Category: Uncategorized

I’ve been watching Generation Tech lately, a YouTube channel that covers Star Wars lore, usually revolving around the space tech and tactical doctrine or logistical efforts of the wars in the stars, i.e. what is an AT-AT’s purpose on the battlefield and why is it designed the way it is (answer: It’s a platoon-size IFV designed to be level with the skyline in most cities so that rebels can’t get above it, which makes it more intimidating, as per the Tarkin Doctrine).

One thing that keeps coming up in an annoying way is that the Star Wars galaxy has a million or more habitable planets. This leads to things like the Imperial Navy having 25,000 star destroyers at its height, and that’s for a navy that’s stretched thin, only able to directly occupy 2.5% of the planets they nominally govern. While it makes sense that they can’t mobilize the whole navy to go to Endor because they need some for patrol duty and occupation of trouble spots, it makes the battle seem pointlessly small if the entire Rebel Navy is there (Palpatine expects the Rebellion to be pretty much finished if they lose), is badly outmatched by the Death Squadron of star destroyers, and Death Squadron is struggling to reach 0.1% of the size of the total Imperial Navy.

Star Wars generally uses planets as though they were small countries: They have a single important city on them surrounded by lots of sparsely populated rural or frontier countryside. There’s clearly lots of small towns and countryside on Naboo, but the only major city we ever hear about is Theed, which is a pretty mid-size city. There’s probably other cities on the planet, but not many, and no megalopolises like Tokyo or New York City that we have here in our one-planet civilization. The Organa family might live in the countryside apart from major cities (medieval and Renaissance aristocrats did this, and they’ve kind of got that vibe), but certainly what we see of Alderaan (a core world!) suggests that there are like twelve cities total on this planet and they’re all pretty small, with forests for child princesses to impulsively wander off into within easy walking distance. Kashyyyk doesn’t seem to have any major cities at all, just small towns and villages peppered across the planet. Coruscant leaps to mind as an exception, where the parts we see are a small country centered on the Senate and Jedi Temple but there’s definitely a very densely populated entire planet beyond that, but there’s not many planets like that.

And if planets are like small countries, there’s only 200-ish countries on Earth, and even accounting for the fact that some of those countries are big countries that would be represented by multiple planets in Star Wars (space California and space Texas could be smushed together into one planet, but you wouldn’t expect it to be), 1,000 habitable planets is more than enough to cover everything Star Wars needs to. Yes, this means only a puny fraction of planets are habitable, but that was already the case. The Star Wars galaxy is about the same size and shape as the Milky Way, which means there are hundreds of billions if not trillions of planets in it, which means even the highest numbers given for the habitable number (“millions”) is well under 0.01% of the total. Since habitable planets are super rare no matter what, let’s cut them down to a number that’s both big enough that we’ll never plausibly outrun it when making up new planets yet also small enough that it’s believable that a battle for one planet matters.

The only reason Star Wars media is anywhere near exceeding a thousand named planets is because people keep inventing new planets unnecessarily. The mainline movies, TV shows, and video games do not do this, and while I’m less familiar with them, I can’t imagine the novels or comics are doing it that much. There’s like a thousand issues of Star Wars comics total, across all of time, so in order to run out of planets they would have to be introducing about one new planet per issue, which they don’t, and there’s only about 400 novels. Novels are more likely to introduce a new planet or even several than comics because they’re longer stories, but also a lot of the “novels” are actually YA books which visit fewer planets because they aren’t that much longer than a single issue of a comic. Plus, most of these stories reuse some of the 100+ planets already established rather than make new ones up.

And at this point an emphasis on reusing planets instead of making new ones up would serve Star Wars pretty well. It’s pretty telling of JJ Abrams’ flaws as a Star Wars creator that he felt like he needed to invent new planets, but the best he could come up with was Jakku, which was Tattooine with a “welcome to Jakku” sign slapped on. The only reason Jakku needs to be a separate planet from Tattooine is because of its backstory, which was not referenced by the movies in any way. Jakku is a mid-rim planet between Endor and Coruscant and the site of a battle between the Rebellion and the Empire in the aftermath of Endor, which is why there’s all these defeated Imperial wrecks lying around.

As far as I can tell this backstory was created by people trying to contrive reasons why this place that is clearly identical to Tattooine could justifiably be a new planet, and all they could come up with is that it is located in a different part of the galaxy. Sometimes that’s justified. Hoth and Rhen Var are both ice planets, Rhen Var has mountains but so what? Hoth could’ve had some mountains on it far away from Echo Base. But Rhen Var also has Sith ruins on it, which means it’s in ancient Sith space, like Yavin 4. If they put those ruins on Hoth instead of making a new planet, that means ancient Sith space covered most of the galaxy. If they moved Hoth to be closer to Yavin 4, that means the Rebels’ secret backup base was pretty close to their original secret base which makes them look like a regional nuisance, not one side of a Galactic Civil War. Rhen Var legit needs to be a new place even though it’s an uninhabited ice planet just like Hoth, purely because of the astrographic implications of Hoth having a Sith ruin on it.

Jakku, though? Sure, if the Battle of Jakku happened between the Battle of Endor and the Rebellion capturing Coruscant then it needs to be in the Mid Rim. But there’s no reason for the timeline to be like that, and Jakku has really strong Outer Rim vibes and is also clearly just Tattooine with a crashed star destroyer on it. So here’s a better backstory for Rey on Tattooine: The Battle of Endor is in the year 4 ABY. Rey was born in 15 ABY. The Battle of Starkiller Base is exactly 30 years later, 34 ABY, when Rey was about nineteen (about the same age as Luke at the Battle of Yavin). This is all from the existing canon, so how can we fit a Battle of Tattooine in to leave a star destroyer lying around for Rey to scavenge there by 34 ABY?

