I think they pretty much have the same amount actually
dp has random move ai lol, idk why people think it's hard
So I looked into this
Specifically it’s the offensive move selection that’s random. Pic related is from the leaked DP source code that the community got its hands on in 2020. You can see a whole chunk of the code is commented out (and there are actually comments made by game freak explaining why it’s commented out at the bottom but I wasn’t able to properly translate the text, if any Japanese speaking web knows what it’s saying feel free to translate it)
This piece of code, that would only execute for offensive moves that can hit you, would check if the move isn’t the strongest one and increase the move’s "preference" score by one. It would also increase the move’s score by 2 if it’s super effective, giving the AI a more offensive playstyle if it can be SE on you. Basically, it was the code that would determine which offensive move the AI uses to do maximum damage.
…and it’s commented out, meaning the AI will pick its offensive moves at random when it decides to attack, as long as they can hit you. To oversimplify, it sees every move as a OHKO move, do it’s not actually trying to deal maximum damage, just to hit you, with anything. So the status move logic is untouched and it won’t use moves you’re immune to, but when it decides to use an offensive move it will pick it at random. This is exclusive to DP and was fixed in Platinum.
(1/2)
I actually tried to uncomment the code and compile the game to see if it fixes it. Turns out, with the code not commented out, the AI will ONLY use its most effective offensive move, it will never click status moves whatsoever. So that’s why they commented it out, it was bugged.
…but the bug was easy to figure out. As I said before, this code is never executed for status moves. For offensive moves it gives +1 to the strongest one. But then… that gives it priority over the status moves as well, right? Right. So I changed it to instead give -1 to every offensive move that is NOT the strongest. And just like that, it worked. The AI was just like in Platinum. Literally a 2 lines fix and they were too moronic to figure it out so they commented out the whole thing. Game freak moment.
Code commented out (final game) : https://emalm.com/?v=-LvLU
Code uncommented (bugged - not shown here but geodude and onix didn’t set up rocks either and just spammed rock throw) : https://emalm.com/?v=IUh1i
Fixed : https://emalm.com/?v=AZQ2M https://emalm.com/?v=631Ir https://emalm.com/?v=kc4CS https://emalm.com/?v=VBtoy (4 parter because desmume)
I also tested with a Geodude and Cranidos does use pursuit in this case, as it should since Geodude resists headbutt
>would check if the move isn’t the strongest one and increase the move’s "preference" score by one
I meant would increase it if the condition, isn’t met, so, if it is in fact the strongest move
I actually tried to uncomment the code and compile the game to see if it fixes it. Turns out, with the code not commented out, the AI will ONLY use its most effective offensive move, it will never click status moves whatsoever. So that’s why they commented it out, it was bugged.
…but the bug was easy to figure out. As I said before, this code is never executed for status moves. For offensive moves it gives +1 to the strongest one. But then… that gives it priority over the status moves as well, right? Right. So I changed it to instead give -1 to every offensive move that is NOT the strongest. And just like that, it worked. The AI was just like in Platinum. Literally a 2 lines fix and they were too moronic to figure it out so they commented out the whole thing. Game freak moment.
Code commented out (final game) : https://emalm.com/?v=-LvLU
Code uncommented (bugged - not shown here but geodude and onix didn’t set up rocks either and just spammed rock throw) : https://emalm.com/?v=IUh1i
Fixed : https://emalm.com/?v=AZQ2M https://emalm.com/?v=631Ir https://emalm.com/?v=kc4CS https://emalm.com/?v=VBtoy (4 parter because desmume)
I also tested with a Geodude and Cranidos does use pursuit in this case, as it should since Geodude resists headbutt
Good stuff, thanks. Is early Battle Factory AI (battles 1-14) from Platinum/HGSS the same as DP AI?
Not the guy you replied to. From what I've noticed the battle factory AI in rounds 1-2 is completely random, like it'll use one of its four moves at random.
This.
And it's genuinely difficult since the e4 can't be re-challenged with their old teams if you begin the sevii islands bullshit. By the time you finish the half ass rocket plot and return the ruby/sapphire, your team is maybe 2-3 levels higher than what you just finished previous the e4 with
B2W2 challenge mode is ruined by the fact that the opponents are effectively underleveled due to a bug
DP’s semi random move selection makes the AI harder to predict, it is actually known to be harder by the speedrunning community because of that, but speedrunning =/= casual gameplay and the randomness will save you from a KO move a lot of the time.
