Fire Emblem: Radiant Dawn Category

kirbymastah

Edit history:

kirbymastah: 2014-12-09 09:59:24 am

kirbymastah: 2014-12-09 09:58:52 am

kirbymastah: 2014-12-09 09:58:03 am

kirbymastah: 2014-12-09 09:53:12 am

kirbymastah: 2014-12-09 09:52:45 am

kirbymastah: 2014-12-09 09:52:12 am

kirbymastah: 2014-12-09 09:52:05 am

2014-12-09 09:50:59 am

<(^_^)>

The most commonly RTA speedran category for Fire Emblem: Radiant Dawn is on Easy Mode, with transfers. This is the fastest category and is essentially pure any%, as it uses the fastest difficult mode, as well as a transfers file to boost character stats, allowing for faster/safer strats. I'm thinking about submitting a solid run to SDA once I get one, but I mostly want confirmation on this from the SDA staff.

My question is would this kind of category be allowed for submission on SDA? Fire Emblem: Radiant Dawn (RD) is a direct sequel to Fire Emblem: Path of Radiance (PoR). A completed file from PoR can be transferred into RD (details about the transfer itself can be found here - http://serenesforest.net/radiant-dawn/miscellaneous/data-transfer/ ), but in a speedrunning context, this doesn't allow for more glitches or modes or any extra content; it just merely boosts several characters' stats.

The transfer file currently used by RTA speedrunners of RD is available here - https://drive.google.com/folderview?id=0BwjwG7unziUvaDREMXFIeDhYSWM&usp=sharing

Important to note: the agreed-upon used transfer file as of now is a hacked file from PoR which basically caps almost every character's stats, even though many of those stats are impossible to get without hacks. However, the hacked file is from PoR itself, not RD, and it allows for the fastest possible completion of RD and very neat strategies for RD.

Currently, the easy-mode categories are split into "easy mode + no-transfers" and "easy mode + transfers", which IMO is the most logical choice. If we weren't allowed to use a hacked PoR file for the transfers category, then comes the question of: "what kind of file is ok?" In my opinion, it's pretty much impossible to objectively define what transfer file should universally be used if a "perfect file" were not allowed, as there are way too many options and only so many stat-boosters. Some stat caps can only be reached via stat boosters, and stat-boosters are limited, so it's practically impossible to find an objective "middle-ground acceptable transfers file" (in other words, you can't have two characters both cap speed if they both require a certain number of limited stat-boosters). Thus IMO the transfers file used should be either all or nothing:

- go all-out with a hacked transfers file to go as fast as possible, or
- don't use a transfers file at all

...which is two categories already as mentioned earlier

To note: whatever decision made here by SDA staff will likely not affect the FE:RD speedrunning community's decision to use a hacked transfer file. We want to go fast, we don't like increasing our reset counter by x20 due to more level-up RNG and riskier/slower/less-interesting strats. Because of this, easy-mode no-transfers is practically a dead category now since there's basically no benefit or interesting thing about running it compared to easy-mode transfers Tongue

Thread title:

The most commonly RTA speedran category for Fire Emblem: Radiant Dawn is on Easy Mode, with transfers. This is the fastest category and is essentially pure any%, as it uses the fastest difficult mode, as well as a transfers file to boost character stats, allowing for faster/safer strats. [b]I'm thinking about submitting a solid run to SDA once I get one, but I mostly want confirmation on this from the SDA staff.[/b]

My question is would this kind of category be allowed for submission on SDA? Fire Emblem: Radiant Dawn (RD) is a direct sequel to Fire Emblem: Path of Radiance (PoR). A completed file from PoR can be transferred into RD (details about the transfer itself can be found here - http://serenesforest.net/radiant-dawn/miscellaneous/data-transfer/ ), but in a speedrunning context, this doesn't allow for more glitches or modes or any extra content; it just merely boosts several characters' stats.

The transfer file currently used by RTA speedrunners of RD is available here - https://drive.google.com/folderview?id=0BwjwG7unziUvaDREMXFIeDhYSWM&usp=sharing

Important to note: the agreed-upon used transfer file as of now is a hacked file from PoR which basically caps almost every character's stats, even though many of those stats are impossible to get without hacks. However, the hacked file is from PoR itself, not RD, and it allows for the fastest possible completion of RD and very neat strategies for RD.