In the immediate aftermath of the Galactic Civil War, the New Republic fought on-and-off regional wars with Imperial remnants, especially in the Outer Rim. The Outer Rim had always been mostly de jure independent and after consolidating the Core and some key Rebel planets like Mon Cala, the New Republic was mostly willing to play nice with Imperial remnants in other parts of the galaxy, especially the Outer Rim, but sometimes those Imperial remnants poked the New Republic, resulting in a small war. In 14 ABY, ten years after the Battle of Endor, an Imperial remnant fleet caught up in one of these regional wars had fled to Tattooine and made their last stand against New Republic forces there. Because of the remote location of the battle, the New Republic never bothered salvaging the Imperial wrecks. They didn’t want Tattooine, they just wanted to make sure this specific fleet would not make incursions into Republic space again, so job done, they leave.

Scavengers descend on the planet to salvage advanced, military-grade Imperial technology like turbolasers and AT-AT cannons and twin-ion engines. Rey’s parents meet during this gold rush and have a kid in 15 ABY. Two or three years later, the really valuable stuff is gone, and Rey’s parents leave the planet. Unable to afford a kid now that the gold rush years are over (and having squandered all the money they made during the gold rush on living large), they leave her as well. The Imperial wrecks are still full of lighting fixtures and regular old power conduits like the kind they sell for twenty credits at Space Target, because the gold rush scavengers walked right past all of those to get to stuff that cost tens of thousands, hundreds of thousands, or millions of credits to build, which could be salvaged for a year’s worth of pay even at pennies on the dollar. What’s left is still enough to make a respectable living on if your standard for “respectable” is moisture farmers and cantina rats, but it requires spending all day scavenging half-functional parts worth just one or two credits each. Nobody wants to live on Tattooine and the people who already live there mostly already have jobs that are less dangerous or more profitable (in the latter case, because of crimes), so even fifteen years later there’s still some salvage left if you know where to look (star destroyers are big), but the last few scraps are finally starting to get picked clean. Open the curtain on Rey in 34 ABY.

I think Star Wars creators are in the habit of adding in one-off planets because this helped make the galaxy feel very big and believable for a long time. Han Solo makes a one-off comment about a bounty hunter on Ord Mantell and it helps make the galaxy feel like there’s more to it than just what was purpose-built for our protagonists. But we’re long past the point where adding in more and more planets makes the galaxy feel believably large and have reached the point where new planets make the galaxy feel unbelievably large, so large that it’s difficult to see how our protagonists could possibly be having an impact on it. The Galactic Senate chamber is not big enough for a million worlds, the Rebel fleet at Endor is not big enough to stand a ghost of a chance against 25,000 star destroyers even given a descent into chaos following the death of the Emperor, the Battle of Christophsis can’t possibly be significant for a war between factions that each control hundreds of thousands of planets.

Back in Empire Strikes Back, Star Wars had already established the significance of individual planets (roughly equivalent to a small country), but had only named Tattooine, Alderaan, Dantooine, Yavin IV, Hoth, Bespin, and Dagobah. So when Han Solo needs a throwaway line about an encounter with a bounty hunter shaking him up, it’s a good idea to drop a one-off reference to Ord Mantell, some planet we’ve never heard of before and which there are currently no plans to use in any movie or spin-off material. There’s at least a few hundred and probably about a thousand planets out in the galaxy, when we’ve named a grand total of seven of them, the odds that any random planet referred to would be one of those seven is struggling to reach 1% even before considering how remote and backwater most of those planets are. But that was a long time ago. We have hundreds of Star Wars planets now, so it’s not weird that when one comes up, it tends to be one we’ve already heard of.

Idle sci-fi worldbuilding idea: A measurement system based on the speed of light would make a reasonable amount of sense. It would still be nailed to Earth units of time, but if we assume that a healthy human sleep/wake cycle is 24 hours long regardless of what the lighting situation is (has anyone actually studied this?), then we’ll carry that length of day in our biology even as we leave behind the planet that evolved it into us. And light distances mostly conform really well to easily managed 1-100 scales for most human use cases.

A light nanosecond is almost the length of one foot. That means it’s a good length for measuring the size of things that are close to human scale. Human height is measured in a single digit number of light nanoseconds. It is slightly annoying to express small lengths like we usually use inches or centimeters for as hundreds of light picoseconds, but people who use meters have the reverse problem where you have to use decimal points all the time because most humans are between one and two meters tall and there’s never any riots demanding a return to imperial measurements.

A light microsecond is a really good unit of distance for overland travel. The average human can walk a light microsecond in a couple of minutes, and can walk about sixteen light microseconds per hour, so human walkable distances largely go from a scale of one (for something that can just barely be described as going for a walk rather than going down the street) to one hundred (for something that will require several hours, approaching the limit of what a reasonably fit but untrained human can walk if they set aside the day for it).

Car speeds largely go from 100 microseconds/hour for residential distances to 500 microseconds/hour (US speed limits usually cap out around 430 microseconds/hour, but most people end up going ~450 anyway, and it’s easy to imagine a society built around this measurement system using 100-500 as “car speed”). This isn’t quite as snug as nanoseconds to microseconds because now instead of a 1-100 scale that keeps commonly used distances to double-digits. If you’re in any kind of hurry, you’ll take the car for anything that would take more than 5-10 light microseconds, and the upper limit of a day trip in a car is well over a light millisecond. A light millisecond is about 186 miles, which is far past the limit of being a quick car ride but well under the limit of how far a car can take you in a day.

