Suikoden II

Omnigamer

Edit history:

Omnigamer: 2014-04-08 12:47:27 pm

Omnigamer: 2014-04-08 12:11:46 pm

2014-04-08 12:10:12 pm

All the things

Posting this so that somebody can grab it before it disappears. JP runner with 6:33 by our timing using a non-glitched route. I will try to save the files when I get home at some point, but I'm not sure exactly when that will be.

I am quite frustrated that the JP runner is able to get such a time while still making very silly mistakes and otherwise unoptomized things. Many, many times he will select the wrong option from a choice and watch 30 seconds worth of cutscene as a result. I don't know where the gains come from, but I imagine it's some effect from turbo, a faster console, or both.

EDIT: Also note the softlock around 1:05:35. Silly Nanami.

BadPotato

Edit history:

BadPotato: 2014-04-16 11:33:01 am

BadPotato: 2014-04-15 08:37:09 pm

BadPotato: 2014-04-15 08:36:47 pm

BadPotato: 2014-04-15 08:35:20 pm

BadPotato: 2014-04-15 08:34:25 pm

BadPotato: 2014-04-15 08:34:05 pm

BadPotato: 2014-04-15 08:32:39 pm

BadPotato: 2014-04-15 08:30:52 pm

BadPotato: 2014-04-15 08:29:49 pm

BadPotato: 2014-04-15 08:28:44 pm

BadPotato: 2014-04-15 08:28:15 pm

BadPotato: 2014-04-15 08:27:44 pm

BadPotato: 2014-04-15 08:27:24 pm

BadPotato: 2014-04-15 08:27:06 pm

BadPotato: 2014-04-15 08:26:44 pm

2014-04-15 08:26:19 pm

Quote from Omnigamer:

I am quite frustrated that the JP runner is able to get such a time while still making very silly mistakes and otherwise unoptomized things. Many, many times he will select the wrong option from a choice and watch 30 seconds worth of cutscene as a result. I don't know where the gains come from, but I imagine it's some effect from turbo, a faster console, or both.

I didn't watch the whole thing yet, but faster console sure help a bit. At 3:23:18, we can notice random music glitch too.

Also, to extends your thought with your recent testrun about the Kindness Rune glitch on the PC version, they probably had to refactor every use of some sort of custom PSX integer in the code, to replace them by signed long(or similar) to get the PC build working or something like that.

That being said, it's clear that some work can be done with the chinese edition, so we can have a better understanding of the mechanics:

Stuff the check out:
-Going "out of screen border" at castle lvl3.
===>I didn't post a video about this at the time(just this crappy video), but it's possible to have Riou character in normal field movement sprite(with the ability to run) on the world map. I remember that I couldn't do much with this, but I believe it's possible to access restricted area with this, if you go to the right (x,y) coordinate. Now that we managed to get the PC setup, it should be easier than ever to debug this.
-Wrong warp at Greenhill?
-Equip item that you're aren't supposed to able?
-Get the exacts stats % in battle, avoid encounter, etc...
-Softlock with Amada skip, still happen?
-Does softreset only keep old level of the characther, wasn't there something about an opened chess in the kobold forest that wasn't to be? I saw this a while ago on suikosource and it should have been verified.
===> What if there's more than that?
-...

I guess the save file isn't exactly processed the same way on the PC version, but I'm sure that they should be kind of the same.

Maybe I could help at some point with this.
edit: typo

[quote="Omnigamer"]I am quite frustrated that the JP runner is able to get such a time while still making very silly mistakes and otherwise unoptomized things. Many, many times he will select the wrong option from a choice and watch 30 seconds worth of cutscene as a result. I don't know where the gains come from, but I imagine it's some effect from turbo, a faster console, or both.[/quote]
I didn't watch the whole thing yet, but faster console sure help a bit. At 3:23:18, we can notice random music glitch too.

Also, to extends your thought with your recent testrun about the Kindness Rune glitch on the PC version, they probably had to refactor every use of some sort of custom PSX integer in the code, to replace them by signed long(or similar) to get the PC build working or something like that.

That being said, it's clear that some work can be done with the chinese edition, so we can have a better understanding of the mechanics:

Stuff the check out:
-Going "out of screen border" at castle lvl3.
===>I didn't post a video about this at the time(just this [url=https://www.youtube.com/watch?v=rQjsM7foTEI]crappy video[/url]), but it's possible to have Riou character in normal field movement sprite(with the ability to run) on the world map. I remember that I couldn't do much with this, but I believe it's possible to access restricted area with this, if you go to the right (x,y) coordinate. Now that we managed to get the PC setup, it should be easier than ever to debug this. 
-Wrong warp at Greenhill?
-Equip item that you're aren't supposed to able?
-Get the exacts stats % in battle, avoid encounter, etc...
-Softlock with Amada skip, still happen?
-Does softreset only keep old level of the characther, wasn't there something about an opened chess in the kobold forest that wasn't to be? I saw this a while ago on suikosource and it should have been verified.
===> What if there's more than that?
-...

I guess the save file isn't exactly processed the same way on the PC version, but I'm sure that they should be kind of the same.

Maybe I could help at some point with this.
edit: typo

Thegamemaster5181990

2014-04-20 06:02:42 pm

I LOVE SUIKODEN

Hello everyone. my name is joseph and i am streaming my infamous Suikoden 2 New Game Plus Glitch!
Check It Out On Twitch.tv/thegamemaster5181990

This will change the speedrun for this game. forever.
Im planning on running a world record run for this game and any help is appreciated.
Have A Great Day!

Thegamemaster5181990

2014-04-29 05:41:22 pm

I LOVE SUIKODEN

Quote from Lag.Com:

Yeah, you don't see me starting topics too often do you?

This gem of an RPG deserves a run, and though I can't record a 'legit' one, I'm certainly going to make one on an emulator and share it out. Normally I just do this for my own enjoyment and work it out myself, but I think there is a little too much planning involved in this game, so this is a call for help.

No, I have no idea how long the resulting run will be.

Most important points:
* Optional characters to get
* Trading - could be abused to make silly amounts of money
* Runes to buy/equip
* Tough boss battles

Any useful information (assume I am able to read GameFAQs) is appreciated.

Quote from Lag.Com:

Yeah, you don't see me starting topics too often do you?

This gem of an RPG deserves a run, and though I can't record a 'legit' one, I'm certainly going to make one on an emulator and share it out. Normally I just do this for my own enjoyment and work it out myself, but I think there is a little too much planning involved in this game, so this is a call for help.

No, I have no idea how long the resulting run will be.

Most important points:
* Optional characters to get
* Trading - could be abused to make silly amounts of money
* Runes to buy/equip
* Tough boss battles

Any useful information (assume I am able to read GameFAQs) is appreciated.

Tomorrow At 1pm Pacific time
I will Be Running Suikoden 2 with a glitched% World Record Run!
Check it out on my twitch!
Twitch.tv/Thegamemaster5181990
I failed my first run because the beginning was messed up.
but now i got it down perfectly.
Cheesy

I love the suikoden Community.
Thank you all for being so awesome!
Best game series ever!

[quote="Lag.Com"]Yeah, you don't see me starting topics too often do you?

This gem of an RPG deserves a run, and though I can't record a 'legit' one, I'm certainly going to make one on an emulator and share it out. Normally I just do this for my own enjoyment and work it out myself, but I think there is a little [i]too[/i] much planning involved in this game, so this is a call for help.

No, I have no idea how long the resulting run will be.

Most important points:
* Optional characters to get
* Trading - could be abused to make silly amounts of money
* Runes to buy/equip
* Tough boss battles

Any useful information (assume I am able to read GameFAQs) is appreciated.[/quote]
[quote="Lag.Com"]Yeah, you don't see me starting topics too often do you?

This gem of an RPG deserves a run, and though I can't record a 'legit' one, I'm certainly going to make one on an emulator and share it out. Normally I just do this for my own enjoyment and work it out myself, but I think there is a little [i]too[/i] much planning involved in this game, so this is a call for help.

No, I have no idea how long the resulting run will be.

Most important points:
* Optional characters to get
* Trading - could be abused to make silly amounts of money
* Runes to buy/equip
* Tough boss battles

Any useful information (assume I am able to read GameFAQs) is appreciated.[/quote]

Tomorrow At 1pm Pacific time
I will Be Running Suikoden 2 with a glitched% World Record Run!
Check it out on my twitch!
Twitch.tv/Thegamemaster5181990
I failed my first run because the beginning was messed up. 
but now i got it down perfectly.
:D I love the suikoden Community. 
Thank you all for being so awesome! 
Best game series ever!

Thegamemaster5181990

2014-05-04 03:06:35 pm

I LOVE SUIKODEN

http://www.twitch.tv/thegamemaster5181990/c/4173436

I beat suikoden 2 using a memory card corruption glitch. and i beat the game in 15 Minutes.
Check it out. this is the world record now because it is not using a gameshark or any cheat device and
i apologize for the harsh language (i was up for 2 days) but otherwise check it out.
its really cool Cheesy

ill also beat it again to show you guys how to do it yourself. and explain how it works in super detail.

Omnigamer

2014-05-10 01:06:29 am

All the things

OK, I think I've routed a pure magic route for the Chinese version since Kindness Underflow doesn't exist. I tried to balance for usefulness and safety, but there are a few cases where things can be uncertain.

Pest Rat - Get Rina above lvl32 before trying
Riou: Thor -> Bolt
Luc: Thor -> Bolt
Rina: Unite Glitch Water

Luca - Level up Zamza and Nina to ~38 for lvl 4 MP
Phase 1: Go all out with Flik, Camus, Eilie, Stallion, Freed Y, and Viki*
Phase 3: Hero, Miklotov, Nina, Luc, Zamza, Rina. Destruction with lvl 4 Lightning. 2T

Bone Dragon -
Abuse the hell out of shredding in both item and magic form. 2T if 4 users.