Currently, the easy-mode categories are split into "easy mode + no-transfers" and "easy mode + transfers", which IMO is the most logical choice. If we weren't allowed to use a hacked PoR file for the transfers category, then comes the question of: "what kind of file is ok?" In my opinion, it's pretty much impossible to objectively define what transfer file should universally be used if a "perfect file" were not allowed, as there are way too many options and only so many stat-boosters. Some stat caps can only be reached via stat boosters, and stat-boosters are limited, so it's practically impossible to find an objective "middle-ground acceptable transfers file" (in other words, you can't have two characters both cap speed if they both require a certain number of limited stat-boosters). Thus IMO the transfers file used should be either all or nothing:

- go all-out with a hacked transfers file to go as fast as possible, or
- don't use a transfers file at all

...which is two categories already as mentioned earlier

To note: whatever decision made here by SDA staff will likely not affect the FE:RD speedrunning community's decision to use a hacked transfer file. We want to go fast, we don't like increasing our reset counter by x20 due to more level-up RNG and riskier/slower/less-interesting strats. Because of this, easy-mode no-transfers is practically a dead category now since there's basically no benefit or interesting thing about running it compared to easy-mode transfers :P

Onin

2014-12-09 01:21:49 pm

If the hacked file is impossible to achieve for real, is there a limit to the stats you'd be physically able to achieve in PoR? Couldn't you create a transfer file with the best humanly possible stats, rather than the capped stats that are otherwise inachievable?

kirbymastah

Edit history:

kirbymastah: 2014-12-09 02:01:08 pm

kirbymastah: 2014-12-09 02:00:14 pm

kirbymastah: 2014-12-09 01:59:18 pm

kirbymastah: 2014-12-09 01:58:31 pm

kirbymastah: 2014-12-09 01:56:42 pm

kirbymastah: 2014-12-09 01:54:17 pm

kirbymastah: 2014-12-09 01:53:22 pm

kirbymastah: 2014-12-09 01:51:48 pm

kirbymastah: 2014-12-09 01:51:28 pm

kirbymastah: 2014-12-09 01:50:05 pm

kirbymastah: 2014-12-09 01:44:17 pm

kirbymastah: 2014-12-09 01:43:51 pm

kirbymastah: 2014-12-09 01:43:38 pm

kirbymastah: 2014-12-09 01:43:15 pm

2014-12-09 01:42:42 pm

<(^_^)>

Quote from Onin:

If the hacked file is impossible to achieve for real, is there a limit to the stats you'd be physically able to achieve in PoR? Couldn't you create a transfer file with the best humanly possible stats, rather than the capped stats that are otherwise inachievable?

Having a file with the "best humanly possible stats" is impossible to really define objectively, since there are limited stat boosters and you'll also have to assume that everyone gets a level up in every stat every level. Nobody would really want to go through the trouble of actually playing through PoR and doing that, so you'd hack that anyways if there actually were some "theoretical humanly possible best file"

For example, let's say I have two characters A and B. Without any stat boosters, their stat cap for speed is 30. Assume they're both recruited at level 15 with 24 speed. The max level in PoR is level 20, so the maximum feasible speed from level-ups alone is 29 for both of them (24+ 5 more levels), meaning they're both 1 point away from speed. Hypothetically, let's say there is only one speedwing in the entire game (this is an item that permanently increases one character's speed by 2). Which character do you choose for the transfer file? Now apply this question to over 30 characters and 7-8 stats, with 2-3 of most stat boosters. Then you get into the question of is it feasible to cap everyone's level at 20, with the limited amount of weapons and gold (even though the game vomits a ton of it) and healing items for boss abuse?

Note that a character's stat in a transfer file only applies if that stat for that character hit maximum AND that character is at level 20. So transferring a character with 25 speed (cap for this example is 30) makes no difference from transferring that same character with 29 speed.

IMO You really can't define an objective "best humanly possible stats" since there's just so much variation, so many characters in PoR (and a lot of them do matter in RD), and so many stats to worry about. As I mentioned, it's basically impossible to objectively define a "possible middle-ground", and personally I'd rather just go either all out with a hacked file or don't use a transfer file at all.

The main difference using no/bad transfers vs. a hacked transfer is that it cuts out a lot of stupid meaningless RNG in RD, especially in character level-ups. It doesn't really add any interesting strats at all in RD, and makes the more interesting+faster strats viable because of transfers. There are a few exceptions but that's what the no-transfers category would be for, if someone wants to deal with resetting over level-ups and worse hit-rates much more often. Ideally a no-transfers run would probably only be like 2-3 minutes slower than a transfers run in a almost-3-hour run; it's likely that it's even as low as one minute but I doubt a no-transfer time would get as close because of how much more RNG is introduced.

A few examples where no/bad transfers merely adds RNG and nothing very interesting strategy-wise:

Chapter 1-5 - Ilyanna has a significant chance of missing two soldiers without transfers, and if she misses once, that's run over. Her chances are very minimal with transfers, and the strategy is exactly the same.
Chapter 1-6-1 and 1-6-2 - Even with a ton of stat boosters, Jill needs two level ups of strength and three level ups of speed without transfers by 1-6-2 to kill a boss in one round of combat, adding a ton of unnecessary RNG. She's forced to kill several more enemies to help her reach this benchmark, wasting significant time, and if she still doesn't make this benchmark after killing that many enemies (and it's very common), you lose at least a minute, meaning run over. With transfers, she only needs one level up of speed, which is reasonable to not even bother BEXPing her or grind her.
Chapter 1-8 - Sothe has a pretty reasonable chance of dying from several mages without transfers. With transfers, those chances are basically 0%, and the strategy is exactly the same.
Chapter 3-4 - Ike has a reasonable chance of dying from being attacked by several enemies, even at level 20 tier 2, without transfers. With transfers, he will likely have capped everything at this point (besides resistance) and will almost never die. The strategy is exactly the same. Other strategies would waste a turn and in this chapter, lose significant time.
Chapter 3-6 - Zihark has a reasonable chance of dying without transfers, and no chance of dying with transfers. This is, again, more meaningless RNG that doesn't change any strategy, and could mess up unit ordering down the road (though it isn't run killing)
Chapter 4-2 - Tanith needs 29 str to pull off a 5-turn strategy in this chapter. Without transfers, the chance of her missing this benchmark is significantly higher with the current route since she needs four levels of strength, and this would either be run over or you would have to use bonus EXP on her, wasting significant time and adding a lot more RNG. With transfers, she only needs two levels of strength and that's a 2% chance of missing that benchmark by the time she needs it in this chapter

[quote="Onin"]If the hacked file is impossible to achieve for real, is there a limit to the stats you'd be physically able to achieve in PoR? Couldn't you create a transfer file with the best humanly possible stats, rather than the capped stats that are otherwise inachievable?[/quote]

Having a file with the "best humanly possible stats" is impossible to really define objectively, since there are limited stat boosters and you'll also have to assume that everyone gets a level up in every stat every level. Nobody would really want to go through the trouble of actually playing through PoR and doing that, so you'd hack that anyways if there actually were some "theoretical humanly possible best file"

For example, let's say I have two characters A and B. Without any stat boosters, their stat cap for speed is 30. Assume they're both recruited at level 15 with 24 speed. The max level in PoR is level 20, so the maximum feasible speed from level-ups alone is 29 for both of them (24+ 5 more levels), meaning they're both 1 point away from speed. Hypothetically, let's say there is only one speedwing in the entire game (this is an item that permanently increases one character's speed by 2). Which character do you choose for the transfer file? Now apply this question to over 30 characters and 7-8 stats, with 2-3 of most stat boosters. Then you get into the question of is it feasible to cap everyone's level at 20, with the limited amount of weapons and gold (even though the game vomits a ton of it) and healing items for boss abuse?

Note that a character's stat in a transfer file only applies if that stat for that character hit maximum AND that character is at level 20. So transferring a character with 25 speed (cap for this example is 30) makes no difference from transferring that same character with 29 speed.

IMO You really can't define an objective "best humanly possible stats" since there's just so much variation, so many characters in PoR (and a lot of them do matter in RD), and so many stats to worry about. As I mentioned, it's basically impossible to objectively define a "possible middle-ground", and personally I'd rather just go either all out with a hacked file or don't use a transfer file at all.

The main difference using no/bad transfers vs. a hacked transfer is that it cuts out a lot of stupid meaningless RNG in RD, especially in character level-ups. It doesn't really add any interesting strats at all in RD, and makes the more interesting+faster strats viable because of transfers. There are a few exceptions but that's what the no-transfers category would be for, if someone wants to deal with resetting over level-ups and worse hit-rates much more often. Ideally a no-transfers run would probably only be like 2-3 minutes slower than a transfers run in a almost-3-hour run; it's likely that it's even as low as one minute but I doubt a no-transfer time would get as close because of how much more RNG is introduced.

A few examples where no/bad transfers merely adds RNG and nothing very interesting strategy-wise:

Chapter 1-5 - Ilyanna has a significant chance of missing two soldiers without transfers, and if she misses once, that's run over. Her chances are very minimal with transfers, and the strategy is exactly the same.
Chapter 1-6-1 and 1-6-2 - Even with a ton of stat boosters, Jill needs two level ups of strength and three level ups of speed without transfers by 1-6-2 to kill a boss in one round of combat, adding a ton of unnecessary RNG. She's forced to kill several more enemies to help her reach this benchmark, wasting significant time, and if she still doesn't make this benchmark after killing that many enemies (and it's very common), you lose at least a minute, meaning run over. With transfers, she only needs one level up of speed, which is reasonable to not even bother BEXPing her or grind her.
Chapter 1-8 - Sothe has a pretty reasonable chance of dying from several mages without transfers. With transfers, those chances are basically 0%, and the strategy is exactly the same.
Chapter 3-4 - Ike has a reasonable chance of dying from being attacked by several enemies, even at level 20 tier 2, without transfers. With transfers, he will likely have capped everything at this point (besides resistance) and will almost never die. The strategy is exactly the same. Other strategies would waste a turn and in this chapter, lose significant time.
Chapter 3-6 - Zihark has a reasonable chance of dying without transfers, and no chance of dying with transfers. This is, again, more meaningless RNG that doesn't change any strategy, and could mess up unit ordering down the road (though it isn't run killing)
Chapter 4-2 - Tanith needs 29 str to pull off a 5-turn strategy in this chapter. Without transfers, the chance of her missing this benchmark is significantly higher with the current route since she needs four levels of strength, and this would either be run over or you would have to use bonus EXP on her, wasting significant time and adding a lot more RNG. With transfers, she only needs two levels of strength and that's a 2% chance of missing that benchmark by the time she needs it in this chapter

