All Gens Past Gens Research Thread

In RBY it is very much possible that a move does 0 damage (i.e., "misses"), as the damage is actually floored prior to applying weakness or resistance, and the floor (2) only accounts for a single resistance. I even recall being personally very salty about this on original hardware, as I wanted to poison Brock's Onix to death but my poison stings kept missing >:(

A Crystal_ youtube video on the topic

That said, maybe somebody can point to a difference in link battle calculations (as opposed to in-game) that indeed floors damage properly.
Poison Sting vs Onix is 1/4 effective. So, minumum damage before weaknesses or resistances is 2, halved and floored is 1, halved and floored is 0. What I'm saying is that according to the PS damage calculator, a move that is only 1/2 effective can cause 0 damage, even though 2 halved and floored is 1.
 

Jorgen

World's Strongest Fairy
is a Forum Moderator Alumnusis a Community Contributor Alumnusis a Contributor Alumnusis a Past SPL Champion
Ah, my bad. I had read "x2 resistance" as having two stacked resistances (so 1/4 damage), rather than a net resistance associated with a factor of 2 (i.e., 1/2 damage).
 
Led by the difference in what a few different pages report on it, I decided to test out some of the interactions that Fling has with a target if the attacker is holding a Mental Herb. I can't test things out in SwSh, but I can test them in Ultra Sun, and so I did. I figured it'd be worth posting my results here.

For a bit of context, pages like Pokemondb and Serebii mention that throwing a Mental Herb at a foe using Fling heals their Infatuation, which is correct.
However, Bulbapedia reports that on top of that, it can also clear up a few other statuses too.
Effectively, Bulbapedia's article for Fling was quite correct about that. I recorded a video showing it off.
 
What are all the things that baton pass can pass? This information is currently incomplete on Bulbapedia.
All stats modifications including atk, def, spa, spd, spe, acc, evasion, and critical hit ratio.
Also, it pass all non-status condictions you're suffering, like: Torment, Confusion, Attract, Substitute, Traping, Partial Traping, Minimized, Taking Aim, Perish Song, Curse Ghost-type, Nightmare, Telekinesis, Aqua Ring, Ingrain, Leech Seed, and Magnet Rise.
 
I dont know if this is already known and I've been ignorant of it this whole time but I encountered an odd mechanic in game in pokemon diamond that I've never seen on showdown. Basically what happened is that I was playing the battle tower for fun and u-turned into my scarfed flygon. However, when I went to click earthquake, it told me the choice scarf allowed only the use of u-turn! I replicated this a few more times to verify, one of which I've captured on video that I cannot figure out how to attach to this post because the file is too big. Hard switching in or coming in after a k.o. made the scarf work as usual so I dont believe it to be some freak cartridge malfunction. My theory is that because flygon is also carrying u-turn, and technically u-turn was the last move used on your side before the turn was reset, the game gets confused and decides flygon must have gone for u-turn and locks you in. If anyone is interested in verifying this on their own consoles to confirm that would be amazing. Also if this is already well known and I'm just an idiot, just yell at me or mods can delete it. And if someone could tell me how or where to share the video that would be helpful
 

Inspirited

There is usually higher ground.
is a Contributor Alumnus
https://www.smogon.com/forums/threa...t-before-posting.3663703/page-47#post-8741217
Deoxys in gen 3 has been on my mind for awhile, Fc finally brought it up in a bug reports thread and DaWoblefet ended up thinking about this in a way that raises many questions that I hadn't thought of until now.

Gen 3 questions that may affect a potential future PR thread about Deoxys:

1. If they exist, what are battle mechanic differences between Ruby/Sapphire (RS aren't super important tbh), FireRed/LeafGreen, and Emerald?
I asked pret about the differences between FRLG and Emerald's battle engines and the couple of answers I got said nothing is different between these two versions.

2. Assuming there are differences between the different carts, which engine is used when a link battle occurs between Emerald and one of FRLG or RS?
I assume it is the older game as this is the only answer that makes intuitive sense to me and I'm pretty sure this is the case in later generations.

