FE13 Lunatic RNG-abused LTC

[Kuroi Tsubasa Tenshi]

Oops, yeah, I somehow got it in my head that the bonuses from C to A were Hit, Attack, Attack, not Attack, Hit, Attack.

[ruadath]

Took a look at this briefly and noticed that we're spending more money than we have. After selling the 2 bullions, buying 5 Str Tonics, forging the Javelin and the Iron Sword, and buying 2 more Javelins and a Hand Axe, I'm down to 450G (with potentially another 1250 from the Secret Book). After that, we have to sell weapons to make cash, and a +3 Mt forge on the Beast Killer is nearly 5k gold. I suppose this could be remedied by just pulling the Large Bullion out of Renown (or selling other Renown items), but do you have any better ideas?

Note that one possibility is to give the Beast Killer a +1 Mt forge (~750G) and then rig a Gaius DS when attacking the cav. Sumia will dish out 30 of the 36 damage, and Gaius will do 7+8+2-8-1=8 damage to finish him off.

Edited May 23, 2016 by ruadath

[Kuroi Tsubasa Tenshi]

With the strat as it is, yeah, you're probably gonna just have to go for the DS. That's really too much money to invest into something as specialized as the Beast Killer at this point in the game. It won't be helpful again until C11 and by then, you'll both have a lot more gold and Robin throwing around high magic damage.

[ruadath]

I'm certain I've seen the Dark Mage move to the west spot at some point, but after doing ~20 test resets (by hand), he doesn't seem to want to move away from the southern spot as things currently stand. And sadly, the game pairs up Gregor with Nowi when the mage attacks, so Robin can't kill on the DS...

Fortunately, all hope might not be lost. Since Gregor is next to Robin, he gets a 10% crit boost from her (for a 10% crit rate). If we give him a +2 Mt forged Levin Sword, then a crit (21 damage) + a DS from Nowi (15 damage) takes care of the mage. But this is like a ~2% total chance of happening, and the strain on both my pocket and ability to RNG-abuse might be getting to the point where it is too much to handle...

Edited May 23, 2016 by ruadath

[Kuroi Tsubasa Tenshi]

How imperative is it that Gregor be above Robin? If you swap him and Nowi around, he'll probably still go after Gregor, but will be forced to move to a position that doesn't body block the Myrm. Of course, this then brings up the issue of needing to kill him on EP2, which might make things a lot more dangerous for Fred.

[ruadath]

This would be an excellent idea if it wasn't for the fact that swapping Gregor and Nowi puts Gregor out of attack range (Robin ends up on the spot just above Nowi's deployment square), so the mage will probably just do the same thing but attack Nowi. I don't think it actually makes things terribly dangerous for Fred because he has a high dodge rate (outdoor fighter is awesome), and he can still easily kill him with the Hand Axe.

Maybe I'm overlooking something silly though and the the other one (of Gregor and Nowi) is still in the mage's attack range, but I don't think so.

[Kuroi Tsubasa Tenshi]

No, you're right. I got it wrong and counted that you could move one space farther than you could. If he attacks Nowi, that still gives you better odds than Gregor, though. Nowi can counter at 2 range and thanks to Robin's Solidarity, she'll have a 6% crit chance (I'm assuming Gregor will pair to her too and his Steel Sword attack would be enough to matter if he DSs). That's three times more likely than the Gregor Levin Sword strat and costs you no money. It's still not great, but it's something.

[ruadath]

Nowi can attack at 2-range? For some reason I assumed that she couldn't lol. OK, that's definitely good to know. She does 15 damage, so critting would immediately kill the Dark Mage. Gregor in a pair up attack does 16 damage (with the Steel Sword), which is 5 short of what it needs to be. He can actually be ferried (and has a turn free to use, unlike Nowi since she has to drop Robin South) a Str Tonic, which puts him 3 damage short. So with a +3 Mt forge to the Steel Sword, this becomes significantly more reliable, although this is another 2.5k (closer to 3 with the Tonic) that I have to shell out. Odds of success for that are roughly 30%, which isn't so bad (actually it's less because Nowi can miss, but let's overlook that for now).