AlecK47

2014-12-09 02:04:58 pm

As somebody familiar with the game and (to some extent) the route, and having actually put some thought into this matter recently, I have to say that if something is theoretically achievable is actually rather significant, even if the save used is hacked in order to save countless hours of effort and sanity. Using a "perfect" import file that cannot be achieved without hacking or cheating in some form is at best highly questionable, and I think most people can agree on that (otherwise I doubt you'd even be asking here Tongue

). As for which characters to give stat boosting items in order to hit caps requires some thought, sure, but let's face it: if some late game units (Lucia, Geoffrey, etc.) don't cap every single stat it's not going to affect the run in any great way. Most units will be able to get most or all of the stats they really need within bounds theoretically possible within the game anyway, and most of the units I can think of that would have trouble getting more than a couple stats capped aren't really used anyway.

My thoughts were that perhaps a PoR save could be made just before the final level that caps everyone's level (or even leaves them 1 exp off level 20 so you can cap as many or as few as you like) with the maximum possible stats without items, and all possible stat-boosting items in the convoy. That way, if strats or preferences change it wouldn't take much effort to tweak the file actually used and still have a "possible" transfer file. I'm not versed in how the hacked files are most easily made, but I believe this would be possible.

As somebody familiar with the game and (to some extent) the route, and having actually put some thought into this matter recently, I have to say that if something is theoretically achievable is actually rather significant, even if the save used is hacked in order to save countless hours of effort and sanity.  Using a "perfect" import file that cannot be achieved without hacking or cheating in some form is at best highly questionable, and I think most people can agree on that (otherwise I doubt you'd even be asking here :P ).  As for which characters to give stat boosting items in order to hit caps requires some thought, sure, but let's face it: if some late game units (Lucia, Geoffrey, etc.) don't cap every single stat it's not going to affect the run in any great way.  Most units will be able to get most or all of the stats they [i]really[/i] need within bounds theoretically possible within the game anyway, and most of the units I can think of that would have trouble getting more than a couple stats capped aren't really used anyway.

My thoughts were that perhaps a PoR save could be made just before the final level that caps everyone's level (or even leaves them 1 exp off level 20 so you can cap as many or as few as you like) with the maximum possible stats without items, and all possible stat-boosting items in the convoy.  That way, if strats or preferences change it wouldn't take much effort to tweak the file actually used and still have a "possible" transfer file.  I'm not versed in how the hacked files are most easily made, but I believe this would be possible.

Molotov

2014-12-09 03:16:52 pm

I'm not sure how this could possibly be acceptable for an SDA run. Using shady means to get impossible stats that give an unfair advantage in the next game is still cheating. It's tedious to change the transfers file to match the current best strategies, or include Radiance leveling plans in the overall Radiant Dawn run, but that's just how it is if you want to run that category. Even if all it does is minimize resets, you're still getting an extra advantage over someone who doesn't use a hacked transfer file. Seems clear to me, at least.

(Personally, I don't even see how "less RNG" could be considered an argument. Either brute force it or plan around it. That's what everybody else does.)

Vulajin

2014-12-09 03:36:54 pm

From a community standpoint, I can understand that it makes the game much more enticing to run with a modified transfers file for the reduced RNG and the lowered frequency of resets. Particularly, for a marathon run, I can see where this would be desirable.

From SDA's standpoint, a modified transfers file is effectively fruit of the poisonous tree. If you couldn't accomplish it legitimately in PoR, then it seems doubtful that it should be permitted in a file that you use as an input for RD. I could see an argument being made that, as far as RD is concerned, any set of bytes that sufficiently resembles the PoR save format would be a valid input (which would make it similar to a password or a fixed seed in Minecraft). (I am not stating an opinion in favor of or against this hypothetical argument.) Can you think of any other games where this sort of thing is not only possible, but permitted and used within the community?

AlecK47

2014-12-09 03:42:23 pm

I've started a bit of analysis of my own as to what I'd do, which I'll post in the main RD thread when I'm done, but the only sticking point that might remain from the situations Kirby mentioned would be Sothe in 1-8, because he'd have lower luck (by 16, without any items - that should give some indication how ridiculous transfers can be). Other than that, it nerfs Tauroneo (already mostly good enough anyway, though he'd probably get some items), and Laguz - of which only the royals are very important, and I'd argue that they don't need any help (though Ranulf does cap strength without help). I honestly see a lot of the items going to Tauroneo and everything else being just minor convenience.

In other words, I can't justify using a save hacked to that extent at all, even in a practical sense for any one level's strategy or one unit's usefulness. Maybe for fun, but it honestly doesn't change nearly as much as the copious bonuses to strength, skill, speed and such that nearly every character can get without any items.

