Vykan_12

First new WIP in 5 months, hopefully Braster will continue the TAS in a more timely manner :/

I decided to watch all the videos in one go at 1.5 to 3x speed. You know you're playing a hard difficulty when it take THREE units to take out a HEALER. I don't really like that the prevailling theme in this run is to have 9302489320789 units chip away at enemies, often with shoddy accuracy javelins until multiple ridersbanes show up (even then), but that's not your fault, more just bad gameplay design. Not exactly my cup of tea but I respect the fact that you'd actually tough out a run like this. P.S. Can't wait for that ballista + Wrys 1 turn on Medeus ;)

The test run is finished, check the OP for the playlist link. The sum of the video times comes out to roughly 39 minutes, so I'd expect an optimized run to achieve a time a little under 45 minutes since I used savestates to iron out mistakes in the test videos. Here are some notes I made about resource allocation: My main concern for single segment runs is that MU can't even be slightly below average on str or def at many key points in the game. My other concern is hitrates, as there are a lot of crucial attacks from my units that only yield an 85ish display hit. The str & hit issue I could actually get around by making more forges in earlygame. Since I managed the rout with only 1 again staff, that frees up a thief staff to acquire a bullion (L) in chapter 10. There are a couple other minor planning mistakes to iron out such as giving Palla a silver lance, only to have her give it back to convoy, unused, but I'll be sure to correct those. I'll probably need a solid week of practice before attempting full-game runs since the game's memorization burden is insane.

That C24 video was awesome! I now pronounce you king of the rescue-drops.

If it's any help, this is the best turn count I've ever achieved in C24 (skip to 5:24): Do you have a video of your 8 turn strat? Only found an old 12 turn on your channel, so it must be unlisted. Would save a lot of time seeing what approach you use vs trying to come up with something from scratch. And do you care about plausible RNG or could I rig as many crits as I need to get the theoretical low turn count?

Couple new videos. Prologue-VIII New Strat Chapter 9 Chapter 10 Chapter 10x Chapter 11 I also made a lot of minor adjustments to my previous strats such as selling Arran's Bullion in C2 in order to be able to afford door keys in C5. I also devised an alternate strat in C6 using Yumina instead of Bishop!Sirius which costs a turn but makes the base much simpler. I'll be sure to record that next time. I must say, this game is ridiculously fast paced, which makes it very stressful to speedrun. There are no breaks whatsoever to have time to think or relax other than acquiring a star shard. Item management is also a nightmare since you only have 5 slots instead of Radiant Dawn's 7, enemies are constantly dropping items and a couple slots are lost for things like vulneraries, star shards, pure water, silver/vip card, etc. I'm glad I've almost reached the warpskip portion of the game because this game is very challenging to rout. It's almost as though the enemies are spaced out in such a way as to prevent you from using cheap strats, and it exponentially increases the number of possibilities to consider just in how you space your own units. I'm a bit stuck on chapter 12 since Marth gets ambushed by 3 wyverns and Darros' posse of enemies. The strat I've worked out so far is to give MU 2 turns to clear the top-middle of the map before moving Marth, and have Palla take out the Pisces dragon so MU can take on the top-left portion of the map with Marth. But it's easier said than done when both Marth and Palla are 3HKOed by pretty much everything on the map. I actually have to give Marth, MU AND Palla a specific star shard (+2 spd, str and def, respectively) to make the strat work. It turns out that it's faster to kill Elrean than to recruit him since you don't have to bother with the base at all to prepare Wendell.