Lucia Rd 3 -
Rune Unite glitch her to death with 3x Thor and something else

Seed & Culgan -
Rune Unite Glitch with 3x Scorched Earth, Luc uses Thor

Beast Rune -
Take out Rune with Luc Thor, 2x Bolt of Wrath.
Proceed to take out both heads by using 6x Earthquake.

Now to actually test the sucker. Probably won't happen any time soon, however. I want to test it against my Kindness route just to come up with a rough comparison.

OK, I think I've routed a pure magic route for the Chinese version since Kindness Underflow doesn't exist. I tried to balance for usefulness and safety, but there are a few cases where things can be uncertain.

Pest Rat - Get Rina above lvl32 before trying
Riou: Thor -> Bolt
Luc: Thor -> Bolt
Rina: Unite Glitch Water

Luca - Level up Zamza and Nina to ~38 for lvl 4 MP
Phase 1: Go all out with Flik, Camus, Eilie, Stallion, Freed Y, and Viki*
Phase 3: Hero, Miklotov, Nina, Luc, Zamza, Rina. Destruction with lvl 4 Lightning. 2T

Bone Dragon - 
Abuse the hell out of shredding in both item and magic form. 2T if 4 users.

Lucia Rd 3 -
Rune Unite glitch her to death with 3x Thor and something else

Seed & Culgan -
Rune Unite Glitch with 3x Scorched Earth, Luc uses Thor

Beast Rune - 
Take out Rune with Luc Thor, 2x Bolt of Wrath.
Proceed to take out both heads by using 6x Earthquake.

Now to actually test the sucker. Probably won't happen any time soon, however. I want to test it against my Kindness route just to come up with a rough comparison.

Mishrak

Edit history:

Mishrak: 2014-06-20 11:43:46 am

Mishrak: 2014-06-20 11:40:03 am

Mishrak: 2014-06-20 11:37:54 am

Mishrak: 2014-06-20 11:36:57 am

Mishrak: 2014-06-20 11:36:50 am

2014-06-20 11:36:06 am

I'm in the process of going over your video and your notes (updating them), Omni, in hopes that I can actually learn to run it. I'm still only about as far as Sindar in terms of really analyzing it, but I have read most of this thread so I have at least a cursory knowledge of what you do. I have played with early Kindness and Double Beat in a casual run too.

My plan is to do a segmented run to learn the strats and then when I'm comfortable with that, do single segment runs.

Two questions I have:
1) Skipping Double Beat and grabbing Violence at Rockaxe on the way to Kindness.
-> If you're not getting Shiro or Kennison, and you're using Destroyer Jowy and powering leveling Nanami and Riou up at strategic places, Double Beat doesn't even really get used until Abomination anyway, and you'd have Violence by then. That's the assumption right?
-> Is it still worth it to kill the rabbit battles on your first trip through North Sparrow? I'd be worried that exp might be too low, but maybe the Mist Monster will equalize that. You could still kill them for the exp and DB chance on your way back, but just no longer stop to farm. Or you could ignore them all together and just Let Go if the experience won't be an issue.

2) Turbo - that's referring to a turbo controller to speed through dialog right? What are your thoughts on that? I'll be running the game on a PS2 without any disc speed modifications and using a standard dual shock PS2 controller. Turbo seems goofy to me.

I'm in the process of going over your video and your notes (updating them), Omni, in hopes that I can actually learn to run it.  I'm still only about as far as Sindar in terms of really analyzing it, but I have read most of this thread so I have at least a cursory knowledge of what you do.  I have played with early Kindness and Double Beat in a casual run too.

My plan is to do a segmented run to learn the strats and then when I'm comfortable with that, do single segment runs.

Two questions I have:
1) Skipping Double Beat and grabbing Violence at Rockaxe on the way to Kindness.
-> If  you're not getting Shiro or Kennison, and you're using Destroyer Jowy and powering leveling Nanami and Riou up at strategic places, Double Beat doesn't even really get used until Abomination anyway, and you'd have Violence by then.   That's the assumption right?
-> Is it still worth it to kill the rabbit battles on your first trip through North Sparrow?  I'd be worried that exp might be too low, but maybe the Mist Monster will equalize that.   You could still kill them for the exp and DB chance on your way back, but just no longer stop to farm.  Or you could ignore them all together and just Let Go if the experience won't be an issue.

2) Turbo - that's referring to a turbo controller to speed through dialog right?  What are your thoughts on that?  I'll be running the game on a PS2 without any disc speed modifications and using a standard dual shock PS2 controller.  Turbo seems goofy to me.

Omnigamer

Edit history:

Omnigamer: 2014-06-20 11:55:14 am

Omnigamer: 2014-06-20 11:54:04 am

2014-06-20 11:53:50 am

All the things

JP RPG runners all use Turbo so it makes things difficult to compare. You can get performance on par with turbo by mashing X O L1 and L2, but it's still not perfect.

On your first question, that's still a tricky subject. Double-beat isn't really necessary until the Luca battle. It does speed up each boss fight from SD Sword onward, but the benefit isn't really needed until Luca. You can still match the number of turns on all the bosses, but it takes a lot longer to set up appropriate magic and characters. Violence helps early on, but it's harder to set up a proc on the early bosses. It takes ~40 seconds to go out of your way on the way to Kindness Early to get Violence, just as a heads-up. Double beat is nice because it's consistent damage and stacks with Violence later on. If you don't get double beat early, next place you can get it is from Lo Wen after Neclord. It's not entirely necessary, but it does simplify the final boss battles a lot. In the ideal case, (ie for segmented) you'd get a double beat on first rabbit encounter and then be done with it.

Either way, I encourage you to keep looking into it and give it a try yourself.

EDIT: Also, my earlier notes are pretty junky, so I apologize for that. I was working on a compressed version that just sticks to key details and is up-to-date with current strats, but it isn't done yet. I will post it when I finish it off.

JP RPG runners all use Turbo so it makes things difficult to compare. You can get performance on par with turbo by mashing X O L1 and L2, but it's still not perfect.

On your first question, that's still a tricky subject. Double-beat isn't really necessary until the Luca battle. It does speed up each boss fight from SD Sword onward, but the benefit isn't really needed until Luca. You can still match the number of turns on all the bosses, but it takes a lot longer to set up appropriate magic and characters. Violence helps early on, but it's harder to set up a proc on the early bosses. It takes ~40 seconds to go out of your way on the way to Kindness Early to get Violence, just as a heads-up. Double beat is nice because it's consistent damage and stacks with Violence later on. If you don't get double beat early, next place you can get it is from Lo Wen after Neclord. It's not entirely necessary, but it does simplify the final boss battles a lot. In the ideal case, (ie for segmented) you'd get a double beat on first rabbit encounter and then be done with it.

Either way, I encourage you to keep looking into it and give it a try yourself.

EDIT: Also, my earlier notes are pretty junky, so I apologize for that. I was working on a compressed version that just sticks to key details and is up-to-date with current strats, but it isn't done yet. I will post it when I finish it off.

Mishrak

Edit history:

Mishrak: 2014-06-20 12:09:12 pm

Mishrak: 2014-06-20 12:08:48 pm

2014-06-20 12:08:26 pm

I appreciate your response.

That's a good point on Double-Beat and Violence. I think throwing that into the mix when I'm just starting to learn is probably a bad idea. I'll follow your run/notes till I'm comfortable enough with it to think clearly about the game and then I can decide if I want to swap to it.

Also re: your notes - No need to apologize. In fact, it's forced me to read more and pay closer attention to your video so I can see what you're actually doing and why. It's been a fun learning experience. It's something that separates speed running an RPG from a platformer (Super Metroid) - the skill is in the knowledge of the game and running an optimized route versus reactions and precise timing.

Delicious

2014-06-21 12:06:13 am

Double Beat is on Gijimu, not Lo Wen

Omnigamer

2015-02-12 01:00:38 pm

All the things

Hello all, long time no post.

I've been working on a complete re-route for the middle and ending of the run, with some interesting successes and failures. My general notes from testing are too disorganized to share here, but here is a brief list of the things I tried along with some pros and cons:

No Double Beat
This was one of the main reasons for searching out a new route, and overall I think I've cracked the benefits of double beat by other, non-random means. The main benefit of double beat is just a straight 2x damage amp, and it can go to any character with an extra rune slot. The 2x stacks with Violence and it is repeatable ad nauseum (ie, no use limits). This is great, but what I really needed was efficiency. Double Beat made things simple, but in many of the cases it was overkill. So I started looking for a better way. Removing double beat also makes for a much more consistent early game since experience conditioning will always use the same enemy mobs.

Kindness Early
Another point of contention that came up is Kindness Early. I explored new strats for going through SDS, Abom, and Rat without Kindness at all vs having Kindness without Double Beat, and in the long run Kindness Early still makes sense. The difference isn't big, but it is there. It also eases up some of the opportunities to attach and swap runes as needed.

Kindness Bearer
This is also something I struggled with and only last night settled on something reasonable. Kindness is very powerful not just because of the damage, but also because it carries to any other attack runes or unite attacks. So I started to focus pretty strongly on these things in the hunt for new strategies. The old strategy relies on Nanami as the Kindness Bearer up through Bone Dragon, which she is decently suited for since you have to farm deaths regardless of whether she has Kindness. Her Tech is also quite good, so she has no trouble hitting things once she gets to ~Level 45. She loses quite a bit of power from not having Double Beat however. Sheena, as the backup, also works decently enough but he still relies on the amp from only double beat and double-strike, capping his damage output without Violence at about 3x. 3k damage is still a lot for many bosses, so instead I tried to find alternatives that mixed in magic usage to help out and get a "cleaner" finish. I narrowed the list down to these candidates:

Zamza
+Self-perpetuating Kindness bearer. Will likely kill himself after using a powered up Fire Dragon, thus only the first death is "out of the way."
+Automatic Violence procs. Using Fire Lizard once will deal approximately 2k damage, so with a Fire Emblem this is mitigated to less than 500 damage returned. At level 70 (which I routed him to be) this procs Violence but leaves him about 200 HP to spare.
+No switching necessary. Put it on him once and it will stay on the whole game.
+Quick to recruit and set up.
-Fire element damage. This isn't terrible, but means some specific bosses need to be worked out differently.
-Somewhat inconsistent battles. After the first Fire Dragon, he'll have somewhat low HP. The high level gives him high enough defensive stats to survive most attacks after, but it is still somewhat risky.
-Fairly low TECH. This needs to be supplemented by attaching Thunder Amulet(s) or else he risks not being able to attack some bosses.
-Poor Armor selection. Being limited to Robes is a bit of a killer in terms of other ways to increase his TECH.