I've started a bit of analysis of my own as to what I'd do, which I'll post in the main RD thread when I'm done, but the only sticking point that might remain from the situations Kirby mentioned would be Sothe in 1-8, because he'd have lower luck (by 16, without any items - that should give some indication how ridiculous transfers can be).  Other than that, it nerfs Tauroneo (already mostly good enough anyway, though he'd probably get some items), and Laguz - of which only the royals are very important, and I'd argue that they don't need any help (though Ranulf does cap strength without help).  I honestly see a lot of the items going to Tauroneo and everything else being just minor convenience.

In other words, I can't justify using a save hacked to that extent at all, even in a practical sense for any one level's strategy or one unit's usefulness.  Maybe for fun, but it honestly doesn't change nearly as much as the copious bonuses to strength, skill, speed and such that nearly every character can get without any items.

Vykan12

2014-12-09 04:41:25 pm

Uwee Hee Hee

I think it's important to first establish whether or not transfers would be allowed in an SDA run to begin with. The way me, gwimpage and kirby use transfers is in a 0:00:00 save file that already has them pre-loaded (ie don't need to use a GC save card), but this invites the possibility that the 0:00:00 save file was hacked in subtle ways, such as decreasing certain enemy stats by 1, changing steel lances to iron ones, etc.

I would hope that transfers are ok if the transfer file is created legitimately and uploaded for the judges to verify themselves, and that the submitted run clearly shows a new file being created with the transfers being used.

JackintheBox333

2014-12-09 05:36:46 pm

Professional Shaq Fu Speedrunner

This is one of the only other games I can think of that involves transfer information... so I'm gonna bring up Golden Sun. Why? Because in Golden Sun: The Lost Age, it allows you to transfer save data from Golden Sun via a link cable or via a password. Only thing is that the passwords have been completely broken. You can write whatever password you want and get whatever you want, whether that be items or the lvl of the characters you're transferring into Golden Sun: The Lost Age. This includes at least one beta item (the Infinity +1 sword whose name escapes me at the moment) that was never properly implemented into Golden Sun and cannot be obtained legitimately, but was implemented into its sequel. Other than the beta items, the levels and items can be legitimately obtained. So would a prewritten password without the beta item be a legitimate category for that game for SDA's purposes? (The 30ish minute password typing notwithstanding)

AlecK47

Edit history:

AlecK47: 2014-12-10 01:50:24 pm

2014-12-09 06:25:39 pm

Using a legitimate mechanic in any given game in order to speed it up should be allowed, but if it's abusable (in the sense of cheating, that is), that abuse should be disallowed. I see no reason why a transfer file that has been created with hacks/cheats shouldn't be fine so long as it only achieves results possible without said hacks/cheats, but if it were ultimately decided that it all would have to be made the old fashioned way, I'd understand. That is, of course a significant problem, and could make life harder on verifiers, but if the save is uploaded for all to see and confirm as legit (or otherwise) by those terms it would help a hell of a lot. AFAICS those criteria should be exactly the same for both FE:RD and GS:tLA.

Also, to clarify something: the 0:00 save in RD could be created with a GCN memory card, and once created would have the transfer data on it just as it would with a hacked save because the 0:00 save is created when you start the game and remains until deleted or saved over, meaning the use of a GCN memory card wouldn't be necessary beyond the initial creation of the file anyway.

Edit: as promised, I posted my thoughts in the "main" RD thread (insofar as there is one here): https://forum.speeddemosarchive.com/post/fire_emblem_radiant_dawn_take_two_14.html#fire_emblem_radiant_dawn_take_two_14

Using a legitimate mechanic in any given game in order to speed it up should be allowed, but if it's abusable (in the sense of cheating, that is), that abuse should be disallowed.  I see no reason why a transfer file that has been created with hacks/cheats shouldn't be fine [i]so long as it only achieves results possible without said hacks/cheats[/i], but if it were ultimately decided that it all would have to be made the old fashioned way, I'd understand.  That is, of course a significant problem, and could make life harder on verifiers, but if the save is uploaded for all to see and confirm as legit (or otherwise) by those terms it would help a hell of a lot.  AFAICS those criteria should be exactly the same for both FE:RD and GS:tLA.

Also, to clarify something: the 0:00 save in RD could be created with a GCN memory card, and once created would have the transfer data on it just as it would with a hacked save because the 0:00 save is created when you start the game and remains until deleted or saved over, meaning the use of a GCN memory card wouldn't be necessary beyond the initial creation of the file anyway.

Edit: as promised, I posted my thoughts in the "main" RD thread (insofar as there is one here): https://forum.speeddemosarchive.com/post/fire_emblem_radiant_dawn_take_two_14.html#fire_emblem_radiant_dawn_take_two_14

Vykan12

2014-12-12 02:40:12 pm

Uwee Hee Hee

Since nobody's posted for a couple days, I would like to offer my perspective for why I wouldn't be opposed to a hacked transfer file.

