to continue our blitz of devblog posts, imma talk about difficulty modes.
but first, I would like to present you with this:
this is a chronologically organized list of every major battle up until e18.
if you think it looks cool now, wait until the e19 stuff gets added.
this was inspired by me trying to edit trainer pbs data while having absolutely no idea where/when you'd fight people who appear more than once. it's nice if you want a quick reference for when/where you fight people throughout the game.
i'll probably add more data in as i need it, but for now it's just a neat little v1.0 version of a nice reference banner.
but this came about as i was working on difficulty modes! which were another thing i just...suddenly decided to do a few days ago.
some people tend to have a hard time with reborn. other people think it's too easy. i am really going to try and make both of those groups happy.
details are kind of scarce at the moment, but i can at least talk about the general idea behind them.
the easier mode (which i've taken to calling reborn lite) is effectively your standard reborn playthrough but with a little bit more leeway with how you play. trainer levels aren't quite so high. their IVs aren't quite so good. you aren't quite so money starved. it will still be significantly harder than a canon playthrough, but if you're particularly in it for the story, lite mode should help smooth over some of those harder fights.
the harder mode (which ame cheekily called reborn dark) is the standard reborn experience but with the difficulty knob broken off. it basically takes the rate at which things like IVs, EVs, etc are added and pushes it way forward, adding in some additional bullshit when you get to the late game.
i really want to emphasize that regardless of the difficulty, the game isn't really changing: team composition is staying largely the same, with any modifications being minor at best. the difficulty changes are more about how hard the game pushes back, rather than what the game pushes you with.
i've also been working for, like, 11 hours today and feel like i'm just rambling. if you have questions, ask 'em.
gosh, this post wasn't even funny. what am i doing out here.
hi also i edited the pic up there with an updated version
Preposterous! Not in my Reborn! Perish the thought!
G R E E T I N G S, Rebornians. Over the years I have carefully crafted Reborn's AI from the ground up (shoutout to ame who actually designed it tho, she the real mvp). With each episode the AI has only gotten smarter, and better at playing pokemon. In particular with the advent of 18.2, our AI has really stepped up it's game with some major bugfixes, outplaying you* so hard, it really makes you FEEL like you're a shit pokemon trainer.
*by 'you' I mean me, I'm terrible at pokemon and the AI has consistently been better than me since like, E18, let alone 18.2
And with the constant AI improvements, we are often asked 'Hey, how come the AI is so smart when battling me, but so garbage when it's my partner? Is this just designed that way to make the game harder?'
To which we say 'No, of course not! We don't believe in that kind of unfair game design, AI shouldn't cheat and we would never artificially dumb it down in situations where it should be helping the player!'
...We'd never do that on purpose.
So have you ever had an experience in Reborn where your AI partner has an all hitting move like, say, Earthquake, and uses it when it would KO your mon? Or doesn't use it when it would barely damage your mon and seriously hurt the opponent?
Because in E18 we added code specifically to deal with that, and have your partners make smarter choices! And guess what? They still didn't make smart choices and we were sad.
So let me tell you a little something about loops. Loops are great. They let you check through a bunch of cases with only a single block of code!
I could write 'for i in 1..4' to go through four different cases (funny that, just as many cases as a mon has moves!) with a single bit of code!
But do you know what is near i, both in the alphabet and on the keyboard? j.
And do you know what happens if you loop through i and accidentally type a j somewhere? Why, it checks whatever j is, which is probably something else from earlier in the code, nothing to do with poor old i.
And that can have the side effect of making AI real unpredictable. Like, for example, lets say your AI wants to use Earthquake. But part of the AI for that is to check how much damage it does to it's partner.
But oh no, I typed a j instead of an i in that check? So instead of checking how much damage Earthquake would do, it checks another move. Could be any of them. Like say, a status move. So it'll think no damage. And Earthquake away no matter how hard that'll hurt your mon. Oops.
But now this is fixed. We have officially cancelled j. Except for like, the places it's meant to be. In those places we've cancelled i. So for E19, you can finally look forward to the partner AI being just as smart as the opponent AI! Hopefully that should make some of the paired fights in the game a lot easier.
