Battle Tree Discussion and Records

[paperquagsire]

I get that's a joke, but I just thought I'd clear up for anyone wondering that that was a development error most likely and kommo-o was meant to have shell smash as a real move probably, and it was removed

EDIT: Worldie beat me to it

 

[Smuckem]

100 pages! Great work everybody, we seem to have a ton of momentum built up to carry into USUM Tree...high fives all around.

I suppose that, now that the thread has its Lansat Berry, so to speak, I'm thinking the next goal is 125 pages before USUM release.

You may carry on the current argument now.

 

[Worldie]

Do not worry, once I'm done messing up with Java, Josh C. 's shiny Aromatisse will take his baptize on the Btree, going for that 100 or more.

 

[Hiddenfreezer]

it's just something some people say to convince themselves they can't win because of it, rather than having teams with gaping holes to certain pokemon, moves, or trainers

And again this garbage. I never said I lost to the AI cheating having Ice or Fairy Types in my example, I literally lost to a fighting team as I had Slowbro, a very good mon against most Fighting Types, due to a misplay, which is in the original post. But don't worry, you are at home with your not-being-able-to-read here.

I said the Anti-Dragon(Ice, Fairy Types) leads were very, very common. However Worldie made a good point that the possible mons in this range are Ice Type specialist, so that makes more sense why there appear often. That would explain my anectodal experience.

About the Fairy specialist from Worldie, how is Mimikyu a problem when Haxorus OHKO and outspeeds it always? The problem mons from that list is MegaMawile, Whimsicott and Primaseal for Haxorus itself. Latter is dealt with Chansey. Former, depending on the EV it runs, means a change of MegaBro's EV so that it outspeeds 0EV Variants(if needed). If its a fast variant, my only hope +85% EQ damage, meaning that Chansey could potentially weaken it with Stoss and Haxorus can deal with it. I admit losing Chansey against a Fairy specialist like that would be absurd.
I don't know enough about the AI of the enemy about how Whimsicott LeechSeed spam works. Do some people here stall it out in any feasible way?

 

[paperquagsire]

And again this garbage. I never said I lost to the AI cheating having Ice or Fairy Types in my example, I literally lost to a fighting team as I had Slowbro, a very good mon against most Fighting Types, due to a misplay, which is in the original post. But don't worry, you are at home with your not-being-able-to-read here.

I said the Anti-Dragon(Ice, Fairy Types) leads were very, very common. However Worldie made a good point that the possible mons in this range are Ice Type specialist, so that makes more sense why there appear often. That would explain my anectodal experience.

Alright sorry about that, you're right, can't read :p

But yeah there's quite a lot that is 'good' haxorus, and haxorus isn't good on many pokemon due to having only dragon stab, have you considered maybe a different lead?

Edit: dragonite would be pretty good against chesnaught, could even run some z-fly set or something if you want, or just lum berry

 

[Hiddenfreezer]

But yeah there's quite a lot that is 'good' haxorus, and haxorus isn't good on many pokemon due to having only dragon stab, have you considered maybe a different lead?

I admit that Haxorus is very subpar, I only chose it because it has electric/grass resist, can sweep with DD and has Mold Breaker against some annoying abilities.
My initial starter should be Landorus-T with a fitting Z Move("funnily" enough Steelium Z might still be good if Ice-Type spam happens again). My traded Lando was sadly hacked though, which sucks. If I use hacked Pokemon I can be immediately go to Misty Terrain + Sturdy Googles Shedinja :/


Probably I should be lame and try Durant teams or something in the beginning, dunno. Have no QR-Reader though, have to breed it myself :/

 

[paperquagsire]

I admit that Haxorus is very subpar, I only chose it because it has electric/grass resist, can sweep with DD and has Mold Breaker against some annoying abilities.
My initial starter should be Landorus-T with a fitting Z Move("funnily" enough Steelium Z might still be good if Ice-Type spam happens again). My traded Lando was sadly hacked though, which sucks. If I use hacked Pokemon I can be immediately go to Misty Terrain + Sturdy Googles Shedinja :/


Probably I should be lame and try Durant teams or something in the beginning, dunno. Have no QR-Reader though, have to breed it myself :/

You can even try some durant+haxorus team, I remember in early Maison someone did that with brick break, shadow claw, protect, and ddance for unresisted coverage. Wouldn't say it's optimal but you wouldn't have to breed a new sweeper

EDIT: found the team, as I said not very optimal

Just an update that I've passed 50 with the Haxorus after breeding a perfect one. Here's the set so far

Durant @ Choice Scarf
Ability: Truant
EVs: 252 HP / 252 Spe / 4 SDef
Jolly Nature
Entrainment
Crunch
Guillotine
Iron Head

Greninja @ Life Orb
Ability: Protean
EVs: 252 SAtk / 252 Spe / 4 HP
Timid Nature
Surf
Ice Beam
Grass Knot
Dark Pulse

Haxorus @ Lum Berry
Ability: Mold Breaker
EVs: 252 Atk / 252 Spe / 4 HP
Adamant Nature
Protect
Dragon Dance
Shadow Claw
Brick Break

Thinking about swapping the Greninja for Garchomp but I don't feel comfortable with two physical attackers. Greninja also has better SE coverage and all moves are 100% accurate. Any suggestions or input would be appreciated!

