Gen III Battle Frontier Discussion and Records

[Actaeon]

Short update on the monotype challenge: I was so impressed with Stantler on the Christmas team I decided to see if my Mono Normal could get any better by using this kind of setup/archetype. The Xmas Suicune is probably unparallelled with Stantler support, but I thought CurseLax was probably still going to do a decent job, provided I play carefully with Substitute (similar to using Mono-Steel Registeel). And not having to use Delibird is of course a big plus. So I came up with this, and it hasn't lost so far:

Stantler @ Sitrus Berry
Ability: Intimidate
EVs: 252 HP / 228 Def / 20 SpD / 4 Spe
Bold Nature (+Def, -Atk)
IVs: 0 Atk
- Thunder Wave
- Sand-Attack
- Thief
- Skill Swap

Same old moves, but this Stantler is physically oriented because Blissey is on the team and it doesn't have to paralyze Espeon/Jolteon/Sceptile/Starmie necessarily. Instead, this one survives a 170+ MM two times unless it deals more than 103 damage two times in a row, which is only a (2/16)^2=1/64 chance, not factoring in misses or crits. As a bonus, it survives Cross Chops and stuff, of course. The physical bulk allows it to Thief better as well against many Quick Claw users. Otherwise, this thing does the same as it does on the Christmas team.

Blissey @ Leppa Berry
Ability: Natural Cure
EVs: 8 HP / 252 Def / 68 SpD / 180 Spe
Bold Nature (+Def, -Atk)
IVs: 0 Atk
- Growl
- Substitute
- Softboiled
- Mud-Slap

Growl further cripples opponents, guaranteeing Snorlax' setup against otherwise powerful Strugglers. It wins PP wars too. Substitute for hax prevention and scouting misses, and Mud Slap is for when Stantler can't get enough Sand-Attacks off, also preventing Focus Punch sometimes. Calm Mind or Psych Up are viable alternatives, as well as Thunder Wave, but most of the time Special Attackers don't plow through because their accuracy is reduced and/or they are paralyzed.

Snorlax @ Leftovers
Ability: Thick Fat
EVs: 164 HP / 100 Atk / 92 Def / 152 SpD
Careful Nature (+SpD, -SpA)
IVs: 30 SpD
- Curse
- Substitute
- Return
- Hidden Power [Steel]

Wins pretty much every time it gets to +6, it's CurseLax what can you say.

Problems so far:

• Marowak is dangerous and almost a reason to run Roar on Stantler; you can Sand-Attack it three times and then hope to stall it out with Blissey. Luckily, the Bonemerang version doesn't have Swords Dance.
• Espeon and Gardevoir can Calm Mind up and outspeed my squad, but they often end up missing way too many moves. Don't be afraid to use Blissey's Leppa Berry for Substitute PP instead of Softboiled.

Curious if you think this is better than my tricky Linoone/Porygon2/(Vigoroth/Tauros) team that's also not that bad.

 

[Actaeon]

I was toying around with Mono Steel again, because I'm still not satisfied with the number of times I can setup Registeel.
It made me wonder what the probability distribution of the opponent's Accuracy stage is if you want to cripple the Accuracy of a paralyzed Pokemon with a Substitute/Sand-Attack staller like Skarmory, possibly with Protect to increase PP usage. How many Sand-Attacks will it get off on average against a threatening opponent, e.g. a Thunderbolt or Flamethrower-user? Does Brightpowder give better odds than Leftovers if Protect is used?

I was too lazy to implement Markov chains here, as my HP, the Substitute toggle and both my PP as the user's PP are hard to implement without enormously increasing the number of states. So I opted for a simulation; according to a well-known normal approximation to the binomial distribution, the theory of confidence intervals guarantees a estimator precision of around 0.03% with more than 95% confidence if the average of around n=10^7 (that's ten million) simulations is used as the estimator of the actual probability.

What's the strategy? If Protect is enabled, the battle always starts with Protect. Then you use Substitute and on free turns (when they miss or get FP'ed while you have a Substitute up) you use Sand-Attack. If you can use Protect with 100% certainty, you do and otherwise, you don't. If you run out of Protect PP you can still Substall and if both Sub and Protect run out of PP, you'll just spam Sand-Attack. If you're too low on HP, you'll use Protect and Sand-Attack. If the crippler has Leftovers, it might be able to use Substitute again later.