Eeeeeeexcept the fact that this cuts both ways. The AI is also now much better at being a partner, regardless of which side it is on,, so this fix will also mean that AIs will act a lot smarter about not sacking their own mons in doubles so like... Y'know, have fun with the AI being better in literally every double battle in the entire game! ❤️❤️❤️
Of course, this wouldn't be a devblog post without at least a LITTLE teaser for E19. And sure, I've told you about an AI update that it will bring but, because I'm such a great guy, you can have a little more. So I leave you with a screenshot of some actual dialogue from E19 with no context... I wonder what character would say this? 🤔
we have a functioning site. probably. maybe? whatever. you should be able to stay logged in while visiting this page and you can like and subscribe to posts.
feed me i need attention plz gib
to celebrate this momentous occasion, i would like to offer you proof that we haven't just been working on nothing all this time.
oh my! it's a picture! but what could it be a picture of????? who knows????!!?!??!?!?!?
hello. it is i, cass, six-time winner of the Most Likely to Not be Ame award.
today i am going to let you in on a little secret.
you may have noticed that it's been over a year since our last release.
now, in that time, we've gotten some stuff Done™️
especially the ai.
now, y'all are still out there running that old e18 ai
and lemme tell you.
it is garbage.
the ai always thinks it's faster than you. it sends out types that your pokemon are super effective against. item use occasionally scores 12000 on a scale that effectively peaks around 100. it sucks. it has been annoying the hell out of me for months.
now, as you might also have noticed, it is the season for the Giving of Gifts.
so we're putting out a little holiday update.
it is not e19.
i shouldn't really have to say that. we have progress bars. we keep saying that it won't be out for quite a while yet. it will also 100% not be a surprise. it will take months of testing before it's ready.
but a little package of ai updates and small fixes?
we can do that in a week.
so today i'd like to announce the e18.2 release. it's an aipdate!
there's a good deal of stuff that isn't in this.
last week i stared ame in the face, and was like "y'know what would be great? let's put out a mini release for the holidays as a surprise present!"
and she was like "hhhHHHHhhhhhhhhhhhhhhhhhhhhhhh ok fine"
the number of h's in that quote is inversely proportional to what I was able to convince her to add. a very limited testing period means that we can't really check for fatal bugs terribly well.
but there is some nice new stuff!
- A Fuckload ™️ of AI updates.
- A Smalload ™️ of trainer updates.
- HP bars move faster.
- The ability to forfeit trainer battles.
- Multiple active save slots
- New animations
it's not a whole lot, but we really wanted to get something out there for y'all.
hope you all enjoy!
holy shit it's done
like we haven't bugtested it yet so there's probably some kinks to work out
been working on this thing for five months
oh, right, post. sorry. *ahem
hi there i am cass and today i'd like to tell you about how i built a tower.
it is a long story.
(editor's note: so there's basically two parts to this- there's the story part where I talk about the development of the tower, and there's the part where I talk about the tower itself and what exactly you get to do in it. i am very worried that the first part is boring. thus, i would like to direct those who want to get to the good stuff down to where I say "it's time for the good stuff". there are some line breaks near it. you can find it. i believe in you.)
you may recall my original Postgame Announcement where I said this: "this FITE will include a Battle Tower- style system. i'd be more specific about it but, uh, we, kinda, sorta, maybe, uh, like, y'know, don't....know what we're doing for it?"
you see, this was around the time that i was finishing off some of the theme team and boss rush trainers and had taken exactly one glance into the default essentials implementation to see what exactly it was that I had to work with.
now, there are a lot of files that define pokemon sets and trainers for various battle tower rulesets. there's, like, the fancy cup, the pika cup, the little cup, the pokecup. no one cares about those, though. the Big files are bttrainers and btpokemon. bttrainers is what defines the actual people that you fight. btpokemon is what defines the pokemon that they use. each trainer has a list of sets that they can pull from btpokemon- for example, default trainer zero is defined to pull from sets 0-65.
this probably all sounds pretty fine so far, right? well.
the first set from btpokemon is this:
you might notice a few issues.
the first is that there's not a great way to define forms and abilities.
the second is that a lax incense is kind of an annoying cheese item.