On the other hand, light milliseconds do lend themselves really well to measuring the distances for flights. A distance of less than one millisecond is generally too short to be worth going to the airport for, an NYC to Singapore flight is about fifty milliseconds, and the circumference of Earth is about 134 milliseconds. It’s hard to doublecheck the shortest commercial flight distance because the low end of that scale is taken up by trips that are well within driving distance but there’s water in the way and trips are infrequent enough that it’s not worth it to build a bridge. London to Paris is slightly more than one light millisecond and they built a tunnel under the English Channel because people were getting annoyed at having to take a plane for that distance. It’s definitely rare for any single-trip distance on Earth to be anywhere near the 100 milliseconds at the top of a 1-100 scale, on account of the Earth is a sphere so if you’re going 80 milliseconds it’s basically guaranteed that you could go the other way around and get there in 60 or less.

One full light second is about the distance of the Earth to the moon, and intrasolar distances are frequently measured in light minutes in the inner solar system and light hours in the outer solar system because those are already the most convenient units to use. The inner solar system keeps to 1-100 light minute scale while the entire solar system is about 22 light hours across, so even in the outer solar system where the distance from one object to another might be, depending on orbits, hundreds or even low thousands of light minutes away, distances in light hours are always manageably low numbers.

As you get into sci-fi intergalactic distances, you quickly run into the problem that the galaxy is a heccin chonker 100,000 light years across, and that’s not unusably large the way trying to measure galactic distances in miles or kilometers is (although metric has the advantage of being able to scale up to petameters and exameters), but it does massively exceed the 1-100 scale I’ve been trying to keep to. Of course, to actually use that space you must necessarily be using FTL travel of some kind, at which point you are making up how fast things go and may as well invent some hypertech excuse why light speed travel is measured in kiloyears on major trade spines but goes down to lightyears per hour (or per day or per week, depending on how isolated you want solar systems to be) outside of the main hyperspace routes.

Fractions of the speed of light are a futuristic, science-y kind of measurement tied to a fundamental law of physics and a unit of time that, while arbitrary, is probably pretty deeply tied to human biology, and it works really well for measuring the size of things at roughly human scale in light nanoseconds, walking distances in light microseconds, flight distances in light milliseconds, distances within solar systems in light seconds, minutes, and hours depending on the exact region, and distances between solar systems in light years (although that last one’s a freebie because the propulsion systems and frequency of landmarks worth caring about is up to the author anyway).

It’s annoying how car distances straddle the microsecond/millisecond line, but it’s otherwise very usable and fairly easy for readers to translate, easy enough that it might not be frustrating in use – provided that your plot doesn’t deal with the ugly car distance microsecond/millisecond overlap, because while I think people can quickly grok “walking distance is measured in microseconds, in-atmosphere flights are measured in milliseconds, and they both go on a scale of 1-100,” the hundreds of microseconds to whole milliseconds scale of car travel means they’ll start trying to convert to miles or kilometers in their head or treat the distances as white noise. And while you can make atmospheric flights stretch up to 100 milliseconds by making planes faster (while the bottom of the scale remains the same because it’s controlled by the point at which a car is slow enough to justify a plane, not the point at which the plane is too fast) and just not use the top 50 points of the scale because Earthlike planets aren’t that big and it’ll still work, you can’t do the same for cars because humans aren’t getting any faster so you cannot make a 300 microsecond trip reasonable walking distance, but cars are already too fast to reasonably be capped at sub-millisecond distances. Getting around that would require significant worldbuilding oriented around making this distance system reader-friendly enough to be usable.

I’m not really going anywhere with this, I have no plans for this measurement system and am not working on anything remotely Star Trek-ish enough to bother using it. So this is one of the random research project posts.

I’ve been watching Delicious in Dungeon lately. Ten out of ten, no notes, except that name. It doesn’t roll off the tongue very well in English. The anime is generally very well translated overall (I’m watching on Netflix, and if it has a subs option, I couldn’t find it at a glance, and I don’t care enough about subs vs. dubs to spend more than five seconds looking, so fuck it, dubbed it is), and even does a good job of using current English slang and idioms in a way that feels fairly natural, though I worry that as early as the 2030s people saying “cringe” is going to make this feel like a period piece.

But for some reason they went with “Delicious in Dungeon” instead of something nearly identical but with better flow, like “Delicious Dungeon” or “Dungeon Delicious.” “Delicious in Dungeon” doesn’t come up directly in the story except that the narrator will sometimes say the title at the end of an episode as a bookend, which works with any title that has the word “dungeon” and “delicious” in it (and if you called it “Dungeon Cooking” instead, all you’d have to do is slightly rework the narration to fix the segues). I wonder if it’s supposed to have the same cadence as “Dungeons and Dragons?” But if so, it doesn’t, the syllables are wrong and using the “in” instead of the “and” throws off the usual “it’s like D&D but” title where you do two alliterative nouns separated by an “and.” Starships and Sorcerers doesn’t have the cadence, but the [noun] and [noun] format makes it recognizable and the nouns used signal some kind of space fantasy setting. If you really want the D&D reference, go with Dumplings and Dragons.

And it reminds me of K-On! The title itself is untranslated (I don’t know if K-On! even means anything in Japanese? It doesn’t sound like a full Japanese word), but the name of the rock band that the main characters are in is named “After School Tea Time,” after the fact that they meet as a club after school and spent a lot of the early episodes using their club as basically an excuse to fuck around having tea and snacks after school. But whereas “After School Tea Time” sounds like an anime with an overly literal translation, “Tea After School” is a good name for a band.

Both of these shows have perfectly good English translations but for some reason they fall apart on proper nouns.

Also this blog updates on Fridays now because my Patreon updates on Wednesdays and I want a Discord bot to be able to post both of them to the same channel without doubling up on two posts the same day. And also if you want to read my thoughts on TTRPGs and game design, those are on Patreon these days. I am now a professional game designer, so I don’t mind charging roughly $0.25 per post for my thoughts on that subject, whereas this blog is mostly my dumping ground for video games and anime and other stuff where I am just a guy.