I'm trying to think of ways to make the NE corner less RNG-reliant but I can't think of anything that's not expensive. The best idea I've come up with so far is to replace Lonqu with Sully; if you give Ricken a Mag Tonic and Sully a Str Tonic (and the Steel Lance), they are 5 short of ORKOing the myrmidon without a crit; so if you're willing to distribute +5 in forges among both weapons you could increase the reliability of the strategy by about 4x, but I think the cost of that is somewhere around 5k... (in which case you might as well just forge up the Beast Killer).

[Kuroi Tsubasa Tenshi]

Yeah, Dragonstones were actually really good for offense in Awakening, thanks to being 1~2-range. Fates nerfed that by reducing the range 1, but at the same time, some of the additional stats, in conjunction with the Nohr Prince(ss) line makes it really good for tanking as compensation.

If the Beast Killer forge was too expensive, I don't know if forging the Steel Sword, Elwind and Steel Lance are going to be affordable, either. I guess the sword might be worth keeping Nowi's odds from being so bad, but I'm not so sure about the other forges. I guess one minor tweak you could do is use Miriel instead of Lissa to ferry Chrom. She gives +2 Skl as part of her pair-up bonus, which is 1 extra crit.

[ruadath]

In addition to that, if we give Chrom a Str Tonic, he has 13 (Base str) + 9 (Killing Edge Mt) + 1 (Weapon Rank) + 2 (Tonic)= 25 Atk, which means he can ORKO the mage even if he doesn't crit (although this means dodging another Nos).

[Horakthi]

Can't respond in full detail right now (will edit soon) as I'm trying to get something done within the hour, but:

You're misunderstanding what I'm saying. It's pointless and inefficient (and probably not even possible) to have a script run the whole map at once if the strat has a reasonably likely chance of failing on the first turn. Thus the more sensible thing to do is to run the turn 1 (or whatever) script many times, generate a ton of bookmarks at the start of turn 2 (where the first turn has been cleared successfully), and then iterate from there.

Um no, lol. You don't understand my argument. In Awakening, you HAVE TO script run the whole map at once because the RNs in Awakening stay the same if you bookmark. You CANNOT change the RN of a map by making a bookmark and reloading it. You HAVE TO script run the whole map at once.

The automated resetting tool you describe would work for Radiant Dawn, which you have already attempted to do, though. Why? Because the RNs of the map do change when you make a suspend state and reload the map. But this is impossible with Awakening, because you cannot use your automated resetting tool on a bookmark. You'd get the same results over and over again with your automated resetting tool because every bookmark has the same RNs upon loading!

Therefore, we can infer that you are lying about the automated resetting tool. There is no such thing. I am surprised KTT hasn't pointed this out given that he has more knowledge than probably anyone in the world about Awakening's mechanics, but w/e.

Further, you refuse to give me a printscreen of your tool, which makes it even more obvious that you're lying. Your run is not only illegimate as a proper LTC run because it uses tools to begin with, but it's also even more illegitimate since you won't tell us what tools you are actually using. My guess is you use the Awakening save editor to hack your unit's stats to make RNG abusing a lot easier. Given your previous history of lying (I looked into this), and the ease with which you can get 1/100000 clears, I'm surprised anyone takes any of your runs seriously at this point.

[Kuroi Tsubasa Tenshi]

What's with the hostility? And given the title of the topic, I don't think he ever pretended that this was anything other than the TAS equivalent of an LTC.

In any event, his description of what he's doing makes perfect sense and could even be done (very inefficiently, due to the normal bookmark limit) with a physical copy. He's running the turn 1 script and then having the game bookmark anything that makes it to turn 2. Anything else is discarded. Then he runs the turn 2 script and anything that makes it to turn 3 is bookmarked. Anything else is discarded. This continues until he finds a save with an RN string that lets him complete the map. If all the bookmarks get culled, he has to start over and have the first script produce a bunch more successes to iterate off of. It's a really brute-force method, but still makes way more sense than running a full map script on something that has a high chance of failing on turn 1.