3. I'm fairly positive I know the answer to this but might as well: When a link battle occurs between say Emerald and FireRed with both player using Deoxys, does Deoxys retain its FireRed (Attack) form for the FR player and Emerald (Speed) form for the E player regardless of which cart is hosting / older (etc)?
I'm pretty darn sure the battle would have a Deo-f and a Deo-e for their respective players.
 
https://www.smogon.com/forums/threa...t-before-posting.3663703/page-47#post-8741217
Deoxys in gen 3 has been on my mind for awhile, Fc finally brought it up in a bug reports thread and DaWoblefet ended up thinking about this in a way that raises many questions that I hadn't thought of until now.

Gen 3 questions that may affect a potential future PR thread about Deoxys:

1. If they exist, what are battle mechanic differences between Ruby/Sapphire (RS aren't super important tbh), FireRed/LeafGreen, and Emerald?
I asked pret about the differences between FRLG and Emerald's battle engines and the couple of answers I got said nothing is different between these two versions.

2. Assuming there are differences between the different carts, which engine is used when a link battle occurs between Emerald and one of FRLG or RS?
I assume it is the older game as this is the only answer that makes intuitive sense to me and I'm pretty sure this is the case in later generations.

3. I'm fairly positive I know the answer to this but might as well: When a link battle occurs between say Emerald and FireRed with both player using Deoxys, does Deoxys retain its FireRed (Attack) form for the FR player and Emerald (Speed) form for the E player regardless of which cart is hosting / older (etc)?
I'm pretty darn sure the battle would have a Deo-f and a Deo-e for their respective players.
I can be wrong but what i remember with my time playing RSE in cart:
1- Exist a little diference if you use the Wireless Adapter to play but since we emulate the link cable ones, there is no diference at all.

2- the engine used is always the recent cart. The comand order is E > FRLG > RS. The older cart just follow the order from the new one.

3- the sprite is locked for each cart because of hardware limitation. E players always saw Deo-S, LG players saw Deo-D, etc.
 

Inspirited

There is usually higher ground.
is a Contributor Alumnus
3- the sprite is locked for each cart because of hardware limitation. E players always saw Deo-S, LG players saw Deo-D, etc.
Here is some video evidence that might suggest otherwise as Xenon is clearly playing one of a FireRed or LeafGreen ROM and normal Deoxys shows up for him. I can make a strongish case for that Deoxys being Attack form as well given some of the damage it dealt though not knowing where the other 256 EVs and the nature boost on the Jirachi are make it hard to prove (they aren't in SpD though). This could be an old ROM specific quirk unfortuately. There is clearly something going on with the sprites at any rate. I also found some evidence in the current disassemblies that point to what you say being true. This stuff really makes me wish I had never sold my second GBA so I could test this for myself.
 

Plague von Karma

Banned deucer.
There is clearly something going on with the sprites at any rate.
As far as I know, Deoxys shows up with a normal sprite in Link Battles, but keeps the stats. You can tell what form it is simply by recognizing the opponent's GBA cartridge.

EDIT: I'm at my PC now, so I can give more information.

Due to being able to recognize someone's cartridge, you could say this is a rudimentary form of team preview. It sounds complicated to simulate what version the opponent is playing, and iirc Showdown's policy states that the version used is Emerald. This is a bit of a conundrum since it means that ADV Ubers has essentially been played wrong for over 10 years. I'll prolly help with whatever PR thread happens with proper research, it's been a long time coming anyway.
 
Last edited:
Today I brought up the Pokemon XD rule menu in the Deoxys thread, and it doesn't seem like a lot of people know about it. Its options aren't even mentioned on Bulbapedia. I recorded a quick clip going through it for DaWoblefet and ended up making a slightly more detailed video shown here. Personally I have not actually played XD's battle mode and I'm not sure how well Dolphin even works with it, so there are no mechanic tests, just scrolling through the menu itself to show its options.

By default you have 6 rules, with 3 presets of "Anything Goes", "Max. Lv 50", and "Max. Lv 100" along with 3 editable rule slots. Apparently, Pokemon Colosseum also has a similar rule editing feature according to the Pokemon Colosseum manual.

On page 1 you can set level range restrictions, as well as add a combined level limit, like the cups from Stadium had. There are also settings for a forced number of Pokemon and time limits per battle/move selection, counted by minutes and seconds respectively. Again I don't actually know how any of this plays out ingame, but for some mechanics, we can refer to the Colosseum manual (page 45). Apparently the player with the fewest fainted mons will win, and if they're the same it compares HP totals to determine a winner. It also says running out of time will automatically select the first move slot, just like the main games do.