Diablos III and Immortal are unworthy successors to the Diablo title. Diablo IV is reportedly much better (I haven’t tried it, and for that matter I haven’t tried Immortal, but for Immortal to be good, nearly everyone who ever wrote or talked about it would have to be lying), but it’s tied to plot points introduced in Diablo III and Immortal, which makes me worry that it would be a jarring lurch to go into straight from Diablo II. I hate the feeling that I’ve skipped a chapter, and while some installments in an ongoing series are so superfluous that you totally can just avoid them (the Assassin’s Creed series, for example, is an episodic series whose only ongoing plot is the stupid modern day frame stories that are best off ignored in the first place), it’s not clear to me if Diablo IV is that. I’d like it if someone took the new plot elements introduced in Diablo III and Immortal, cleared out the other 80% of the game retreading ground covered by Diablo I and II, and then presented it in a style consistent with those first two games rather than being World of WarCraft-ified in a quest to make all of Blizzard’s franchises more like the one that makes the most money. That game doesn’t exist, and so long as I’ve got 100+ games in my backlog, that’s enough to disqualify the otherwise (reportedly) quite good Diablo IV from my wishlist.

Hades is not that game. It isn’t literally a spiritual successor to Diablo II, but it’s closer to being that than Diablo III is, which is surprising since Diablo III is an actual numbered sequel and Hades was at no point even trying to be a Diablo successor. Diablo III had a few new ideas but squandered them with a combination of poor writing and slavish devotion to the settings and characters of the first two games. Diablo is a series where you go into a cool place, kill cool monsters, and watch the numbers go up as a result. The loot progression is critical to the gameplay loop, but what the loot actually is makes fairly little difference. The new weapons and armor you get are rarely that engaging on their own, but it does feel cool when you equip a new weapon and are now cleaving through monsters at a noticeably accelerated rate. And that’s not to say it’s a mindless process, either. While grabbing whatever has the biggest number will carry you through the base game, there is a lot of depth to build optimization in Diablo and especially Diablo II. Rather, what I’m saying is that where that build optimization comes out viscerally is not in what hat your character is visibly wearing, but purely in the rate of monster destruction.

And the reason that works well in Diablo and especially Diablo II is because the monsters and locations you’re in visually progress and the narrative accelerates towards a climactic confrontation (provided you’re playing with the Diablo II expansion’s Act V – Act IV is actually kind of lame, you just show up at Diablo’s house and beat him up and the Hell levels are just long enough for the momentum of the Act III/IV transition cut scene to wear off and then there’s not really anything else to pick up the slack in the story). Newer games’ “new locations” are locations from the old games with pretty minimal edits, and their new monsters are mostly the same. Even when locations do change significantly, it’s only to change into another old location. The mountain from Diablo II was reduced to a crater at the end of that game, so it is now the caves level from Diablo I.

The shift in art style in Diablo III was also a bad idea. I remember people making fun of the fan backlash over it and those people were and are stupid hacks. Making Diablo look like a gothic-themed expansion for World of WarCraft is no better an idea than World of WarCraft looking like an orc-heavy Diablo expansion. The unification of all of Blizzard’s games into a single mono-art style is bad. That it was the critics pushing for homogeneity in art against fans calling for more diversity is a bizarre reversal of what’s usually one of the critics’ redeeming qualities: A greater appreciation for the new. We can’t give the fans too much credit here, because retaining the old Diablo style would obviously not have been new, but at least it would’ve avoided winnowing down the old and actually shrinking the diversity of art.

And on the other end: Hades is a pretty good successor to the Diablo series. There’s definitely enough of a tonal mismatch that it is not a good idea to take it as literally a stealth Diablo sequel. In a Diablo game, Hades would be a world-devouring arch-horror and Zagreus’ quest would be to stop him, whereas in Hades the titular antagonist is a stubborn old man who must be worn down through a relentless defiance, but the ultimate resolution is reconciliation. I really like how well it’s handled, the game doesn’t present Zagreus as flawless but also doesn’t mind presenting a scenario where one side is more in the wrong and has to give more ground in a reasonable compromise than the other, and since the whole theme of the game is about wearing someone down over time, Hades’ slow crumbling before Zagreus’ relentlessness feels earned and plausible, not like the story reached act three so the writers flipped a switch and Hades instantly became more reasonable after a moment of emotional catharsis. It dawns on Hades slowly that no, really, Zagreus is never giving up on this, and it takes him a while to update his behavior on that realization.

Anyway, all that’s badly out of step with the tone of Diablo, but otherwise it’s a pretty snug fit. A gothic atmosphere, dungeon levels full of gory, infernal imagery, a host of ghastly and skeletal monsters, a heavily stylistic visual presentation that makes the game look like a playable film of a specific medium. Most people think of Diablo and Diablo II as just looking like themselves, but that distinctive art style comes from trying to look like claymation as best as the available tools permit. Hades looking like a 2D animated film that you can play is exactly the kind of visual update that Diablo III should have had – moving on from the old to do something new, rather than a relatively niche old thing being assimilated into the Borg cube of a bigger, more profitable old thing. Likewise, the mechanics are recognizably a realtime action dungeon crawler, but the details of how upgrades work and the new Roguelike elements significantly overhaul exactly how you play. The locations have Diablo III’s problem of being a bit of a retread – Tartarus is basically the crypts from Diablo I, Phlegethon is the pre-requisite Hell level, and while Elysium is new and interesting, Diablo III also managed exactly one interesting new location in its last act – but the new monsters and art style make them feel more distinct from older Diablo games than Diablo III’s rendition. Also, to be clear, I’m not marking Hades down for Tartarus and Phlegethon being vaguely thematically similar to Diablo series dungeons, because it’s not actually a Diablo game, just pointing out that Hades’ advantage over Diablo III as the accidental third installment in the series is more slight here than in other aspects.

Hades isn’t usable as the “real” Diablo III even with a few script edits to change the proper nouns around (especially not for purposes of being the missing bridge to Diablo IV), but if Diablo III had been good, it would’ve had more in common with Hades than with the Diablo III we got.