The main benefit of a hacked transfer file is that it allows you to get transfers that are theoretically possible but incredibly unlikely in practice. Haar is a perfect example of this. For Haar to get a strength, skill, speed and defense transfer requires that he gains these stats on each of his level-ups (with the exception of strength, he can miss that once). That amounts to a net 5.67% chance PER LEVEL-UP. Now this can be easily achieved using bonus experience since you can reset on any bad level-up. However, getting a 5% chance 9 times in a row is extremely tedious, requiring on average 180 resets. Similarly, for Geoffrey to get str, skl, spd and def transfers requires an average of 112 resets. The most egregious example is Sothe, whose transfers are one-to-one, unlike other units who simply get +2 in each capped stat (and +5 for hp). For Sothe to get the best possible transfers, he basically needs as many perfect level-ups in a row. The odds of a perfect level-up are 0.19%, and this must be carried over 19 levels. Now the units I mentioned are a bit of a special case since they are pre-promoted units (and Sothe) with very few levels to gain. Indeed, lower levelled characters like Ike and Oscar cap many important stats on average and don't require such levels of tantamount abuse. Nonetheless, most units still require some degree of abuse in stats like hp, def and luck.

The more important consideration is that bonus exp is a finite resource and therefore cannot be used to repeatedly achieve single digit % outcomes. If a unit like Haar were to level-up twice in a map from combat experience, suddenly you're dealing with a 0.25% chance, and I'm sure nobody in their right mind would want to re-do a chapter 400 times just to get two level-ups. When you take into account having to level-up multiple units per chapter, the odds become astronomically low, to the point where you might as well be playing on emulator and manipulating RNG. That could also be considered cheating, although by that logic mapping out an RNG route in the game boy advance FEs with emulator and carrying that out on console would equivalently be cheating.

This isn't even mentioning the biggest concern of all, namely that combat exp is basically finite but theoretically infinite through extremely tedious boss abuse. Bosses sit on special tiles that recover 10% of their hp each turn, so you could in theory attack a boss, wait a few turns and attack again. This is extremely slow and tedious, taking hours for a handful of level-ups. So in theory every unit in the game could be levelled up to 20/20 or --/20 and get transfers, but in practice only a team of 10-15 units ever cap their level.

By the way, this sheds light on a flaw with "realistic" transfer files: The amount of transfers you achieve is limited only by your patience and ability to endure extreme tedium, which doesn't reflect at all on skill level, precisely what a speedrun is supposed to display.

Last thing I'll mention, the extent of hacking to make such a file is not very demanding. In fact, you could create such a file, pass it off as a normal save file and nobody would be able to tell the difference. Basically, use codes to cap the level of the units you want to have transfers, and hack in infinite-use stat boosters. After the stat boosters have been used, drop them. And voila, a legimitate looking file that achieves thousands of hours of work in a few minutes.

I still wouldn't mind using a transfer file made the usual way since only 10 units really need transfers (Ike, Titania, Zihark, Jill, Sothe, Tanith, Ulki, Haar, Geoffrey, Elincia), which you can easily raise on a regular playthrough. It would suck to lose transfers on one-hit wonder type units (ex/ Ilyana to avoid the forge in 1-5) but that's life. At the same time I would prefer using a hacked file that achieves theoretically possible stats since it saves enormous time from doing something silly like manipulating every level-up in the game with cursor movement (FE9's RNG is affected by the cursor similar to FE6-8). Plus, the more transfers you have, the more realistic it is to have a normal or hard mode single segment run, which I'm sure any FE fan would like to see.

Since nobody's posted for a couple days, I would like to offer my perspective for why I wouldn't be opposed to a hacked transfer file.

The main benefit of a hacked transfer file is that it allows you to get transfers that are theoretically possible but incredibly unlikely in practice. Haar is a perfect example of this. For Haar to get a strength, skill, speed and defense transfer requires that he gains these stats on each of his level-ups (with the exception of strength, he can miss that once). That amounts to a net 5.67% chance PER LEVEL-UP. Now this can be easily achieved using bonus experience since you can reset on any bad level-up. However, getting a 5% chance 9 times in a row is extremely tedious, requiring on average 180 resets. Similarly, for Geoffrey to get str, skl, spd and def transfers requires an average of 112 resets. The most egregious example is Sothe, whose transfers are one-to-one, unlike other units who simply get +2 in each capped stat (and +5 for hp). For Sothe to get the best possible transfers, he basically needs as many perfect level-ups in a row. The odds of a perfect level-up are 0.19%, and this must be carried over 19 levels.  Now the units I mentioned are a bit of a special case since they are pre-promoted units (and Sothe) with very few levels to gain. Indeed, lower levelled characters like Ike and Oscar cap many important stats on average and don't require such levels of tantamount abuse. Nonetheless, most units still require some degree of abuse in stats like hp, def and luck.

The more important consideration is that bonus exp is a finite resource and therefore cannot be used to repeatedly achieve single digit % outcomes. If a unit like Haar were to level-up twice in a map from combat experience, suddenly you're dealing with a 0.25% chance, and I'm sure nobody in their right mind would want to re-do a chapter 400 times just to get two level-ups. When you take into account having to level-up multiple units per chapter, the odds become astronomically low, to the point where you might as well be playing on emulator and manipulating RNG. That could also be considered cheating, although by that logic mapping out an RNG route in the game boy advance FEs with emulator and carrying that out on console would equivalently be cheating.

