Moglar5k (Magolor9000 on Youtube) holds the world record score in Ketsui: Kizuna Jigoku Tachi, Cave’s 2003 dystopian corporate-military bullet hell. Legendary for its focus on aggressively dispatching enemies with proximity chaining and lock-on shot, as well as its insanely difficult Omote and Ura 2nd loops, Ketsui truly demands excellence of play. In today’s interview, Moglar discusses his own practice methods, enjoying a game for thousands of hours, and how to one-credit clear 2-loop shmups.
April 2021
CAT: Hey Moglar! I was wondering if you have any wisdom to impart on learning skills, at Ketsui or in general. I’ve been making some inroads lately at Rolling Gunner, but I feel like as a general thing, deeply learning a new skill is something I find very difficult. I have drawn for years and years, and yet I sometimes feel way behind people who took it up two years ago, for example.
MOG: Think a lot, both when you are playing/drawing/whatever and when you aren’t as well. When you are playing, look out for things that you could be doing better and try to get a better understanding of everything. When you aren’t playing, you can consider what you want to do or learn next, come up with new ideas, and think about upcoming goals and how you can go about achieving them.
Don’t force yourself to use the absolute fastest seeming practice methods you can think of if it’s not as fun. I don’t think there are a lot of methods that will be huge killers to progress speed, so long as you don’t find yourself using one of them just stick with what’s fun. The less fun you are having the worse you will perform since you won’t put in as good of an effort. In the end you won’t be learning faster and will have lowered motivation to some extent.
The main ways I’ve noticed that people really screw themselves with progress rate is by either not challenging themselves enough and not learning new things, or the opposite, doing things that are way beyond what they are capable of and ending up with completely unacceptable consistency. I think the best is to try to be in the middle of those as much as you can, although as long as you aren’t super far on one end you should always keep getting better the more time you put in.
February 2023
CAT: Before achieving the Ketsui world record you tackled Touhou, innovated speedrunning tech in Hollow Knight, and uploaded a sub-20 minute Super Meat Boy any% run to Youtube—which you called “bad.” Have you always pushed games to their limit? Or was there a certain game that started this skill snowball rolling?
Eh… maybe a bit much. You can answer that if you want. But I guess what I’m wondering is, you had a pretty solid thing going, speedrunning platformers like Celeste and Hollow Knight. Why switch to shmups? And why Ketsui?
MOG: Well, I first started watching speedruns when I was 11ish. For a while I was mainly just watching a lot of them. I had a short period of this with shmups as well. I never streamed til I was 13 but I messed around with some speed stuff a bit before that. There were a couple of noteworthy things; one was doing strat hunting in Toad’s Treasure Tracker. I would mess around with that game and tell the one guy running the game at the time some of my findings in his Twitch chat.
The other was Kirby Triple Deluxe. I spent a good amount of time making a route for the game. You have to collect a certain amount of collectibles in each world to open the boss, so that was the main routing difficulty. After routing everything I did a couple runs. The first one got a 2:15-2:16 I believe, and the 2nd was a very slightly better run at 2:15. The best time on speedrun.com was 2:18, so I beat that, which is pretty cool. Although the best now is 1:53, so it wouldn’t be too good nowadays.
At 12 I got S rank in Splatoon, and was one of the first people to beat the now very well-known Pangeapanga’s Skyzo level in Super Mario Maker, beating it a couple hours after upload, if I remember correctly. Never ended up playing any of his future levels too much, but I messed around trying to set good times on a bunch of random levels in that game.
A few months after I turned 13 the punch card glitch was found in Undertale. I really loved that game when it came out, but only watched speedruns of it because it didn’t seem interesting enough to run myself. But when that glitch was found I immediately started running the game. Undertale is still definitely the most impressive thing I’ve done in speedrunning, and it was the first time I ever put out any video of my play.
So, the answer to that question would be that yes, I was pretty interested in pushing games and challenging myself in them from quite a young age. But I can’t say always—I wasn’t pushing games when I was 3.