I don’t know why I don’t announce my Kickstarters on this blog. It’s not like I’m putting out content so dense that a post talking about one of my actual main-job projects would require me to push something else back. So let’s talk about Tactical Doctrine: Rat King! Tactical Doctrine is a series where I try to build an expanded Monster Manual one month at a time, with an emphasis on creating complete armies for different enemy factions. Tactical Doctrine: Troll King focused on goblins, orcs, and trolls, and Rat King is focusing on ratlings and sewer monsters. The Rat King doesn’t just have a ratfolk and a boss ratfolk, he has a ratfolk heavy and shaman and rat swarm and they all fight together. There’s different units with different roles who work together, following a tactical doctrine, hence the name. Heavies and rat swarms hold the line, skirmishers shoot from afar, and the shaman throws on debuffs.

You might wonder why, and the answer is because the Rat King is the guy who gets the Rogue class, and I want to get three tactical classes out as fast as possible so you can have an absolute bare minimum size tactical party. Troll King had the Barbarian, so Rat King has the Rogue. Tactical classes come with paired abilities where using one also exhausts the other, which means you have a lot of options but using them comes with more consequences than just attrition, so you need to put some thought into which one you pick. The goal is to add decision making into individual combats, a place where D&D has usually struggled (D&D is really a character building game, and even more really than that, a party building game).

There’s also rules for solo play. A lot of people have characters and adventures they can’t find a group for, and solo play rules are meant to finally make those dreams come true. It’s a smaller audience, but I feel like I’ve pretty much completely mined out standard D&D 5e content. I basically never get players wondering about a character concept that’s not already playable in Chamomile’s Guide to Everything. Rat King’s solo rules is for mysteries, which give instructions on how to turn a city gazzeteer (standard in most sourcebooks) into an Arkham Horror board to gather up clues, catch serial killers, and thwart conspiracies.

Midnight Fight Express is a game where you are a criminal who left the life but now has to come back for one last job to kill every other criminal in the city and thus save the city from crime. It’s a third-person action game whose combat system is basically Arkham but worse, although not so much worse that it isn’t any fun, and in which the combat is the whole game, there’s none of the rest of Arkham gameplay.

There’s a lot of variety in enemy appearances but relatively little in how they fight. The Bozos who serve as standard street criminal fodder at the beginning are basically the same as the Warriors and Earth Smashers who get into a gang war that you have to fight through are basically the same as the Death Bunnies who are knife-wielding crime strippers. There are lighter enemies and heavier enemies and some enemies have guns, which are very annoying to deal with but if you go straight for them to beat them to the death and take their gun before they run it out of ammo, you can unload the gun on their friends, which is good for some free kills. So it’s not like there’s no enemy variety at all. The Bozos at the start of the game have all of this stuff, though, and the only noticeable difference with most later enemy types is that their melee weapons have higher durability and their guns have more ammo and a higher rate of fire. The drops improve, and the number of enemies goes up so the game does get harder (managing large crowds is much harder than fighting enemies in twos and threes), but most enemies are Bozos with a different model. There are some exceptions to this, the rat mutants spit out plague bile that leaves a lasting hazard on the ground and the zombies will pretty much instantly kill you if they get into melee with them so you have to use guns, but these kinds of enemies are the exception rather than the rule.

As that rundown of the enemies suggests, this game gets much more gonzo than it seems very quickly. It was always firmly in pulp territory, opening with a drone saying “literally every criminal in the city is in on a scheme to take over the city and only you can stop them! There’s some now!” And then you segue from there into street fights with violent criminals. It takes about 40 or 50 minutes in before you reach the rat mutants in the subway tunnels, though, and that’s the first sci-fi enemy. By the end it turns out the criminal takeover isn’t something generic like killing all the uncooperative civil servants to install a corrupt cadre in bed with the mafia, but rather killing everyone and uploading their brains into murderbots, because murderbots require brains for some reason, and apparently it’s not a problem if the sourced brains come from civilians you just murdered like six hours ago. Despite how crazy the plot quickly gets, the central character relationship between the protagonist and the hacker piloting the drone who’s guiding you to foil various components of the evil Operation Neo Dawn works pretty well. It’s by-the-numbers, but well-paced and well enough delivered that I felt some actual feelings about the two characters when I reached the end of the game. For a game that’s primarily a vehicle to kill spec ops squads with a katana, that’s more than was necessary.

Perhaps most importantly, it’s only about five or six hours long. I accidentally left the window open overnight so I have no idea how long I spent playing it exactly, but it was short enough that I hadn’t worn out on its gameplay when I got to the end, and that each beat of the plot and character arc came fast enough that I didn’t get sick of that, either. I keep bringing this up because I keep running into games that do it well and games that do it poorly: Five hours of all-thriller no-filler gold star content is better than ten or fifteen hours of mediocre hit-and-miss content.

RTS campaigns are basically extended tutorials for how to play multiplayer. This is of necessity – there’s so much stuff in an RTS and the way it interacts with other units/buildings is so complex that it’s best to introduce things one unit at a time, and by the time you’ve done that, there’s not really any room left for anything else. In fact, StarCraft’s Terran campaign doesn’t even do a particularly good job of introducing some of the units because its first two stages are gobbled up with very basic tutorials for the game’s controls and resource gathering. Figure out what firebats are good for yourself, because their intro mission is focused on base-building (although it’s not that hard to piece together that they trade well with Zerglings, especially in a bunker – and to the extent that it’s hard to figure that out, it’s mostly because that’s an extremely niche role and one they’re not even that good at).

