Let’s Run

So Awesome Games Done Quick 2013 is done, and with it the SDA marathon events have raised about 3/4 of a million dollars for charity. That’s really amazing, and on one level I’m really proud that I’ve been able to participate, even if it’s on some small, nigh-useless level. I had felt, for many months prior to my trip this year, that general speedrunning communities and activities, and even the charity angle on the whole thing, might not really be enough to make my trip worthwhile. I was really going just to see friends, my family, and Yuzo Koshiro.

None of that’s changed from my trip, really. But it doesn’t have to. Every year I meet more and more people that I’m excited to go see again, so while my sister’s family won’t be in the area next year, and Magfest may not have quite as amazing of a headliner (though I can think of a few I’d like to see at least as much as Koshiro), I’m really hoping that I can make the trip work out again next year, even if it’s not as long.

This was my fourth Magfest, and I’m still astounded by how quickly the event has grown. In 2012 the event moved to the gigantic Gaylord National in National Harbor due to space constraints at the previous venue, and I couldn’t imagine how many people were there. This year didn’t seem that much different, but according to the organizers the jump was from 6000 to at least 9000! By comparison, in 2009…there were apparently only around 1350.

I’m not really much of a convention-goer; I wouldn’t even say I don’t like them, but for the most part I wouldn’t choose them over other things I could be doing with the time and money. Fortunately, Magfest lines up very closely with something else I like to do, and I’ve always had an amazing time, so it’s not usually a hard choice.  Hanging out is fun enough, but I’m amazed by how many people show up to play music and the amount of obscure or largely inaccessible games that people bring out every year. This year I was only exposed to a few games that really caught me off guard or that I just hadn’t gotten to play before, but they sure didn’t disappoint.


This is a picture of me, for some reason.

My “favorite” was Pac-Man VR, a hilarious little entry in the Virtual Reality craze of the 90s. You put on the helmet, stand inside a small enclosed ring on a platform so you don’t stumble off and hurt yourself or the precious machine, then get a small 1-handed grip for the controller. You’re Pac-Man, in first person, and the controller has one button, which is used to walk forward since the game would be impossible to control if you were forced to always move like in the original game. Also, since you can’t see ghosts that might be behind you (or outside your general field of vision period) the game doesn’t use a standard life system. Instead, you have a short amount of time to collect 1/4 of the dots on the field, and if you’re successful you get another equal period of time for the next 1/4, and so on. As such, the main challenge of the game after the adjustment period is over is not getting lost trying to make your way to the last few dots. The final one is marked with a nice arrow, but since you can’t see the overall layout it’s not easy to tell what the best way to get there is. I never saw anyone finish the first level so I’m not sure how the later levels change the formula up, but it was great to watch (both the player in the game, and the tv which had a separate line out) and play for a bit.

I also spent my traditional time at the Cave shmup setups, and while, between Magfest, emulation and the fantastic array of console ports, I’ve played nearly all of their games at least a bit, there was still one gap in my experience: Ibara Kuro. The original Ibara is a sequel of sorts to Battle Garegga, which I consider my favorite game of all time, but the “Black” version is essentially a remix by another programmer that changes up a lot of the mechanics. The game is quite a bit easier than the original if you play conservatively,  but trying to take advantage of the scoring mechanics tends to place you in between impenetrable walls of bullets. It’s not a very forgiving game and I didn’t have time to learn anything tricky, but I really enjoyed it nonetheless. I think I embarrassed my friend by singing the soundtrack constantly, but what else was I supposed to do? It was so loud in there you could barely hear it!

I also met some guys out there who play King of Fighters XIII and some other fighting games, and a few hours of KoF even when I hadn’t spent any time practicing recently was really nice. I had really forgotten how much I liked the game, and I’m always surprised by how much better I do against players who aren’t local for the most part. I feel like a lot of the guys in Denver have “figured me out” at this point, and it’s hard to beat them without coming up with new one-time gimmicks, but on the road the fact that I play less common characters and do have a lot of bad tricks can help a lot. In the end I didn’t ever actually play Zerp, who plays two of the same characters that I played for a long time, Elisabeth and Shen. That was kind of a disappointment, but he showed up a couple more times while I was at AGDQ since he also knows Josh and a couple other guys, and ended up introducing me to Chaos Code as well, which is a weird Taiwanese fighting game that mashes up some “anime” mechanics with more KoF-ish type basics. It definitely feels a bit weird and “cheap” (in the production sense) but it’s got some fun backgrounds and a weirdly addictive soundtrack, and I’m pretty excited to share its goofiness with some of the local players in Denver.