In addition to that, if we give Chrom a Str Tonic, he has 13 (Base str) + 9 (Killing Edge Mt) + 1 (Weapon Rank) + 2 (Tonic)= 25 Atk, which means he can ORKO the mage even if he doesn't crit (although this means dodging another Nos).

For such a low cost, I'd say the improvement to the odds is worth it, especially since Nos Dark Mages aren't very accurate in the first place.

[ruadath]

Um no, lol. You don't understand my argument. In Awakening, you HAVE TO script run the whole map at once because the RNs in Awakening stay the same if you bookmark. You CANNOT change the RN of a map by making a bookmark and reloading it. You HAVE TO script run the whole map at once.

Exactly my point. That's why you start off by doing turn 1 many times (each from a different reset), and then take a bookmark from each one. Then when you go on to do turn 2 (for the sake of argument, often it's convenient to split one turn into multiple phases but whatever), if a particular bookmark doesn't give you the desired result you want, you discard it and move on to the next one. This is an algorithmic equivalent of just resetting from scratch each time, it's just significantly easier to code and runs faster because you're not wasting time doing pointless stuff.

Further, you refuse to give me a printscreen of your tool, which makes it even more obvious that you're lying. Your run is not only illegimate as a proper LTC run because it uses tools to begin with, but it's also even more illegitimate since you won't tell us what tools you are actually using. My guess is you use the Awakening save editor to hack your unit's stats to make RNG abusing a lot easier. Given your previous history of lying (I looked into this), and the ease with which you can get 1/100000 clears, I'm surprised anyone takes any of your runs seriously at this point.

I can give screenshots if you want, but what do you want to see? Like I said, I haven't packaged the thing, so there isn't any interface. All I can give you is screenshots of code and/or make a gif of the memory changing really fast (viewed in a memory watching program like CheatEngine; it actually is kind of fun to watch lol), neither of which are particularly enlightening. I'm not sure what the Awakening save editor is, though I can tell you that it's unnecessary to edit the save files (I've been hacking in the Renown items as I need them since I was too lazy to do a casual playthrough first and unlock everything).

Also regarding the RD incident you linked to, yes I did hack stats (really just one stat, because all of the dodges I needed to occur except one occurred) there, and that was when I did all my resetting by hand and needed to get a 1 in 10,000 odd clear. After that happened, I took a break (for a couple of months? I don't remember the length) from the playthrough to develop these tools so that I wouldn't have to resort to doing anything like that again.

Edited May 24, 2016 by ruadath

[SSJDennis]

You gave up on this run? Or you like to break Fates better at the moment? Did you ever finish the FE7 run? Or... I like your "Shove" idea in Fates.

[ruadath]

Yeah, it's the quintessential problem of having too many runs going on at once...

I haven't necessarily given up on this run, it's just that it's really annoying to do because of the ridiculous RNG abuse coupled with money management vs. reliability issues (which go away if you just throw in all the renown items for forge capital, but not sure I want to do that). Fates (at least Conquest) has comparatively significantly less RNG abuse (with the exception of Chapters 7 and Chapter 9, I think pretty much every strat I use could be replicated by hand, not counting the stuff that goes on in My Castle) and has more interesting map objectives (not only rout/defeat boss) which make it more fun to plan out.

I will probably come back to this eventually, maybe even sooner than I think. But I'm probably going to wrap up my Conquest run first.

EDIT: Also, as for the FE7 run that is still going on... I've been focusing attention on redoing certain chapters for more optimal results. Since I limit my RNG abuse in GBAFE, things are very difficult to plan out, and making small changes often requires redoing whole chapters in order to get everything to work out properly..

Edited July 1, 2016 by ruadath

[ruadath]

Just a thought I had the other day for if/when I eventually return to this:

Chapter 1 should be improvable to 2 turns by rigging the Event Tile in the Prologue to yield an Orsin's Hatchet. This wouldn't hurt Robin's EXP gain at all since she handles the enemies who attack the fort.