But starting in Terran 3, you learn how bunkers and vulture mines work for base defense with a defensive mission, Terran 4 is a no-build mission that doesn’t even require effective micro but it introduces the concept of no-build missions, I guess, then in Terran 5 you face an enemy base on an island with defenses laid out so as to encourage you to send wraiths over to clear a landing zone and then dropships to bring in a ground force to clear out the AA perimeter around the enemy base, on up to Terran 9 when you first unlock battlecruisers. I think there might be some upgrades you can only get in Terran 10, but either way, you spend pretty much the entire Terran campaign unlocking one unit at a time, with scenarios laid out so that the most effective strategy involves using that unit. They’re often even laid out so that the winning strategy of the previous scenario doesn’t work, forcing you to figure something else out, like in Terran 6 where the enemy’s air defenses are so thick that the wraiths-and-dropships plan is pretty doomed, but using your new goliaths works much better.

Starting out with the smallest, weakest units and then slowly unlocking the bigger ones is definitely the most satisfying way to structure the campaign. But so far as effectively teaching the mechanics goes, it’s the reverse of how you want it to work. The most straightforward Terran strategy is the one you unlock last in the campaign: Build a fleet of battlecruisers and roll over everything. Presenting battlecruisers as the capstone means that you come out of the campaign thinking of battlecruisers as just better than wraiths. They aren’t weak against ground units the way wraiths are, they’re bulkier and do more damage in general, and while they can’t cloak, their Yamato cannon lets them destroy missile turrets and weaken photon cannons from a distance, allowing them to counter their own counters. Wraiths are also a counter to unescorted battlecruisers, but it only takes one science vessel to reveal the cloaked wraiths, at which point wraiths only trade well in tremendous numbers.

There are things that can defeat battlecruisers, of course. Protoss arbiters with stasis fields can split battlecruiser fleets up, freezing some to reduce the number of scouts or dragoons required to get the upper hand. A few high templar psi storms can take out an entire battlecruiser fleet at the cost of just 5-6 high templar, and the battlecruisers will have difficulty getting away from the storms because of how slow they are. Zerg defilers can use dark swarm to make hydralisk swarms immune to air fire, and their queens can use ensnare to reduce already slow battlecruisers to a crawl, making it even harder for them to get away from the shielded hydralisks. Enemy Terrans can make cheap (albeit high tech, but no more so than battlecruisers) ghosts with lockdown to allow wraiths to trade favorably with more reasonable numbers or just give the ghost-user the edge with their own battlecruiser fleet. You can make nine battlecruisers and twelve ghosts for the same vespene cost (and fewer minerals and supply) as twelve battlecruisers, but the ghost-supported fleet will likely win the confrontation at the loss of less than half their ghosts and no battlecruisers (and if the enemy fleet doesn’t have comsat or science vessel support, cloaked ghosts will probably get away unscathed).

StarCraft doesn’t do a terrible job of demonstrating this in the one mission they have after introducing battlecruisers for the first time. In Terran 10, General Duke uses a ton of ghosts, which makes it unviable to use unsupported battlecruiser fleets to crack his base. Except he also uses a ton of nuclear strikes (also used with ghosts), which are expensive as Hell if they fail, so if you know how to defend against that, he’ll run himself out of vespene on futile nukes and you can roll over him with a battlecruiser fleet anyway. That map also has Arcturus Mengsk on it, but I honestly don’t know what his strategy is supposed to be. He uses a lot of wraiths and they do cloak, but nowhere near enough to trade well with a battlecruiser fleet, so as long as you bring a science vessel, you’re good. I guess he does teach players that battlecruiser fleets need a single science vessel supporting them to counter the wraiths.

But it’s still very easy to come away from the StarCraft Terran campaign thinking that the Terran strategy basically amounts to “build battlecruisers to win,” and it would be more clear that this wasn’t the case if battlecruiser fleets were one of the first things you got, rather than one of the last. Massed battlecruisers are the most straightforward path to victory, so start people out on that, and then show how that strategy can be countered and then the counters to those counters and so on.

While it would probably weaken the singleplayer campaign as an experience unto itself, the campaign would do a better job of teaching StarCraft if you started out making battlecruiser fleets, and then steadily got introduced to enemies with counters to this most obvious tactic, thus slowly making it more clear what the other units are for. Your first two missions would still be movement and base-building tutorials using marines and firebats because those units are easy to build, and the third could still be a defensive tutorial, although that one should definitely unlock siege tanks instead of vultures. It’s much easier to see how siege tanks are a vital part of Terran defense than vulture spider mines, like, sure you can see what a spider mine does no problem and you’ll want to try the new unit out, but the conclusion I and, I’m confident, most other new players came to after Terran 3 is that vulture spider mines don’t really do anything that a good bunker complex couldn’t do better.

The fourth (or fifth, if we’re still having a no-build mission for four despite a radically redesigned plot necessitated by this new order of introducing units) would give you a fully developed Terran base with all units locked except marines, firebats, and battlecruisers, and a massive Zerg (or whatever) base to sic the battlecruisers on. Put the base on an island so you’re forced to use battlecruisers to wipe them out. Maybe include dropships to also introduce the need to create expansion bases to feed the resource-intensive battlecruiser fleets. Have the plot dialogue stress (in the same way that things like building damage and creep got stressed in dialogue in early Terran missions from the game as it is – draw the player’s attention to its importance without beating them over the head with it) the power of battlecruisers, but also that they’re slow, both literally and in the sense that they’re at the top of the tech tree with a slow build time and a high resource cost so it takes forever to amass a fleet. “Fortunately, these mindless Zerg have given us all the time we need to bring our battlecruisers into position. Now we are unstoppable!”

Then in the mission after that, you either disallow battlecruisers completely for plot reasons or stress to the player that the enemy has more resources than you and you can’t win a straight-up fight, so you’ll have to be sneaky, while giving them access to ghosts and wraiths for the first time. Use dialogue (like what Kerrigan says in Terran 5 from the game as it is) to stress how detectors work, and the importance of taking them out. Either leave the player with siege tanks so they have an answer for missile turrets or else really focus on stealth by making the enemy defenses so heavily air oriented that they basically don’t even have bunkers and you can use ghosts to take out the turrets for your wraiths.

