Pokémon General Discussion

[The Host]

Roman Demigod? is that a Percy Jackson reference?

and just so this is still pokemon-topic: how soon does the third version of a generation usually come out?

[Flipz]

If Generation III is the norm, after the remakes.  If Generation IV is the norm, before the remakes.

--Flipz, being supremely unhelpful since 1991. 

[Magical Girl Mimi]

ZLS: Entirely so. =P Been reading book 2 of the second series recently. And yeah, you're right about the second part of your post. >>;

Flipz: So true. =P Funny how everyone used that as evidence that HGSS would come out, and then it went and was released anyway, just to prove them wrong. Well, and please the fans, of course, can't forget that. =P

[The Host]

If Generation III is the norm, after the remakes.  If Generation IV is the norm, before the remakes.

--Flipz, being supremely unhelpful since 1991. 

In either case a new game should be coming out soon, hopefully. I would actually almost prefer RSE remakes to Grey, since I haven't done an actual runthrough of Emerald since 05-06. Plus if RSE remakes come with the old Battle Frontier that will make it better.

But I don't think the DS can handle good graphics, 649 pokemon and a proper, 7 facility battle frontier. Any remakes will probably come out on the 3DS. I'll be lucky if gray/grey manages to come out on the DS.

[Magical Girl Mimi]

I'm pretty sure whatever is next after Gray, or possible even Gray, will be on the 3DS. The DS is really only around now because a lot more people have them than the 3DS... But really, once a main-series Pokemon game(Or even Mystery Dungeon) comes out for the 3DS, that'll change in a flash.

[The Host]

Just read that the new TCG theme decks are called Fast Daze and Furious Knights. Based on none other than Aceelgor and Escavalier. I actually thought they were quite good titles.

I'm pretty sure whatever is next after Gray, or possible even Gray, will be on the 3DS. The DS is really only around now because a lot more people have them than the 3DS... But really, once a main-series Pokemon game(Or even Mystery Dungeon) comes out for the 3DS, that'll change in a flash.

and so I will need to buy a 3DS soon.

[Magical Girl Mimi]

Do it. I mean, the price drop put it into literally the same range as the DSi, and will be much cheaper than the Vita assuming it doesn't get a price drop too... Plus, good games are out even now. No Pokemon main series games yet, but who knows; that might change next year. =P

[The Host]

Most likely I'll get it at Christmas, mainly for Mario games (kart and SMB3D), Zelda, and future pokemon games. Plus it should allow me to (finally) use wi-fi for pokemon White.
169.99 is actually a really good price.

[Magical Girl Mimi]

I agree, it got me 20 free games, even if 10 of them are still to come. =P

And that's true, I'm wondering what the new Zelda game will be, since it's been revealed that an actual new game is in the works...(And I hope there's no stylus control... While it worked very well for PH, and likely ST[Haven't played Spirit Tracks yet], it won't work on the mismatched screens of the 3DS.) But, that's not the place for this, since this isn't the Zelda topic, even if I momentarily forgot about that. >>;

[The Host]

Picked up a 3DS today since TRU had a sale... now I just need to figure out how to connect it to my wi-fi, which seems to be trying as hard as it can to prevent me from connecting...

also, is Pokedex 3D free to download if I ever get internet connecions to work on the 3DS?

[Magical Girl Mimi]

Yeah, Pokedex 3D is free.

As for the DS game stuff, there's a separate menu you have to configure. And possibly a third in-game, not sure; I haven't tried that yet, nor have I had a need.

[Flipz]

You know what I'd like to see Nintendo take a stab at?  An optional "App" for the Dsi/3DS that, when enabled, tracks Pokemon players' battle moves and their effectiveness.  Basically, the app itself would record every match and send the data of Pokemon used, moves used, hold items used, IVs, EV training, etc. to a central server.  That central server, in turn, would crunch the numbers and analyze the way that the fanbase uses Pokemon in battle; that digitized information, created in a way that computers can understand, would be used to alter the A.I. of the ingame trainers, making the experience more realistic.

Note that this doesn't necessarily mean all of the early-game trainers will be uber-hax and kill you constantly; once the A.I. knows how to battle well, it will also be able to pick out poor battle strategies.  Each trainer would have a set skill "stat" of some kind that would allow them to choose a strategy; in-game Trainers who have a low skill stat set by the programmers would be more likely to choose a poor strategy or a strategy that is good but is ill-suited for their Pokemon, while those with a high skill stat would be able to choose more advanced strategies, though the system would also have a random aspect to it that would allow them to "slip up" once in a while.

THIS would be a welcome addition to Gen VI, wouldn't it?

[The Host]

it would be cool if the AI were updated to be more challenging. Especially Elite Four members and gym leaders. but an app like that would be challenging to add I think, although I don't know enough about that kind of stuff to be accurate.

[Magical Girl Mimi]

Oooh. I so wish they'd have some sort of increasing/decreasing difficulty. While it's fun smashing through grunts like they're a stone beneath a steamroller, it would be fun to actually have a challenge for once. =P

[Flipz]

it would be cool if the AI were updated to be more challenging. Especially Elite Four members and gym leaders. but an app like that would be challenging to add I think, although I don't know enough about that kind of stuff to be accurate.