 

[Coeur7]

EXCISED

 

[PikaAbuser]

Coeur7 I don't think anyone is specifically referring to you on this matter. I don't why you are taking it personally. Just more generally on the act of doing a multi-team streak then reporting in as a single-team streak.

As for the AI cheating. There are definitely some anecdotal evidence going around. But I've posted video before on AI's Quickclaw activating 3-4 times in a row.
Here is an video from earlier post, not sure if I took it down already: WCVG-WWWW-WWW8-GD5J
But something like this is easy to test, just find a Quick Claw opponent and stall to see how many times it activates, if the percentage is way above what it's supposed to be (20%?), then we can assume that there is RNG manipulation going on, which probably would also apply to other things the AI does too.

 

[PrivateJoker]

Or another way to test it could be to ask people who run these facilities for thousands of battles, and ask them if it seems like Quick Claws activate more often than they should. You'll likely find that the answer is "no" from any and every person with a decent streak on the leaderboard, and anyone who says "yes" hasn't actually played many battles in the Tree, and cites one, possibly even two isolated incidents as evidence that the AI obviously cheats.

 

[Worldie]

About the Fairy specialist from Worldie, how is Mimikyu a problem when Haxorus OHKO and outspeeds it always? The problem mons from that list is MegaMawile, Whimsicott and Primaseal for Haxorus itself. Latter is dealt with Chansey. Former, depending on the EV it runs, means a change of MegaBro's EV so that it outspeeds 0EV Variants(if needed). If its a fast variant, my only hope +85% EQ damage, meaning that Chansey could potentially weaken it with Stoss and Haxorus can deal with it. I admit losing Chansey against a Fairy specialist like that would be absurd.
I don't know enough about the AI of the enemy about how Whimsicott LeechSeed spam works. Do some people here stall it out in any feasible way?

I am glad you do have some plans for Xio, because yes you would need it.

For your own personal interest, this are the 2 mimikyu sets Xio has:

Mimikyu-3 will 1hko your Haxorus turn 1, so if you ever run into a Mimikyu in your streaks past battle 30, bare it in mind.
..... why did they put a Rocky Helmet on Mimikyu-4, GF pls.

Now, aknowledging that you are not claiming the AI cheats or similar and probably just vented out some frustration due to bad RNGs and a misplay, I am quite sure you noticed that due to the amount of shitstorm that happens every other month since the existence of facilities of "AI hax" claims, we do not take such claims lightly at all and are very aggressive in denying those.

We can guarantee you that the AI does not attempt to purposely throw Ice or Fairy types at you, and it's more as you have noticed of an issue with a weak/mediocre lead that has too many weak matchups.

If you have lack of Pokemon to play/breed with, feel free to pass by Discord, most of us have healthy reserves of literally everything and do not mind sharing more eggs.

 

[PikaCuber]

I guess I may as well ask, in case it ever happens to anyone else:

Assuming both sides are down to their last Pokemon, if I KO my opponent but also KO myself through recoil, does that still count as a win?

And I really don't like Cobalion3.

 

[Worldie]

I *THINK* that since the B-tree follows essentially the same ruleset as VGC and Battle Spot as far as mechanics go, that the person who goes KO first loses, so KOing a opponent while also dieing yourself at same time to recoil should count as a win.

I don't think any of us has ever been in said situation, though, to verify.

 

[PikaAbuser]

Or another way to test it could be to ask people who run these facilities for thousands of battles, and ask them if it seems like Quick Claws activate more often than they should. You'll likely find that the answer is "no" from any and every person with a decent streak on the leaderboard, and anyone who says "yes" hasn't actually played many battles in the Tree, and cites one, possibly even two isolated incidents as evidence that the AI obviously cheats.

That wouldn't be a test, that would just be a survey with typical reporting biases. By actually running an actual test, there would be conclusive empirical data so the topic can finally be laid to rest.