the third is that it's a fucking sunkern.
this is not a one-off thing. sets 11 and 12 are:
has anyone ever woken up in the morning and thought "hello yes I am a Real Person and I definitely want to fight a Silcoon and a Cascoon"
now, the default sets are organized roughly in order of difficulty. so, sure, you might assume, the lower sets are horrifically disgustingly bad, but they must get better eventually! there are 882 sets, after all!
well, here's set 854:
first of all, this set is kind of silly. who uses hyperbeam as an actual serious move? let alone the only way to deal damage. but then, you see dragon dance. dragon dance is a good move, right? boosts attack and speed! great for sweeping!
but the only attacking move is Hyper Beam, which is a special attack. so dragon dance doesn't even help!
the realization hits.
it doesn't work now.
but it used to.
because all of these movesets
are for gen 3
irl pic of me upon realizing this:
like omg it's like these were made out of pure, concentrated, ass.
so even if we were okay with godawful pokemon, the mechanical changes since gen 3 limits the sets so severely that even the hardest sets would be a cakewalk. that'd be unacceptable for any game, but y'all are playing reborn. y'all have beaten shelly. you're here for a challenge. you do not want to fight a dragonite that uses fucking hyper beam.
thus, btpokemon had to be remade from scratch.
so i sat down. for months. going through pokemon by pokemon, moveset by moveset. it was grueling. i've almost memorized every single pokemon's smogon tier. eventually i had to enlist the help of ame's patreon so i didn't die under the monotony of the whole thing. they cranked out ~500 sets which, combined with 1000+ that I had written, reached a total of almost 1600 sets. and these sets are good. I cut a lot of sets that were really similar to each other, there aren't any sets that use bullshit items like bright powder or quick claw, and there aren't any sets that just kind of sit there and harden until they get killed. these sets are made to kill you and kill you good. and, honestly? it's probably still on the lower end of what we could put out. there's probably more creative sets, especially towards the end, that i just missed out of exhaustion. i could've probably drawn from a wider pool of items. hell, we don't even have z-moves and megas implemented. (hey, that's why this is v1.0, y'know.)
but 1600 is a lot. if a trainer could pick any set with the only limitation being that you can't pick the same species twice, you would have 203,114,206,600 ways to fight that trainer. and yes, that trainer does exist! it's a clown.
if each trainer was allowed to have four pokemon, the number of possibilities would increase to over 160 trillion. TRILLION. gosh, look at all of that content.
speaking of content: trainers.
the original trainer system, as i mentioned earlier, draws sets based on their line number in the PBS. this is a good method for picking and choosing your sets very specifically. however, it has a pretty key flaw. as I mentioned earlier, we could easily put out more sets. and i definitely plan to. but if we do, where would we put them? want to organize by difficulty? sorry, fucko. the moment you change a single line, you have to change every line after that for every trainer. it's a goddamn nightmare. to fix this, Nerd Captain Marc of the Cello wrote up some scripts to change how trainers drew set from by line number to by pokemon species.
and it was a miracle.
now, all the trainers from before were coded for sets made in gen 3, and we have massively expanded the set list from there. constructing sets based around individual trainers would have taken a lot of decision making on each individual pokemon they should be able to use. in order to make set creation easier, i extracted some data from btpokemon. i grabbed the number, species, types, total number of sets, and added a value for the relative difficulty of every single pokemon on the list and threw it all in a spreadsheet. i added some checks to eliminate pokemon of certain types, and used the sheet to manipulate the list into only spitting out pokemon that matched certain types and only pokemon of a certain difficulty level.
after this i could whip out trainer sets like they were nothing.
see this trainer here?
had to type in six words and click maybe 10 things and boom. four lines of pokemon.
there's still a few problems with this method. it doesn't really play nicely with alolan forms with completely different types. if i make a mistake somewhere on the spreadsheet, i have to correct it 20 times. the difficulty could probably be stratified a little better. but, like, this is still definitely the preferred way to have to remake 300 trainers.
next time this will absolutely be handled by a script. i'll just have to plug in some types and difficulties and boom. every trainer completed while i sit and play more three houses. what a good game.