If the opponent runs out of Attacking PP (a variable in my program), it is considered to be at -6 Accuracy automatically because that's probably the result. The program doesn't take into account actual damage calculations; it just assumes that the Substitute always breaks (otherwise you can use better strategies and don't need this probability), so keep in mind Flamethrower against Skarmory for example.

This is the program, feel free to customize and use it for your own purpose (although I certainly didn't write an elegant code). I recently updated this so Pressure, Leftovers, Paralysis and some other things can be toggled on and off. I noted that Smogon sadly doesn't copy the indentations in the code.

SubAccuracyStall.py

import numpy as np
import sys, os
from tqdm import tqdm

MonteCarlo = 10000000

#######################################################################
#ONLY PUT TRUE IF YOU WANT TO TEST THE OUTPUT (SET MONTE CARLO TO 1)
log = False

#MODIFY TO -1, -2, ETC, IF OPPONENT HAS BEEN CRIPPLED BEFOREHAND
INITOpponentAccStage = 0

#BASE ACCURACY OF THE MOVE, I.E. PUT 0.75 FOR THUNDER
INITMoveAccuracy = 1

#ONLY MODIFY TO 0.9 IF Crippler IS HOLDING BRIGHTPOWDER
INITCripplerEva = 0.9

#A PERFECT LEFTOVERS NUMBER IS 1 MODULO 16, E.G. 177 AND 353
INITCripplerHP = 169

#MAX TOTAL PP OF THE ATTACK(S) TO BE STALLED OUT
INITAttackingPP = 100

#PUT FALSE IF YOU DON'T WANT TO USE PROTECT IN BETWEEN TURNS
INITProtectToggle = True

#PUT FALSE IF YOU DON'T WANT LEFTOVERS (E.G. WHEN BRIGHTPOWDER IS HELD)
INITLeftoversToggle = False

#PARALYZED OPPONENT, ONLY PUT FALSE IF YOU OUTSPEED WITHOUT PARALYSIS
ParalysisToggle = True

#SET TO 2 IF YOUR CRIPPLER HAS PRESSURE
PPUsedOnMove = 1
#######################################################################

#DON'T CHANGE THIS
INITSubToggle = False
INITProtecting = False
INITProtectPP = 16
INITSubstitutePP = 16

#NUMBERS THAT DON'T CHANGE INSIDE THE LOOP
SubstituteHP = int(INITCripplerHP / 4)
LeftoversHP = int(INITCripplerHP / 16)

# DISABLE PRINT
def blockPrint():
sys.stdout = open(os.devnull, 'w')

# RESTORE PRINT
def enablePrint():
sys.stdout = sys.__stdout__

# DETERMINE WHETHER THE ATTACK HITS IF NOT FULLY PARALYZED
def threshold(OpponentAccStage):
if OpponentAccStage == 0: return MoveAccuracy*CripplerEva*1.0
elif OpponentAccStage == -1: return MoveAccuracy*CripplerEva*0.75
elif OpponentAccStage == -2: return MoveAccuracy*CripplerEva*0.60
elif OpponentAccStage == -3: return MoveAccuracy*CripplerEva*0.50
elif OpponentAccStage == -4: return MoveAccuracy*CripplerEva*0.43
elif OpponentAccStage == -5: return MoveAccuracy*CripplerEva*0.36
else : return MoveAccuracy*CripplerEva*0.33

# TRIES TO EXECUTE ATTACK, BUT FIRST CHECKS WHETHER A FULL PARALYSIS OCCURS
def attack(OpponentAccStage, CripplerEva):
global CripplerHP
global SubToggle
global AttackingPP
global Protecting
if np.random.rand() < 0.25*(1 if ParalysisToggle == True else 0):
print("Opponent is fully paralyzed!")
elif Protecting == True:
print("Opponent used Attack but Crippler protected itself!")
AttackingPP = max(AttackingPP - PPUsedOnMove,0)
elif np.random.rand() < threshold(OpponentAccStage):
if SubToggle:
print("Opponent used Attack and broke Crippler's Substitute!")
SubToggle = False
else:
print("Opponent used Attack and Crippler fainted!")
CripplerHP = 0
AttackingPP = max(AttackingPP - PPUsedOnMove,0)
else:
print("Opponent used Attack and missed!")
AttackingPP = max(AttackingPP - PPUsedOnMove,0)