I started playing and watching Touhou a bit in early-mid 2017, since I knew they were supposed to be hard bullet hell games, which sounded pretty cool. Later that year the Touhou player Yatsuzume hosted Satis (also known as semenfairy, cooldood69, etc.) after a stream. He was playing Crimzon Clover Unlimited, which was my first introduction to arcade shmups. Afterwards for a bit Crimzon Clover was the main thing I was playing, but Ketsui ended up being #2. Ketsui had the coolest looking score runs which probably sparked some interest, but it was also just the most fun to play out of everything. Bit after mid-2018 Ketsui became the main thing, and I just kept going with it. It’s an extremely fun game that always left me with tons of things to learn and improve upon.
In the future I’ve been wondering how much I’ll play shmups vs. speedrunning. I don’t think there are really that many shmups with scoring that suit my taste honestly, but Progear, Dragon Blaze, Esprade, and Space Bomber at least all seem pretty likely for me to enjoy a lot.
Another thing is that Hollow Knight is ridiculously popular now. If I end up getting into Hollow Knight Silksong a lot, it wouldn’t be surprising if I could get enough views to make lots of money from it. I’d imagine it would be harder to retain good viewers with shmups vs. some other speedruns, so that could sway what I do potentially.
Shmups can also be really boring socially compared to speedrunning, which is definitely another factor that could keep me a bit more on the speedrunning side in the future. Also, I guess you say I “had a pretty good thing going,” but it’s not like I had a lot of viewers. There wasn’t really a monetary incentive to stick to it, so it’s not like there was anything that would make me stay with speedrunning instead of playing shmups. I just wanted to play fun games.
Although on the other side of things, I kind of wonder if most speedruns might feel a bit too simple after playing Ketsui for so long. Hopefully not, though.
CAT: Previously you told me the main impediments to skill gain are playing games that are too easy or too hard. So I find it very fitting to learn that some of your first speedrun experience was with Captain Toad and Kirby Triple Deluxe. These are super easy games the way most people play them. Some gamers even have an attitude of “I won’t play that, that’s a baby game.” But you were playing them in a super hard way.
Even so, I keep thinking in my head that there are “best” ways to practice, like some secret sauce or golden ratio between stage practice and full runs or something. But you’re basically saying “yeah, all practice is pretty good, so don’t worry about it and just do what feels fun.” In the end, does it all just come down to putting a lot of hours in a certain game?
MOG: It’s mostly about putting in the time, but there is more to it than that. For scoring I think it’s pretty important to have a good balance of learning new strategies while not trying to do more than you can handle at the time. Whenever I see people improving really slowly, they’re usually on one extreme end here. I’m not sure if doing more than you can handle is as bad as barely learning at all long term, although it seems to be the far more common of the two. When I was trying to improve my Ura scores in Ketsui, I would try to learn a couple new strategies at a time and get them down really well, so that I could improve my routes by a couple million while only slightly reducing my odds at getting into the loop.
If someone is just playing survival for gaining overall shmup skill, playing games that aren’t harder than their current hardest clear, or playing stuff that is way too hard would slow down your speed I think. Honestly the best way to improve here is probably just playing a game that is really hard to 2-ALL from pretty early and sticking with it. The game will automate the whole process I mentioned earlier. You will be continuously pushed by the game without it ever completely destroying you, and you don’t need to put any thought into game choice.
Ketsui and Dodonpachi Daioujou are probably especially good choices—going for Omote and then stepping up to Ura, or doing Daioujou Black Label 2-ALL and then doing White Label. By the time you’re finished you’ll have done one of the toughest clears in the genre. But if you did this then most games wouldn’t really give you a satisfying challenge anymore. So, enjoyment wise, I don’t think it would make sense to force yourself to do Ketsui or Daioujou just for skill improvement.
CAT: Hmmm… I think I get where you’re coming from. But I can already hear the gnashing of teeth if I don’t follow up on that answer. You’re saying basically to stick to the middle balance by pushing yourself just past where you currently are. And then you keep pushing, and bit by bit your skill floor will rise until it’s well beyond where it was before. At the same time I think a lot of people will say, “Hey wait a damned minute, Moglar. You’re telling me to take things easy and just push a bit at a time, and then you tell me to start 2-ALLing Daioujou?” It sounds a bit like saying they should grind slime knights, but actually what they want to do is beeline straight for the demilich. Can you clarify this approach a bit? For somebody already struggling, how should they approach these games—some of Cave’s hardest?