I suppose I should clarify a few things. The final run shall be a single segment real-time run of normal mode, no DLC or lunatic clear bonuses (like full male reclassing). It's done on an english patch on emulator using a USB gamepad for controls so it wouldn't set any "official" records. For that I'd have to buy a DS and a Japanese cartridge, which would be a huge pain and barely worth it. I might have a crack at TASing the game too if I can fully figure out how the RNG works. Hmm, that's more plausible than I thought, didn't take into consideration the staff rank boos from promotion. The combination of warp, rescue, thief and barrier should easily amount to over 15 uses, although sometimes I need multiple staff users in the same map playing different roles. I'd actually be more worried about getting Yumina 6 level-ups in order to promote her in time. It seems warp, rescue and thief each give 50 exp, minus a small deduction as levels increase, up to 5 once level 18 is reached. 13 staff users would give 6.5 levels, more than enough to cover the gradual exp degradation factor. Yeah, guess it's feasible after all. Way to complicate my life with more planning decisions. Also, does anyone know if it's possible to 2 turn C10 without using rescue? I could only manage a 3 turn since Feena has trouble crossing water.

When I was working on a Tales of Phantasia TAS, a guy named Amaurea actually dissambled the rom in order to determine the exact formulas the game uses for things like damage, RNG, etc. I wonder if the same thing could be done to determine AI targetting priority.

I recorded Prologue to Chapter 8 with a time of 18:58 so far. Each chapter averages 1-2 minutes and I'm surprised the run can manage to be this fast-paced before warp even shows up. Links are below and I included the playlist link in the original post of the thread. Prologue Chapter 1 Chapter 2 Chapter 3 Chapter 3x Chapter 4 Chapter 5 Chapter 6 Chapter 6x Chapter 7 Chapter 8 It looks like I'm going to have to devote one hammerne use to the rescue staff, although I should manage enough warps off the remaining 2 hammerne charges. Depending on how chapters 11-14 go, I'm not even sure I have to invest that much in Palla since MU is managing fine on his own so far. For chapter 9, I haven't looked into it in great detail but I'm undecided as to whether I want to recruit Etzel or not. He could be really useful in C11 if the mag user who rescues Marth ever gets attacked along the way. I watched through them all and it gave me some great ideas, particularly that it was viable to skip Julian. I have a set of notes regarding that below in the spoiler tag. The main difference between speedrunning and LTC is that I won't have time to get anyone from C to B rank staves, which means no again staff. All that changes is having to use the warp staff more often.

You should re-record Prologue to Chapter 4 since the framerate in those videos are awful. Look forward to watching through your H2 and H3 LTC videos regardless.