# EMPTY VECTORS FILLED FOR MONTE CARLO FREQUENCY EXPERIMENT
SurvivalVector = []
AccuracyVector = []
PPVector = []

if log == False:
blockPrint()

# MAIN FOR-LOOP THAT COUNTS FREQUENCY OF EVENTS
for i in tqdm(range(MonteCarlo)):

# MAIN LOOP OF "SUB(/PROTECT)-FISHING" WHILE SAND-ATTACK IS USED ON FREE TURNS
OpponentAccStage = INITOpponentAccStage
MoveAccuracy = INITMoveAccuracy
CripplerEva = INITCripplerEva
CripplerHP = INITCripplerHP
ProtectToggle = INITProtectToggle
LeftoversToggle = INITLeftoversToggle

SubToggle = INITSubToggle
Protecting = INITProtecting
ProtectPP = INITProtectPP
SubstitutePP = INITSubstitutePP

AttackingPP = INITAttackingPP

turn = 1

while CripplerHP > 0:
print("---------- Turn", turn, "----------")

if ProtectToggle == True and Protecting == False and ProtectPP > 0:
print("Crippler used Protect!")
Protecting = True
ProtectPP = ProtectPP - 1
attack(OpponentAccStage, CripplerEva)

elif SubToggle == True or CripplerHP <= SubstituteHP or SubstitutePP == 0:
Protecting = False
print("Crippler used Sand-Attack!")
OpponentAccStage = max(OpponentAccStage - 1,-6)
print("Opponent's Accuracy fell to stage", OpponentAccStage)
attack(OpponentAccStage, CripplerEva)

else:
Protecting = False
print("Crippler used Substitute!")
SubToggle = True
SubstitutePP = SubstitutePP - 1
print("Crippler's HP was reduced to", CripplerHP - SubstituteHP)
CripplerHP = CripplerHP - SubstituteHP
attack(OpponentAccStage, CripplerEva)

if LeftoversToggle == True and CripplerHP > 0 and CripplerHP < INITCripplerHP:
print("Crippler restored", min(LeftoversHP,INITCripplerHP - CripplerHP), "HP with its Leftovers!")
CripplerHP = min(CripplerHP + LeftoversHP,INITCripplerHP)

print("Crippler now has", CripplerHP, "HP.")

if AttackingPP == 0:
print("The opponent is out of Attacking PP!")
OpponentAccStage = -6
break

if OpponentAccStage == -6:
break

turn = turn + 1

# SUMMARY
print("========== END OF CRIPPLING ==========")
if CripplerHP == 0:
print("Crippler has fainted.")
else:
print("Crippler is still alive and has", CripplerHP, "HP left.")
print("The opponent has", AttackingPP,"/", INITAttackingPP, "Attacking PP left.")
print("It's Accuracy was reduced to stage", OpponentAccStage,".")
print("Crippler has", ProtectPP, "Protect PP and", SubstitutePP, "Substitute PP left.")

AccuracyVector.append(OpponentAccStage)
SurvivalVector.append(1 if CripplerHP > 0 else 0)
if CripplerHP == 0:
PPVector.append(INITAttackingPP - AttackingPP)


enablePrint()
print("---------------------------------------------------")
print("The probability distribution of the Accuracy stage:")
for x in range(-6,0):
print("Stage", x, ":", round(AccuracyVector.count(x)/MonteCarlo,3))
print("---------------------------------------------------")
print("Crippler has a", round(sum(SurvivalVector)/MonteCarlo,3), "chance of surviving.")
print("If it faints, this is the distribution of PP used by opponent:")
for x in set(PPVector):
print(x, "PP:", round(PPVector.count(x)/len(PPVector),3))
print("On average,", round(sum(PPVector)/len(PPVector),1), "PP is used before Crippler faints.")

Here's some sample logs with different mechanics enabled. The Accuracy of the move used was 100% in these battles. The program can be toggled to not print these logs if you want to do large simulations.