Finally, there was the man himself, Yuzo Koshiro. I honestly had no idea what his performance was going to be like, although it was billed as the first of two acts for a “dance party.” Sadly I was correctly left to assume this meant he wouldn’t be pulling out any of his more prog rock/metal/fusion type stuff, but the set certainly didn’t even begin to disappoint. He ended up DJ’ing about an hour of game music from across his career, starting with some Actraiser music before moving into some heavier and more “rave”-y stuff like Streets of Rage and Shinobi. The climax involved many tracks from the Wangan Midnight Maximum Tune arcade games, which I was especially excited about. I’m not at all familiar with the games, but I basically always love singing in game soundtracks and wish there was more.

I ended up having to cut Magfest short on both ends and would’ve really loved to stay even a little longer, but it was still just a great time and I didn’t regret going out to make sure I got a nice Sunday afternoon and evening with my sister’s family. After that, it was time to get to AGDQ.

I feel like a lot of the event actually speaks for itself if you watch the stream, although anyone who’s been to a major fighting game tournament or anything else knows how completely different being there is from being a long-distance spectator. The atmosphere was definitely a lot more subdued overall than a competitive event, but a lot of the same passion was there. Speedrunning may be less of a commitment per game in many cases (although the top players on Super Metroid and similar popular games have certainly poured a lot of their lives into their titles), but I was still struck by the enthusiasm with which players absorbed new tricks and asked for help on a staggering variety of runs and games.

I can’t say I felt that same drive, even in the middle of all this. For me learning about the basic speed tricks or concepts in a game is really fun, but it’s kind of like trying out combos for characters I don’t play in fighting games. It’s not fundamentals, or sometimes even useful, it’s just a fun little unique thing to try out.  I don’t want to rehearse a whole game in most cases, I just want to see something new and fresh. But it all ties into my greater understanding of the “performance” aspect of speedrunning and other superplays. It’s not just about knowing how the game works and playing it a lot; there’s a lot of important “setups” or other combo-esque sequences needed to really excel. The players know exactly what the desired outcome is, and the drama stems from human error and the unpredictability of programmed behavior. This doesn’t change my outlook towards these kinds of plays overall, but it does make the idea of doing them a lot more approachable now that I’ve started to understand the methods and scale of work better.

The other facet of performance is, of course, the player and their personality. For the most part, viewers respond best to people who are fun to watch along with their games. Whether these are people like Mr. K or Cosmo, who bring a warm friendliness to match their esoteric and detailed knowledge, someone like PJ, whose enthusiasm and temper hold steady through even the most stunningly painful games, or the over-the-top personalities like tri-hex and Flicky who often dominate the stream’s attention even if they’re not playing (for better or worse), the players are definitely a big draw not just for me, but the viewers at home. Of course, some carefully cultivated mystique can go a long way too, but it’s usually best for someone like Siglemic who’s got a game that’s sufficiently popular and flashy to entertain the viewers on its own.

I’ve been around the block a few times with these events, and I knew coming in that there wasn’t some kind of life-changing experience coming. But as my interest in fighting games has grown faster than my interest in speedruns, there’s still a lot for me to think about in terms of what this trip and the community is “worth” to me, and there’s not an easy answer to that. I’ll be back in the future, I’m sure, but with my sister moving out and a lot of other stuff unresolved in my life, I’m not quite sure when that future will be.


That is, if something is nothing.

“The better a game is for speedrunning, the better it is in general.”