It wouldn't be THAT hard to record matches over the Internet, and it would only be moderately difficult to translate metagame strategies into computer-readable formats.  The hard part would be analyzing them.  Here's how I would go about it:

A team of about 20-30 employees would then read the Serebii Pokemon of the Week info and other sources to learn about the strategies in human-understandable language, then they would search through recordings to see that strategy in action.  (Since the battles would be stored as sequences of actions (a la Brawl replays), it would be perfectly easy for the team to do a simple search for Pokemon X using move Y.)

Once a number of replays showing a given strategy in use are identified, the team would assign the strategy some sort of ID number or other tag that the computer would be able to read and learn.  This tag would also identify which Pokemon can legally use the strategy.  Repeat until most ordinary strategies are recorded in the system.

Here comes the tricky part: the team would then sit down with a list of strategies and watch them in use, then judge their overall effectiveness and difficulty to counter.  This highly subjective rating would be averaged across all employees who rated a given strategy and then compiled into the database, essentially giving each strategy a "difficulty rating."  From there, a simple code will assign each strategy a probability of use in various lookup tables (each balanced for a certain level of difficulty).  When the player encounters a Trainer, that trainer's file contains a variable that tells it which lookup table to randomly select a strategy from, as well as any relevant modifiers (i.e. early-game trainers' random "dice rolls" would carry a heavy penalty, ensuring they only select strategies that are easy to counter).  If a given strategy is illegal for the Pokemon that Trainer has, the game will select a new one.

This is, essentially, the same A.I. currently used for in-game Trainers, just tweaked a bit to pseudo-randomly choose series of moves, instead of just random moves.

Also note that, to prevent in-game Trainers from ONLY using really effective battle strategies, the team would have to come up with various "dud" strategies and place them into the database and lookup tables.  This would have the effect of representing poor training or poor prediction, and would serve to randomize moves a little to keep things interesting for players.  TL;DR not every Rattata should be a FEAR Rattata. 

[The Host]

I HAVE WIFI CONNECTION!!!!!!! FINALLY!
 
i will soon post info regarding friend codes and such. i can finally use the dream world, and trade online and everything!!!   

lol editing this from 3ds. never happening again cuz typing is rediculous.

[Magical Girl Mimi]

Congrats. =D I might start playing again soon; but possibly not. I have enough other things to do. D=

[The Host]

don't worry. as i said in the 3ds topic, i still need to set up wi-fi for ds games...
when grey comes out though... now i'm actually hoping grey is for 3ds, or at least has a 3ds version.

[Magical Girl Mimi]

Even if Grey was released for the DSi like Black and White, it'd still support WPA.

However, I want Grey to be released for the 3DS, if only for Persistent Friend Code. =P

[Flipz]

it would be cool if the AI were updated to be more challenging. Especially Elite Four members and gym leaders. but an app like that would be challenging to add I think, although I don't know enough about that kind of stuff to be accurate.

It wouldn't be THAT hard to record matches over the Internet, and it would only be moderately difficult to translate metagame strategies into computer-readable formats.  The hard part would be analyzing them.  Here's how I would go about it:

A team of about 20-30 employees would then read the Serebii Pokemon of the Week info and other sources to learn about the strategies in human-understandable language, then they would search through recordings to see that strategy in action.  (Since the battles would be stored as sequences of actions (a la Brawl replays), it would be perfectly easy for the team to do a simple search for Pokemon X using move Y.)

Once a number of replays showing a given strategy in use are identified, the team would assign the strategy some sort of ID number or other tag that the computer would be able to read and learn.  This tag would also identify which Pokemon can legally use the strategy.  Repeat until most ordinary strategies are recorded in the system.

Here comes the tricky part: the team would then sit down with a list of strategies and watch them in use, then judge their overall effectiveness and difficulty to counter.  This highly subjective rating would be averaged across all employees who rated a given strategy and then compiled into the database, essentially giving each strategy a "difficulty rating."  From there, a simple code will assign each strategy a probability of use in various lookup tables (each balanced for a certain level of difficulty).  When the player encounters a Trainer, that trainer's file contains a variable that tells it which lookup table to randomly select a strategy from, as well as any relevant modifiers (i.e. early-game trainers' random "dice rolls" would carry a heavy penalty, ensuring they only select strategies that are easy to counter).  If a given strategy is illegal for the Pokemon that Trainer has, the game will select a new one.

This is, essentially, the same A.I. currently used for in-game Trainers, just tweaked a bit to pseudo-randomly choose series of moves, instead of just random moves.

Also note that, to prevent in-game Trainers from ONLY using really effective battle strategies, the team would have to come up with various "dud" strategies and place them into the database and lookup tables.  This would have the effect of representing poor training or poor prediction, and would serve to randomize moves a little to keep things interesting for players.  TL;DR not every Rattata should be a FEAR Rattata. 

*cough*comments?*cough*