Meg
+Trick is a very fast (time-wise), very powerful Attack Rune with x3 damage amp. She also has an extra rune slot for Double-Strike to bring it up to an instant 4.5x amp. Triggering Violence gives it a second turn power of 13.5x, which will kill any boss in the game instantly.
+Her HP and MDEF growth mimic Nanami's. This means that I can almost always set up a Luc Empty World -> Violence until after Tinto.
-Longish recruitment time. She's a little out of the way, and takes approximately 1:20 to recruit on the way to Matilda. She also requires Castle lvl 2, which is not a necessity otherwise (although I'd probably get it anyway for Jeane during Kiba).
-Initial level does not work well for leveling during Matilda, nor do I have good opportunities to level her or kill her off before Luca.
-Her TECH is only OK, but is not strong enough without being leveled to take on Luca. It wouldn't make sense, and would rely on significant luck to actually hit Luca.
-One use only. This means that if I'm going to use it, I need to make sure it hits. This can be somewhat helped by Thunder Amulets, but she only has two accessory slots and one of them should be a Jizo.
-Not useable during Luca or the Worm due to level/deaths.

Oulan
+Angry Dragon syncs well with Fury/Violence. x2 Damage for free, and reuseable when Berserk is great (meaning x3 amp total).
+Beefy. Lots of HP to work with, and decent defenses.
-Beefy. Makes it that much harder to ensure Violence procs.
-Recruitment time is about a minute, but there's also no easy way to farm deaths on her.

Shin
+Strong 3x amp attack. Takes a somewhat longish time to perform, but 3x does serious damage to any enemy.
+Amazing TECH and LUCK. These feed into more crits, but also mean I don't have to get TECH boosting items in order to make him effective.
+Decent armor selection. Gotta get that Master Garb.
-No open rune slots. This means further damage amp has to come organically (ie, from Battle Oath or items) and he is not compatible with Violence.
-One use only. Not as big of a deal since he's almost guaranteed to hit, but it does limit some continuous damage for some bosses.
-Difficult level to work with. He is not capable of going up to a decent level from Matilda alone.
-Not useable during Luca or the Worm due to levels/deaths.

Millie
I didn't get too far in routing her out because the convenient times to recruit her pretty seriously hamper Jowy's experience conditioning for Sindar. She would compare similarly to Oulan though.

Flik + Nina Unite Attack
I put this in my list, but I haven't actually though about how it would work yet. It's a 2.5x Amp, but using two characters is a unique requirement and I'm not sure how Unite Attacks play with Violence or double-strike yet. Probably not worth it in the long run, but I'll think through it more soon.

Luca Strategies
Luca is still a big pain in my side overall, but I worked out a couple of alternative strats based on the Solo Riou bug and Kindness+Violence without double beat. Interestingly they all seem to take around the same amount of time, so it just depends on what's more convenient for the given route. All three take 2:30-3:00 to take Luca out; for comparison, Double Beat strategies could take him out in 1:15 without any luck needed.

Exertion - By far the simplest strategy. Go out of your way to grab an exertion rune in Kobold Village, slap it on Riou and then just wail away on Luca.

Violence + Fire Lizard - Matches up well since Riou should have Fire Lizard on already, but a little more risky. It's more consistent with a Fire Emblem equipped. Basically self-proc Violence and rely on Jizos and Mega Medicines to keep you going. Crits make it unpredictable though.

Nanami Violence w/o DB - This relies on the same strat as with Double Beat, but requires at least one additional turn to make it feasible. Use Luc to proc Violence twice and just have her lay into him. Some support healing/damage from Riou is necessary.

L'Ren Optimization
Changing up the damage sources also led to a pretty significant re-route of L'Ren. Most battles are still cake, but the one thing I wanted to try to work in was a 1-turn Beast Rune kill. At this point, it doesn't look possible without significant reorganization. The main issue is that the Beast Rune's Rune is resistant to all non-Unite forms of damage, so the 3300 is effectively 6600 HP. If I had a way to kill it in just one attack, I could kill both heads immediately after through a combination of Unite Magics and Kindness damage, but it just doesn't seem possible without gimping the followup damage. For whatever reason though Unite Magic isn't halved against the Rune, so that's still fine to use. The new strategy also nixes Stallion in favor of Sierra for a guaranteed last member with a lvl 4 MP to use with Unites. Stallion can get a level 4 MP by that point, but on average only after he hits level 58 or 59 (which is after Seed and Culgan).

=====================================

Those are the highlights of my testing so far. At this point, I think I've settled on a route that at least from drawing up all the costs seems to be fastest, and approximately 3 minutes faster than the Double Beat route. The goal is to use Nanami as the Kindness Bearer through Luca, but then swap to Shin as a raw damage source for the rest of the game. This makes some of the boss planning very interesting, and things aren't quite finalized yet. I am refining my short-version notes to reflect the necessary steps, and I'll post them soon.

All of the above still needs a moderate amount of actual testing, but as things stand I don't think you can get much tighter optimization than what I discussed in the above paragraph. I still need to flesh out some more early game items, but we're getting toward a final route.

Hello all, long time no post.

I've been working on a complete re-route for the middle and ending of the run, with some interesting successes and failures. My general notes from testing are too disorganized to share here, but here is a brief list of the things I tried along with some pros and cons:

[b]No Double Beat[/b]
This was one of the main reasons for searching out a new route, and overall I think I've cracked the benefits of double beat by other, non-random means. The main benefit of double beat is just a straight 2x damage amp, and it can go to any character with an extra rune slot. The 2x stacks with Violence and it is repeatable ad nauseum (ie, no use limits). This is great, but what I really needed was efficiency. Double Beat made things simple, but in many of the cases it was overkill. So I started looking for a better way. Removing double beat also makes for a much more consistent early game since experience conditioning will always use the same enemy mobs.

[b]Kindness Early[/b]
Another point of contention that came up is Kindness Early. I explored new strats for going through SDS, Abom, and Rat without Kindness at all vs having Kindness without Double Beat, and in the long run Kindness Early still makes sense. The difference isn't big, but it is there. It also eases up some of the opportunities to attach and swap runes as needed.

[b]Kindness Bearer[/b]
This is also something I struggled with and only last night settled on something reasonable. Kindness is very powerful not just because of the damage, but also because it carries to any other attack runes or unite attacks. So I started to focus pretty strongly on these things in the hunt for new strategies. The old strategy relies on Nanami as the Kindness Bearer up through Bone Dragon, which she is decently suited for since you have to farm deaths regardless of whether she has Kindness. Her Tech is also quite good, so she has no trouble hitting things once she gets to ~Level 45. She loses quite a bit of power from not having Double Beat however. Sheena, as the backup, also works decently enough but he still relies on the amp from only double beat and double-strike, capping his damage output without Violence at about 3x. 3k damage is still a lot for many bosses, so instead I tried to find alternatives that mixed in magic usage to help out and get a "cleaner" finish. I narrowed the list down to these candidates:

[u]Zamza[/u]
+Self-perpetuating Kindness bearer. Will likely kill himself after using a powered up Fire Dragon, thus only the first death is "out of the way."
+Automatic Violence procs. Using Fire Lizard once will deal approximately 2k damage, so with a Fire Emblem this is mitigated to less than 500 damage returned. At level 70 (which I routed him to be) this procs Violence but leaves him about 200 HP to spare.
+No switching necessary. Put it on him once and it will stay on the whole game.
+Quick to recruit and set up.
-Fire element damage. This isn't terrible, but means some specific bosses need to be worked out differently.
-Somewhat inconsistent battles. After the first Fire Dragon, he'll have somewhat low HP. The high level gives him high enough defensive stats to survive most attacks after, but it is still somewhat risky.
-Fairly low TECH. This needs to be supplemented by attaching Thunder Amulet(s) or else he risks not being able to attack some bosses.
-Poor Armor selection. Being limited to Robes is a bit of a killer in terms of other ways to increase his TECH.

[u]Meg[/u]
+Trick is a very fast (time-wise), very powerful Attack Rune with x3 damage amp. She also has an extra rune slot for Double-Strike to bring it up to an instant 4.5x amp. Triggering Violence gives it a second turn power of 13.5x, which will kill any boss in the game instantly.
+Her HP and MDEF growth mimic Nanami's. This means that I can almost always set up a Luc Empty World -> Violence until after Tinto.
-Longish recruitment time. She's a little out of the way, and takes approximately 1:20 to recruit on the way to Matilda. She also requires Castle lvl 2, which is not a necessity otherwise (although I'd probably get it anyway for Jeane during Kiba).
-Initial level does not work well for leveling during Matilda, nor do I have good opportunities to level her or kill her off before Luca.
-Her TECH is only OK, but is not strong enough without being leveled to take on Luca. It wouldn't make sense, and would rely on significant luck to actually hit Luca.
-One use only. This means that if I'm going to use it, I need to make sure it hits. This can be somewhat helped by Thunder Amulets, but she only has two accessory slots and one of them should be a Jizo.
-Not useable during Luca or the Worm due to level/deaths.

[u]Oulan[/u]
+Angry Dragon syncs well with Fury/Violence. x2 Damage for free, and reuseable when Berserk is great (meaning x3 amp total).
+Beefy. Lots of HP to work with, and decent defenses.
-Beefy. Makes it that much harder to ensure Violence procs.
-Recruitment time is about a minute, but there's also no easy way to farm deaths on her.