On page 2 there's bans like Species Clause and Item Clause, as well as the ability to turn on hold items. If you press A on "Restricted", you get to pick and choose all the hold items you want allowed, with Soul Dew off by default. I noticed that for some reason Cleanse Tag is in this menu despite having no battle effect.

Page 3 contains rule mods like revealing Deoxys formes (referred to as disabling DEOXYS's camouflage ability), the classic Sleep Clause, and also disabling Skill Swap for some reason. I am not sure if this makes it fail or just bans it, but I'd assume fail. I also don't know what you get if you trade a Deoxys to XD, but I suppose it must have the data and models for each form if it has this rule. While looking up "Deoxys's camouflage ability", I found a guy on serebii forums from 2005 saying that a Deoxys from FireRed looked like Normal forme but with Attack forme stats. As a sidenote, while I was writing this post Plague von Karma pointed out in the Deoxys thread that Sleep Clause counts Rest in XD.

Page 4 is the place for basically blocking "cheap" strategies. There's self-KO clauses for losing with Explosion/Selfdestruct and having Destiny Bond/Perish Song fail when used by the last mon. This page also has Freeze Clause and lets you make Dragon Rage/SonicBoom fail for Little Cup.

Finally, page 5 is just where you can reset the rules to defaults. When you're looking at a preset rule there isn't even a page 5, so technically there are only really 4 pages for rules.

So that's the Pokemon XD rule menu. I don't think this will be that relevant to competitive, but I think it is neat and it's documented now.
 

Plague von Karma

Banned deucer.
I was just about to get to this!
As a sidenote, while I was writing this post Plague von Karma pointed out in the Deoxys thread that Sleep Clause counts Rest in XD.
This Sleep Clause exists in every "Stadium" game; as a result, in the RBY Community, we tend to call this "Stadium Sleep Clause", which is how it's implemented on PS. A few notes on this Sleep Clause;
  • If you use Rest while a Pokemon has already been put to sleep on your end, it still works.
  • If a Pokemon that has been put to sleep by the opponent wakes up while you have a Pokemon using Rest, Sleep Clause remains active. Vice versa also applies.
  • This basically means that as long as you can get a Rest off, you can invalidate sleep moves. You can imagine what this does from a metagame perspective.
I believe that the Sleep Clause works this way because of how the checks work. I think it's something like "if the opponent has sleeping mon, the sleep move fails" while Rest doesn't have a check for this at all. Since the game can't differentiate the way the Pokemon went to sleep, it just sees it as any other situation. Just my thoughts.

--

And some minor addendums to the above post, just to add context to the rules.
Again I don't actually know how any of this plays out ingame, but for some mechanics, we can refer to the Colosseum manual (page 45). Apparently the player with the fewest fainted mons will win, and if they're the same it compares HP totals to determine a winner. It also says running out of time will automatically select the first move slot, just like the main games do.
I believe that this happens in Stadium 2, but Beelzemon 2003 would absolutely know this better than I do.

and also disabling Skill Swap for some reason. I am not sure if this makes it fail or just bans it, but I'd assume fail.
Don't hold me to it, but I believe Skill Swap was banned in some Nintendo-run tournaments, maybe due to fears of Slaking? I believe this may be part of that. Regardless, that rule makes the move fail. Official Gen 3 competitive is very poorly documented compared to Gens 1-2 and 4-8, so this context may not be quite right.

having Destiny Bond/Perish Song fail when used by the last mon.
This was also innately run in Stadium 2 as part of Self-KO Clause, and I believe they made it so you forfeit if you use Perish Song last in Nintendo Cup 2000.
 
Don't hold me to it, but I believe Skill Swap was banned in some Nintendo-run tournaments, maybe due to fears of Slaking? I believe this may be part of that. Regardless, that rule makes the move fail. Official Gen 3 competitive is very poorly documented compared to Gens 1-2 and 4-8, so this context may not be quite right.
https://bulbapedia.bulbagarden.net/wiki/Rule_variants#Pok.C3.A9mon_Masters_2005

Pokemon Fiesta was the first tour, in 2004, to start playing Doubles as the official format. The Skill Swap + Slaking combo was a fearfull one in RS due to the lack of Pokemon diversity, but this was solved in FRLG. However, I think GF just wanted to play safe with Slaking in case it becomes problematic for tours (yep, they didn't know what they are doing).
 