Leftovers + SubTect

---------- Turn 1 ----------
Crippler used Protect!
Opponent used Attack but Crippler protected itself!
Crippler now has 169 HP.
---------- Turn 2 ----------
Crippler used Substitute!
Crippler's HP was reduced to 127
Opponent used Attack and broke Crippler's Substitute!
Crippler restored 10 HP with its Leftovers!
Crippler now has 137 HP.
---------- Turn 3 ----------
Crippler used Protect!
Opponent used Attack but Crippler protected itself!
Crippler restored 10 HP with its Leftovers!
Crippler now has 147 HP.
---------- Turn 4 ----------
Crippler used Substitute!
Crippler's HP was reduced to 105
Opponent used Attack and broke Crippler's Substitute!
Crippler restored 10 HP with its Leftovers!
Crippler now has 115 HP.
---------- Turn 5 ----------
Crippler used Protect!
Opponent used Attack but Crippler protected itself!
Crippler restored 10 HP with its Leftovers!
Crippler now has 125 HP.
---------- Turn 6 ----------
Crippler used Substitute!
Crippler's HP was reduced to 83
Opponent used Attack and broke Crippler's Substitute!
Crippler restored 10 HP with its Leftovers!
Crippler now has 93 HP.
---------- Turn 7 ----------
Crippler used Protect!
Opponent used Attack but Crippler protected itself!
Crippler restored 10 HP with its Leftovers!
Crippler now has 103 HP.
---------- Turn 8 ----------
Crippler used Substitute!
Crippler's HP was reduced to 61
Opponent used Attack and broke Crippler's Substitute!
Crippler restored 10 HP with its Leftovers!
Crippler now has 71 HP.
---------- Turn 9 ----------
Crippler used Protect!
Opponent used Attack but Crippler protected itself!
Crippler restored 10 HP with its Leftovers!
Crippler now has 81 HP.
---------- Turn 10 ----------
Crippler used Substitute!
Crippler's HP was reduced to 39
Opponent is fully paralyzed!
Crippler restored 10 HP with its Leftovers!
Crippler now has 49 HP.
---------- Turn 11 ----------
Crippler used Protect!
Opponent is fully paralyzed!
Crippler restored 10 HP with its Leftovers!
Crippler now has 59 HP.
---------- Turn 12 ----------
Crippler used Sand-Attack!
Opponent's Accuracy fell to stage -1
Opponent used Attack and broke Crippler's Substitute!
Crippler restored 10 HP with its Leftovers!
Crippler now has 69 HP.
---------- Turn 13 ----------
Crippler used Protect!
Opponent used Attack but Crippler protected itself!
Crippler restored 10 HP with its Leftovers!
Crippler now has 79 HP.
---------- Turn 14 ----------
Crippler used Substitute!
Crippler's HP was reduced to 37
Opponent is fully paralyzed!
Crippler restored 10 HP with its Leftovers!
Crippler now has 47 HP.
---------- Turn 15 ----------
Crippler used Protect!
Opponent used Attack but Crippler protected itself!
Crippler restored 10 HP with its Leftovers!
Crippler now has 57 HP.
---------- Turn 16 ----------
Crippler used Sand-Attack!
Opponent's Accuracy fell to stage -2
Opponent used Attack and missed!
Crippler restored 10 HP with its Leftovers!
Crippler now has 67 HP.
---------- Turn 17 ----------
Crippler used Protect!
Opponent used Attack but Crippler protected itself!
Crippler restored 10 HP with its Leftovers!
Crippler now has 77 HP.
---------- Turn 18 ----------
Crippler used Sand-Attack!
Opponent's Accuracy fell to stage -3
Opponent used Attack and broke Crippler's Substitute!
Crippler restored 10 HP with its Leftovers!
Crippler now has 87 HP.
---------- Turn 19 ----------
Crippler used Protect!
Opponent used Attack but Crippler protected itself!
Crippler restored 10 HP with its Leftovers!
Crippler now has 97 HP.
---------- Turn 20 ----------
Crippler used Substitute!
Crippler's HP was reduced to 55
Opponent used Attack and missed!
Crippler restored 10 HP with its Leftovers!
Crippler now has 65 HP.
---------- Turn 21 ----------
Crippler used Protect!
Opponent used Attack but Crippler protected itself!
Crippler restored 10 HP with its Leftovers!
Crippler now has 75 HP.
---------- Turn 22 ----------
Crippler used Sand-Attack!
Opponent's Accuracy fell to stage -4
Opponent used Attack and missed!
Crippler restored 10 HP with its Leftovers!
Crippler now has 85 HP.
---------- Turn 23 ----------
Crippler used Protect!
Opponent is fully paralyzed!
Crippler restored 10 HP with its Leftovers!
Crippler now has 95 HP.
---------- Turn 24 ----------
Crippler used Sand-Attack!
Opponent's Accuracy fell to stage -5
Opponent is fully paralyzed!
Crippler restored 10 HP with its Leftovers!
Crippler now has 105 HP.
---------- Turn 25 ----------
Crippler used Protect!
Opponent used Attack but Crippler protected itself!
Crippler restored 10 HP with its Leftovers!
Crippler now has 115 HP.
---------- Turn 26 ----------
Crippler used Sand-Attack!
Opponent's Accuracy fell to stage -6
Opponent used Attack and missed!
Crippler restored 10 HP with its Leftovers!
Crippler now has 125 HP.
========== END OF CRIPPLING ==========
Crippler is still alive and has 125 HP left.
The opponent has 9 / 30 Attacking PP left.
It's Accuracy was reduced to stage -6 .
Crippler has 3 Protect PP and 9 Substitute PP left.