 

[PrivateJoker]

That wouldn't be a test, that would just be a survey with typical reporting biases. By actually running an actual test, there would be conclusive empirical data so the topic can finally be laid to rest.

stalling a single Quick Claw user is not "conclusive empirical data", it's an insufficient sample size. Even if all 3 of your mons were tanky stall based sets, the Quick Claw user would struggle itself to death before you gathered enough.

 

[Darth_Scorpion]

I guess I may as well ask, in case it ever happens to anyone else:

Assuming both sides are down to their last Pokemon, if I KO my opponent but also KO myself through recoil, does that still count as a win?

And I really don't like Cobalion3.

I believe I have seen this in my own runs, and I think Worldie is correct - in a 1v1 situation, the Pokémon that is KO'd first loses even if the second faints due to recoil.
This may also work against you of course...

 

[paperquagsire]

I guess I may as well ask, in case it ever happens to anyone else:

Assuming both sides are down to their last Pokemon, if I KO my opponent but also KO myself through recoil, does that still count as a win?

And I really don't like Cobalion3.

Sorta just confirming what worldie said, but yeah whoever dies first wins, I once lost to a metagross and rough skin still killed it on the final blow, but AI fainted second so they won

 

[Altissimo]

As for the AI cheating. There are definitely some anecdotal evidence going around. But I've posted video before on AI's Quickclaw activating 3-4 times in a row.
Here is an video from earlier post, not sure if I took it down already: WCVG-WWWW-WWW8-GD5J
But something like this is easy to test, just find a Quick Claw opponent and stall to see how many times it activates, if the percentage is way above what it's supposed to be (20%?), then we can assume that there is RNG manipulation going on, which probably would also apply to other things the AI does too.

I think it's safe to say at this point that you have close to no idea what you're talking about when it comes to the Battle facilities.

 

[Smuckem]

I think it's safe to say at this point that you have close to no idea what you're talking about when it comes to the Battle facilities.

No reason to bash him for it though, he's still pretty new at this (if the post count is anything to go by). Everyone learns, just give it some time.

PikaAbuser so who's next on deck for Project: Mega Awesome?

 

[Worldie]

Speaking of AI cheats, I wanted to mention an actual proved AI cheat.

If you are at work or in office, the AI of Battle Tree will inevitably generate several Sightseeker, Dancer or Golfer trainers in a row, just to remind you that your life sucks.

 

[Pencilcheck]

The RNG (random number generator) that determines if a quick claw is activated is randomized, and only 20% of the time it will pick. I personally can't and don't know what RNG they use. But it is possible that sometimes RNG might be a bit bias based on the sources it is generating from, but in the long run quick claw don't always active if you count every turn in the battle that the pokemon is on the field, or even count the ones that you never fought. Remember statistics should count everything.

 

[Worldie]

As people pointed out on Discord, it's definitely possible that if the RNG generator used to generate AI trainers is comparable to the one used to generate wild pokemon and gifts, so possibly based on date, hour, player ID, or frames, or similar things, it does not surprise that occasionally there could be "repeats" of similar sets in short terms.

In fact several of us reported that not-too-rarely you could run in the same trainer or trainer archetype 2-3 times over 10 matches for example . However, even in such situations, it wasn't generally "A archetype that counters", rather, the same archetype, which generated both edges of super easy matches or relatively hard.

Unfortunately, there'd be no realistic way to prove if the RNG generation is indeed linked to fixed numbers, possibly even manipulable in same way as RNGing Pokemon, without actually hacking the game, but it is definitely a possibility nontheless.

 

[Smuckem]

Unfortunately, there'd be no realistic way to prove if the RNG generation is indeed linked to fixed numbers, possibly even manipulable in same way as RNGing Pokemon, without actually hacking the game, but it is definitely a possibility nontheless.

Well, I assume Team Rocket Elite can suss this out, but we'll just have to wait and see if he ever takes an interest in such a project.

Also, as noted on the Discord, I think several of us could just start logging every Trainer we run into on a given day (and by extension, during a given streak) and see if any noticeable patterns pop up. Since some of us can't seem to crack into the leaderboard despite everything we try, and some of us can breeze to 500 without blinking, we would be capably able to note any trends on both the pre-Legend and post-Legend sides of this 'issue'.

 

[Worldie]

To be honest the whole "not crack into the leaderboard despite everything" might have something to do with us insisting in using trash for the sake of it instead of just being realistic and running 4 uber/ou level threats :P

I'm also still contemplating if attempting another go at getting Vileplume to 100 or not.

 

[Pencilcheck]

Also your pokemon's team or move's presence will greatly alter the sequence of those trainers popping up as well.
But I do recognize that some people here that has a lot of streaks sometimes will lose to a battle to a team combo that I will meet before 10 or 20 battles. Either they are super lucky, they play on an old AI that prob didn't adapt as quickly, or they switch teams... :P