@doipy hooves @Plague von Karma I did a brief investigation of competitive Clauses in Gens 3 and 4; however, since I don't have those games I don't know how they're enforced. Anyway, in the case of Pokémon Stadium 2, if you run out of time during your Selecting Action Phase, the computer chooses Move 1 for your pokémon, which is assigned to the C Up button, if that move doesn't have PPs, Move 2 (C Right) is chosen; again, if that move also doesn't have PPs, Move 3 (C Down) is chosen; if Move 3 doesn't have PPs, Move 4 (C Left) is chosen; finally, if none of your moves have PPs, Struggle is chosen.

Regarding Self KO Clause in Stadium 2, if you use Destiny Bond and Perish Song, the moves simply fail; however, you can continue playing without being penalized. We don't know how Nintendo enforced this rule in Nintendo Cup 2000, but the Japanese and Smogon players have recently agreed that if a player's final pokémon uses Destiny Bond or Perish Song in Gold, Silver or Crystal, he should only be penalized if his pokémon dies due to the effects of those moves, so if a Gengar uses Destiny Bond but is not killed after that, his trainer can continue playing. In the case of Perish Song, when the player's last pokémon activates the move, he has three more turns to win the match, because at the end of the third turn Perish Song's effect will activate and will kill his Pokémon giving him a loss for breaking Self-KO Clause.

Now let's see what happens in Stadium 2 when the time runs out. If the time runs out and both players have the same number of remaining pokémon, the players whose pokémon have the biggest sum of Hit Points percentages wins, if this number is the same, then the match is declared a Draw (I've tested this in the real game). Let's say that both players only have one pokémon left, one is a Snorlax and one is a Chansey and both have full HPs, the match ends in a tie despite the fact that Chansey has 703 HPs and Snorlax has 523 (the game only takes into account HPs percentages, not the actual numbers). I don't know how this rule is enforced in Pokémon Stadium 1 because I don't have a second Transfer Pak to test Event Battle mechanics, but as soon as I get one I'll test them and let you know.

By the way, the Japanese and Smogon communities have been building the Gen 2 Poké Cup Suggested ruleset, you can see it here if you're interested.
 
Last edited:
I have made a very important RBY discovery! It is completely possible to EV train a Pokemon in game in RBY such that they gain EVs in all but one stat.

In RBY and GSC, EVs are 0-65535 instead of 0-255 and use a special formula to determine how much stats are affected by EVs. When a Pokemon KOs another Pokemon in RBY and GSC, the KO-er adds the base stats of the KO'd Pokemon to its EV stats, causing all of its EVs to rise together instead of gaining EVs in specific stats like in later generations. For this reason, it was thought to be impossible to EV train a Pokemon such that one stat remained at 0 EVs while every other had 65535. This, as it turns out, is wrong.

I finally worked out how EVs are gained in battle in RBY after realizing that battle participants & EXP share splits up how much EVs are gained from KOing a Pokemon.

Without EXP Share:
Battle Participant EVs gained = Base Stat ÷ # of Battle Participants

With EXP Share:
Battle Participant EVs gained = Base Stat ÷ 2 ÷ # of Battle Participants
Party EVs gained = Base Stat ÷ 2 ÷ # of Battle Participants ÷ # of Party Members

The Party EVs gained with EXP Share equipped is the critically important formula. It is technically bugged, resulting in 25% or more Experience / EVs being lost if more than one Pokemon participate in a fight while EXP Share is equipped.

To test this, I wanted to see if I could EV train a perfect 15/15/15/15 Mewtwo with 0 EVs such that its Special stat stays at 341 to avoid the "Special Drop Overflow" glitch while maxing out all other stats. I then looked up a Pokemon with its Special base stat lower than all other stats, and found Caterpie. Lowest base stat is Special at 20, second lowest is Attack at 30. I then looked up how many battle participants and party members I needed to abuse integer divides to divide Special into 0, and Attack into 1. I found that 2 participants and 6 party members worked for what I needed. Mewtwo must only receive shared party EXP, so it was not a battle participant for the fight against Caterpie.