This isn't even mentioning the biggest concern of all, namely that combat exp is basically finite but theoretically infinite through extremely tedious boss abuse. Bosses sit on special tiles that recover 10% of their hp each turn, so you could in theory attack a boss, wait a few turns and attack again. This is extremely slow and tedious, taking hours for a handful of level-ups. So in theory every unit in the game could be levelled up to 20/20 or --/20 and get transfers, but in practice only a team of 10-15 units ever cap their level.

By the way, this sheds light on a flaw with "realistic" transfer files: The amount of transfers you achieve is limited only by your patience and ability to endure extreme tedium, which doesn't reflect at all on skill level, precisely what a speedrun is supposed to display.

Last thing I'll mention, the extent of hacking to make such a file is not very demanding. In fact, you could create such a file, pass it off as a normal save file and nobody would be able to tell the difference. Basically, use codes to cap the level of the units you want to have transfers, and hack in infinite-use stat boosters. After the stat boosters have been used, drop them. And voila, a legimitate looking file that achieves thousands of hours of work in a few minutes.

I still wouldn't mind using a transfer file made the usual way since only 10 units really need transfers (Ike, Titania, Zihark, Jill, Sothe, Tanith, Ulki, Haar, Geoffrey, Elincia), which you can easily raise on a regular playthrough. It would suck to lose transfers on one-hit wonder type units (ex/ Ilyana to avoid the forge in 1-5) but that's life. At the same time I would prefer using a hacked file that achieves theoretically possible stats since it saves enormous time from doing something silly like manipulating every level-up in the game with cursor movement (FE9's RNG is affected by the cursor similar to FE6-8). Plus, the more transfers you have, the more realistic it is to have a normal or hard mode single segment run, which I'm sure any FE fan would like to see.

Molotov

2014-12-12 03:40:29 pm

I'm kind of confused now. You're saying that you hack 20/20 and assume +1 per level up, then add in your limited boosters? If so, I'm not sure why this is an issue at all (outside of it being unrealistic, which probably isn't a problem). You'd probably only create problems if you start ignoring some mutually exclusive stuff, like A and B both getting +6 speed when you only have 3 Speedwings total.

Vykan12

2014-12-12 04:29:53 pm

Uwee Hee Hee

Precisely. Assume each unit used gets a perfect level-up for each of their level-ups, and the limited stat boosters go towards the units who need them to cap stats even after perfect levelling (Ex/ Janaff's str, most of Tauroneo's stats, etc). This is different from blindly giving full transfers to all units since most units can't cap luck and it's impossible for all 3 laguz royals to have transfers since you can only recruit one of them.

As me and Gwimpage like to put it, our preference is for max transfers that are theoretically possible.

UraniumAnchor

Edit history:

UraniumAnchor: 2014-12-12 04:44:25 pm

2014-12-12 04:43:51 pm

Not a walrus

I didn't read the full thread yet, but just in case it wasn't covered, transfers in general are acceptable in the same way that NG+ is acceptable. We'd decide what to label it when such a run comes in.

I'll bring it up with the staff whether or not "theoretically possible, given infinite time" hacked files would be alright.

Hypothetically, if we did, then you'd want to document in the run comments exactly how you WOULD get it, so that people could verify the math more easily.

AlecK47

Edit history:

AlecK47: 2014-12-12 06:44:57 pm

2014-12-12 06:42:07 pm

Vykan basically summed things up as well as I could have hoped to myself, with the additional information about the way the hacked transfer files are made and the philosophy used in their making.

Quote from Vykan12:

\As me and Gwimpage like to put it, our preference is for max transfers that are theoretically possible.

This is my preference exactly as well, and I frankly don't give a crap whether the transfer file is "legit" so long as it is within the bounds of possibility without such methods. The way kirby phrased it, I was afraid that the file would contain transfers that would be impossible (everybody gets everthing, including luck, even laguz royals or some such), and that I would actually take issue with.

edit: The transfer file is also not necessary to check which units get what bonuses - you can see the stats at levelups and BEXP screens and cross reference their base stats without transfers - but being able to see the file to know what it is would make the work a little easier.

Vykan basically summed things up as well as I could have hoped to myself, with the additional information about the way the hacked transfer files are made and the philosophy used in their making.

[quote="Vykan12"]\As me and Gwimpage like to put it, our preference is for max transfers that are theoretically possible.[/quote]This is my preference exactly as well, and I frankly don't give a crap whether the transfer file is "legit" so long as it is within the bounds of possibility without such methods.  The way kirby phrased it, I was afraid that the file would contain transfers that would be impossible (everybody gets everthing, including luck, even laguz royals or some such), and that I would actually take issue with.

edit: The transfer file is also not necessary to check which units get what bonuses - you can see the stats at levelups and BEXP screens and cross reference their base stats without transfers - but being able to see the file to know what it is would make the work a little easier.