[u]Shin[/u]
+Strong 3x amp attack. Takes a somewhat longish time to perform, but 3x does serious damage to any enemy.
+Amazing TECH and LUCK. These feed into more crits, but also mean I don't have to get TECH boosting items in order to make him effective.
+Decent armor selection. Gotta get that Master Garb.
-No open rune slots. This means further damage amp has to come organically (ie, from Battle Oath or items) and he is not compatible with Violence.
-One use only. Not as big of a deal since he's almost guaranteed to hit, but it does limit some continuous damage for some bosses.
-Difficult level to work with. He is not capable of going up to a decent level from Matilda alone.
-Not useable during Luca or the Worm due to levels/deaths.

[u]Millie[/u]
I didn't get too far in routing her out because the convenient times to recruit her pretty seriously hamper Jowy's experience conditioning for Sindar. She would compare similarly to Oulan though.

[u]Flik + Nina Unite Attack[/u]
I put this in my list, but I haven't actually though about how it would work yet. It's a 2.5x Amp, but using two characters is a unique requirement and I'm not sure how Unite Attacks play with Violence or double-strike yet. Probably not worth it in the long run, but I'll think through it more soon.

[b]Luca Strategies[/b]
Luca is still a big pain in my side overall, but I worked out a couple of alternative strats based on the Solo Riou bug and Kindness+Violence without double beat. Interestingly they all seem to take around the same amount of time, so it just depends on what's more convenient for the given route. All three take 2:30-3:00 to take Luca out; for comparison, Double Beat strategies could take him out in 1:15 without any luck needed.

[u]Exertion[/u] - By far the simplest strategy. Go out of your way to grab an exertion rune in Kobold Village, slap it on Riou and then just wail away on Luca.

[u]Violence + Fire Lizard[/u] - Matches up well since Riou should have Fire Lizard on already, but a little more risky. It's more consistent with a Fire Emblem equipped. Basically self-proc Violence and rely on Jizos and Mega Medicines to keep you going. Crits make it unpredictable though.

[u]Nanami Violence w/o DB[/u] - This relies on the same strat as with Double Beat, but requires at least one additional turn to make it feasible. Use Luc to proc Violence twice and just have her lay into him. Some support healing/damage from Riou is necessary.

[b]L'Ren Optimization[/b]
Changing up the damage sources also led to a pretty significant re-route of L'Ren. Most battles are still cake, but the one thing I wanted to try to work in was a 1-turn Beast Rune kill. At this point, it doesn't look possible without significant reorganization. The main issue is that the Beast Rune's Rune is resistant to all non-Unite forms of damage, so the 3300 is effectively 6600 HP. If I had a way to kill it in just one attack, I could kill both heads immediately after through a combination of Unite Magics and Kindness damage, but it just doesn't seem possible without gimping the followup damage. For whatever reason though Unite Magic isn't halved against the Rune, so that's still fine to use. The new strategy also nixes Stallion in favor of Sierra for a guaranteed last member with a lvl 4 MP to use with Unites. Stallion can get a level 4 MP by that point, but on average only after he hits level 58 or 59 (which is after Seed and Culgan).

=====================================

Those are the highlights of my testing so far. At this point, I think I've settled on a route that at least from drawing up all the costs seems to be fastest, and approximately 3 minutes faster than the Double Beat route. The goal is to use Nanami as the Kindness Bearer through Luca, but then swap to Shin as a raw damage source for the rest of the game. This makes some of the boss planning very interesting, and things aren't quite finalized yet. I am refining my short-version notes to reflect the necessary steps, and I'll post them soon.

All of the above still needs a moderate amount of actual testing, but as things stand I don't think you can get much tighter optimization than what I discussed in the above paragraph. I still need to flesh out some more early game items, but we're getting toward a final route.

Lag.Com

2015-02-13 10:17:19 am

sda loyalist

I'm excited to see such an amazingly optimised run!

Omnigamer

Edit history:

Omnigamer: 2015-02-13 11:39:17 am

2015-02-13 11:38:55 am

All the things

I looked at other amped attack users as well, but the only ones that may be possibilities I left off the list for this or that reason. Wakaba is one of them, and takes 1:30 to recruit, but she suffers a lot of the same problems as Meg and has lower damage output. Kasumi/Valeria/Anita are all also possibilities, but the point in time that you get them is not helpful for farming deaths. At the end of the day, the burden lies a lot more on the setup time rather than the end battle savings.

I opened the debugger last night to poke around at a few things.

-Viki has a 2% chance to teleport you incorrectly. I think it is constant throughout the game.
-Escape chances vary by map region and your relative party's level. If you don't satisfy "Let Go" conditions, when you try to run the game checks to see if your party's level is greater than 2x, 1.5x, or just 1x the enemy's party level. It assigns a constant based on that, then compares it with a random value between 0 and 99. Common constants I've seen are typically between 18% and 30%, but I don't have a good way to look up what they are for each area. North Sparrow Pass at least seems to be 30% fail rate regardless of party levels.
-I double-checked the Hit% formula and I'm still pretty sure it is correct.

I looked at other amped attack users as well, but the only ones that may be possibilities I left off the list for this or that reason. Wakaba is one of them, and takes 1:30 to recruit, but she suffers a lot of the same problems as Meg and has lower damage output. Kasumi/Valeria/Anita are all also possibilities, but the point in time that you get them is not helpful for farming deaths. At the end of the day, the burden lies a lot more on the setup time rather than the end battle savings.

I opened the debugger last night to poke around at a few things.

-Viki has a 2% chance to teleport you incorrectly. I think it is constant throughout the game.
-Escape chances vary by map region and your relative party's level. If you don't satisfy "Let Go" conditions, when you try to run the game checks to see if your party's level is greater than 2x, 1.5x, or just 1x the enemy's party level. It assigns a constant based on that, then compares it with a random value between 0 and 99. Common constants I've seen are typically between 18% and 30%, but I don't have a good way to look up what they are for each area. North Sparrow Pass at least seems to be 30% fail rate regardless of party levels.
-I double-checked the Hit% formula and I'm still pretty sure it is correct.

Vaslof

2015-02-19 05:51:25 am

6h 43min run by neoslash72: http://www.twitch.tv/neoslash72/c/6139087

Omnigamer

Edit history:

Omnigamer: 2015-02-19 11:33:19 pm

2015-02-19 10:55:36 pm

All the things

I did a lot of testing over the last couple days to try and determine timing differences between different versions or platforms for the game. I tested several specific, repeatable scenarious:

-The amount of time between when the screen freezes and the menu text appears for the first battle against soldiers during the prologue.
-The amount of time the "Buddy Attack" Unite attack takes against the full set of 4 enemies. Timing ends when their damage values disappear.
-Using Buddy Attack again against the lone commander.
-Time taken from the last dialog box disappearing before the prologue movie until Viktor's face appears above the text box.
-25 round trips into Muse and back to the World Map.

My results are summarized in this doc:

https://docs.google.com/spreadsheets/d/1mjT6y-sXj-mWebF1UvvM1VVH341rPR7NpfIm5PG4jhM/edit?usp=sharing

Basically, FDS and PSTV are king. Anything with FDS is head and shoulders above anything else, but PSTV by itself fairs on par with FDS for either PS2. Don't even bother with PS1 or playing with disc on PS3.

Some other observations from this data:
-For what it's worth, PS3 with Disc is pretty close to playing on the fat PS2. In that regard, it gets the emulation pretty accurate; the discrepancies are very minor.
-The different scenarios test different effects and how well the platform handles them. Battle Opening deals with rendering and processing ability. Unite attack deals with special moves with many sprites or few sprites, so you can see the impact of sprite count on lag. The only platform unaffected by this difference is PSTV with FDS. The cutscene is a long sequence of events that should be matched to music and has no user interaction, so this is purely on how well the game can load up the scenes and display them. The muse round trips show an aggregate time from transitioning between overworld and a town/dungeon, which deals with loading and swapping different scripts and resources.
-Digital versions, understandably, are more consistent on the Muse test. This is simply due to the fact that it doesn't have to wait for a disc seek, which can take a variable amount of time.
-While the given data indicates that PS3 PSN version is a superior choice to any of the disc-based but non-FDS versions, I have noticed significant lag in L'Ren that is not present in other versions. Thus, these results shouldn't be considered all-encompassing, as there are still cases that may benefit/penalize particular versions.
-The PSTV FDS is the only FDS variety that does not play some music tracks at double speed.

Some other things that I'd like to test are certain magics, but I didn't have a way to test it across all versions. At the very least I can narrow the field and come up with some setups to look at, but it will take a lot to dissuade from using PSTV FDS at this point. I still need to give it a full and proper playthrough, but this could be a potentially huge amount of time. The speedup on the PS2 versions (no FDS) can also probably help to explain some of the speed discrepancies between my older runs (PS3 w/ Disc) and JP runs (slim PS2) but there's a lot more that could be at play there.

Not included are the Chinese PC version and the Suikoden I & II remake for PSP. While I'd like to compare them, they're different enough that they may as well be unique games. Both have the potential to be significantly faster than any of the standard versions, but the language barrier and technical changes make them a challenge to work on.

EDIT:

Forgot to mention I also did some checking on diagonal movement. Every step you can take against a diagonal edge saves 4 frames vs running to the same spot without the slope (8 frames if walking). This applies equally to World Map movement. So it can add up a fair bit if movement is optimized to make the most of the diagonals. I'll try to work out optimized pathing when I get to the point of weekly practice.

I did a lot of testing over the last couple days to try and determine timing differences between different versions or platforms for the game. I tested several specific, repeatable scenarious:

-The amount of time between when the screen freezes and the menu text appears for the first battle against soldiers during the prologue.
-The amount of time the "Buddy Attack" Unite attack takes against the full set of 4 enemies. Timing ends when their damage values disappear.
-Using Buddy Attack again against the lone commander.
-Time taken from the last dialog box disappearing before the prologue movie until Viktor's face appears above the text box.
-25 round trips into Muse and back to the World Map.

My results are summarized in this doc:

https://docs.google.com/spreadsheets/d/1mjT6y-sXj-mWebF1UvvM1VVH341rPR7NpfIm5PG4jhM/edit?usp=sharing

Basically, FDS and PSTV are king. Anything with FDS is head and shoulders above anything else, but PSTV by itself fairs on par with FDS for either PS2. Don't even bother with PS1 or playing with disc on PS3.