Fighting Caterpie With EXP Share, 2 Battle Participants and 6 Party Members:
Party HP EVs gained = 45 ÷ 2 = 22 ÷ 2 = 11 ÷ 6 = 1
Party Attack EVs gained = 30 ÷ 2 = 17 ÷ 2 = 7 ÷ 6 = 1
Party Defense EVs gained = 35 ÷ 2 = 15 ÷ 2 = 8 ÷ 6 = 1
Party Special EVs gained = 20 ÷ 2 = 10 ÷ 2 = 5 ÷ 6 = 0
Party Speed EVs gained = 45 ÷ 2 = 22 ÷ 2 = 11 ÷ 6 = 1

I confirmed afterwards via a save editor that the 0 EV'd Mewtwo gained 1 in all stats, EXCEPT the Special stat. A Pokemon needs 63001 EVs to get max stats in the GameBoy games, and vitamins only get you to 25600. That means with an EXP gain of 1 in all stats except the one I want to stay at 0, it would require finding, switching to a second Pokemon and KOing 37,401 Caterpies. This allows me to get a Mewtwo with 341 Special and max stats everywhere else, albeit by VERY patiently committing genocide on the Caterpies of Viridian Forest.

Now, say I also wanted a 0 Attack EV Chansey so I can minimize self inflicted Confusion damage. I simply look up a Pokemon with Attack as its lowest stat, figure out how to divide just Attack into 0, and voila:

Fighting Magikarp With EXP Share, 1 Battle Participant and 6 Party Members:
Party HP EVs gained = 20 ÷ 2 = 10 ÷ 1 = 10 ÷ 6 = 1
Party Attack EVs gained = 10 ÷ 2 = 5 ÷ 1 = 5 ÷ 6 = 0
Party Defense EVs gained = 55 ÷ 2 = 27 ÷ 1 = 27 ÷ 6 = 4
Party Special EVs gained = 20 ÷ 2 = 10 ÷ 1 = 10 ÷ 6 = 1
Party Speed EVs gained = 80 ÷ 2 = 40 ÷ 1 = 40 ÷ 6 = 6

By overfishing the oceans and abusing the power of teamwork, I can get my formerly thought to be illegal 0 Attack EV Chansey and laugh at the poor Lapras trying to Confuse Ray it and seeing Chansey barely scratch itself. And then my 0 Attack EV Alakazam, 0 Attack EV Starmie, 0 Attack EV Jynx... There probably is more efficient setups to make EV training slightly less of a grinding nightmare, but the simple fact it is possible at all means there is no such thing as "illegal EVs" in RBY.

I am not yet sure whether the same integer division abuse works in GSC, but it should at minimum work for 0 Speed EVs. Shuckle only has a base Speed stat of 5, so KOing a Shuckle with 6 battle participants should yield 0 Speed EVs and atleast 1 in all other stats.

I can get image or video proof if it is required.

Edit: It also occurred to me that Experience gain uses the exact same code as EVs. There had been a question at some point as to whether or not you could EV train Level 5 Pokemon for Little Cup, and whether the Daycare reset experience. I don't recall what was determined about the GSC daycare, but the exact same methods used to gain 0 EVs in one stat could also be used to gain 0 Experience while still gaining EVs, making it possible albeit super tedious to fully EV train a Level 5 Pokemon (or level 2 Rattata) entirely within RBY. I'll do a test later to confirm.
 
Last edited:
@Enigami Excellent discovery chum, way to go! By the way, I wanted to ask, is it possible for Pika Cup pokémon to get max Effort Values on all stats with the method mentioned above? Pika Cup is the lowest level format of Gen 1, it has pokémon of levels 15-20. I ask, because if it is possible to max EV those mons with Gen 1 methods, that would mean that the player could also use the Gen 2 Daycare trick as an alternative to make the process faster.
 

AM

is a Community Leader Alumnusis a Community Contributor Alumnusis a Tiering Contributor Alumnusis a Contributor Alumnusis a Battle Simulator Moderator Alumnusis a Past WCoP Champion
LCPL Champion
Greetings, recently been thinking why hasn't Hydreigon been banned from BW OU since it literally holds the title of completely uncounterable (at least in BW OU), which makes me realize its one of very few uncounterable Pokemon that doesn't get banned to Ubers.