Like I said, I could think of more than one thing this could mean, not to mention that not every game (including plenty that I knew the person who said it to me liked) is appropriate for speedrunning anyway. But it wasn’t just meant as some cliche “the more you put in, the more you get out” thing, nor a suggestion that games be “convenient” for speedrunners (though games in which the character moves unreasonably slowly or that feature long stretches of nothing interesting still don’t do so well by this mantra, perhaps rightly). The idea is that a good game features interesting, often varied ways to achieve goals that are satisfying to play out, and it’s one that I’ve come to agree with overall.

That may sound a bit much like a “a game is good if it’s fun”-style tautology, but the game we were talking about at the time was Sega’s turn-based military shooter Valkyria Chronicles. I’ve talked to a lot of people who were annoyed by the game’s overbearing focus on executing missions quickly, as the ranking you get at the end of each mission (which is tied into various kinds of rewards) is based entirely on how many turns it took you to achieve the goal. There’s basically no benefit whatsoever to avoiding casualties or inflicting tons upon the other side, which sounds great for speedrunning, right? Unfortunately, most missions have pretty basic “capture a point (surrounded by a few guys)” or “kill a major enemy” type goals, so as far as we could tell the optimal strategy on at least 3/4 of the levels in the game involved running a single unit (usually the main heroine, who has solid if not top damage against both single targets and closely grouped enemies) down a pretty obvious path (as there’s usually only a couple routes that offer any sort of cover) over a couple of turns, then assassinating the target(s) for the finale. I’m not sure if the game has evolved past that point as I haven’t really followed it, but if it hasn’t it’s pretty much the exact kind of thing we would consider “not that interesting.”

All of this is just a lead in for me to admit that Devil May Cry 3 isn’t the greatest game if your main goal is just to beat the game as fast as you can, either. It’s still a good game, easily the most enjoyable I’ve ever spent a significant amount of time speedrunning, and I absolutely consider it one of my favorites games in general. But like Valkyria, rushing through obscures many of the possible facets of playing the game. The difference is that DMC3 actually keeps track of those other aspects and rewards you for mastering them, making the most fun and rewarding way to play the game a lot more than just mashing through everything, and it does this all while presenting a variety of ways to play that offer different advantages in those elements.

The assessment the game gives you at the end of each playthrough of a mission rates you in five different categories:

  • Time. Faster play means a higher ranking, hopefully not surprisingly. You can still afford a certain number of mistakes and detours, and due to the other categories the latter are often necessary for a high composite ranking (which is a benefit, as after a few times learning where everything is it’s not difficult to move directly through objectives and play quickly). Either way, this is a category that rewards precise, purposeful play, like you commonly see in speedrunning.
  • Stylish Points are DMC3’s main way of assessing your offensive skills. Using your best one or two attacks over and over isn’t considered “stylish,” so you have to mix it up to build and maintain your stylish rating (and there’s benefit to taking a breather from attacking to taunt enemies, as well). The rating also decreases significantly if you get hit, so maneuvering around attacks is also important. The more enemies you kill with a high rating, and the longer you have it up there in general, the higher your stylish “score” is at the end of the mission.
  • Orbs are the game’s currency, obtained when enemies die and occasionally other hittable objects and places in each level. Enemies who die when your stylish bar is higher drop more orbs, which means this is fairly closely related to the previous criterion, but sometimes this isn’t enough to cover the spread and you’ll have to find some optional enemies to pick up some extras.
  • Damage Taken is the game’s primary defensive rating. Getting hit is bad. This is a fairly generous goal on lower difficulties, but on “Dante Must Die” the game expects you to avoid everything. The other side of this is that everything is possible to avoid getting hit by! That’s good.
  • Items primarily include healing items and other things that aren’t too helpful if you aren’t getting hit anyway. However, one of the items you can buy and use is a “Holy Water” which deals a huge amount of damage to nearby enemies. While you obviously don’t get Stylish rating for using a menu, there are many fights where you can only attack on certain “cycles” of the enemy’s pattern. By using Holy Water you can increase the damage you land in each cycle by a large amount and reduce the amount of chances you have to get hit (not to mention the enemies who become instantly stunned when you use it, thereby allowing you to hit them even more and faster). As a result, items are viewed as a pretty big crutch by the developers and using any type of consumable will raise your item score above 0 and prevent you from getting an “S” rank in the category.