Test Run Youtube Playlist I'm working on a speedrun of FE12, normal mode classic difficulty. Technically casual would be faster but it just feels sloppy letting units die knowing fully well you'll get them back the very next chapter, and cheapens a lot of the strategy in the game. I'm also going for the good ending since beating the game in C20 after fighting Harden is really lame. Getting all the star shards is very time consuming, but very much worth it. I haven't been playing this game that long, in fact I've only completed the game twice, so I have a lot of questions, some based on gameplay and others based on stylistic choices. -Would you guys be opposed to using DLC to get a faster time? Maturity drop on Marth or MU and rainbow potions would make the run a lot easier, especially in the earlygame. Personally, I'm all for anything that reduces the amount of RNG I have to deal with in a single segment run. -Since customizing MU counts against your in-game time, I just go with the defaults. Anyone particularly opposed to this? Having an MU named Chris is kind of lame but a necessary evil I guess. -How would you feel about making enemy phase skips and just enemy movement skips separate categories? Being able to skip enemy phase has an enormous effect on speedrun strategy, such as avoiding player phase attacks to minimize the number of level-up screens ever seen. At the moment I am using enemy phase skip but am interested in both approaches. -What's the best way to go about reclassing MU, and various other key units? Generally I have MU as a cavalier, then switch to a DracoKnight upon promotion, and only switch back to Paladin in bow-heavy maps. I noticed MagicBarrier reclassing his MU to berzerker in his reverse lunatic playthrough, any particular reason he does that? He also mentioned swordmaster being better than hero in one of his video descriptions, why is that? -Since you can get a lot of arms scrolls throughout the run, and since DracoKnights can use axes, is it worth getting MU from E axes to B or A rank? -My strategy for dealing with Gharnef is to recruit Katarina, then give her 3 spirit dusts & 2 speedwings, some of which is bought in the C21 secret shop. This allows her to ORKO with starlight, and thus 1 turn the map with 2 warps. Is there perhaps a better way of handling Gharnef I'm not aware of? Speaking of warps: -Almost every map I warpskip requires 2 warps because Marth doesn't have enough def to survive an enemy phase, especially when Glower sorcerors are in range. Is there some clever way around this? I can load 3 dracoshields & seraph robes onto Marth after visiting the C21 secret shop, but that's already very late into the game. There are also some maps that require 3 warps to 1 turn because the throne is hidden until I open a door. -How would you allocate funds for weapons & forges? Currently I'm planning the following (roughly): +2 mt rapier (Chapter 1) +4 mt steel sword (Chapter 2, for Marth) Buy a couple steel lances & javelins (Chapter 3) Buy 3 dragonpikes & wyrmslayers in the secret shop, some of which will get +2 Mt forged (Chapter 11) Buy a bunch of stat boosters in the secred shop (Chapter 21) I don't really need to buy silver lances since so many are dropped throughout the game, same goes for pure water and door keys. On my last test run funds were really tight, so I plan to thief staff for a couple Bullion (L)s. -How does the DS touchscreen compare to using the regular controls? Since I am playing on emulator I am using a USB gamepad to control everything. -How much is known about the game's RNG? I know it's based on the clock in some way. That's all I can think of for now. I have a set of notes I update frequently here. The beginning part in normal font are what I've planned out move by move up until chapter 6. The remaining text in green are more general observations from two test runs, stuff like what each village contains and when reinforcements show up. I'll try and get some videos up soon so you can give me some more meaningful feedback. So far I've managed to beat all 8 prologue chapters in ~5:30 and estimate a full run would take about an hour.

+1 for Shadow Complex. One of the best Xbox Live games ever made.

If you worry about the ethics behind your purchases, you'll have trouble buying food from the grocery store, no less a cell phone or any type of gadgetry.

Bertin told me this via skype yesterday: "He [Mikwuyma] told me he was told by at least 4 people that I was not practicing or I wasn't preparing for it correctly and that he feels uneasy about giving me runs at the marathon and 'don't be surprised if I don't get a run'" That seems pretty definitive to me. I think the punishment's too harsh, as he should really only be banned from running FE at AGDQ for being so poorly prepared for Sacred Stones.

Congrats to Molotov, that run was absolutely phenomenal. That definitely redeemed Fire Emblem's reputation in the marathon. Edit: Mentioned it on reddit.

The only major adjustment to real-time is that you can't afford to RNG burn as much for obvious reasons. As a general rule, you want to avoid RNG burning at all unless you really need to, such as landing a crit on a boss when the map goal is defeat boss.

Wow, this is great news. Can't wait to watch it! Today? You didn't specify. Even though it's the longest, it's also the fastest paced (except for games that can skip enemy phases altogether, but eh). It's also the only game that has battle saves to circumvent risk and a BEXP system that guarantees you characters are never RNG screwed, both of which dramatically reduce the odds of a run getting mercy killed. As for its length, consider that Starcraft 2 made it through this year with an estimate of 3:30. Sure, Radiant Dawn isn't as popular as a Final Fantasy game, but it also definitely has an audience. When I was doing sub 2:55 attempts I would consistently draw in 75-100 viewers without a raid on a channel with less than 500 followers at the time. Bertin also once told me that when he streamed Sonic R he would draw in 300 viewers, compared to 800 for Sacred Stones. I've even had a number of people come in and tell me they've never played FE but can still appreciate RD runs because of how fast the menuing is. Also consider that there are now 3 people who run RD (myself, kirbymastah and gwimpage) and another 3 who are learning the game (heero_fred, rushinassassin, superluigi) and it's clear that this isn't an obscure game that only I speedrun anymore.