I would say Platinum is the hardest official game. I mean, Cynthia literally has three OU mons and one Uuber mon on her team. And one of the remaining two has no weaknesses. Clearly they went hard with this one, it’s still easy because it’s pokemon, but it’s clearly in a different league compared to the other champion fights
They have as many Pokémon as in every other game outside of gens 5 and 6, and the held items they have are mostly trash. Sure, it’s better than holding nothing, but remember, they’re also underleveled.
They're higher level, levels matter in Damage formula.
6 months ago
Anonymous
Hardly so. A few levels will never make a difference, legitimately the random damage roll is gonna matter more lmao
6 months ago
Anonymous
back that claim up with math
6 months ago
Anonymous
The guy you replied to is right, the range of damage has more variance than the small small damage boost you get from being an increased level.
Iris' level 61 Hydreigon using Focus Blast will do 85-101 damage to a Pokemon with 100 sp.def. At level 58 (normal mode), it would do 82-97.
You'd only notice a difference in damage if you calculated the rolls beforehand, which nobody outside of speedrunners or nuzlockers would do.
6 months ago
Anonymous
So it does make a noticeable difference, got it.
We can now put this moronic "Chalenge Mode is easier" discussion to grave, I would hope.
6 months ago
Anonymous
Nobody ever said Challenge Mode was easier.
6 months ago
Anonymous
This "nobody" says it every single time Challenge Mode is brought up.
6 months ago
Anonymous
Is this "nobody" in the thread with us right now?
6 months ago
Anonymous
At the very least
>Challenge B2W2
kek
this guy is a "nobody" even if they didn't say it outright.
6 months ago
Anonymous
That was probably yawngay, the whole posting a picture without actually explaining things is kind of his signature move.
6 months ago
Anonymous
I’m concerned by the fact that you thought I was saying challenge mode was easier, and tried to use moves dealing 3 more hp worth of damage as an argument instead of the obvious fact that they have one more pokemon each
6 months ago
Anonymous
This is the way the conversation goes every single time. >BW2 Challenge Mode is brought up >Someone posts the image >Counter-arguments are brought up regarding improved movesets, more Pokemon etc. (Is there info on it having a better AI I wonder) >No it's easier because you get more EXP
repeat ad-infinitum
6 months ago
Anonymous
Yes discordgays shitpost against gen 5 (and 4), what else is new.
Special Attack stat set to 165 for both Hydreigons
10/16 rolls still out damages a perfect roll of the 57
This includes stat increases which don’t happen in B2W2, only the level itself is higher, the stats stay at the original level.
6 months ago
Anonymous
>This includes stat increases which don’t happen in B2W2, only the level itself is higher, the stats stay at the original level.
Can you fricking read? >Special Attack stat set to 165 for both Hydreigons
6 months ago
Anonymous
Nevermind I thought you just copypasted data from the calculator.
You’re right, it does make a difference, but it won’t matter at all in normal gameplay and will get lost in the damage roll, while your own pokemon will actually have a higher level, so clearly the attack will deal less damage, but it’s true that the held items somewhat make up for it.
6 months ago
Anonymous
Those 4 levels alone account for a damage variance of 12-14 per roll.
Assuming you're higher level for your team, bigger gap for every level you're below.
Even with an alleged error in coding, won't believe it without the de-compile showing all exact stats instead of a shitty YT nuzlockers "seismic toss" research, Challenge mode is stronger
6 months ago
Anonymous
There's no realistic way for it to be an error even. It's intentional.
6 months ago
Anonymous
6 months ago
Anonymous
Both this and the original "test" happen to fall into the same issue.
Hugh's starter probably has same IVs regardless of difficulty, and Iris has straight 30s on all of them.
This test wouldn't necessary get the same results everywhere in the game because of that, the E4 go from 24 to 30 IVs on all but their ace, which always had 30s.
