Early Game HM Rout Maps

[ruadath]

I want to try a little experiment and see if I can make improvements to the low turn counts for the early game HM maps, namely Chapters 4-6. I think I can do the first one in 1 turn, I'm reasonably conifdent I can do the second one in 2 turns (I did it in Lunatic, although a weaker Fred may be problematic, but I'm hoping this will be offset by lower enemy stats), and have some ideas for 2-turning the last one.

I started a HM file last night and I'm at the end of Chapter 2. I'm too lazy to make a stat box right now (I'll probably do one after the next chapter), but Robin's already level 10, Chrom is like 2.30, and Fred is at ~1.70. I'll let you know how things go; I expect progress to be relatively quick and am hoping to be done in a couple of days (maybe less).

[ruadath]

Chapter 4

Preparations: Big 3000G +9crit forge for Orsin's Hatchet (grabbed from Renown) plus a Str Tonic for Fred. Didn't really need to spend that much on the forge, but it gave him 9% crit on Marth and a sizeable crit on other people so worth it I guess. Deploy Sully next to Chrom, Sumia on the eastern spot, and everyone else wherever

Turn 1: People move to certain spots and kill people on the EP (just look at the image below). Chrom and Stahl are off the screen SW. Chrom has to crit kill (w/ Rapier) the barb who attacks him, Fred has to crit-kill Lucina and either crit or get a DS on the knight and fighter that attack him; he easily kills the mages. Sumia takes out a fighter that attacks her; I was planning on a double plus Robin DS, but Solidarity resulted in a crit which I don't mind. Robin easily slaughters the knight and Hand Axe guy who attack her.

UNIT     CLASS          LEVEL  HP STR MAG SKL SPD LCK DEF RES  WEXP
Chrom    Lord            4.08  22  10   1  10  11   8   9   1  D Swd
Robin    Tactician      11.77  28  16  15  13  18   7  14  12  E Swd C Tome
Fred     Great Knight    2.62  29  14   2  13  11   7  15   3  D Swd B Lnc D Axe
Sully    Cavalier        2.82  BASE                            E Swd D Lnc
Stahl    Cavalier        2.90  BASE                            D Swd E Lnc
Sumia    Pegasus Knight  3.07  20   8   4  13  13  10   6   8  D Lnc

C Supports:
Robin x Frederick
Robin x Sully
Frederick x Sully

Gold: 1970

Renown Items Claimed: Orsin's Hatchet

13 uses of Orsin's Hatchet remaining. 24 uses of the Javelin. Looking pretty good I guess!

[ruadath]

Sigh... the Prologue map nominally looks like it can be cleared in 2 turns (pair up Fred with Robin, move forward 6 spaces, drop Fred forward; then pair up Robin with Fred, move him forward and drop her in range of the boss), but it looks pretty annoying to actually carry out.

I'm going to try and convince myself that the odds are low enough that it isn't practical for me to carry out lol. Since I'm not really looking to do a full-blown LTC here I suppose it doesn't really matter, but I do care about things being in proper context, and maybe having a weaker Chrom or having less opportunities to build up Robin x Chrom support points will matter.

[Gradivus.]

Ignore the prologue for now, you risk overcomplicating this experiment tbh. Chrom and Robin will marry automatically after C11 if their support is the highest among Chrom's. You can easily keep this under control because the support rank is only determined by how many convos you activate.

Edited May 15, 2016 by Gradivus.

[ruadath]

Yeah, I think you're right here, I'm probably just going to ignore it. The odds look pretty bad since I think Robin would need Fred to DS or need to proc a Thunder crit on each of the eight enemies, not to mention dodging (she doesn't get a chance to use Vulneraries like in the standard 3-turn clear) so yeah. Probably doable on NM if anyone cares about that mode (I don't).

Thanks for the input and hope you enjoy the next couple of Chapters! I'll try and get at least Chapter 5 done by the end of today.

[SSJDennis]

Do like you did with Shadow Dragon, just carry on and think back about chapters later. That way you'll actually make a clear file. Its more impressive to beat the game in 60 turns, while it could've been 55, then just telling everyone it could be 55, but never achieve it.