Crow!

2014-12-19 09:09:05 am

What's that gemma?

From a viewer's standpoint, rather than a rules standpoint, if I were presented with a run of Radiant Dawn that used a transfer file and a moderately longer one which didn't, I'd watch the run with a fresh start rather than the transfer file one. If the difference were like a factor of 2 in length, I might change my mind.

kirbymastah

Edit history:

kirbymastah: 2014-12-19 12:52:04 pm

2014-12-19 12:44:54 pm

<(^_^)>

Quote from Crow!:

From a viewer's standpoint, rather than a rules standpoint, if I were presented with a run of Radiant Dawn that used a transfer file and a moderately longer one which didn't, I'd watch the run with a fresh start rather than the transfer file one. If the difference were like a factor of 2 in length, I might change my mind.

Uh this isn't exactly related since this is, again, about the definition. Also, it seems like you don't really know what kind of difference a transfers file makes by saying that <_< Unfortunately for you, nobody runs no-transfers because nobody likes to do it and transfers is more convenient in every way possible. I guess if you prefer watching someone battle save 5 extra times and maybe reset in the middle of the run, be my guest Tongue

I'm fine with what's currently proposed, I was rather hasty with my initial post honestly.

Vykan12

Edit history:

Vykan12: 2014-12-22 02:39:46 pm

Vykan12: 2014-12-22 02:38:47 pm

Vykan12: 2014-12-22 02:38:18 pm

2014-12-22 12:17:19 am

Uwee Hee Hee

Transfers aren't used to save time so much as they are used to eliminate RNG, which means a much higher %age of runs actually reaching the finish, and less hair-pulling over things like Jill not having enough str on the 1-6-2 bosskill. On normal and hard, transfers are practically required for single segment to be realistically achievable, as far as I can tell.

Edit: Yeah, transfers make an enormous difference on normal mode. In 2-1 Brom & Nephenee can actually double enemies, which is the difference between the chapter being hitrate hell to being reasonably consistent. In 2-2 spd transfer Mordy allows for an amazing 3-turn strat. In 2-3 Geoffrey & Kieran don't risk dying anymore, and the horseslayer guy's hit is reduced by 6, which is a huge deal considering his hitrate is now in the single digits and the 2RN system skews low hitrates to astronomically low levels. In 2-E Haar can still be BEXPed to 19.88 and caps everything but luck. Without transfers his spd would already be an issue for part 3. In 3-P, spd transfers on Ike & Titania mean they double every non-SM enemy on the map instead of just mages. This isn't even touching on how much of a difference transfers Sothe makes in early part 1 when all your other units completely suck, and I haven't even looked beyond 3-P yet.

Tl;dr transfers on normal (and by extension, hard) are the difference between having near 100% consistent strats and like 50-75% consistent ones. If, for example, you had to rely on Ike, Titania, Haar, Jill, Tanith, etc on getting spd blessed each run, you'd never see the end of a run since the cumulative odds of finishing from that alone are already like 3%. Another way of putting it: The game has 43 chapters. If you use 80% reliable strats on say, half of those, you would finish 0.92% of your runs. Reliability is EVERYTHING in Radiant Dawn single-segment.

Transfers aren't used to save time so much as they are used to [b]eliminate RNG[/b], which means a much higher %age of runs actually reaching the finish, and less hair-pulling over things like Jill not having enough str on the 1-6-2 bosskill. On normal and hard, transfers are practically required for single segment to be realistically achievable, as far as I can tell.

Edit: Yeah, transfers make an enormous difference on normal mode. In 2-1 Brom & Nephenee can actually double enemies, which is the difference between the chapter being hitrate hell to being reasonably consistent. In 2-2 spd transfer Mordy allows for an amazing 3-turn strat. In 2-3 Geoffrey & Kieran don't risk dying anymore, and the horseslayer guy's hit is reduced by 6, which is a huge deal considering his hitrate is now in the single digits and the 2RN system skews low hitrates to astronomically low levels. In 2-E Haar can still be BEXPed to 19.88 and caps everything but luck. Without transfers his spd would already be an issue for part 3. In 3-P, spd transfers on Ike & Titania mean they double every non-SM enemy on the map instead of just mages. This isn't even touching on how much of a difference transfers Sothe makes in early part 1 when all your other units completely suck, and I haven't even looked beyond 3-P yet.

Tl;dr transfers on normal (and by extension, hard) are the difference between having near 100% consistent strats and like 50-75% consistent ones. If, for example, you had to rely on Ike, Titania, Haar, Jill, Tanith, etc on getting spd blessed each run, you'd never see the end of a run since the cumulative odds of finishing from that alone are already like 3%. Another way of putting it: The game has 43 chapters. If you use 80% reliable strats on say, half of those, you would finish 0.92% of your runs. [b]Reliability is EVERYTHING[/b] in Radiant Dawn single-segment.