We stop watching an image intensely as soon as we have understood what it represents. The moment an image registers as a concept, its existence as an image is over.
—Volker Siebel, Painting Pavements
Proving Grounds of the Mad Overlord
Over the last couple weeks, I've been slowly chipping away at Wizardry: Proving Grounds of the Mad Overlord, specifically the PSX Llylgamyn Saga remake by Soliton Soft (1998), which I've been told is the easiest way to experience the original trilogy with party transfers, and the Digital Eclipse remake of the first game (2023). I'm not very far at all in either version, only finally breaking into floor 2 on the PSX version last week.
Considering the fact that you kind of need two parties in Wizardry 1 (a ‘main one’ and a ‘rescue one’ for when you invariably have to retrieve the bodies of the main one), this means I've been concurrently raising ~four crews of six adventurers[1] all named A[2]. The power difference between a level 1 and level 3 character in this game is gigantic & required to even think about leaving the foyer of the first floor of the dungeon, and my patience for grinding has worn really thin, so it's been a lot of really short bursts of gaming where I get to contemplate in detail the fundamental appeal of fragging orcs and kobolds over and over.
Maybe I'm being reductive here, but the strategic depth of early Wizardry 1 is really quite narrow—nobody has any interesting abilities yet, so the only real decision for the first couple hours is whether to keep grinding or go back home and heal. That Wizardry became such an enduring series leads me to believe at some point it was enough to just have a computer running something kinda AD&D-ish for you. The promise of tabletop RPGs always was that they would induce some kind of Lord of the Rings stuff on the table entirely through rolling dice and comparing numbers on spreadsheets[3], so why not give the boring math part to a computer and let the player focus on feeling like Aragorn?
Everyone's read Niinou Kazuya's “How to use your imagination” by now. I've probably linked it like three times on this site already. What I'm finding that the PSX version of Wizardry 1 is the one where I'm having more of those flashes of narrative, whereas the PC version is the one I'm playing as an abstracted systems management game Mechanically, both the Wizardry 1s I'm playing are pretty much identical[4], so it's mainly a difference in aesthetic wrapper[5]. The thought I keep coming back to is that it's exactly all the lush and well-crafted visual art in the PC remake that is getting in the way of those moments of auto-fiction. This is probably the best-case scenario to dig into this kind of phenomenon since I like how the new and shiny version looks, unlike nearly every big HD remake made these days[6]. The town environment in particular is really nice and atmospheric, the character portraits are great too. And yet! It feels more like I'm touching the mechanics directly than the PSX version does, in a way that's not completely attributable to “quality of life” things like the menus being faster.
Here's a specific instance of this: when I'm making a character in PSX Wizardry 1, I have no idea what they look like. As I move through the assembly line, for example, DWARF-GOOD-(like 15ish VIT and 11 PIE)-PRIEST, I only get something resembling a clay block-in in the shape of whatever the statline suggests. Equipment shopping clarifies this block-in a little as I imagine them trying on different armour pieces, and by the time my new character gets to ~level 3 I'll probably have spent enough time zoning out during level grinding that I've materialized some kind of specific dwarf in my head. This might happen faster if I'm feeling particularly indulgent and have given the G-PRI DWARF a proper name at character creation. In the PC remake, you pick a character portrait at the time of creation. Here's what your character's face looks like. Again, they're good portraits, it's just that I don't have to think about my character's appearance anymore, it's done for me. The beautifully moody, perma-sunsetting eurofantasy castle town is the only resolution of the idea of “The Town” I'm allowed. The model of the tavern is lovely, but it does not look like the “Gilgamesh's Tavern” that shows up in my mind when playing the PSX version. To be entirely fair to the art team that worked on this remake, I don't think any actually physically resolved image could look like the one I see in my mind, exactly because a mental image does not have to be resolved in any way. Of course, PSX Wizardry 1 has illustrations as well, notably by Suemi Jun, but the in-game ones are low resolution and the high resolution ones exist in “outside” material like box art and manuals, so there's always plenty of room for the mind to roam and invent.
having to think about it
MechWarrior 2 has been on my mind again. Partially because I played it for the first time around this time last year, partially because I've been getting into tabletop Battletech, and partially because we just passed the 30th anniversary of MechWarrior 2: Mercenaries. The game came out during a real wild west era for 3D graphics, and as such, there's something like 36 different versions of it out there for various configurations of graphics card and operating system.
The 1998 Titanium Trilogy rerelease is an interesting example of the state of graphical remasters in the 90s, straining against development issues, the limits of the game engine and 3D APIs of the time to present something that is definitively more detailed but way uglier than the original DOS release. In fact, I don't think any of the ~35 additional releases of MechWarrior 2 looked better than the original DOS version, but that really did not stop Activision from trying as long as it made some amount of business sense. In the years since then, graphical APIs got more consolidated & videogame companies adjusted their posturing several times over, so “beloved older game rerelased on latest hardware” no longer feels as scattershot as it did with MechWarrior 2.