Also, not related to that, I got a laugh from Iris' Hydreigon having Adamant nature with all special moves on Easy/Normal. It gets Lonely on CM instead, which is still not ideal, but at least it's not losing Sp. Atk for that one.
6 months ago
Anonymous
You’re free to not believe it but it’s very real.
Doesn’t make challenge mode harder than regular mode though, since regular mode had E4 members with only 4 mons and the last gym leaders had only 3, which is just stupid.
If anything it make the easy mode harder than the normal mode, since everything is overleveled
Btw gaygonHG got told in his Twitter thread
Yes they have better natures and IVs which does matter, but the stats are still calved as underleveled.
6 months ago
Anonymous
>more Pokemon don't make it harder than normal mode
?
6 months ago
Anonymous
Shit I fricked up i meant the opposite lole
Overleveling doesn’t make challenge mode easier than regular mode since important trainers have more pokemon which obviously makes them harder
6 months ago
Anonymous
>Doesn’t make challenge mode harder than regular mode though
Oh you're right it's easier when they have more mons on their teams and all have held items and they all have higher level damage formula calcs
Lol
>If anything it make the easy mode harder than the normal mode, since everything is overleveled
See pic
6 months ago
Anonymous
So Easy Mode is actually slightly easier than Normal Mode, because of the damage calc including the lowered levels.
6 months ago
Anonymous
Huh I didn’t know about the easy mode thing. I wonder why it’s inconsistent between easy mode and challenge mode like that
That may in fact be intentional because what would make easy mode easy otherwise
6 months ago
Anonymous
This was part of why it has to be intentional. They'd have to frick up on so many different levels to make this accidentally work the way it does.
6 months ago
Anonymous
The easy mode exp thing is probably intentional but the stats calculation clearly isn’t
6 months ago
Anonymous
No reason why it wouldn't be intentional. Again it would require a series of stupid mistakes to make it work like that on accident.
I know GF is incompetent, but that would take too much skill for them to do.
6 months ago
Anonymous
>No reason why it wouldn't be intentional
So that the game wouldn’t lie to the players by having lying levels?
It’s honestly not that far fetched of a bug and I’m not sure why you think it is. The stats are just calculated before the increased levels are applied.
6 months ago
Anonymous
Why would a game for kids make it seem harder than it actually is?
The issue can't be just with that though, the custom trainers (Gym Leaders and E4) are entirely separate, are they not?
6 months ago
Anonymous
Apparently their levels are determined dynamically which explains this happening
You could make an argument towards Challenge Mode being easier than Normal Mode, given that your Pokemon would gain more experience fighting higher level Pokemon, thus being a higher level by the end of the game.
Also you get more prize money in Challenge Mode.
What gamefreak thinks of as difficulty is much more complex than how players think of it though.
For example, things like DP having less overall pokemon in the dex, and every pokemon being used by at least 1 trainer so you can see every pokemon without necessarily needing to find it in the wild, make it "easier" to complete the dex than in some previous games. Things like having moves be color coded, having the physical special split, having double the TMs, and having more high powered moves overall, make it much "easier" for players to manage their team's movesets. Things like the AI being a bit worse on average, and the average level of non-gym trainers being quite low (such that it's closer to the level of wild mons than in some past games), make it much "easier" to find opportunities to train up weaker team members than in some previous games.
Most people on this board simply do not consider the full breadth of what an actual game designer would consider to be "difficulty", they aren't simply gauging how close a gym leaders team composition is to a smogon set, but rather they're legitimately asking themselves whether the various aspects of the game were designed to feel easier or harder overall. This includes a great number of facets, many of which have more to do with balancing what to communicate the player directly vs what to make them figure out for themselves than with battling itself.
TLDR quotes like this make perfect sense if you actually understand RPG game design and aren't a compgay.