so. it's time for the good stuff.
you, dear player, will find the battle tower on the upper level of the nightclub. it's a little more like a battle arena than a battle tower, but, y'know, details.
currently, the tower is only setup for singles. the sets are specialized for singles, the scripts function with singles, and we really have no idea how well anything works with doubles. it's ok, that's why this is v1.0.
but it's okay because boom:
it's an actual arena and not just some fuckin floor of a tower. we got seats. we got colors.
and we got
(also please excuse the text we're still working on it)
that's right, kids. every seventh battler is not a randomly generated trainer, but a precoded, randomly chosen boss.
i won't spoil the list of bosses for you.
but you saw which one i fought up there, right?
we live in a world where they are in the boss list.
anything can happen.
...well, not anything.
there are simply some toys you just aren't allowed to bring into the battle tower:
pokemon with a base stat of 671 or higher
kyogre and groudon
terrain change abilities
the ability power construct
the ability beast boost
other things are cool though. feel free to bring marshadow in with you. idc.
you'll also have the option to fight with or without fields. fields would be randomly generated with each trainer, throwing yet another layer of RNG into the mix. we understand you might not be into that. so we won't make you deal with it.
...so imma be honest here, i've spent so much time on the battle tower and on pokemon and on bosses and on trainers and on this post that i literally don't know what to talk about anymore. so the rest of this post is just going to be an ongoing Q & A. feel free to drop a question in the comments! i'll respond there too, and then copy my answer into here. i'll also add in questions and answers that i think of along the way.
Q & A:
Q: Cass, do you plan to add in Gen 8 when it comes out?
A: fuck no. i am not redoing all of this because some new move came out and fucking broke everything.
Q: Okay, sure, but would you consider adding minor things like moveset changes?
A: so our current position is that, for all intents and purposes, gen 8 will have no impact on the gameplay at all. only doing a partial implementation of the new mechanics and movesets would be confusing to players since you'd never know if something was included or not, and a complete implementation would add a lot of work to e19. and, like, i'm sure y'all actually wanna play this some day. it also gives a slight advantage to the pokemon that are in swsh over the ones that aren't.
so, y'know, maybe when the gen 4 remakes come out and they have all the pokemon in it. kinda like in the GBA days.
Q: Madame Meganium is a boss!? Gosh, who else are you going to have there!?
A: mr bigglesworth
Q: Which Patreon submission made you groan the loudest, but you still put it in?
A: toss up between
Q: Are y'all gonna incorporate items or some other kinda rewards into the battle tower?
A: you bet we are. we just...haven't gotten that far into thinking about it yet.
Q: Does Madame Meganium's ownership of other Pokémon count as slavery?
A: no. they are a team united behind the common cause of kicking your ass.
Q: Aside from battle tower, battle factory is probably my favorite format in the battle frontier so will that ever be implemented considering all the new sets you created?
A: so i mentioned that it took a lot of work to get the battle tower functioning despite the basic framework being there.
battle factory does not even have a framework to start off with.
we also think it'd be cool
but it'd be so much additional work that it's not really worthwhile.
Q: Just so I'm absolutely clear on restrictions, the Mega-Z Ring is pretty much disabled during Battle Tower matches for BOTH the player AND the AI correct? Meaning no Mega Evolution for either side and no Z-Moves for either side?
A: okay, this is a good question and i should've said this earlier: z-crystals and megas do not work with team generation. they are banned for you mainly because they aren't an option for the generated trainers.
Q: I know it will take still a lot of time but will there be a lvl cap for the pokemon for example 50 or 100?
A: both of those will be the level cap.
Q: Will it only be 6vs6 or will we be able to choose the number of pokemon we want to battle against for example 3vs3?
A: the battles will also be 3v3 only since a.) making the battles too long would make the climb to get to a boss significantly more difficult, and b.) if we did 6v6 i'd feel legally obligated to tell you how many combinations there were- and the 4v4 calculation literally took an entire day. 6v6 would take three years to calculate.
Q: Why does Madame Meganium wear specs?
A: madame meganium is a professional who requires glasses to read small print.
Q: Also, what how can a tower/arena fit on top of such a smol building?