The game averages your five rankings and gives you a overall rating. There are two kinds of rewards; first, gaining an “SS” on every mission on a given difficulty (which means earning an S in each of the 5 categories, as opposed to an “S” average gained from, for example, 4 S’s and one A) gives some cool unlocks like overpowered characters. Second, the game gives you a certain amount of bonus orbs after each mission, and the higher your rating is, the larger that bonus is. As a result, it’s hugely rewarding to earn those SS rankings on a few early missions in a speedrun, as more buying power helps you gain useful upgrades (such as the double jump) earlier. That said, it’s not really worth going out of your way to kill anything in a speedrun, obviously, but it does stand out as a bit of a tactical concern.

In fact, it almost sounds too risky for a marathon…get hit once on level 1 and you’re behind the curve for the rest of the game? I’m pretty sure this is why Flicky set the category to New Game +, to save all that menu/resource management time and distill the game to pure action for the viewers. To balance out we’re playing on Very Hard, where the enemies can still kinda put up a fight and don’t die immediately (DMD isn’t a good option since it’s got kind of a nasty tendency to make fights really long and stressful). Although it’s not the best speedrun possible, it should be a solidly fun run to play and watch, without much risk of run-ruining disasters. But with everything needed already purchased there’s not a lot of strategy left for Vergil, just practice and memorization. He can access all of his options at once, and while Dante can’t, he still benefits a lot since he has more purchases to make and strong abilities that don’t open up on a first play until later in the game (particularly Devil Trigger and his last weapon). I think that’s okay. I don’t really want this run to be more grinding than it already is.

Vergil’s inclusion as a playable character in the game was meant as a bonus for the “special edition” rerelease, so naturally he’s kind of overpowered. Like in Marvel vs Capcom 3, he’s not rounded like Dante, just really strong on offense with some teleporting shenanigans to boot. He has three weapons, and gets to carry them all the time: a fast, basic sword like Dante’s Rebellion; a large, slower katana; and the melee weapon that Dante gets in the main game after defeating Vergil for the second time. Unlike Dante, he doesn’t get any guns, but his summoned swords are way more powerful (at least in terms of killing things quickly) than any of them anyway. You can mash them constantly when you’re not getting hit, which is a huge damage increase against everything, especially during times when you otherwise have to back off or are too far away to do normal attacks. As a result, pretty much the only important purchases speed-wise for Vergil are Devil Trigger upgrades and sword damage increases, as his basic abilities are more than enough to carry everything else. Obviously if resource is not a concern anyway then you may as well buy most of the other upgrades and get a reasonable amount of health as well (but not too much! There’s a room on Mission 15 (out of 20) where it generally saves time to kill yourself as Vergil because of a jump that he can’t do reliably without killing all the enemies in the room), although ironically I think Rapid Slash is pretty useless in DMC3 and makes the game harder to play because it means you have to be less lazy in certain situations (due to the input it overrides).

Anyway, the main concern for Vergil playthroughs is physically mashing out summon swords while doing everything else you need to. In the Japanese community it’s customary to use turbo controllers to automatically handle this, but SDA doesn’t allow the use of features like that if they’re not on first-party controllers, so the main challenge for me has been to find a button layout that’s, if not comfortable (setting sword to a trigger is really the only option, but even then pressing it 1000 times in two hours is pretty severe), then at least possible to work with. In the end I went with a pretty default layout, but with summon swords on the left trigger (which usually activates Devil Trigger), and Devil Trigger on R2 (since there are also long periods of time where you have to hold that down, because it charges up an explosion that does tons of damage). So I can mash triggers with my left hand to throw swords and change weapons, and hold down triggers with my right hand to charge explosions and manage targets, while doing main movement and attacks with the analog stick and buttons. It all kinda works, just barely.

I still wish I had an arcade stick for this game though.