Some other observations from this data:
-For what it's worth, PS3 with Disc is pretty close to playing on the fat PS2. In that regard, it gets the emulation pretty accurate; the discrepancies are very minor.
-The different scenarios test different effects and how well the platform handles them. Battle Opening deals with rendering and processing ability. Unite attack deals with special moves with many sprites or few sprites, so you can see the impact of sprite count on lag. The only platform unaffected by this difference is PSTV with FDS. The cutscene is a long sequence of events that should be matched to music and has no user interaction, so this is purely on how well the game can load up the scenes and display them. The muse round trips show an aggregate time from transitioning between overworld and a town/dungeon, which deals with loading and swapping different scripts and resources.
-Digital versions, understandably, are more consistent on the Muse test. This is simply due to the fact that it doesn't have to wait for a disc seek, which can take a variable amount of time.
-While the given data indicates that PS3 PSN version is a superior choice to any of the disc-based but non-FDS versions, I have noticed significant lag in L'Ren that is not present in other versions. Thus, these results shouldn't be considered all-encompassing, as there are still cases that may benefit/penalize particular versions.
-The PSTV FDS is the only FDS variety that does not play some music tracks at double speed.

Some other things that I'd like to test are certain magics, but I didn't have a way to test it across all versions. At the very least I can narrow the field and come up with some setups to look at, but it will take a lot to dissuade from using PSTV FDS at this point. I still need to give it a full and proper playthrough, but this could be a potentially huge amount of time. The speedup on the PS2 versions (no FDS) can also probably help to explain some of the speed discrepancies between my older runs (PS3 w/ Disc) and JP runs (slim PS2) but there's a lot more that could be at play there.

Not included are the Chinese PC version and the Suikoden I & II remake for PSP. While I'd like to compare them, they're different enough that they may as well be unique games. Both have the potential to be significantly faster than any of the standard versions, but the language barrier and technical changes make them a challenge to work on.

EDIT:

Forgot to mention I also did some checking on diagonal movement. Every step you can take against a diagonal edge saves 4 frames vs running to the same spot without the slope (8 frames if walking). This applies equally to World Map movement. So it can add up a fair bit if movement is optimized to make the most of the diagonals. I'll try to work out optimized pathing when I get to the point of weekly practice.

Omnigamer

2015-02-23 12:31:32 pm

All the things

Quick update. I'm still working toward fine-tuning a few things, but I think I've pretty well settled on the main route at this point. The unfortunate news: it looks like getting double beat is still the theoretical fastest way to proceed. The good news: I've optimized a lot of the fights and otherwise such that its gains are fairly small; the difference of double beat is likely to make the difference of only a minute or so assuming you can get one with only 30 seconds of farming (aka, one battle). Thus I'm going to be skipping it for the sake of consistency in my upcoming runs, since the loss of a minute can be offset further by better overall consistency of practice.

Also, PSTV seems like it will have a huge impact on the end time by itself. In 40 minutes of practice, I was ahead by about 4 minutes for a fairly average performance; extrapolated out to 6.5 hours, this means it could save up to 40 minutes by itself. This on top of the other route changes means that times under 6 hours can be expected soon, with potential to be in the 5:45 range with good luck and all the route changes. That's a much more palatable time for a run.

The last piece I wanted to mention is the consistent strat I developed for the Beast Rune. The biggest problems with the Beast Rune stem from the fact that it starts with the barrier in place and that the Rune itself cannot be killed in a single attack. Thus, I wasn't able to find any reasonable way to pull off a 1-turn kill with an any% party. My other strats for killing it in two turns were also moderately inconsistent, due to the left leg's ability to cast status effects and disable certain party members. Then I stumbled on a way to get a guaranteed 2T kill against the Beast Rune by instead switching up and focusing some damage on the Left Leg itself. The strategy is fairly complex and requires getting Mother Earth and forcing certain members' speed stats to be within specific ranges, but it is otherwise consistent and doesn't allow revives. This is a huge help since it means I don't have to rely on Sheena and the Earth Shield. I need to spend a little bit of time and write out the exact requirements, but the final party turns into this:

Rina - Uniter, BG + Fire + Water
Shin - Kindness
Flik - Lightning + Rage
Sierra - Darkness
Riou - Earth + Lightning
Luc - Mother Earth + Water + Lightning

Quick update. I'm still working toward fine-tuning a few things, but I think I've pretty well settled on the main route at this point. The unfortunate news: it looks like getting double beat is still the theoretical fastest way to proceed. The good news: I've optimized a lot of the fights and otherwise such that its gains are fairly small; the difference of double beat is likely to make the difference of only a minute or so assuming you can get one with only 30 seconds of farming (aka, one battle). Thus I'm going to be skipping it for the sake of consistency in my upcoming runs, since the loss of a minute can be offset further by better overall consistency of practice.

Also, PSTV seems like it will have a huge impact on the end time by itself. In 40 minutes of practice, I was ahead by about 4 minutes for a fairly average performance; extrapolated out to 6.5 hours, this means it could save up to 40 minutes by itself. This on top of the other route changes means that times under 6 hours can be expected soon, with potential to be in the 5:45 range with good luck and all the route changes. That's a much more palatable time for a run.

The last piece I wanted to mention is the consistent strat I developed for the Beast Rune. The biggest problems with the Beast Rune stem from the fact that it starts with the barrier in place and that the Rune itself cannot be killed in a single attack. Thus, I wasn't able to find any reasonable way to pull off a 1-turn kill with an any% party. My other strats for killing it in two turns were also moderately inconsistent, due to the left leg's ability to cast status effects and disable certain party members. Then I stumbled on a way to get a guaranteed 2T kill against the Beast Rune by instead switching up and focusing some damage on the Left Leg itself. The strategy is fairly complex and requires getting Mother Earth and forcing certain members' speed stats to be within specific ranges, but it is otherwise consistent and doesn't allow revives. This is a huge help since it means I don't have to rely on Sheena and the Earth Shield. I need to spend a little bit of time and write out the exact requirements, but the final party turns into this:

Rina - Uniter, BG + Fire + Water
Shin - Kindness
Flik - Lightning + Rage
Sierra - Darkness
Riou - Earth + Lightning
Luc - Mother Earth + Water + Lightning

Omnigamer

2015-03-09 08:49:56 pm

All the things

Some minor updates:

-Final game plan has to change slightly since Flik is removed from the party after Bone Dragon. Instead, the goal is to keep one Shredding left over during L'Ren and use that from Sierra instead.
-Similarly, cut down on the runes to attach. Flik just stays with Lightning unless I figure out a better point in the story to throw something on him. Rina doesn't need anything other than Fire, but I still take her along during Bone Dragon to fetch some levels.
-I swear during practice I was able to Instant Kill the Gold Wolf with Sierra doing Darkness Lvl 3, but I haven't been able to reproduce it. Made me wonder about the actual IK resistances for other bosses though.
-I found a minor skip (!) for one of the battles in the Greenhill Liberation. You can skip the very last soldier group by just hugging the building on the right. Now I'm going to go back through and check to see if there are ways to skip any other forced encounters.

mrprmiller

2015-03-09 08:59:49 pm

Everyday is puppies and sunshine...

Keep going. I'm watching this like a hawk. Hehehe. Cheesy

Omnigamer

2015-03-17 11:33:47 pm

All the things

So I need to decide how I should proceed. My test runs for it on PSTV indicate that I should be shooting for a 6:15 time or so, with the upper limit around 6:10. That's pretty good, but still a significant time sink.

Meanwhile, I recently acquired the digital version of the Suikoden I + II collection. Based on some meager testing, it's faster yet. The main reasoning is the diagonal movement, but most of the screen transitions seem to be faster than even PSTV with FDS, so that's attractive. It also carries the "triple speed" magic and unites of the PC version, which is a nice bonus. I estimate an additional 15-20 minutes can be saved with that version, not counting any skips that might be possible due to the diagonal movement. The only downside is no Kindness Early, which I can live with. Having to learn the Duel quotes will also suck, but is reasonable.

But to that end, it's functionally a different game. The diagonal movement, faster magic, and general changes to some screens make it incomparable to other versions. So I think I will probably stick with PSTV w/ FDS for the time being as it is still the "stock" game. I might to a for-fun run of the Collection version afterwards, but it's not something I have plans for long-term.

That said, if some other Collection-specific skips and timesavers are found, that makes it that much more tantalizing. It's just that so much would have to be re-tested for it that I feel like it would burn me out if I started on it. It is probably the fastest overall version of the game, however. It's just not worth enough to switch if the savings are only 10-15 minutes instead of 30-60.

So I need to decide how I should proceed. My test runs for it on PSTV indicate that I should be shooting for a 6:15 time or so, with the upper limit around 6:10. That's pretty good, but still a significant time sink.

Meanwhile, I recently acquired the digital version of the Suikoden I + II collection. Based on some meager testing, it's faster yet. The main reasoning is the diagonal movement, but most of the screen transitions seem to be faster than even PSTV with FDS, so that's attractive. It also carries the "triple speed" magic and unites of the PC version, which is a nice bonus. I estimate an additional 15-20 minutes can be saved with that version, not counting any skips that might be possible due to the diagonal movement. The only downside is no Kindness Early, which I can live with. Having to learn the Duel quotes will also suck, but is reasonable.

But to that end, it's functionally a different game. The diagonal movement, faster magic, and general changes to some screens make it incomparable to other versions. So I think I will probably stick with PSTV w/ FDS for the time being as it is still the "stock" game. I might to a for-fun run of the Collection version afterwards, but it's not something I have plans for long-term.

That said, if some other Collection-specific skips and timesavers are found, that makes it that much more tantalizing. It's just that so much would have to be re-tested for it that I feel like it would burn me out if I started on it. It is probably the fastest overall version of the game, however. It's just not worth enough to switch if the savings are only 10-15 minutes instead of 30-60.