A: it goes down instead of up. it's like an inverse tower.
Q: How did you manage to get bosses added in?
A: Script Maniac Nerdcello whipped out some cool new methods to get that shit up and running.
Q: What kind of things will we get from the battle tower?
A: good things. like that stuff earlier in the game where you were all like "man i so wish i had more of those" yeah well you're gonna get more of those k
Q: Are the battle tower bosses only fightable in the battle tower?
no one got my monaco pun last time so no jokes in the title
hi nerds. it is i.
a lot of you have been asking questions. questions like "is ame ok?" and "when's e19 coming out?" and "why hasn't there been a post in a while?" and "will gen 8 be in reborn?".
i can help.
Q: Is Ame okay?
Q: When's e19 coming out?
Q: Why hasn't there been a post in a while?
A: because the thing i actually want to talk about is battle tower, buuuuuuut it's kinda mind-numbing, so, like, maybe next month.
Q: Will Gen 8 be in Reborn?
A: ...look we gotta release e19 someday.
...maybe i should go into the first two questions a little bit more.
Q: Is Ame okay?
ame is doing okay! she is made of pain, as one might expect from someone recovering from surgery, but she is definitely getting better. she can almost sit!
Q: When's e19 coming out?
e19 is gonna come out... when it does.
i realize that i have now given the same answer twice, so let me explain.
currently, the limiting factor on postgame is ame and i. we gotta make a whole postgame that is, like, actually fleshed out! and thoughtful! and boy that takes some effort.
now, you might ask, "but cass. the dog quests were basically finished in, like, a day! the rest of postgame can't take that long, can it?" to which i say au countraire, dear reader (in a radiant demonstration of my ability to google what french worlds mean).
you see, we have only done early postgame. we haven't gotten to real postgame yet. early postgame is a bit less-involved compared to the stuff that is yet to come. and when it comes to late postgame, ame and I are currently made of busy. if ame were not dealing with the whole "surgery ow everything hurts" thing, she would be dealing with Starlight, which i'm gathering she'd really much rather work on. meanwhile, i have grad school. which is, itself, basically surgery.
once she has a starlight demo out, we will be flying. probably. until then, quest stuff's gonna be a little slow.
in the meantime, though, i have some other mechanics updates! with a small amount of math.
kay, so, y'all know that mining thing? y'know, the infinite source of oval stones and iron balls?
In a previous post, I vaguely hinted at the existence of new items by literally showing you pictures of them existing. well, there are new items! they're the evolution stones from the later gens. now you'll be able to get hot new items right after you beat shade.
but wait, that's not all! i also took the opportunity to rebalance the item appearance rates.
you see, originally, the chance of finding a certain item was:
(drop chance) / 2924
2924 is a stupid number. anything involving the number 2924 become stupid as well. thus, this original drop chance formula is mega stupid.
so i fixed it. that bottom number is 500 now.
here's how the item odds look after the tweaks:
the big thing here is that the shitty stuff is less common, and the good stuff is more common.
this is especially the case for fossils, amplifield rocks, and odd keystones: they're all several times more common than they used to be. heart scales will also appear a lot more frequently since they tend to be easily missable. to compensate for this, some of the other items' percentages have been dropped, including the oval stone and the iron ball. they will be missed. by no one.
these item odds are subject to change during testing, but the point is that there are changes! woo! progress!
i don't know how to end this post. don't forget to click like and subscribe?
so there are some features that ame had thought she might or might not eventually implement into the game, probably very late in development after everything else was done
unfortunately, ame is very impulsive
and thus, she decided to do them immediately.
as of episode 19, new players starting the game will be given an option for 'special instructions', or the ability to enter a text password during the intro. the passwords available for this are listed in the readme, but i'll talk about their effects here:
- Changes Reborn's level cap to a hard cap in which Pokemon no longer gain EXP once they hit it, a la Rejuvenation
- Where (type) can be replaced by any of the 18 types, guarantees that all randomized early game events, including the mystery egg, will generate with Pokemon of the chosen type when possible. Also adds an early game Rock and Ice type so those types have something before Julia.
