So, to some extent, it’s a bit by the numbers; but I do appreciate how its plot asks fundamental questions about what the concept of family means. Family as people you’re related to by birth, but also family as people who choose to care about each other.


Looking at the two together, though, clearly movies that draw on Japanese mythology press my buttons, at least if they do so with a focus on sprits and nature. Which I think is another example of what I was talking about above: enjoying the process of exploring a conceptual space that’s relatively new to me, just in a less abstract way than the intellectual themes I talked about earlier.

Of course, movies aren’t just vehicles for plot and themes: they’re something you see and hear. And both of these movies have bits that are visual spectacles: the entire bath house in Spirited Away has, as its job, to put on a show, and the parade in Pom Poko is really something. And, aurally: Joe Hisaishi is one of my favorite film composers, and Itsumo Nando Demo from Spirited Away is one of my favorite pieces of his.


So yeah, they’re good movies. I probably should branch out more (though, don’t get me wrong, I don’t spend anything like a majority of my movie time watching anime), but there’s something there. And there’s certainly nothing wrong with enjoying exploring lovingly crafted spaces…

her story

November 6th, 2017

(Spoilers for Her Story follow; if for some reason you just want to know my opinion and are thinking of buying it, I’m very glad I played it, so if you’re on the fence, give it a try.)

I am very glad to have played Her Story shortly after playing Tacoma: both games tell stories that feel a lot more familiar outside of games than inside of games, both use interactive techniques to good effect when telling their respective stories, but the interactive techniques and the subsequent effect on how I experience the stories are significantly different.

Tacoma feels like a copiously annotated story. That story unfolds over the course of three days, which you learn about by seeing six key points during those three days; and, during those three points, you can look at the story from a few different perspectives, and are presented with some specific pieces of textual information informing each of those points and perspectives. And there’s subsidiary back story available: extra scenes you can watch about each character, and physical spaces for the ship and the characters that you can inspect, some with further textual information.

Her Story also makes it clear that there’s a linear story going on, but instead of progressing through that story linearly, the game almost immediately allows you to navigate on your own. I’m not even sure what a good metaphor is for the experience: a crystal, with views from different facets? A palimpset, reconstructing a text? Or maybe the best metaphor isn’t actually a metaphor at all, just a description of what’s going on: you’re conducting a murder mystery, trying to piece together what happened from the clues that you come across (that you notice!) and from the unreliable subjects you’re interviewing.


In Tacoma, you could say that the game mechanics focused on perspective, reifying that concept in a changeable viewpoint on a three-dimensional (or, really, four-dimensional) space. In Her Story, in contrast, the travel occurs along a one-dimensional space; and that, in turn, means that the navigation alone is less interesting from a game point of view. So the game has you navigate via conceptual controls instead of thumbsticks, reifying those concepts in the form of search terms that allow you to dip into portions of the timeline in an unpredictable fashion.

Or at least it seems unpredictable from the outside; one of Her Story’s most impressive accomplishments is how it uses what seems like an unpredictable method for controlling how you navigate the timeline and nonetheless ends up with a story development that’s satisfying in a surprisingly traditional way. Because, when reading a novel (a mystery novel, perhaps), I start out getting a picture for the basics of the setting and the problem that it’s presenting; then I start understanding the possible solution space, and thinking about how it might unfold, and what surprises might be in store; then I come across some twists that lead to new levels of depth and predictions; and eventually it all comes together. And, somehow, I went through that same experience while playing Her Story, despite the player’s behavior being aleatory from the designer’s point of view.


(Here’s where the spoilers begin in earnest, for people who want to stay away.)


Concretely: I started out just trying to get a feel for the situation, assuming that I was trying to piece together the events that led to the murder. I searched words that seemed important in the initial interview segments, leaning a bit towards proper nouns.

I’m not sure exactly when I realized that there were two different women appearing in the interviews: I must have heard Eve speak a few times before I realized that she existed. I think it might have been when I heard the name of the midwife, searched on that name, and then heard the whole story about their birth? But at any rate I transitioned quite gracefully into a second act of the game, which mostly centered around learning how the two sisters grew up, but also (from a gameplay point of view) had me asking questions like which sister was speaking during which days.

At some point I happened across a clip where there was a guitar sitting on the table, with no explanation whatsoever. So then I had to search for the term “guitar”, which led me to the first part of the song, and then I quickly found the second part of the song. If I’m remembering correctly, this was the transition into the third act of the game for me, trying to understand the sisters’ points of tension with each other better, and also trying to figure out what happened with Hannah’s parents.

And then I learned about what had happened between Eve and Simon; and eventually about Simon’s death; by then I’d seen the vast majority of the clips, so after a bit more searching of random words I’d jotted down, I declared victory.


In other words: I experienced a very satisfying unfolding of the story, broken down into four coherent acts, with significant parts of the story remaining hidden for quite some time, only appearing once I had the context to appreciate them. And yet all of this came out of a game with a random access interface, driven by search terms!

I still don’t know how the game did that, and how much I got lucky. I imagine quite a lot of it isn’t luck: presumably there are key words that don’t occur in the initial clips? I’d certainly be interested in seeing a graph whose vertices are the clips and whose edges are words shared between clips; does that turn up clusters that are dramatically meaningful?

But of course it’s not just a graph theory puzzle, for a few reasons. If you search a popular term, you don’t see all the clips; so we’d have to reflect that in the graph. (And of course restricting the clips you see in that situation by time order means that, all things being equal, you get more Hannah and less Eve.) And people don’t search words at random: I’m sure I’m not the only person who gravitated towards names and other proper nouns at the start, and in general people are going to search for words that seem meaningful. Finally, people aren’t restricted to searching for terms they’d heard: e.g. I searched for “guitar” not because I’d heard the word spoken but because I saw one.

So, somehow the game manages to balance all those considerations and still help the plot unfold. And I think it does that without cheating; it does say something about one volume being corrupted, but it said that at the start of the game and still says that at the end of the game, so I don’t think the game has been been hiding anything from me, or at any rate that it hid anything that hasn’t remained hidden?


I say “I” above when talking about my experience with the game, but I wasn’t playing it alone: I was at the keyboard but I was displaying it on the TV. Liesl watched a fair amount, and Miranda seemed basically just as involved as I was: at a lot of key moments I was following Miranda’s suggestions for what to type.

The game worked very well in that mode: we could talk about what we thought was going on, Liesl and Miranda both noticed things that I didn’t (e.g. I think Liesl was the first person to notice the tattoo), and the words I searched were mostly words that had been spoken whereas the words Miranda suggested were mostly thematically appropriate ones that may or may not have been spoken recently. So, between the three of us, we jumped around more and saw more stuff; yay for games that support that sort of shared experience.


Her Story of the most interesting games I’ve played this year. I won’t say that I want to play a whole bunch of games using this mechanic, but maybe actually I do? Certainly it’s a reminder to not stay stuck in a rut; and it feels like there’s some sort of deep lesson in the game about how to guide players’ experiences without prescribing.