Omnigamer

Edit history:

Omnigamer: 2015-04-01 11:52:44 am

Omnigamer: 2015-03-31 01:48:24 pm

2015-03-31 01:00:54 pm

All the things

I would like to start attempts soon, however I have some issues with the Beast Rune strategy. The main problem is that the current strats assume that Blazing Camp gets a speed modifier of 1.3x, which it did under my initial tests. However I have had several runs and otherwise make it to that point where the speed modifier does not trigger, and I cannot seem to figure out why. I have three different files at this point, 2 of which correctly apply the speed modifier and then one that does not. The only differences between them is a marginal difference in the speed stats of Flik, but adding any amount of extra speed equipment to him does not change the outcome. I have tried a bunch of things, but nothing seems to affect it yet... which means I'll have to do some digging in the code. This is quite literally the last thing that needs to be completed before the route is effectively complete, so I'm anxious to find a way to make it consistent. The alternative is re-working the strategy to get the same result but not require a particular order, but I'm not sure that can be done without excessive setup.

EDIT: May have found a setup that should work regardless of speed buffs. I will test it out and see what happens, but the damage is close.

EDIT2: New strat does work, and may in fact save up to a minute over the course of the run. It requires changing some party members from Tinto onwards though, so I need to fix up the notes some more.

I would like to start attempts soon, however I have some issues with the Beast Rune strategy. The main problem is that the current strats assume that Blazing Camp gets a speed modifier of 1.3x, which it did under my initial tests. However I have had several runs and otherwise make it to that point where the speed modifier does not trigger, and I cannot seem to figure out why. I have three different files at this point, 2 of which correctly apply the speed modifier and then one that does not. The only differences between them is a marginal difference in the speed stats of Flik, but adding any amount of extra speed equipment to him does not change the outcome. I have tried a bunch of things, but nothing seems to affect it yet... which means I'll have to do some digging in the code. This is quite literally the last thing that needs to be completed before the route is effectively complete, so I'm anxious to find a way to make it consistent. The alternative is re-working the strategy to get the same result but not require a particular order, but I'm not sure that can be done without excessive setup.

EDIT: May have found a setup that should work regardless of speed buffs. I will test it out and see what happens, but the damage is close.

EDIT2: New strat does work, and may in fact save up to a minute over the course of the run. It requires changing some party members from Tinto onwards though, so I need to fix up the notes some more.

Omnigamer

Edit history:

Omnigamer: 2015-05-26 01:12:37 pm

2015-05-19 08:24:16 pm

All the things

Hey all, my notes are somewhat finished. You can view them here:

https://docs.google.com/document/d/1xrFfLEfSzuV9bFYcMLhduMuwK3bdBCK5lLDoo5-oxiA/edit?usp=sharing

There's still a couple things I want to iron out, but for the most part I think the route is pretty solid. Some of the remaining pieces:

-I want to eliminate some of the menuing to move around Shredding before/after Bone Dragon, but I haven't come up with anything that still works out consistently. The main issue is that Riou is just a little lacking on damage to take out Bone Dragon, and there's not really anybody else who has adequate speed and also is in the final party.
-The strat for Seed and Culgan can probably use a little tweaking. The main things will be avoiding use of Empty World, but I do like the rest of it right now since I can use it to tell which spells do and don't have the speed buff.
-Revised the Mist Monster "consistent" strat to an actually consistent strat, based on my misfortunes during attempts the other day.
-There's still some wiggle room for the Beast Rune strat. The current strat goes for consistency regardless of what the buffs are; it can be adjusted based on knowledge of the buffs to be faster and redistribute speed items accordingly. I decided against it while making the notes because I felt like all the extra info-gathering and decision-making would hinder pure execution.
-Another consideration is picking up the Wind Hat in Tinto Mines to further augment speed and make Beast Rune consistent, but it's not used in the current route and takes some ~20 seconds to pick up.
-There's the possibility for some further loading zone skips that need to be worked out, for example in Two River. May or may not turn into something useful long-term.
-There's still room for Double Beat to make a return, but it only really affects the early-game.

Anyway, comments are enabled on the document. Feel free to leave some or make a copy of the notes for your own uses.

EDIT: One more minor change that isn't yet reflected in the notes, but we no longer need to get Stallion at all. The time he saves is slightly less than the time it takes to recruit him, and this gives the possibility to not get Gilbert at all.

Hey all, my notes are somewhat finished. You can view them here:

https://docs.google.com/document/d/1xrFfLEfSzuV9bFYcMLhduMuwK3bdBCK5lLDoo5-oxiA/edit?usp=sharing

There's still a couple things I want to iron out, but for the most part I think the route is pretty solid. Some of the remaining pieces:

-I want to eliminate some of the menuing to move around Shredding before/after Bone Dragon, but I haven't come up with anything that still works out consistently. The main issue is that Riou is just a little lacking on damage to take out Bone Dragon, and there's not really anybody else who has adequate speed and also is in the final party.
-The strat for Seed and Culgan can probably use a little tweaking. The main things will be avoiding use of Empty World, but I do like the rest of it right now since I can use it to tell which spells do and don't have the speed buff.
-Revised the Mist Monster "consistent" strat to an actually consistent strat, based on my misfortunes during attempts the other day.
-There's still some wiggle room for the Beast Rune strat. The current strat goes for consistency regardless of what the buffs are; it can be adjusted based on knowledge of the buffs to be faster and redistribute speed items accordingly. I decided against it while making the notes because I felt like all the extra info-gathering and decision-making would hinder pure execution.
-Another consideration is picking up the Wind Hat in Tinto Mines to further augment speed and make Beast Rune consistent, but it's not used in the current route and takes some ~20 seconds to pick up.
-There's the possibility for some further loading zone skips that need to be worked out, for example in Two River. May or may not turn into something useful long-term.
-There's still room for Double Beat to make a return, but it only really affects the early-game.

Anyway, comments are enabled on the document. Feel free to leave some or make a copy of the notes for your own uses.

EDIT: One more minor change that isn't yet reflected in the notes, but we no longer need to get Stallion at all. The time he saves is slightly less than the time it takes to recruit him, and this gives the possibility to not get Gilbert at all.

Omnigamer

2015-06-03 06:22:14 am

All the things

As a heads-up, I completely figured out the trading system. It's somewhat complicated, so it's still in pseudo-code. Copy-pasta from suikosource below:

Code:

SHOP PSEUDOCODE

cur_time = Get_Minutes();
last_accessed = shop.last_accessed;

if(last_accessed > cur_time)
	cur_time = 1;
	
time_dif = cur_time - last_accessed;

if(time_dif < shop.refresh_period){
	Skip_Updates();
	return;
}
else{
	for(i=0;i<shop.total_items;i++){
		
		new_quantity = shop.item[i].current_quantity;
		
		if(item[i].refresh_period <= time_dif){
		
			if(item[i].standard_quantity > shop.item[i].current_quantity){ //try to normalize towards standard
				new_quantity += time_dif/item[i].refresh_period;
				
				if(new_quantity > item[i].standard_quantity)
					new_quantity = item[i].standard_quantity; //If you overshoot, then just set to standard
			}
			else{
				new_quantity -= time_dif/item[i].refresh_period;
				
				if(new_quantity < item[i].standard_quantity)
					new_quantity = item[i].standard_quantity;
			}
			
			quantity_rand = RNG(-1:1);
			
			temp = quantity_rand * item[i].swing; //multiply sets by RNG
			
			new_quantity = abs(new_quantity + temp - (item[i].swing/2));
			
		}
		
		price_mod = item[i].standard_price * (item[i].standard_quantity - new_quantity);
		
		if(item[i].standard_quantity > 2){
			price_mod = price_mod / item[i].standard_quantity; //if it's a high-stock item, divide price_mod
		}
		else{
			price_mod = abs((price_mod / 3)); //otherwise divide by 3
		}
		
		price_mod+= item[i].standard_price;
		
		if(price_mod > 65000)
			price_mod = 65000;
			
		refresh_cycles = time_dif / shop.refresh_period;
		
		if(refresh_cycles > 4)
			refresh_cycles = 4; //Cap refresh_cycles to 4
			
		final_price = (price_mod - shop.item[i].prior_price) * refresh_cycles;
		
		if(final_price < 0)
			final_price+=3;
			
		final_price = final_price/4; //Divide by 4
		
		price_rand = RNG(0:10);
		
		final_price = price_rand + shop.item[i].prior_price+ price_rand - 5;
		
		other_price = (2/3) * item[i].standard_price;
		
		if(final_price < (other_price/2))
			final_price = (other_price/2);
			
		if(final_price>65000)
			final_price = 65000;
			
		item[i].current_price = final_price;
		item[i].current_quantity = new_quantity;
	}
}

I was missing a few key details in the prior version. Before purchasing for the first time, the initialization code is fairly straightforward. The main idea is that price is set based on the new quantity, which has one of three possible values from the RNG. The "swing" of an item is the amount that it varies as a result of this RNG. It is centered at 0, but the "high-stock" value should be about half of the swing value above the standard quantity, while the "low-stock" value will be 1.5x the swing value below the standard. The average case is .5x swing value below standard. Price follows pretty simply for initialization.

If you're revisiting the shop, however, things get a bit trickier. Items follow a "restocking" or "consumption" pattern if you've bought/sold the good before. You can think of it as the shop gradually moving back towards its standard quantity, either by consuming excess goods or accumulating missing stock. This would normally by awesome, however the shop also keeps track of the last price you purchased your good for and uses it as part of the calculations. This prevents the price from suddenly skyrocketing if you buy up all their goods; the change is gradual, and only accumulates as the shop cycles accumulate. It's critical to note that the shop cycle is what determines the pricing, and not the item cycles.