You can keep going down the tech tree and each additional unit further down will be filling a new niche – goliaths won’t feel like a replacement for marines, nor battlecruisers like a replacement for wraiths, because the weaker units are introduced in missions that spotlight the situations where they’re more useful than the stronger ones.

“Mary Sue” was a big deal back in the 00s. The term was coined in a Star Trek fan ‘zine back in the 70s according to Internet Lore (and while I don’t think anyone’s ever dug up a copy of the original magazine to check, this is recent enough and has a specific enough origin that I’d be surprised if Internet Lore was wrong about this one), but I have no idea how the term was used back then. What I do know is that the term had a heyday during a specific decade and I was an eyewitness to the phenomenon. Everyone reading this is likely vaguely aware of the character archetype, but likely also aware that defining what makes a “Sue” is hard to nail down.

Are they held up as beyond reproach by the narrative? No, the Villain Sue is recognizably cut from the same cloth, and there are characters like Samurai Jack and Superman who are generally portrayed as always doing the right and reasonable thing, and these characters are clearly not what people were talking about when they were discussing Mary Sues.

Are they a critical mass of out-of-place traits in the setting like angel wings, red or purple or otherwise unusually colored eyes, and powers either not present in source material (and not within the reasonable ability of new characters to obtain, i.e. it’s not an X-Men style setting where new people are expected to have never-before-seen powers) or else having a conglomeration of powers that are otherwise mutually exclusive? Not really, while people did start to develop kneejerk reactions to certain eye or hair colors at the height of the phenomenon fifteen to twenty years ago, the fact is that Aang is not a Mary Sue because he can bend all four elements, has not one but two special animal companions, and has distinguishing body tattoos connecting him to an extinguished culture slaughtered by the main villain. Aang isn’t even a little bit of a Mary Sue despite a pileup of so-called “Sue signifiers” that easily push him into at least mild Sue territory if you try to count purely by number of unique and attention-grabbing (or at least, attention-seeking) design elements.

Is it purely a construct of 1) unreasonable standards for fanfic writers to bow to the original IP and canon and 2) rank misogyny? While it is certainly notable that Mary Sue’s name is Mary Sue, and that people even proposed “Gary Stu” as a male alternative and even sometimes tried to define Gary Stu as having different traits, Kim Possible and Buffy Summers were rarely smacked with the Mary Sue label. Indeed, Buffy Summers was frequently held up as the anti-Sue counter to Bella Swan, who did not escape the Mary Sue label despite being an original (legally, at least) character in a book sold at Barnes and Noble. Indeed, precisely because of her prominence in bookstores, Bella Swan was probably the main target of the anti-Sue community throughout the late 00s. There was definitely a general attitude that internet writers were kids playing dress up as real authors or, at the very least, on a lower rung on a career ladder that would hopefully end in becoming a “real author” one day, and a lot of the criticism of Mary Sues was bulwarked by if not rooted in the idea that a character this self-indulgent was unbecoming of prose fiction, which should aspire to publication. But that didn’t stop the consensus from emerging (amongst people who talked about Mary Sue at all, at least – Twilight’s fans were not generally in those communities) that Bella Swan was a Mary Sue, and Twilight unworthy of publication and popularity because of it.

So if Mary Sue isn’t a character without flaws, isn’t (just) a character with a critical mass of attention-seeking traits that people were particularly sick of seeing at the time, and isn’t purely a construct of gatekeeping (even if some amount of gatekeeping caused certain characters and authors to receive unfair scrutiny), then what is it?

I think the inchoate vibe that people were trying to describe with Mary Sue is basically that a Mary Sue narrative feels like it’s the self-perception of someone with Narcissistic Personality Disorder. Internetizens were obviously not broadly familiar with the diagnostic criteria for NPD back then (or now) and weren’t actually judging narratives on a rubric of how narcissistic they seemed on a clinical scale, but narcissism isn’t something psychiatrists invented, it’s a real phenomenon that people can pick up on in their environment without having first learned about it from a book. You wouldn’t expect people to be good at identifying a psychiatric phenomenon like this from pure experience, especially when the average age of the community is sixteen, but the clumsy and amateurish efforts of people to describe Mary Sues isn’t exactly impressive in its precision. People’s reaction to Mary Sue is exactly the poorly articulated and yet clearly felt experience you would expect from teenagers in absolutely no condition to accurately describe a psychological disorder encountering it.

Nor for that matter do I think the people who wrote Mary Sues necessarily had NPD themselves – I expect that was very rarely the case. Mary Sue was a phenomenon mainly of either inexperienced writers or hobbyists making no effort to improve, not a handful of prolific serial abusers. When Mary Sue writers did show signs of narcissism, it was usually teenage narcissism which they grew out of. I think the main cause of why Mary Sue feels like a story told by a narcissist is not because it is literally written by someone with NPD, but because writing is something you do alone, and people’s empathy seems to shut off almost completely when someone else is not physically in the room (or at least audible). It’s not controversial to say that people are much more anti-social typing on their screens than when speaking in person, so I don’t find it a huge reach to imagine people writing alone tend to default towards narcissism even if they’re not like that when interacting with other people.

But what do I mean when I say that Mary Sue stories feel like the world through the perspective of a narcissist?

Well, you know those “Sue tests” designed to tell you if a character was a Mary Sue based on a point total calculated from whether they had rainbow eyes and what their BMI was? The diagnostic criteria for NPD are like that, except there’s only nine of them, you need five to qualify, and they’re way better at describing Mary Sue stories (though not necessarily Mary Sue herself – Mary Sue is how a narcissist describes themselves). It should be noted, though, that a story does not automatically improve the fewer of these traits it has, and that my goal with this post is to describe a phenomenon, not provide writing advice.