But while Vergil is a clear winner in the speed category, he’s really just an efficient killing machine and doesn’t have many opportunities for unnecessary styling. (Once again, this probably sounds familiar if you follow fighting games.) And if you’re not playing for raw speed, then DMC3 is all about style. Dante has a few different ability sets, which the game actually calls “styles,” that affect his game a lot more. One of them is the Trickster, which is weak offensively but can move effectively and makes avoiding damage a lot easier in some situations. On the other hand, you have the Swordmaster, who’s able to deal a lot more damage in most situations and pretty good at building up the Stylish meter. Or there’s Gunslinger if you really want to go crazy with your combos and don’t care about lousy damage and mobility. I know that I’ve barely scratched the surface of the game’s other possibilities, but even knowing they’re there amazes the hell out of me.

Unfortunately, there’s also a fourth main style, which completely breaks the game if you put in the work. I’m not really convinced that playing Royal Guard in a single-segment run is practical, and it’s not at all necessary to SS any mission, but for segmented speedruns I think it obviously wins because it takes out a lot of the back-and-forth flow of boss fights and the evasion involved in not getting killed and instead allows you to block attacks when they come in and use the damage you would’ve taken to hit the enemy in return. The timing is really tight, like most “parry” moves in fighting games, so this is hardly trivial, but even more than Vergil I find that it seems to remove a lot of the cool planning and crazy options that you get in favor of raw execution. I respect that, but I don’t have to like it.

But after all this ranting about how things work I’ve barely even gotten to what I like. Even when you’re just starting out and not pulling off insane action sequences, the game just feels amazing while you’re playing it, with the controls spot on and (aside from a couple of awkward angle shifts) no monkeying with the camera. Each style and weapon has cool options and certain disadvantages, and the game never has anything that feels paint-by-numbers. The combat is a crazy dreamscape where you can do more things than I could ever imagine, and to have that in a single-player framework that also provides a lot of different scenarios to play with is far too rare. (One game I may write about soon, which has a surprising amount of overlap with DMC3’s fans (at least in English-speaking regions), is similar, though!) The art design is fantastic, from the seedy back-alleys on stage 1 to the increasingly distorted world and creepy black-and-white hell of the later stages. The alternate costumes, and Devil Trigger designs by SMT series demon designer Kazuma Kaneko (most reminiscent of his work on Digital Devil Saga) are sick as hell, too. The industrial soundtrack hits all the right notes, especially the shit-talking battle themes. For as much as I like God Hand and Bayonetta, DMC3’s two closest relatives, neither of them have ever gotten me as amped as this game. And neither one lets you fight a demonic chess set with a sniper rifle, a magical ray gun, a three-sided ice nunchaku, and a bat-shooting-witch-scythe-guitar.

Really, it’s just that simple.

Demon’s Emblem: Path of Radiant Stories

Sometimes, when I go to read the SDA forum, I’ll see a post by someone who’s trying to break down a game they haven’t even played. Maybe they’ve watched most of it on youtube or read a guide written by some importer to try and suss out the mechanics before the English release, or maybe they just have a demo and are playing it over and over to try everything they can think of. I’ve certainly tried to do that myself with fighting games and Cave releases during the long wait before they hit consoles, but it’s never really crossed my mind in a speedrunning sense. I really try to walk into any non-arcade game I’m looking forward to as unspoiled as possible, because that first impression is a big deal. And even if I know how a game kinda looks at a high level before I get my first shot at it, I spend some time playing around and figuring out what feels best to me before really digging in. I’m sure that a lot of these runners take time to enjoy their first real playthrough as well (while of course at least taking mental notes on things that seem to be effective, if nothing else), but I’ve never picked up a game that I bought expecting I’d want to do serious speedruns on it. And that’s assuming you can even call my original attempts “serious”…

Fire Emblem: Path of Radiance

I’m sure I wasn’t the only one who played the Melee demo in stores and was blown away by Marth. “Who is this guy? Is he from some Zelda game?” With his shieldless sword stance and his Anime Hair (I didn’t know that’s what it was at the time, I was just twelve and thought it was cool) and his counter-attack (anyone who’s played against me or even knows my mains in most fighting games would know that I love counters) I knew right away I had my favorite character. At least, until I owned a copy years later and unlocked Mr. Game and Watch after countless hours. Anyway, I never had a GBA either so the Gamecube installment was my first crack at this mysterious series that I hadn’t gotten to play before! After playing through the game, I thought it was really cool and seemed like a pretty easy game to run. Before long I got hooked up with SDA’s resident Fire Emblem fan, Molotov. He’d played through the entire series, done or planned runs on most of the games, and was just generally cool to work with.