I have full tables for the trading posts with the standard quantities, prices, sets, and refresh periods. You can check them out here:

https://docs.google.com/spreadsheets/d/1_E96EBPnUoA1bQcUFOUZkVsjeLdW1T5BfzX8v1Bil-8/edit?usp=sharing

As far as analysis of this, I'm still working on it. I at least reproduced something to predict shop prices, but I might just be missing something as far as the quantity RNG. Looking at the code it seems like there should be 3 distinct possibilities, but in practice I only seem to ever see 2. As far as I can tell the math doesn't try to make the RNG value positive, so I'm not sure why it doesn't show up. In any case, this should get us started on determining best trading practices.

As a heads-up, I completely figured out the trading system. It's somewhat complicated, so it's still in pseudo-code. Copy-pasta from suikosource below:

[code]SHOP PSEUDOCODE

cur_time = Get_Minutes();
last_accessed = shop.last_accessed;

if(last_accessed > cur_time)
	cur_time = 1;
	
time_dif = cur_time - last_accessed;

if(time_dif < shop.refresh_period){
	Skip_Updates();
	return;
}
else{
	for(i=0;i<shop.total_items;i++){
		
		new_quantity = shop.item[i].current_quantity;
		
		if(item[i].refresh_period <= time_dif){
		
			if(item[i].standard_quantity > shop.item[i].current_quantity){ //try to normalize towards standard
				new_quantity += time_dif/item[i].refresh_period;
				
				if(new_quantity > item[i].standard_quantity)
					new_quantity = item[i].standard_quantity; //If you overshoot, then just set to standard
			}
			else{
				new_quantity -= time_dif/item[i].refresh_period;
				
				if(new_quantity < item[i].standard_quantity)
					new_quantity = item[i].standard_quantity;
			}
			
			quantity_rand = RNG(-1:1);
			
			temp = quantity_rand * item[i].swing; //multiply sets by RNG
			
			new_quantity = abs(new_quantity + temp - (item[i].swing/2));
			
		}
		
		price_mod = item[i].standard_price * (item[i].standard_quantity - new_quantity);
		
		if(item[i].standard_quantity > 2){
			price_mod = price_mod / item[i].standard_quantity; //if it's a high-stock item, divide price_mod
		}
		else{
			price_mod = abs((price_mod / 3)); //otherwise divide by 3
		}
		
		price_mod+= item[i].standard_price;
		
		if(price_mod > 65000)
			price_mod = 65000;
			
		refresh_cycles = time_dif / shop.refresh_period;
		
		if(refresh_cycles > 4)
			refresh_cycles = 4; //Cap refresh_cycles to 4
			
		final_price = (price_mod - shop.item[i].prior_price) * refresh_cycles;
		
		if(final_price < 0)
			final_price+=3;
			
		final_price = final_price/4; //Divide by 4
		
		price_rand = RNG(0:10);
		
		final_price = price_rand + shop.item[i].prior_price+ price_rand - 5;
		
		other_price = (2/3) * item[i].standard_price;
		
		if(final_price < (other_price/2))
			final_price = (other_price/2);
			
		if(final_price>65000)
			final_price = 65000;
			
		item[i].current_price = final_price;
		item[i].current_quantity = new_quantity;
	}
}[/code]

I was missing a few key details in the prior version. Before purchasing for the first time, the initialization code is fairly straightforward. The main idea is that price is set based on the new quantity, which has one of three possible values from the RNG. The "swing" of an item is the amount that it varies as a result of this RNG. It is centered at 0, but the "high-stock" value should be about half of the swing value above the standard quantity, while the "low-stock" value will be 1.5x the swing value below the standard. The average case is .5x swing value below standard. Price follows pretty simply for initialization.

If you're revisiting the shop, however, things get a bit trickier. Items follow a "restocking" or "consumption" pattern if you've bought/sold the good before. You can think of it as the shop gradually moving back towards its standard quantity, either by consuming excess goods or accumulating missing stock. This would normally by awesome, however the shop also keeps track of the last price you purchased your good for and uses it as part of the calculations. This prevents the price from suddenly skyrocketing if you buy up all their goods; the change is gradual, and only accumulates as the shop cycles accumulate. It's critical to note that the shop cycle is what determines the pricing, and not the item cycles.

I have full tables for the trading posts with the standard quantities, prices, sets, and refresh periods. You can check them out here:

https://docs.google.com/spreadsheets/d/1_E96EBPnUoA1bQcUFOUZkVsjeLdW1T5BfzX8v1Bil-8/edit?usp=sharing

As far as analysis of this, I'm still working on it. I at least reproduced something to predict shop prices, but I might just be missing something as far as the quantity RNG. Looking at the code it seems like there should be 3 distinct possibilities, but in practice I only seem to ever see 2. As far as I can tell the math doesn't try to make the RNG value positive, so I'm not sure why it doesn't show up. In any case, this should get us started on determining best trading practices.

Omnigamer

2015-06-09 10:08:35 am

All the things

I spent a lot of the past couple weeks thinking about True End. Most of the recruits have pretty obvious story points to pick them up during, so you don't waste much time. There are also a few obvious goals to shoot for, namely hitting Castle Level 3 before Gregminster to get all of those recruits without having to make a return trip. There are a few other subtle ones, however, that create some interesting scenarios to compare against. "Interesting" in this case unfortunately also means very difficult to directly compare, but I think I've worked out some routes for the most troublesome characters.

The trouble children are:

-Millie
-Templton
-Hix & Tengaar & Yuzu & Sid
-Yoshino
-Wakaba & Tony

Of these 9, you must recruit at least 5 in order to get to Castle Level 3, as there's only 67 recruits available and 63 are needed to get CL3.

I spent time coming up with creative ways to get Millie in the early game by using Bonaparte's experience to boost Jowy to the right experience, but unfortunately it involved some extra setup and about 30-40 seconds of extra battle time. In the end I just came to the conclusion that if I'm going to get her at all before CL3, it would just be faster to pick her up when I get Viki than to try and finagle a route for it early-game. More on that in a moment.

Templton is also pretty "out-of-the-way", but not as much as you might think. With some good planning he can be pretty reasonable to pick up early-game.

The Hix and Tengaar quest is its own beast, requiring a bunch of back-and-forth to complete it. It also provides a good opportunity for Sid as well though, who would cost 70 seconds to pick up out-of-band. Yuzu shows up after the quest is done for a quick and simple quest. So for the extra effort that would go into getting the quest done early, you wind up with 4 recruits. However, the gains are not entirely clear-cut.

Wakaba and Tony are simple, just a little out of the way. To get them early on with just Stallion incurs both a return trip of ~25 seconds and additional walking time for a second visit later on to get more recruits. Wakaba has that lovely Skill Ring though, which may further factor into boss strategies.

Yoshino is also simple, but with the caveat of requiring Freed Y in your party. To get her from one side trip to Radat requires going before Matilda rather than after Greenhill, which throws some of the other questing convenience for a loop. It's easy to just get her later when you visit for Banner Pass, but then she won't contribute to CL3 without a trek back to the castle.

I'm still in the process of figuring out what the best routes are in terms of costs vs benefits, but the combinations of recruits narrow down to:

H&T&Y&S and Yoshino - skips the trip to Forest Village, and requires some backtracking after Yoshino. Very minor savings to coincide with the trip to the Cave.
H&T&Y&S and Wakaba & Tony - makes a trip to Forest Village instead. Simple, straightforward, however it incurs some extra hidden costs that aren't immediately apparent. May also allow less than CL2 up to Greenhill to skip Emilia until later.
Millie, Templton, Yoshino, Wakaba, and Tony - a trickier path, but has several benefits. Saves visiting Kobold village until Tinto arc.

So about the difficulty of comparing these routes: it is really tricky to define what is and is not extra time. For example, most cases of the H&T&Y&S quest result in about 40 seconds lost to getting to and then returning from the Sewers. However, the time itself can be reduced by having the mirror and teleporting out on the other side of TR. The time can be further negated if you can consider also using that instance through TR to recruit Bob. So it's not entirely cut and dry to just compare the different paths based on the short-term merits. Similarly the trip to Forest Village costs time based on whether you can get everybody in just one fell swoop or not, which is a hard to define amount.

I'm still crunching numbers, but I think the current best overall way to handle this section is to get Millie and Templton immediately after Viki, pick up Yoshino on the way back through Radat, and then get a side visit to Forest Village. This can also be combined with the first half of the H&T quest, as you have convenient visits to SW and Cave on the way, but that has tradeoffs in the initial time spent going out of the way to visit Kobold Village and back. Kobold Village incurs costs when not visiting on the way to Tinto and from not having mirror for both TR and the final return out of Kobold. It does work out better for trading routes to have Kobold village available though, so it may come with some additional risk.

So yeah. Still working on a bunch. Comments welcome.

I spent a lot of the past couple weeks thinking about True End. Most of the recruits have pretty obvious story points to pick them up during, so you don't waste much time. There are also a few obvious goals to shoot for, namely hitting Castle Level 3 before Gregminster to get all of those recruits without having to make a return trip. There are a few other subtle ones, however, that create some interesting scenarios to compare against. "Interesting" in this case unfortunately also means very difficult to directly compare, but I think I've worked out some routes for the most troublesome characters.

The trouble children are:

-Millie
-Templton
-Hix & Tengaar & Yuzu & Sid
-Yoshino
-Wakaba & Tony

Of these 9, you must recruit at least 5 in order to get to Castle Level 3, as there's only 67 recruits available and 63 are needed to get CL3.

I spent time coming up with creative ways to get Millie in the early game by using Bonaparte's experience to boost Jowy to the right experience, but unfortunately it involved some extra setup and about 30-40 seconds of extra battle time. In the end I just came to the conclusion that if I'm going to get her at all before CL3, it would just be faster to pick her up when I get Viki than to try and finagle a route for it early-game. More on that in a moment.

Templton is also pretty "out-of-the-way", but not as much as you might think. With some good planning he can be pretty reasonable to pick up early-game.

The Hix and Tengaar quest is its own beast, requiring a bunch of back-and-forth to complete it. It also provides a good opportunity for Sid as well though, who would cost 70 seconds to pick up out-of-band. Yuzu shows up after the quest is done for a quick and simple quest. So for the extra effort that would go into getting the quest done early, you wind up with 4 recruits. However, the gains are not entirely clear-cut.