Organizing around graphical fidelity as the main metric for visual quality has led videogames to a strange place. When you make it a priority to increase detail across the entire image plane, you're severely limiting the range of one of the fundamental tools for composition. The textured, detailed mechs in MechWarrior 2 stand out beautifully over abstracted colour field environments. Who knows what the art team was thinking about when making that game, but technical limitations forced them to make a choice, one that was in some ways rooted in the kinds of choices painters had been making for centuries. Suddenly being able to put textures on everything is where the trouble starts. Maybe you can pick what gets more area on the texture atlas, but the expectation is set that every surface is going to have something going on. The visual becomes loosened from its (probably unintentional) heritage in visual art. In a couple years' time, someone figures out how to mod in 4K textures and a bunch of people cheer because now the game looks better because more pixels on their screen are different colours at the same time. Bloom comes along as a technically advanced (therefore justifiable by the logic of “pushing the medium forward”) method to bring softness back into the videogame image and gets turned into a joke immediately. I didn't buy a HDTV to not see every piece of gravel on the floor!
Intricate, fine detail has a long history as the go-to aesthetic choice for power. A crystallization of kind of money you need to sustain a craftsperson carving out red lacquerware leaves for months, of the kind of money you need to get the materials for red lacquer in the first place. Videogames are no different. The entire medium, born and matured under capitalism, is pulled in the direction of money, and the money is always going towards more power, more detail, more spectacle. The mass audience for this medium seems enthusiastic about this arrangement, eagerly following the games that promise the most detailed gravel on the market.
Yellow paint, bird shit on climbable ledges, Wolverine smell trails, NPCs telling you to go to Death Mountain three times in a row, glowing markers indicating places you can jump on—these are all desperate measures of a visual design language completely backed into a corner. Every time a released game looks slightly less sharp than the teaser trailer is an unforgivable betrayal to gamers worldwide, so there's no backing down. The problem of too much visual information is solved by adding another layer of visual information. Here's the videogame world, fully rendered, clarified and detailed, every aspect of the image articulated, answered for and equally unimportant, since all information sits on top of it, not within it.
I'm not trying to argue for exclusively minimalist aesthetics. I love a big virtual forest with a bunch of billboard flowers[7]. I love zooming in on my unit of yari ashigaru and seeing that all the guys have slightly different outfits and are stabbing in slightly unique ways. I love an ornamental UI. I just want this ornamentation to come from artistic decision-making. Today's tools make it way too easy to create detail on the cheap. I don't think things like SpeedTree are fundamentally evil or that today's game artists are somehow worse than they used to be, but I do think that a tool that automatically generates a forest makes it easy to not consider whether the forest was necessary in the first place.
the great greebling
Writing and reading are complementary aspects of the same process; much of what appears to be the work of writing is in fact done by the reader in the act of reading. While the writer takes advantage of this, making implications & inviting the reader to do the rest, the worldbuilder–lonely & godlike & in control of (or attempting to be in control of) every piece of footage retrieved from her obsessive creation–induces dependency in the audience, then discovers in the subsequent delirious spiral of self-fulfilling prophecy an excuse to take even more responsibility out of their hands. God’s in her Heaven & all’s right with the “world”.
—M. John Harrison, very afraid
The optimizing nature of capitalism has resulted in at least two solid decades of inventing new types of SpeedTree to make more and more aspects of game development more efficient. Tim Sweeney possibly anticipates Unreal Engine 6 to fulfill the theoretical concept of SpeedGame, a tool allowing users to generate over-the-shoulder third person action games at a rate 7th-gen console developers could only dream of. In 2018, Square Enix's garbage interpolated upscale PC rerelease of Final Fantasy IX received some fan love in the form of a garbage AI upscale mod using ESRGAN to redo all the game's backgrounds “while respecting the original art”. To me this marked an important turn in the war against seeing: small groups of people could now pump out dogwater faster and more efficiently than giant international media companies, using new tools to invent detail where there was none before, in a process that made it easier than ever to pursue a project for which you would usually have neither the skills to start nor the taste to stop.
The promise for the kind of person that enjoys AI upscale mods is that there is no longer any need to try and parse an image you're looking at. It's already fully resolved for you, no thinking required. Is this not the promise of most videogame remasters today? Or of HD graphics in general? Since I'm already being acerbic here, is this not what worldbuilding is all about? The need for everything to make sense, and for that sense to already be made for me.