NTA, this is also why USUM being the hardest game is a ridiculous statement. wide selection of pokemon, strong TMs for free, affection and roto boosts out the ass
is it harder if you ignore all of that? yes. but how many mechanics can you ignore and still fairly judge a game's difficulty?
he probably means generally speaking. gen 4 has less dungeons
Yeah I wouldn’t trust game freak on that anon lol
I think they pretty much have the same amount actually
So I looked into this
Specifically it’s the offensive move selection that’s random. Pic related is from the leaked DP source code that the community got its hands on in 2020. You can see a whole chunk of the code is commented out (and there are actually comments made by game freak explaining why it’s commented out at the bottom but I wasn’t able to properly translate the text, if any Japanese speaking web knows what it’s saying feel free to translate it)
This piece of code, that would only execute for offensive moves that can hit you, would check if the move isn’t the strongest one and increase the move’s "preference" score by one. It would also increase the move’s score by 2 if it’s super effective, giving the AI a more offensive playstyle if it can be SE on you. Basically, it was the code that would determine which offensive move the AI uses to do maximum damage.
…and it’s commented out, meaning the AI will pick its offensive moves at random when it decides to attack, as long as they can hit you. To oversimplify, it sees every move as a OHKO move, do it’s not actually trying to deal maximum damage, just to hit you, with anything. So the status move logic is untouched and it won’t use moves you’re immune to, but when it decides to use an offensive move it will pick it at random. This is exclusive to DP and was fixed in Platinum.
(1/2)
I actually tried to uncomment the code and compile the game to see if it fixes it. Turns out, with the code not commented out, the AI will ONLY use its most effective offensive move, it will never click status moves whatsoever. So that’s why they commented it out, it was bugged.
…but the bug was easy to figure out. As I said before, this code is never executed for status moves. For offensive moves it gives +1 to the strongest one. But then… that gives it priority over the status moves as well, right? Right. So I changed it to instead give -1 to every offensive move that is NOT the strongest. And just like that, it worked. The AI was just like in Platinum. Literally a 2 lines fix and they were too moronic to figure it out so they commented out the whole thing. Game freak moment.
Code commented out (final game) : https://emalm.com/?v=-LvLU
Code uncommented (bugged - not shown here but geodude and onix didn’t set up rocks either and just spammed rock throw) : https://emalm.com/?v=IUh1i
Fixed : https://emalm.com/?v=AZQ2M https://emalm.com/?v=631Ir https://emalm.com/?v=kc4CS https://emalm.com/?v=VBtoy (4 parter because desmume)
I also tested with a Geodude and Cranidos does use pursuit in this case, as it should since Geodude resists headbutt
>would check if the move isn’t the strongest one and increase the move’s "preference" score by one
I meant would increase it if the condition, isn’t met, so, if it is in fact the strongest move
Good stuff, thanks. Is early Battle Factory AI (battles 1-14) from Platinum/HGSS the same as DP AI?
Not the guy you replied to. From what I've noticed the battle factory AI in rounds 1-2 is completely random, like it'll use one of its four moves at random.
dp has random move ai lol, idk why people think it's hard
Idk why campaigngays think any game is hard when a single online battle requires more thought than all npc trainers or champions combined
Pokemon online and competitive is garbage
In terms of elite 4
DP > RS > FRLG
and it's not a close race.
Not even true in the slightest as FRLG has a fantastic rematch postgame
This.
And it's genuinely difficult since the e4 can't be re-challenged with their old teams if you begin the sevii islands bullshit. By the time you finish the half ass rocket plot and return the ruby/sapphire, your team is maybe 2-3 levels higher than what you just finished previous the e4 with
Lol Sinnoh's E4 shits on Kanto's. Even the rematches are a joke
she always was
DPPt and BDSP E4 are the two hardest E4s
Challenge B2W2 is harder if you count challenge modes
>Challenge B2W2
kek
Why even have a hard mode where nothing change?
B2W2 challenge mode is ruined by the fact that the opponents are effectively underleveled due to a bug
DP’s semi random move selection makes the AI harder to predict, it is actually known to be harder by the speedrunning community because of that, but speedrunning =/= casual gameplay and the randomness will save you from a KO move a lot of the time.
I would say Platinum is the hardest official game. I mean, Cynthia literally has three OU mons and one Uuber mon on her team. And one of the remaining two has no weaknesses. Clearly they went hard with this one, it’s still easy because it’s pokemon, but it’s clearly in a different league compared to the other champion fights
Still have more Pokemon and all of them have held items
They have as many Pokémon as in every other game outside of gens 5 and 6, and the held items they have are mostly trash. Sure, it’s better than holding nothing, but remember, they’re also underleveled.