If you have a purely turn-based game, you’re basically looking at four distinct ways to improve your time:

  1. Offense. Essentially, the most effective way to defeat enemies and accomplish goals. In a game like Fire Emblem this often means blitzing an enemy boss or moving aggressively to a key capture point, while moving as few of your units as possible.
  2. Defense. Methods for avoiding dying (as your resources and power are often much lower than in a “regular” playthrough), or, better yet, reducing the amount of actions the enemies get to take.
  3. Randomness. Many of the FE games are heavily deterministic (if you make the same moves, the game will put out the same results), but as I recall PoR doesn’t do this and allows you to try the exact same suicidal moves repeatedly until they work in your favor. Additionally, there’s a significant benefit to aiming for specific stat gains, because FE assigns each unit unique chances to increase each stat on a level-up. The hero might have an 80% chance of gaining a strength point while a magic-type unit could have something more like 30%.
  4. Execution. Turn-based games are often derided because you rarely get combo-video type material, and while it’s true that there are often “all or nothing” scenarios in action games where being slightly off can make or break a run, the sheer amount of commands in a long game makes a huge difference over time.

As a newcomer, I viewed #1 as by far the most important. It certainly makes a bigger impact on your time than the others, but not to the extent that I treated it as. Not to mention that ignoring #3 and especially 4 makes the run just look sloppy and lazy, which is undesirable given the emphasis on a finished product at SDA.

In THIS game, the rules say I can do it as much as I want!

It’s even funnier to me in retrospect because #1 is almost completely trivial. The game gives you a small number (I think you get one guaranteed, and can find a few others) of items which you can use to teach a character a unique, class-based skill. For most characters this is a creative little move that’s entertaining but won’t change your tactics much. If you give this item to Ike, the hero, instead, his attack will randomly (and frequently) turn into a double attack, with the first hit stealing an equal amount of life from the enemy and the second ignoring their armor stat, which usually just straight up kills them unless it’s one of the last couple bosses. The self-healing and his overall power also makes this strategy pretty much immune to long-term bad luck. The enemies will try to slowly chip away at him while his revenge attacks take them out of the picture and cancel out the damage that he’s taken. Since this is a speedrun though, playing for overall okay luck isn’t really that good of an idea unless you’re doing the game in one shot. If you’re playing for keeps, you want to make sure that you get those big hits against big enemies and gain good level-ups to boot. Not one of my specialties, in this case.

The other major strategic element involves the Pegasus Rider character, Marcia. She can fly over most terrain features (rocks, trees, holes, water, etc.), as well as other units, friendly or not, and carry Ike at the same time! Aggressive use often allows him to reach his targets several turns quicker than walking, but she also brings in one of the crucial defense trade-offs. Having a passenger weighs down mounted characters, lowering their effective stats. This makes it harder for them to score damage and causes them to get hit for more damage more frequently. Marcia isn’t a terribly strong character to start with, and if she dies before the last mission you have to reset (she’s just too good, all the way to the end), so using her effectively reintroduces some of the luck mitigated by Super Ike.

All that said, I would pretty much sum up my run with a simple statistic: I recorded about 24 hours of footage, for a run that finished under 4. Don’t follow my example.

Radiata Stories

As always, we got renders instead of the sweet concept art for the box.

Pretty much the last time I had a chance to talk to my man Enhasa, he told me, “the better a game is for speedrunning, the better it is in general.” I had to argue a bit to suss out what he was trying to get at, but that’s not important right now. For now, all I really have to say is that as much as I love Radiata Stories, it’s a pretty goddamned terrible game. I’ve thought about this over and over through the past several years, and to date I have not come across another game in which speedrunning forces you to almost completely ignore the best parts of the game and spend the entire time suffering through the bad parts.