Looking at recreations of mainframe videogames from the 1970s I can be sympathetic to the desire for higher-definition, more-representational graphics. Unfortunately, swept up by the economic system they found themselves in, gamers never got the chance to say “okay, that's enough” until it was too late. Now everything is covered in greebles, pouches, picatinny rails, molle webbing, celtic knots and whatever the hell is going on with Link's tunic in the Ocarina of Time remake.
On the internet, I'm being shown new images at a density and rate that far surpasses anything I'm comfortable with. I'm offered a solution: algorithmic curation that shows me the stuff I'm most likely to want to see. I hate every part of this! I want to use as much of my brain as possible for as long as possible! It is unfortunate that my biggest lifetime hobby, the computer, keeps getting moved further away from that. I'm increasingly feeling crowded out by videogames. Online storefronts are completely fucked. There's too much to look at, and even my looking now has to be doubled. 6 years ago, skimming past a highly-ornamented video game screenshot was enough for me to want to double back and at least admire the craft, but nowadays this instinct is completely exploited by how trivial it is to AI-generate skim-level ornamentation. It's exhausting! All of this mental work before I even actually sit down and play something.
The artistic frontier for the entire medium is now defined by hyper-rich guys who are proud of being unable to remember their own home address. To the kind of guy that gets to decide whether 500+ people have a job tomorrow or not, the only thing that matters is that his business is getting more screentime than the other guy's business.
Dull, thoroughly resolved and overexplained images are the art direction for any current videogame remake, and they are also the visual language of attention capture. Having to look closely & interpret what you're seeing is an energy-intensive act. I can't really binge poetry books or visits to art galleries. It is much, much easier to just mindlessly flip through vaguely stimulating imagery that doesn't require a lot of interpretative work. It's one thing in other media, but the videogame industry seems to actively resent the fact that there are games that leave room for the player's imagination. But what's the point of filling my brain up with even-intensity sludge? Why should I explore a game world if I'm only ever going to see things I don't have to think about? Isn't it fun to use my brain[8]? Why would you remake Shadow of the Colossus in HD?
not letting the computer do everything
Firstly, all these machines are too fast for me, even VHS. Working with AVID is horrible. I say to my editor: ‘what if we tried...’ and before I have finished my sentence it is already done <...> Take, for example, a movie like Natural Born Killers (Oliver Stone, 1994). It's hard to see how the first thirty minutes have to much do with Oliver Stone at all; they probably only exist because it is so easy to edit with an AVID. There is hardly any material resistance against the ideas.
—Harun Farocki, Nine Minutes in the Yard: A conversation with Harun Farocki[9]
Earlier last month I finally got to play my first in-person game of Battletech against a human opponent, previously having only done a couple solo games to learn the rules & a couple MegaMek games against my friend Dima. MegaMek is a really convenient way to play Battletech on the computer, since it does all the calculation (mostly important for gauging shot accuracy) and dice-rolling for you. The trade-off for faster turns is that it's easier to forget how certain game maths & rules work, which to me can result in a really “muddy”, unsatisfying feeling during gameplay. I was really worried that, having gotten used to automation, I'd be a bumbling fool IRL. Fortunately, there were plenty of helpful people around, and the game went by relatively smoothly.
That first game (and my second a couple weeks later) was the most fun I've ever had playing anything Battletech-related aside from maybe some of the more evil missions in MW3: Pirate's Moon. It turns out, at least to me, resolving dice rolls manually is the thing that evokes the fantasy instead of getting in the way of it. Having to actually look and read the map to figure out line of sight & range modifiers instantly changes my relation to the battle. Maybe I'm really lazy. Nothing's stopping me from sitting there and taking in the terrain in MegaMek, but I find it really hard to when I am offered the comfort of just clicking a button and boiling a particular line down to a +4 instantly. Similarily, having to manually resolve dice rolls into hit locations for each of the Firemoth P's 12 MPLs (really fucking scary thing btw) makes me feel my mech in a way the (admittedly really good) cockpit views of MechWarrior never fully manage.
Mapping is the aspect where my two Wizardry 1s mechanically diverge the most. Automapping is probably one of the earliest and most universal “quality of life” developments for DRPGs, and as such, PSX-Wiz1 comes with it enabled by default. I decided to keep it that way. My mostly “original style” playthrough of PC-Wiz1, by contrast, relies primarily on hand-drawn mapping cross-referenced with the occasional DUMAPIC spell. These are obviously two very different ways to interact with & perceive the dungeon map. Mapping by hand involves a more varied skillset and is a lot more pleasurable, even if it introduces risks that aren't there with automapping. Having a little extra physical thing on the side to interact with while I play feels great.