Wakaba and Tony are simple, just a little out of the way. To get them early on with just Stallion incurs both a return trip of ~25 seconds and additional walking time for a second visit later on to get more recruits. Wakaba has that lovely Skill Ring though, which may further factor into boss strategies.

Yoshino is also simple, but with the caveat of requiring Freed Y in your party. To get her from one side trip to Radat requires going before Matilda rather than after Greenhill, which throws some of the other questing convenience for a loop. It's easy to just get her later when you visit for Banner Pass, but then she won't contribute to CL3 without a trek back to the castle.

I'm still in the process of figuring out what the best routes are in terms of costs vs benefits, but the combinations of recruits narrow down to:

H&T&Y&S and Yoshino - skips the trip to Forest Village, and requires some backtracking after Yoshino. Very minor savings to coincide with the trip to the Cave.
H&T&Y&S and Wakaba & Tony - makes a trip to Forest Village instead. Simple, straightforward, however it incurs some extra hidden costs that aren't immediately apparent. May also allow less than CL2 up to Greenhill to skip Emilia until later.
Millie, Templton, Yoshino, Wakaba, and Tony - a trickier path, but has several benefits. Saves visiting Kobold village until Tinto arc.

So about the difficulty of comparing these routes: it is really tricky to define what is and is not extra time. For example, most cases of the H&T&Y&S quest result in about 40 seconds lost to getting to and then returning from the Sewers. However, the time itself can be reduced by having the mirror and teleporting out on the other side of TR. The time can be further negated if you can consider also using that instance through TR to recruit Bob. So it's not entirely cut and dry to just compare the different paths based on the short-term merits. Similarly the trip to Forest Village costs time based on whether you can get everybody in just one fell swoop or not, which is a hard to define amount.

I'm still crunching numbers, but I think the current best overall way to handle this section is to get Millie and Templton immediately after Viki, pick up Yoshino on the way back through Radat, and then get a side visit to Forest Village. This can also be combined with the first half of the H&T quest, as you have convenient visits to SW and Cave on the way, but that has tradeoffs in the initial time spent going out of the way to visit Kobold Village and back. Kobold Village incurs costs when not visiting on the way to Tinto and from not having mirror for both TR and the final return out of Kobold. It does work out better for trading routes to have Kobold village available though, so it may come with some additional risk.

So yeah. Still working on a bunch. Comments welcome.

Omnigamer

Edit history:

Omnigamer: 2015-06-12 01:12:19 am

Omnigamer: 2015-06-12 01:11:11 am

Omnigamer: 2015-06-12 01:10:30 am

2015-06-12 12:52:08 am

All the things

Without having done any actual testing yet, I think I have a completely consistent trading strategy. It centers on the fact that Highway is a fantastic short-sale trading center. What I mean by this is that the shop refresh period is very fast compared to most of the item refresh rate, especially the ones that are already particularly valuable there. So you can buy up the stock and then come back between 12 and 14 minutes later and reap the rewards. It can be expanded up to 15 minutes instead of 14 with some extra trading, but I haven't tuned the actual window for that yet in terms of traveling around.

I'll go into more details as necessary, but the general route is as follows:

-Buy all Mayo and Native Costumes in South Window during your first visit.
-During second visit to South Window (Adlai + others recruiting), buy up Mayo and Costumes until you have 6 and 4, respectively.
-During Forest Village visit (Wakaba and Tony recruiting) buy all Native Costumes.
-On first visit to Highway, if Mayo is high price (~2750) sell all. Otherwise buy all. Buy all native costumes and gold bar. If and only if both Mayo and Costumes are low-stock, buy all Deer Antlers.
-On second visit, exactly 12-14 minutes after first visit, sell everything you have.

At least as far as I've calculated so far, this strategy should net you 50k even with the worst possible RNG sets. The big caveat is making that timing window, which needs to be figured out yet. Another downside is that in a worst-case scenario for inventory space, 24 slots are taken up by items. This can be tuned a little bit and I haven't looked at each individual possibility, but having only 6 slots left for other items could be a bit unfortunate. You can always discard excess, but I don't want to even think about those scenarios right now. In any case, I'll try to think it out a bit more tomorrow when I have more sleep in me.

Without having done any actual testing yet, I think I have a completely consistent trading strategy. It centers on the fact that Highway is a fantastic short-sale trading center. What I mean by this is that the shop refresh period is very fast compared to most of the item refresh rate, especially the ones that are already particularly valuable there. So you can buy up the stock and then come back between 12 and 14 minutes later and reap the rewards. It can be expanded up to 15 minutes instead of 14 with some extra trading, but I haven't tuned the actual window for that yet in terms of traveling around.

I'll go into more details as necessary, but the general route is as follows:

-Buy all Mayo and Native Costumes in South Window during your first visit.
-During second visit to South Window (Adlai + others recruiting), buy up Mayo and Costumes until you have 6 and 4, respectively.
-During Forest Village visit (Wakaba and Tony recruiting) buy all Native Costumes.
-On first visit to Highway, if Mayo is high price (~2750) sell all. Otherwise buy all. Buy all native costumes and gold bar. If and only if both Mayo and Costumes are low-stock, buy all Deer Antlers.
-On second visit, exactly 12-14 minutes after first visit, sell everything you have.

At least as far as I've calculated so far, this strategy should net you 50k even with the worst possible RNG sets. The big caveat is making that timing window, which needs to be figured out yet. Another downside is that in a worst-case scenario for inventory space, 24 slots are taken up by items. This can be tuned a little bit and I haven't looked at each individual possibility, but having only 6 slots left for other items could be a bit unfortunate. You can always discard excess, but I don't want to even think about those scenarios right now. In any case, I'll try to think it out a bit more tomorrow when I have more sleep in me.

Omnigamer

2015-06-18 05:35:19 pm

All the things

Just posting this for posterity. For TE runs, a pretty straightforward and consistent way to get a 1T Beast Rune kill is below, however I think that the savings are not worth it compared to the extra prep time. Anyway, the setup:

Riou: Earth or Mother Earth
Mazus: swap fire for Water (or keep Fire, but susceptible to speed bug)
Tengaar: Give Water and Lightning, level 59+
Flik: Give Water, level 59+, speed equips
Luc: Give Water
Sheena/Meg: Kindness bearer [+ double beat] + double strike

From there, it's straightforward enough. Luc and Flik self-unite Thor to take out the rune, Mazus and Tengaar self-unite Thor to take out Right Head, Riou uses Earthquake, and Sheena or Meg finish off Left Head via Kindness. Speed equipment can shift around a bit, but the point is that Flik and Luc are always fastest, only one type of Rune Unite is used so the speed bug doesn't come in to play, and Sheena/Meg can do a huge amount of damage with good consistency. Rina could be used in place of Flik/Tengaar if a better leveling strategy works out for her.

Caveats:
-Lots more rune swapping. Getting the runes isn't as big of a deal, but each additional swap costs 4-5 seconds.
-Additional power leveling needed. There's an easy path for Tengaar to hit level 59 (almost 60) by Beast Rune by leveling her at Rockaxe, but I haven't thought through anything for Flik yet.
-This doesn't consider any of the changes for the prior battles vs Seed & Culgan and soldier groups. Some may be faster, some may be slower, who knows.
-At best, I think a 1T Beast Rune strategy could save 1.5 min. At least for the above, 30+ seconds is lost from rune swapping, not including any extra travel.
-There is at least some mild additional time from taking different characters and taking the time for leveling them up. In several cases, it probably also means additional time to prepare and use other attacks (ie Empty World). The sum of these things is probably in the neighborhood of 45 seconds.
-It's still risky. Tengaar and Flik are not guaranteed to hit 160 MAG, and if either one of them doesn't make it then the whole strategy falls apart. Without hunting specific extra battles, I don't see a way to power level them above 60.

So that's my basic analysis. Still more to think on it, but just wanted to put it down before I forget it or if other people have ideas.

Just posting this for posterity. For TE runs, a pretty straightforward and consistent way to get a 1T Beast Rune kill is below, however I think that the savings are not worth it compared to the extra prep time. Anyway, the setup:

Riou: Earth or Mother Earth
Mazus: swap fire for Water (or keep Fire, but susceptible to speed bug)
Tengaar: Give Water and Lightning, level 59+
Flik: Give Water, level 59+, speed equips
Luc: Give Water
Sheena/Meg: Kindness bearer [+ double beat] + double strike

From there, it's straightforward enough. Luc and Flik self-unite Thor to take out the rune, Mazus and Tengaar self-unite Thor to take out Right Head, Riou uses Earthquake, and Sheena or Meg finish off Left Head via Kindness. Speed equipment can shift around a bit, but the point is that Flik and Luc are always fastest, only one type of Rune Unite is used so the speed bug doesn't come in to play, and Sheena/Meg can do a huge amount of damage with good consistency. Rina could be used in place of Flik/Tengaar if a better leveling strategy works out for her.

Caveats:
-Lots more rune swapping. Getting the runes isn't as big of a deal, but each additional swap costs 4-5 seconds.
-Additional power leveling needed. There's an easy path for Tengaar to hit level 59 (almost 60) by Beast Rune by leveling her at Rockaxe, but I haven't thought through anything for Flik yet.
-This doesn't consider any of the changes for the prior battles vs Seed & Culgan and soldier groups. Some may be faster, some may be slower, who knows.
-At best, I think a 1T Beast Rune strategy could save 1.5 min. At least for the above, 30+ seconds is lost from rune swapping, not including any extra travel. 
-There is at least some mild additional time from taking different characters and taking the time for leveling them up. In several cases, it probably also means additional time to prepare and use other attacks (ie Empty World). The sum of these things is probably in the neighborhood of 45 seconds.
-It's still risky. Tengaar and Flik are not guaranteed to hit 160 MAG, and if either one of them doesn't make it then the whole strategy falls apart. Without hunting specific extra battles, I don't see a way to power level them above 60.

So that's my basic analysis. Still more to think on it, but just wanted to put it down before I forget it or if other people have ideas.