I guess depending on who you talk to, maybe the GTA games count? It’s the same deal here, although obviously with a much more JRPG flavor. Kicking things is the primary way of interacting with the world; it opens chests, reveals other items hidden in piles of rubbish or under rocks, and lets you pick fights with almost anyone you want. The game gives you a fairly large world to check out, filled up with a large number of NPCs who all lead their own little lives. The heart of the game is trying to hang out with these people and find out what they want so that they’ll become your “friend” and join you in battle. (If that sounds cool to you, play this game.) The game has a clock which dictates what each one does every day, so you can basically stalk them until you find out what you need to do, then try to do it. A few characters have pretty cool questlines, and my favorite plays with the system and your expectations in a really clever way.

Watch closely…

But none of that matters, because out of about 170 characters in the game, the speedrun recruits less than ten. Hell yeah! On top of that, you can’t skip cutscenes, which basically instantly adds over an hour of holding O and mashing X to skip text as fast as possible throughout the game. The scenes themselves are generally pretty funny and engaging…the first time. But as a runner you get to get sick of them instead!

The most actively annoying system in speedrunning the game involves avoiding encounters. Like most RPGs from the past decade enemies appear on the map as you walk around, and you can try to move around them or at least avoid being put in a disadvantage by being ambushed. In most of those games this is really easy for the most part, but Radiata’s dungeons and overworld aren’t comprised of “field” areas, and instead, a spindly series of trails. As a result, you don’t have a lot of space to maneuver around them and it’s often difficult to even pull off bait and switch tactics to work around them. Obviously, there are a lot of runs on things like older Final Fantasy games where you can’t avoid “random” battles at all, but the fact that in this game you can as long as you get lucky means that you should play until you get really lucky. Or until this happens and you’re too mindfucked to continue for the day:

The battle system is really mediocre too, like a slow and super dumbed-down version of Star Ocean 2’s free-moving mashy action. You only ever get to control Jack, the main character, who has a couple of basic defensive moves and a prepared attack string plus a special move that uses your meter. You can give orders and try using formation attacks with your party members, but their effectiveness is mostly tied to how strong those party members are in the first place and otherwise just waste your meter that you could be using to kill things. Character building and equipment are also ridiculously limited so while the game is super easy playing normally there’s not much to abuse like in other tri-Ace games, where even at low levels you can get game-breaking damage if you know how. The only way to really make fights fast and easy is to recruit the best characters, and regardless of which story path you’re on, they aren’t in your small pool of applicants. You’re lucky if your allies don’t kill themselves trying to play hero too much, since often you need them to be alive so that they get hit by supers instead of Jack (whose death causes game over).

Like I mentioned before, I didn’t have to figure out much of the overall strategy in this game. A few weeks after I first beat the game, Molotov found me a Japanese Geocities page created by “ogu_dai”, in which he detailed his checkpoint times and overall strategies for each of the separate story paths. His stuff was designed for single-segment play, so it was a bit more luckproof than good segmenting strategies, but most of it was tremendously useful and saved us at least 100 hours worth of testing. In terms of resource management, there’s basically one bottleneck on the “Fairy” storyline that I was running. You want to buy the best weapon in the first half of the game as soon as possible. This basically means taking a detour in an early dungeon during the part of the game where you don’t have any choice of what to do, to get the best item in the game, so you can sell it.

By “best” I mean that it locks Jack’s non-battle movement speed at the medium range where he does a funny “power walk” animation, and does nothing else.

If that sounds cool to you, play this game.

Combined with the small amount of money you get otherwise during the opening of the game, you can head to the weapon shop as soon as the game lets you and buy a big two-handed sword! This triples your damage, which basically still means it’s godawful until the game hands you the best one-hander you can get outside of the bonus dungeon early in the second half.