1. Grandiose sense of self-importance. Narcissists are obsessed with and exaggerate or fabricate their accomplishments and abilities. Mary Sue becomes an archmage at the age of fourteen. This is one where the age of the writer often punts a story firmly into satisfying this criteria: A fourteen year old wants to write about a character their own age, but also wants to participate in important events. This necessitates writing a prodigy of absurd capability.
2. Frequent fantasies about having or deserving success, wealth, power, beauty, or love. Mary Sue stories often straightforwardly are fantasies about having or deserving these things. One of the most critical signs of a Sue is a character who has these things and complains about being unfortunate and victimized anyway, which is also a thing narcissists do.
3. Belief in superiority. Mary Sue is the chosen one, has special powers other people don’t, sometimes including a genius brain that lets them discover things other people couldn’t and with minimal effort. The Tony Stark archetype of being a super genius whose big brain lets them accomplish things other people just can’t even attempt had very little overlap with the era of Mary Sue discourse, and yet it’s not hard to see that BBC’s Sherlock steadily grew into more and more of a Sue over the course of the series and that the first major warning sign was the way his deductive abilities were portrayed as a superpower unique to him, and not (as in the original stories) something he learned to do and constantly tried to pass on to others, with partial success in some cases – Watson was never Holmes’ equal, but he became a noticeably good detective by learning Holmes’ methods, while Scotland Yard remained buffoonishly incompetent by ignoring them. Doyles’ Sherlock’s abilities were a thing he had a knack for but which anyone could learn, while BBC Sherlock’s abilities are a superpower unique to him and a handful of other special gifted people. Sherlock’s use of this trope is pretty glaring since it’s in such contradiction to the less narcissistic source material, but “the story is about Steve because Steve is the one who happened to bond with the Infiniatrix and get super powers” is a perfectly fine premise.
4. Need for admiration, reacting poorly to criticism, obsessed with what others think of them, and fishing for compliments. Mary Sue isn’t usually like this herself, but other characters are as obsessed with her as a narcissist would want people to be with the narcissist themselves. People who like Mary Sue are good, people who dislike Mary Sue are bad. People who are indifferent to Mary Sue don’t exist – they secretly have an opinion and are hiding it, so they probably dislike her and are evil. Nobody is ever on Mary Sue’s side despite disliking her as an ally of convenience, nobody ever opposes Mary Sue for understandable reasons. At best, there is some lie or delusion that causes them to oppose Mary Sue and they switch sides as soon as the truth is revealed. A Mary Sue story doesn’t depict Mary Sue exhibiting the attention-seeking behavior of a narcissist because the world and the people in it are already obsessed with her.
5. Entitlement. Narcissists believe they deserve more than other people. Mary Sue often gets special treatment from authority figures or allies and then complains anyway because the treatment wasn’t special enough.
6. Willingness to exploit others. Narcissists view other people as tools to achieve their own ends, without any goals of their own. Distinct from psychopaths (meaning, people with anti-social personality disorder, because yes psychopathy is a colloquial term with no official diagnostic meaning but it’s also a useful noun describing people of that demographic), narcissists care a lot about what other people think of them, but they don’t care about what those other people want for themselves. This is an important distinction because while Mary Sue often claimed to be a psychopath or used psychopathy as an excuse (this was in vogue back in the 00s), narcissism is distinct. Other people’s thoughts are important to the Mary Sue narrative insofar as they relate to Mary Sue, but they have no goals or desires outside of that. Other characters don’t react the way real people would to this because the story has a narcissistic perspective, which means other people actually do lack any thoughts or feelings not directly relevant to Mary Sue.
7. Lack of empathy. You only need five out of the nine to qualify for NPD and this is the one that a Mary Sue story most often misses, although somewhat incidentally: The people portrayed by the Mary Sue story are so lacking in internal lives that there’s barely anything to be empathetic for. Sometimes Mary Sue’s combination of entitlement and victim narrative cause her to demonstrate a lack of empathy for other people’s problems (narcissists often throw temper tantrums when someone else shares experiences of victimization because it diminishes the importance of their own victim narrative, and sometimes Mary Sue’s antagonists will reflect this, but that’s rare – and I suspect those stories are written by actual real narcissists). Mary Sue also sometimes gets to tick this box because she doesn’t notice or almost instantly forgets when people sacrifice themselves for her. Usually, though, Mary Sue gets a pass on this one because the other criteria have created a world where there’s no opportunity to demonstrate empathy one way or another. These criteria are descriptions of the story’s perspective, not of its protagonist, but counting the story as lacking empathy because its need for admiration has eroded any internal lives in its supporting characters feels like double-counting.
8. Frequent envy. Mary Sue tends to dodge this one as well, but not always: Sometimes her opponents have hordes of unearned wealth, status, and power which Mary Sue then acquires for herself, which will be portrayed as restoring some natural order of things. Other times Mary Sue always had wealth, power, etc., or acquires them without any comparison to a less deserving antagonist.
9. Arrogance. Mary Sue is often rude and contemptuous of other people. This is partly informed by the popularity of acerbic wit in the 00s, which writers of Mary Sue would sometimes try to replicate while failing to get the actual “wit” part. But also being able to insult people and get away with it is semi-frequently part of the fantasy.

This test works really well for identifying characters that feel like Mary Sues for me. I don’t know how broadly applicable it is – I feel like it’s a reasonable guess that this is what people were getting at when they talked about Mary Sues back in the day, but all I can really say for sure is that it accords very well with how I used that term, and people of the time didn’t generaly think I was using it incorrectly.

I’ve resorted to once-weekly posts to get a decent buffer going again. No particular reason, just didn’t have a whole lot of new thoughts worth writing about recently. Despite the slowdown, however, I have managed to reach my one thousandth post on this blog. And only about 10-20% of those posts were space-filling garbage like this one.