MOG: I mean, you start by getting to 1-3, 1-5, 1-ALLing, looping, then getting to 2-3, then 2-5, etc. You aren’t just going for the 2-ALL immediately. You are going for lower goals that are reasonable enough for newer players and then just continuing until harder goals happen.
It’s not really comparable at all to playing Futari Ultra immediately or something. That mode demands extremely tight gameplay from the get go; Daioujou and Ketsui don’t do that at all. It’s more like playing every Futari mode in difficulty order and then starting Ultra. That would be a rougher leap than the other games would make you do even going from Black Label/Omote to White Label/Ura, but you get the point still.
CAT: Right. I think a lot of newer players rush themselves. They compress a 70-hour 1-ALL journey into 30 minutes, and beat themselves up for not being able to clear after a handful of runs. Learning takes time. And that requires some amount of self-direction of course—I don’t think most modern games foster self-direction at all. Or society at large for that matter. So that’s its own vital learning experience, unless you want to hire a dominatrix to whip your ass until you clear Daioujou. Which… I guess DaringSpino gets that same experience grinding turtle in Daioujou White Label, without spending a dime.
For players looking to tackle a second loop, or Futari Ultra, or Daifukkatsu Black Label Strong, do you have any advice beyond tackling the loop a few stages at a time? You’ve remarked that Ketsui Ura loop was way, way, way, way, way harder than Omote. Did you do anything extra to prepare before jumping in?
MOG: For Ketsui Ura, I put some time into learning the loop right after getting my first Omote clear, but I ended up just going back to improving my Omote score after a bit. The Ura loop was really rough on my first go, but after getting 405 million Omote (personal best was 358m at the time of my first clear) it was more reasonable, and I finished out getting my first Ura 2-1 start clear after a good bit of practice.
For Ketsui, if a survival player is struggling a lot with Ura at first, I’d recommend going for an Omote clear with 2 lives remaining and then try again. If it’s still a bit too much you could continue Omote til you have 3-4 lives remaining. But if you’re having more fun getting pounded by Ura than you would be having with Omote, then just continue with it.
You could try switching to a different game which seems a bit too tough, or by doing your current hardest clear with 1-2 extra lives remaining as preparation. Although if skillset isn’t shared too much between the two games, then maybe it won’t help much.
CAT: Last night I had this practice session where everything was just… awful. Straight up repeated dying to everything in stage 3, almost no score, no cash-ins whatsoever. Confusingly, bewilderingly bad. I definitely swore at my monitor, a lot. I get this impression from your interviews that you’re completely stoic when you play. On Ketsui stage 2 boss grenades: “I don’t remember those being a problem.” On being asked if you ever have difficulty recovering after death: “No, not really.” Am I just projecting, or are you able to remain totally calm even if a world record or personal best crumbles at the last minute?
MOG: It doesn’t bother me much at all if a world record pace run dies near the end. Getting a near world record run is way, way better than dying in the first two stages over and over or whatever. I can definitely get frustrated though, and maybe like once a month or so get really mad.
CAT: Do you have anything more to say about strat creativity? Maybe that’s one thing that’s especially hard to talk about. Probably it comes with having played multiple games at a high level and analyzing others’ runs, and then sort of learning the weak spots at which a game, or even all games, can be attacked. But for instance, with Futari God I watched Iconoclast’s replay and learned that I could rapid shot and double bomb the stage 2 turtle midboss for a quick kill leading to a very lucrative cash-in bonus wave of lobsters and mantas. That’s not something that I think would have ever occurred to me… even in all my practice to “do something weird and different,” I was focused on survival and misdirecting shots. I didn’t even know what I didn’t know. And then afterwards I thought, “well, obviously if you cancel the boss, Cave has to have put SOMETHING there.” It was pretty humbling—that this idea had occurred to somebody, and was maybe even an obvious thing for them to try.