My argument is that Hydreigon since it has no counters, because by definition, a counter is a Pokemon that can switch into it and then beat it, Hydreigon has no counters because every Pokemon in the game gets 2HKO'd on switch in therefore I think Hydreigon should be suspected.
Wrong thread to post this in since this is for old gen mechanic research, question more appropriate for this thread here below.

https://www.smogon.com/forums/threa...-a-simple-answer-mark-ii-roa-edition.3468567/
 
I have confirmed that EV training while gaining 0 Experience is possible in RBY.

0 exp.PNG


By utilizing the integer division exploit mentioned in the above post, you can use Route 1 Pidgeys and Rattatas for your EV training needs. I've confirmed via save editor that my test subject, a level 2 Rattata, gained 0 Experience, while rising its EVs by atleast 1 in each stat after KOing a Level 2 Rattata with EXP Share equipped, 2 Battle Participants and 6 Party Members. I've also found that L5 Magikarps fought via Old Rod + EXP Share + 1 Battle Participant + 6 Party Members yield 0 EXP and 0 Attack EVs, so you can even train for 0 Attack EVs max everything else in low level metagames like Little Cup and Pika Cup.
 
I have confirmed that EV training while gaining 0 Experience is possible in RBY.

View attachment 315847

By utilizing the integer division exploit mentioned in the above post, you can use Route 1 Pidgeys and Rattatas for your EV training needs. I've confirmed via save editor that my test subject, a level 2 Rattata, gained 0 Experience, while rising its EVs by atleast 1 in each stat after KOing a Level 2 Rattata with EXP Share equipped, 2 Battle Participants and 6 Party Members. I've also found that L5 Magikarps fought via Old Rod + EXP Share + 1 Battle Participant + 6 Party Members yield 0 EXP and 0 Attack EVs, so you can even train for 0 Attack EVs max everything else in low level metagames like Little Cup and Pika Cup.
So, you even found a competitive use for gen1 Old Rod...
RIP Lapras trying to parafusion win vs Chansey.
 

Plague von Karma

Banned deucer.
Hiya, EB0LA and some Gen 1 Random Battle players asked me to test a mechanic that was implemented on PS regarding player input removal.

To sum up, if you're frozen directly after using Hyper Beam on the same turn, player input will be removed until the Pokemon is KOed. This is similar to the Haze + Sleep/Frz + Multi-turn move debacle. You effectively have the player input removal segment of Hyper Beam and the permanent nature of freeze wrapped into one status. Ergo, you can't attack, you can't switch, you can't even forfeit.

Skip to the 4-minute mark and you'll see the glitch in all its glory.
 

EB0LA

Banned deucer.
A) Ban FRZ Moves
B) Ban Hyper Beam
C) Mod ("slippery slope")
D) Leave "Uncompetetive" Mechanic

Showdown surprisingly has this implemented correctly btw.

This is not different from Fly/Dig bug bar it's a little less likely.
 
Last edited:

xzern

for sure
is a Tiering Contributoris a Community Contributor Alumnusis a Contributor Alumnusis a Smogon Media Contributor Alumnus
yes hello. with the help of SuperEpicAmpharos and zeefable i am here to report a new finding in adv doubles mechanics.

there is a bug in which, when both of your pokemon are holding quick claw, the claw proc will affect both of the mons.

this mechanic is not described in either bulbapedia, showdown, or anywhere else we looked.
this is somewhat tricky to determine, as gen 3 does not report the claw proc with text. instead, the mon simply moves first. in order to lab this as easily as possible, i brought the slowest mons possible to the first double battle in emerald. in case you are curious of the specifics; each of my mons have 6 speed, while the lowest possible speed for both seedot and lotad at level six is 7. in this video, you can definitively see the claw glitch in question proc three separate times; turns 1, 4, and 5.

whether the 20% chance stacks with both mons is unclear. in a separate trial of 50 turns, the claw glitch only went through seven times. im under the impression that the chance remains at 20%. when both mons are holding the claw, there is never an incident where one moves first and the other doesnt; they always move first together.

sea also tested this with other items of the same caliber; specifically focus band and scope lens. so far only quick claw has yielded this result. furthermore, we confirmed that this only works with two quick claws, not one.
 

Users Who Are Viewing This Thread (Users: 1, Guests: 0)

Top