[ruadath]

Something I forgot to think about (and want to ask you guys so I don't have to think about it): is training Ricken in HM a good idea? Or can the same workarounds for Lunatic (whatever they may be, I think Gradivus mentioned Valkyrie!Robin) apply to HM as well?

[ruadath]

While being sad about getting stuck in the Lunatic run, I realized the 2-turn Prologue actually had significantly better odds than I was expecting due to me failing to account for stuff, so not counting Robin's 2 levels, the odds for success are roughly 1 in 100,000, which is very doable.

Not planning on pursuing this any further for the moment, however.

[Horakthi]

While being sad about getting stuck in the Lunatic run, I realized the 2-turn Prologue actually had significantly better odds than I was expecting due to me failing to account for stuff, so not counting Robin's 2 levels, the odds for success are roughly 1 in 100,000, which is very doable.

Not planning on pursuing this any further for the moment, however.

Are you serious? It's only very doable if you're a hacker. If an attempt takes on average around 3 minutes, that's about 300000 minutes or 5000 hours wasted on attempts to cut 1 turn on the Prologue.

I don't browse or sign on this forum very much apart from things like look at LTC playthroughs and the recent translation patch and such, but it's very sad to see LTC playthroughs nowadays resort to mysterious hacking nowadays because of your strange OCD obsession with hacking for the lowest turncounts that aren't practical even in the slightest on console. At first I thought it was cool, but now it's just getting boring.

Edited May 19, 2016 by Horakthi

[SSJDennis]

For Ruadath 1 in 100,000 is a walk through the park. He did much worse and managed to pull it off. Sure, you will NEVER get that lucky with a regular 3DS.

Would really love to see some videos of these maps working.

[Horakthi]

For Ruadath 1 in 100,000 is a walk through the park. He did much worse and managed to pull it off. Sure, you will NEVER get that lucky with a regular 3DS.

Would really love to see some videos of these maps working.

You mean "for any hacker." Being able to hack doesn't make him anything special at all. 1 in 1 billion is doable for any hacker, easily. I don't see why you have to put him on a pedestal because he can hack games on his emulator? So what? This guy took a day break from his Lunatic LTC just to do a Prologue 2 turn which should take about 5000 hours, or more than half a year. That he is hacking is obvious and I don't see why you think it's impressive that he can hack.

Look, anyone can feel free to do a challenge run involving hacking (is that even challenging? part of the difficulty, it seems to me, from an LTC run is being able to come up with reliable strategies). But just don't pretend that this run is legitimate when it involves hacking. This run still can be cool if you just put "FE13 Lunatic Hacked LTC" or "FE13 Lunatic LTC with Hacking" or something like that. That would be cool too. At least admit to it!

In speedruns which I'm more familiar with, for example, in the SDA, real speedruns don't have tool assistance and tool-assisted speedruns are a different beast altogether. The latter are still cool and interesting, but they're a different category from real speedruns.

Edited May 19, 2016 by Horakthi

[ruadath]

Using an emulator allows you to disable the CPU intensive graphical rendering features and turbo the game so that you can run it faster; that's the whole point of using the emulator (well that and automation via scripting), since under normal conditions Citra isn't fast enough to even run the games at a full 60 FPS.

The 1 in 100,000 odds is also an overestimate. The actual odds are somewhere between 10,000 and 100,000 (logarithmically, which imo is the correct scale to compare RNG abuse) and the above screenshot was obtained without concern for Robin's levels since as I said, I'm not planning on working on this any further.

If you want proof that the clears are being done legitimately, I can just send you (again meaning anyone) the bookmark save file. You can import it into either Citra or a 3DS that has been hacked to a sufficient degree (I don't own one so I can't help you with that), and run the save file and you'll see that it works.

Would really love to see some videos of these maps working.

The Citra developers are getting close to making an official release with audio, filtering, and all the other stuff soon so I'm hopeful that we can get some videos soon too! Right now there are lots of issues with recording the videos properly (the most annoying being the dramatic slowdown from recording since I don't think any of my computers have a graphics card that supports NVIDIA Shadowplay), but also problems with crashing and audio desync, etc.