They're higher level, levels matter in Damage formula.
Hardly so. A few levels will never make a difference, legitimately the random damage roll is gonna matter more lmao
back that claim up with math
The guy you replied to is right, the range of damage has more variance than the small small damage boost you get from being an increased level.
Iris' level 61 Hydreigon using Focus Blast will do 85-101 damage to a Pokemon with 100 sp.def. At level 58 (normal mode), it would do 82-97.
You'd only notice a difference in damage if you calculated the rolls beforehand, which nobody outside of speedrunners or nuzlockers would do.
So it does make a noticeable difference, got it.
We can now put this moronic "Chalenge Mode is easier" discussion to grave, I would hope.
Nobody ever said Challenge Mode was easier.
This "nobody" says it every single time Challenge Mode is brought up.
Is this "nobody" in the thread with us right now?
At the very least
this guy is a "nobody" even if they didn't say it outright.
That was probably yawngay, the whole posting a picture without actually explaining things is kind of his signature move.
I’m concerned by the fact that you thought I was saying challenge mode was easier, and tried to use moves dealing 3 more hp worth of damage as an argument instead of the obvious fact that they have one more pokemon each
This is the way the conversation goes every single time.
>BW2 Challenge Mode is brought up
>Someone posts the image
>Counter-arguments are brought up regarding improved movesets, more Pokemon etc. (Is there info on it having a better AI I wonder)
>No it's easier because you get more EXP
repeat ad-infinitum
Yes discordgays shitpost against gen 5 (and 4), what else is new.
This includes stat increases which don’t happen in B2W2, only the level itself is higher, the stats stay at the original level.
>This includes stat increases which don’t happen in B2W2, only the level itself is higher, the stats stay at the original level.
Can you fricking read?
>Special Attack stat set to 165 for both Hydreigons
Nevermind I thought you just copypasted data from the calculator.
You’re right, it does make a difference, but it won’t matter at all in normal gameplay and will get lost in the damage roll, while your own pokemon will actually have a higher level, so clearly the attack will deal less damage, but it’s true that the held items somewhat make up for it.
Those 4 levels alone account for a damage variance of 12-14 per roll.
Assuming you're higher level for your team, bigger gap for every level you're below.
Even with an alleged error in coding, won't believe it without the de-compile showing all exact stats instead of a shitty YT nuzlockers "seismic toss" research, Challenge mode is stronger
There's no realistic way for it to be an error even. It's intentional.
Both this and the original "test" happen to fall into the same issue.
Hugh's starter probably has same IVs regardless of difficulty, and Iris has straight 30s on all of them.
This test wouldn't necessary get the same results everywhere in the game because of that, the E4 go from 24 to 30 IVs on all but their ace, which always had 30s.
Also, not related to that, I got a laugh from Iris' Hydreigon having Adamant nature with all special moves on Easy/Normal. It gets Lonely on CM instead, which is still not ideal, but at least it's not losing Sp. Atk for that one.
You’re free to not believe it but it’s very real.
Doesn’t make challenge mode harder than regular mode though, since regular mode had E4 members with only 4 mons and the last gym leaders had only 3, which is just stupid.
If anything it make the easy mode harder than the normal mode, since everything is overleveled
Yes they have better natures and IVs which does matter, but the stats are still calved as underleveled.
>more Pokemon don't make it harder than normal mode
?
Shit I fricked up i meant the opposite lole
Overleveling doesn’t make challenge mode easier than regular mode since important trainers have more pokemon which obviously makes them harder
>Doesn’t make challenge mode harder than regular mode though
Oh you're right it's easier when they have more mons on their teams and all have held items and they all have higher level damage formula calcs
Lol
>If anything it make the easy mode harder than the normal mode, since everything is overleveled
See pic
So Easy Mode is actually slightly easier than Normal Mode, because of the damage calc including the lowered levels.
Huh I didn’t know about the easy mode thing. I wonder why it’s inconsistent between easy mode and challenge mode like that
That may in fact be intentional because what would make easy mode easy otherwise
This was part of why it has to be intentional. They'd have to frick up on so many different levels to make this accidentally work the way it does.