Is it possible to take the dice roll back from the computer? At the games club, I'm incentivised to count things out & report results properly because I would like to continue playing games there and not get the reputation of being an asshole, but does that work for a singleplayer videogame? Most of the games on the Magnavox Odyssey rely on the player to keep the score, but it seems clear why videogames have largely moved past this level of trust. Even as someone very interested in a hypothetical videogame where I roll physical dice and tell the computer the results, of course I'd end up saying my priest hit max heal on a pivotal DIOS cast, or that the treasure chest had 1% items in it[10] every time, even if that ended up completely flattening the tension. Is the social responsibility towards another human the only thing keeping me away from optimizing the fun of manual roll resolution?
This kind of ended up being a two-bodied entry, but in the end, I think the problems of excessive visual fidelity and smoothed gameplay are just facets of the same thing—videogames that don't trust me to do anything myself. The “artistic medium” and “competitor to sleep” factions of videogame creation have grown to be extremely at odds with one another. Obviously, this is a wider problem of the funny little historical moment we find ourselves in, and not just of one particular medium. Attention spans have been battlefields for a while, and now other basic cognitive processes are too. On top of that, the rules of engagement are completely compromised. A different potential for the videogame is not going to happen from within existing market structures, and it can't be imagined by minds captured by those structures. To quote Liz Ryerson, “the art form of videogames needs to be decoupled from the market in a way that it has never truly been before.”. It almost certainly can't happen by focusing on videogames alone. Good luck to all of us trying to get there.
Though my gut, trained on later games (Wizardry: Labyrinth of Lost Souls was the start of my trek back into DRPGs after kind of getting walled by Etrian Odyssey 2 as a teen), desperately wants a more varied party, I'm going with FIG/FIG/PRI/THI/MAG/MAG in both games. The problem with my favourites SAM and BIS is that they require a lot of bonus stats to roll, and on top of that, BIS is apparently so slow at learning spells that it's always a suboptimal choice for spellcaster in Wizardy 1. ↩︎
You can always go back and change your name, but you can't change your bonus stats and level-ups. You have to go through with 90% of making a character to see their bonuses in Llylgamyn Saga & name input takes the most button presses by far, and on top of that a high-bonus character isn't even guaranteed to survive first contact with 4 kobolds (and reviving a level 1 character is pretty much always a waste of resources), so the rule of my guild is “you only get a real name if you have >10 bonus stat points and you make it to level 3”. ↩︎
People like to point out D&D's origins from tabletop wargaming as an argument that early D&D was somehow not a narrative game (contrasting it against expectations new players have in a post-actual play world). The thing is, narrative has always been a part of wargames! You're trying to tell me people were writing out spreadsheets for “magic users” and “elves” in the 80s because they wanted to fulfill a desire for pure strategic thought? ↩︎
Aside from streamlining character creation in a way that makes it really easy to get 20+ bonus point monsters in the PC remake, I'd say that version ends up skewing closer to the original game than the PSX remake does, but only slightly. Digital Eclipse's Wizardry 1 is really well made in general, but in particular I love how granular all the “old-school” options are. It reminds me of all the additional work M2 do on their Sega Ages releases on the Switch, letting you pick your own position on the archival←→modernization alignment chart. ↩︎
Speaking of aesthetic wrappers, the tiny Apple II version of Wizardry 1 running in the corner of the PC remake adds a really interesting wrinkle to my experience. Probably intended as a kind of proof of authenticity, it ends up making me feel like all the lush new 3D environments “aren't real”, in a sense. ↩︎
Did you see what they did to Heroes of Might & Magic III? I genuinely got so upset over it even though this is just what “the industry” is going to be for the rest of its lifespan & it's not like anyone can take the original away from me or anything. It just feels insulting. ↩︎
Big Walk[11] and nu Marathon are really exciting to me visually because they are very confidently directed subtractive approaches to current graphical standards. The huge, bright-coloured geometric structures of both games contrast so well (and communicate so clearly!) against the really lush natural environments. ↩︎
Further reading: What's The Point If We Can't Have Fun? by David Graeber. ↩︎
The english version of which is available in the book Harun Farocki: Working on the Sight-Lines, edited by Thomas Elsaesser. ↩︎
The difference in character generation between the PSX and PC remakes of Wizardry 1 kind of feel like a “soft” proof of my own worries here. As it's relatively trivial to re-roll for stats in the PC version, a lot of my characters start out with ~15 more bonus points than their PSX counterparts. What if I didn't even have to roll? ↩︎
During the writing of this piece, House House has revealed that LLMs were used during the final phases of developing Big Walk, which gives me complicated enough feelings that I felt it necessary to footnote in this entry. There's a lot of articles and blog posts written about it currently; I will link this one by Laura Michet and this one by Mike Cook as a breadcrumb encouragement to seek out more if you're interested. This one by Brendan Keogh is also pretty good, though I don't think I agree with all the points he seems to be making. ↩︎