MOG: Figuring out new, more optimized strategies is about 1) Understanding a game’s mechanics 2) Identifying opportunities where they could be used to get more score than current strategies 3) Finding out if it works and/or how to make it work, and 4) Figuring out how to make it doable with high consistency if it works.
It depends on the game and the specific strategy when it comes to which parts will be harder or easier here. I think for Ketsui, identifying opportunities for higher scoring was the main difficulty. Identifying opportunities and making them work is generally going to be hardest, but if they’re pretty obvious then achieving consistency will be the hard part, since the reason why no one was doing a strat before would be the difficulty. I guess you could argue to what extent consistency counts as “strat finding,” but it’s still gameplay innovation anyway. The better you understand a game’s mechanics the more likely you’ll be able to do the rest, so I’m not really sure how to assign a comparative difficulty rating to that. It feels a bit more like a prerequisite.
If a game has been played a lot, you’ll generally have to play longer before finding anything new. Not only will the strats be harder, but any new strats you find will probably not be huge gamechanging scoregains. You will probably be more worried about getting down what is already known before trying to figure out completely new strategies. I didn’t really start finding new stuff until I was at 500m in Ketsui.
Strat and routing skills will be regularly put to the test throughout playing a shmup, although that will generally be vastly simpler than figuring out a new higher scoring strategy. Routing for safety and finding tricks to make things easier is also a thing, but I find new discoveries to be the most interesting and rewarding.
CAT: How did you feel when you finally nailed the world record?
MOG: Well, probably top 2 moments in my life, the other likely being the 551m. When I got the world record I’m pretty sure I had already been up for 10+ hours, but if I remember correctly, I ended up staying awake for another 24 hours after that just enjoying the high.
I am unsure if it was as good as the 551m though. 540m was a major goal of mine due to the top PS4 score being 537m at the time, and I saw it as putting my Ketsui score on a bit of a higher level compared to the best western scores back then. The 551m was less expected than the world record. It happened earlier than expected and the world record happened later than expected, so this could have caused the 551m to be a bit more exciting.
I don’t think I stayed up for 24 hours after the 551m, but I’m pretty sure the high stuck around for like 3 days somehow. You would expect it to mostly go away after going to sleep, but I guess that wasn’t the case that time.
CAT: Wow. I’m really glad to hear that you’ve sort of… loved Ketsui the entire time? Gus finds practicing Futari really boring, and Kayar treats Saidaioujou like a job. So it’s comforting to hear from a top level player who continually enjoys the challenge and learning experience of their chosen game.
MOG: I’d say I have loved Ketsui for pretty much the entire time, yeah. I think it did get a bit less fun over time, but on the other hand it felt more rewarding later on so that mostly made up for it I think.
Although, at this point there doesn’t really feel like there is a whole lot to learn or improve at anymore, so ever since the world record it kinda has been a bit boring honestly, which is probably a big part of why I’ve not played much since. Still, I think I could probably continue enjoying it at least a bit if I focus on trying to get good loop 1 scores. Focusing purely on the no-miss is really not fun even if that would be the main goal either way.
Since I have barely played for so long, though, the motivation is quite low. So we will see if that no-miss ever ends up happening.
CAT: You’ve set the Ketsui world record at a stratospheric 599.7 million. That’s a masterful achievement. It’s also… tantalizingly close to 600 million. You seem ready to move onto other games—does it bother you to leave that last scoring milestone unmet?
MOG: I don’t really care about 600m at all after getting a score so close to it. Not having the no-miss bothers me a bit though. But as time goes on, I care less and less.
Moglar’s Estimated Survival and Scoring Milestones
70 hours: 1-ALL (111 million pts)
180 hours: First Omote entry (191m loop 1, 211m total)
350-400 hours: Omote 2-ALL (353m)
600 hours: Omote 405m with 260m loop 1 (And first try on Ura loop and IKD Special)
850 hours: Ura 2-ALL 441m
950 hours: Ura 511m and Omote 441m
1300-1350 hours: Ura 551m
1400-1450 hours: Loop 1 290m
1550-1600 hours: Ura 574m
1850 hours: World record 599m
2000 hours currently played
Catstronaut Loves Games 