There are four other crucial purchases in the game, but once you get going money itself is hardly an issue. One is the command to have a party member heal you. They are not always intelligent enough to do this on their own. The second is a large stock of “Flee Balls.” Like other RPGs, fighting non-bosses to level up is almost a complete waste of time. Even when a battle loads because you couldn’t avoid getting into it you lose precious seconds, and to make matters worse there’s no “retreat” command like in other games. You have to buy these items and use them whenever you get hosed over just to cut your losses. Third, you need items to poison the enemies that you do have to fight. Almost every boss is vulnerable to poison, and once an enemy is poisoned, it doesn’t go away. The poison deals damage faster than your low-level sword hits do anyway, and doesn’t build the enemy meter while doing so (like hitting or being hit by bosses does). If you can get both that’s awesome, but in many cases it’s not safe enough to do so, either because the enemy super will unavoidably kill you, or because they can just kill you in one combo anyway.

A nice fat stack of Bison dollars.

But, if you’re really outnumbered, the poison doesn’t help that much, as you put yourself at risk just trying to use the items. It’s also difficult to hit your target because you literally “throw” the poison at them, so if someone else is in the way it will never hit. That’s where the last purchase comes in, and despite it being the most broken and useful tactic in the game, it was the last one we found out about. Molotov and I investigated it on our own because ogu_dai never had to use it. (He leveled up more and actually recruited a semi-useful character instead.) There’s a command called “Earthquake” that you can buy late in the game. If you have at least one living party member, all of you fall on the ground, causing it to shake so much that all of the enemies are knocked over. Your team gets up first, so this stunlocks anyone you want, while dealing damage similar to poison (and stacking with poison if you can hit it, obviously), but it costs 4 meter per use. Every time it touches an enemy though, you gain back 1 meter, so if you hit four enemies at once, it’s an infinite. It’s so good against the hardest boss on this path that it’s actually worth using healing items on his weaker cronies just so they won’t die and force you to stop using the infinite on him. Sadly, I don’t have a video, since this is the point where I gave up on recording the run since I couldn’t get the strategy down. Now it would be easy, really the least of my worries starting over.

I know that sounds awesome, and that it is awesome, but it doesn’t change the fact that this game is really annoying and broken for speedrunning, with all of the other ways being much less interesting and hilarious. It’s unfortunate.

Demon’s Souls

Don’t have nearly as much to say about this one, as I wasn’t involved in any sort of strategic development and just copied what other people had already figured out until I gave up on the game before really learning it anyway. Basically, the key feature of the Souls games is the death system. It’s kind of old-school in that it puts you back in a designated respawn point at the beginning of the current level, but you also drop all of your “souls” (which serve as the money used to buy both items and stat increases) on the spot where you were killed (or a short distance away if you jaunted into some kind of bottomless pit). You can also only have one puddle of souls on the ground at once. The idea is that you’ll figure out what you did wrong and find a way to not die in the exact same spot next time, and thereby succeed in regaining your souls (as well as keep the ones you collected working your way back to where you died in the first place). In order to make this actually dramatic, the game autosaves pretty much all the time and prevents you from backing up your save to cheat death.

“I play shmups and other games where you can’t go back all the time, so this’ll be a piece of cake, right?”

Most accurate fan art of all time.


In MAME you can make save states. All of the 360 Cave ports have a training mode where you can start on any stage (and usually any boss) with perfectly calibrated parameters that match your usual/desired progress, which is basically as good as a savestate aside from the more limited starting points. Fighting games have training modes where you can try almost anything you can come up with (unless you need the dummy to do something that you can’t figure out how to record) over and over. Even beatmania IIDX has a practice mode where you can break down individual measures at ridiculously slow speed to see what’s going on.

In Demon’s Souls, you can’t reload your save to repeat a level with the same conditions. If you try to go back, you’ll have more stuff and be stronger. The bosses are dead permanently.

Most of Demon’s Souls is not a hard game when you know the easiest routes and how to build your character. Out of the hour-long run, maybe 10-15 minutes of the game at most has actual ways for you to screw up and lose a lot of time. Every time you want to practice those parts, you have to play most of the other 50 minutes too. Obviously, once it’s crunch time before the marathon, you want to get used to just going through the whole game no matter what happens, but I gave up before I could get to that point.

I’m not going to play Demon’s Souls again. I still haven’t even played the sequel, which I was looking forward to before all this happened. I probably will someday, but I’ve held contempt like this for much longer than a year before. It’s just not time.