The easy mode exp thing is probably intentional but the stats calculation clearly isn’t
No reason why it wouldn't be intentional. Again it would require a series of stupid mistakes to make it work like that on accident.
I know GF is incompetent, but that would take too much skill for them to do.
>No reason why it wouldn't be intentional
So that the game wouldn’t lie to the players by having lying levels?
It’s honestly not that far fetched of a bug and I’m not sure why you think it is. The stats are just calculated before the increased levels are applied.
Why would a game for kids make it seem harder than it actually is?
The issue can't be just with that though, the custom trainers (Gym Leaders and E4) are entirely separate, are they not?
Apparently their levels are determined dynamically which explains this happening
I'm trans btw if that matters
Level 57 Hydreigon Neutral nature 0 EVs does (104, 106, 106, 108, 110, 110, 112, 114, 114, 116, 116, 118, 120, 120, 122, 124) to level 100 Bibarel
Level 61 Hydreigon Neutral nature 0 EVs does (114, 116, 118, 118, 120, 122, 122, 124, 126, 126, 128, 130, 130, 132, 134, 136) to level 100 Bibarel
Special Attack stat set to 165 for both Hydreigons
10/16 rolls still out damages a perfect roll of the 57
This is without including the wise glasses boost that challenge mode gets btw
>level 100 bibarel
SE moves don't count. Show neutral rolls.
(70, 72, 72, 73, 73, 75, 75, 76, 78, 78, 79, 79, 81, 81, 82, 84) challenge mode
(58, 60, 60, 61, 61, 63, 63, 64, 64, 66, 66, 67, 67, 69, 69, 70) regular
Dragon Pulse
165 spl atk
Is that with the wise glasses?
yes
>Aaron sent out Dustox!
Dustox in DP is NU
It’s bad, but not as bad as some other E4 members from other games using untiered, PU or PUBL mons
It’s also not in Platinum
>Bruno sent out Onix!
You could make an argument towards Challenge Mode being easier than Normal Mode, given that your Pokemon would gain more experience fighting higher level Pokemon, thus being a higher level by the end of the game.
Also you get more prize money in Challenge Mode.
The few extra levels won't make a big enough of a difference.
Btw gaygonHG got told in his Twitter thread
I knew there was something about IVs in there, thanks for bringing this back up.
Isn't this more a choke for Blue than Cynthia?
blue fans arent intolerable about him being the strongest champion
the text and graphic on that image contradict eachother, lol
He probably means things like giving Charmander Metal Claw so its easy to deal with Brock even if you pick charmander
Yea RS has crappy movepools and pre split gsmeplay.
DS made almost every Pokémon better with better moves.
bait thread, intended to sow discord
What gamefreak thinks of as difficulty is much more complex than how players think of it though.
For example, things like DP having less overall pokemon in the dex, and every pokemon being used by at least 1 trainer so you can see every pokemon without necessarily needing to find it in the wild, make it "easier" to complete the dex than in some previous games. Things like having moves be color coded, having the physical special split, having double the TMs, and having more high powered moves overall, make it much "easier" for players to manage their team's movesets. Things like the AI being a bit worse on average, and the average level of non-gym trainers being quite low (such that it's closer to the level of wild mons than in some past games), make it much "easier" to find opportunities to train up weaker team members than in some previous games.
Most people on this board simply do not consider the full breadth of what an actual game designer would consider to be "difficulty", they aren't simply gauging how close a gym leaders team composition is to a smogon set, but rather they're legitimately asking themselves whether the various aspects of the game were designed to feel easier or harder overall. This includes a great number of facets, many of which have more to do with balancing what to communicate the player directly vs what to make them figure out for themselves than with battling itself.
TLDR quotes like this make perfect sense if you actually understand RPG game design and aren't a compgay.
NTA, this is also why USUM being the hardest game is a ridiculous statement. wide selection of pokemon, strong TMs for free, affection and roto boosts out the ass
is it harder if you ignore all of that? yes. but how many mechanics can you ignore and still fairly judge a game's difficulty?
>is it harder if you ignore all of that? yes.
It's also harder if I don't ignore all that.
discord post
Masuda has such a punchable face
Sinnoh trainers all have random AI except for Doubles fights because of a mistake. Cynthia is the single most moronic Champion in the series.