One thing that's cool about such a run is that I can tell the audience what a particular level-up will be just before it happens, and on every single level-up in the game. Same thing for crits, misses, etc. It almost makes it look like I'm looking into the future, which has a bit of a wtf factor in itself, especially for those who don't understand what's going on. I'd be sure to explain everything that's going on regardless as I've had plenty of practice using a mic during my streams.

Well, one thing's for sure. If any GBA FE game is even allowed back in the event, it will have to be an RNG planned route, which I think is for the best anyway based on how the game's RNG works. Oh btw, legrandgrand has routed FE7 (similar to Molotov's route but with minor differences) and he's capable of cursor movement as fast as I use in FE8. Be sure to check out his twitch channel and throw him a follow if you like what you see. Keep in mind that french is his first language although he does speak english equally well.

What probably upsets me the most is that people who saw the run now think FE speedrunning is boring even though that run isn't at all representative of what optimized FE speedrunning looks like. Just look at some of the messages in that reddit: I can't exactly go in there and say "the AGDQ run was bad, here's some videos showing how it should've been done" since that would just make me look like an elitist asshole. So in that regard I'm kind of forced to let the FE speedrunning community's reputation get tarnished without being able to talk back in any meaningful way, especially since I already acted rather immaturely in the AGDQ chat during the run. I just hope this whole ordeal blows over soon enough and that I can still get Radiant Dawn into SGDQ or the next AGDQ.

I just got a time of 2:53:06 a couple of hours ago, meaning the sub 2:55 milestone has been crossed. The link to the highlighed video has been edited into the opening post. I think I'm just going to retire from speedrunning this game, after all the pain and suffering easy mode put me through, screw single segment speedrunning any higher difficulties.

Yeah, I did manage to improve my time to 2:58:45 (didn't bother to highlight) and sub 2:55 is well within my capabilities, although I'm feeling really burnt out from the game so it might be a while before I grind out that time.

Chapter 26 Does Elincia have renewal? Why did she heal 2 hp at the start of the turn? If it’s because of the rescue staff, delay giving it to her until next chapter. Oh, and I find it’s better to do everything in base, whereas you gave Marcia savior in base and did the rest in advancement. Any particular reason for that? And I’m kinda thinking someone should take out the bow paladin hassling Elincia who she can then rescue to safety. I don’t know if you have any units capable of pulling that off though as the likely candidates (Titania, Geoffrey, Largo) are probably too heavy. Just a silly idea. Chapter 27 What’s the point of positioning by the general guarding the middle door if you don’t kill him? The physic heal wastes a lot of time. Might want to give pure water to Tanith so that bolting death chance falls to 0. It seems to me you can still pull rescue staff/smite/vigor strats here and keep those extra units safe by luring the starting enemies that chase you down near the door on the left and have like an unarmed Geoffrey chokehold it. Final Just an execution thing, use d-pad left on the Ike drop and you don’t risk fumbling around if you miss the spot you’re supposed to be dropping Ike with the analog stick. --- Well, that looked really solid throughout, a definite improvement over Rushin and SMK/Molotov’s runs. You’ll get sub 2:10 easily with better execution, although I don’t know how much further the time can be brought down beyond that. Sub 2:00 seems like TAS only territory. So when do you plan to crank out some full game attempts?

Did you even read the post I made about execution? >_> The general fundamental is you use the d-pad for most straight moves (ex going full movement range left) and movements where you can take advantage of movement boundaries (ex/ holding d-pad left then d-pad up when there's a nearby wall to the left). The analog stick is used for movements where the d-pad doesn't work well, particularly diagonal movements. If I had to guess, I'd say 1/3 to 1/2 of my movement in RD speedrunning uses the d-pad. It's important you get used to using both since you will constantly be switching from one to the other, depending on movement. I even use the d-pad sometimes to correct a movement that was one square off using the analog stick, but that's a somewhat advanced trick I only learned recently.

Upload it to wiisave instead so everyone can use it