Password: nuzlocke - Prevents Pokemon from being healed when their HP drops to 0. Can be turned off after a loss. In order to account for different types of lockes, no other restrictions are applied.
- Randomly shuffles Pokemon species and moves around for a playthrough. Online play is disabled in this mode. Shout out to @HauDareYou for the initial framework.
- Allows any Pokemon to use any TMX move provided the machine and badge are held. Thanks to @Waynolt for letting us integrate this one.
there are also other hidden passwords that can be found in-game...
any of these passwords can be used with each other, including the monotype ones for dualtype runs. however, i don't promise there won't be some wacky interactions. who knows, really
i hope you will look forward to playing the final game in various ways~
So as you may have seen from the previous Dev blog post, we've begun internally testing the E19 story just to make sure it's all up to standard before moving on fully with the mammoth that is postgame.
As such I have been playing through the story and testing it out, providing my thoughts and critiques along the way.
As we have known for the last... 4? 5? Episodes of Reborn, there is a PULSE Mr. Mime somewhere in Labradorra. So I think it's safe to say there's no spoiler in revealing this is something that is going to be fought.
And so we figured, just to give you a taste of what is to come when we eventually do release the episode, we'll share the commentary I gave on that particular fight:
' So here's my thoughts on the fight so far:
I knew the gimmick for this fight a long time in advance.
PULSE Mime is like the one aspect of E19 I know more about than anything else, I nominally helped with the design, I discussed the moveset and such with Ame, and I wrote the AI for it
And with all this foreknowledge
And with a team that (admittedly accidentally) has the tools to deal with it
against this thing '
That's all for today, I HoPE yOU aLl ArE SUfFerrINg LeSS tHan mE!
so, we've finally finished cleaning up from our several supervised test periods, and we're shipping off to internal. but for those who haven't heard, we are not doing a full test cycle right now-- internal will be the last test. the purpose here is just to make sure that if we for any reason have to drop the project and release as-is, the main story will at least not be horribly bug ridden for y'all. we'll have another full test cycle again (one that will probably span eons) once postgame is complete.
that said! it's been a minute since i talked about some of our new features and changes. let's start with the thing i've been screaming about. hint: it's AI.
we've made a number of improvements and fixes to a lot of facets of it, but there's two things in particular that have really just knackered my knickers:
the first is what we've just started referring to as the 'speed bug'. basically, there was a certain line of code used in default essentials that, when we reworked AI, we reused a lot. it speed like a nice, strapping young line of code, and we had no reason whatsoever to think it might be completely broken. so naturally, it was, and we copied it into 200 places. this caused the AI to basically always think it was faster than the player, even when it definitely was not, which as you might imagine, has caused a lot of issues and a lot of emotional pain. this has now been fixed, unleashing never-before-seen chambers of code into the wild for the first time. that can come with its own issues. basically: aaaaaaa
the second we just discovered a couple days ago. let's call it the 'switch-reversal bug'. basically, when deciding what pokemon to switch in, the parts of the code that checked type effectiveness got switched around. thus, it has been intentionally sending in pokemon with bad type match ups the entire time. this was an easy fix, but also 😰 😰 😰
frankly it's amazing that switching has worked well enough to this point that it wasn't completely obvious
here we are.
those are the big bois but like even aside from those i've been fixing dozens of case issues for the last week or two. it's gonna be good.
or i'm gonna cry.
what else! check this out:
by popular demand, we've merged the multi-move mod into the main project! Huge shoutout to @Waynolt for getting this going and letting us adapt it. We've also made some convenience improvements that you can see in the above gif-- Instead of selecting each pokemon through the menu, you can just hold Ctrl when you select them to automatically add them to the multi-move queue.
We've fixed some other misc things too, the fishing animation that's been missing for years, some PULSEs have been buffed, and we found an issue with hidden power where the player's hidden powers weren't actually properly set to the static 60 BP even though we tried to change that before. Very sorry about that, guys :c
One more update for you today. This one comes from our animator Vulpes, and it's exactly what the move deserved:
That's more or less where we're at. So since we're not moving on to later stages of testing, there's no set end date to internal-- we'll just fix stuff as we go and as we work. And thus, internal will simply evaporate into the distance, as must I.