Brightpowder + Protect failure

---------- Turn 1 ----------
Crippler used Protect!
Opponent used Attack but Crippler protected itself!
Crippler now has 169 HP.
---------- Turn 2 ----------
Crippler used Substitute!
Crippler's HP was reduced to 127
Opponent used Attack and broke Crippler's Substitute!
Crippler now has 127 HP.
---------- Turn 3 ----------
Crippler used Protect!
Opponent used Attack but Crippler protected itself!
Crippler now has 127 HP.
---------- Turn 4 ----------
Crippler used Substitute!
Crippler's HP was reduced to 85
Opponent used Attack and broke Crippler's Substitute!
Crippler now has 85 HP.
---------- Turn 5 ----------
Crippler used Protect!
Opponent used Attack but Crippler protected itself!
Crippler now has 85 HP.
---------- Turn 6 ----------
Crippler used Substitute!
Crippler's HP was reduced to 43
Opponent used Attack and broke Crippler's Substitute!
Crippler now has 43 HP.
---------- Turn 7 ----------
Crippler used Protect!
Opponent used Attack but Crippler protected itself!
Crippler now has 43 HP.
---------- Turn 8 ----------
Crippler used Substitute!
Crippler's HP was reduced to 1
Opponent used Attack and broke Crippler's Substitute!
Crippler now has 1 HP.
---------- Turn 9 ----------
Crippler used Protect!
Opponent is fully paralyzed!
Crippler now has 1 HP.
---------- Turn 10 ----------
Crippler used Sand-Attack!
Opponent's Accuracy fell to stage -1
Opponent used Attack and Crippler fainted!
Crippler now has 0 HP.
========== END OF CRIPPLING ==========
Crippler has fainted.
The opponent has 6 / 15 Attacking PP left.
It's Accuracy was reduced to stage -1 .
Crippler has 11 Protect PP and 12 Substitute PP left.

Brightpowder + Protect success

---------- Turn 1 ----------
Crippler used Protect!
Opponent used Attack but Crippler protected itself!
Crippler now has 169 HP.
---------- Turn 2 ----------
Crippler used Substitute!
Crippler's HP was reduced to 127
Opponent used Attack and broke Crippler's Substitute!
Crippler now has 127 HP.
---------- Turn 3 ----------
Crippler used Protect!
Opponent used Attack but Crippler protected itself!
Crippler now has 127 HP.
---------- Turn 4 ----------
Crippler used Substitute!
Crippler's HP was reduced to 85
Opponent is fully paralyzed!
Crippler now has 85 HP.
---------- Turn 5 ----------
Crippler used Protect!
Opponent used Attack but Crippler protected itself!
Crippler now has 85 HP.
---------- Turn 6 ----------
Crippler used Sand-Attack!
Opponent's Accuracy fell to stage -1
Opponent is fully paralyzed!
Crippler now has 85 HP.
---------- Turn 7 ----------
Crippler used Protect!
Opponent is fully paralyzed!
Crippler now has 85 HP.
---------- Turn 8 ----------
Crippler used Sand-Attack!
Opponent's Accuracy fell to stage -2
Opponent used Attack and missed!
Crippler now has 85 HP.
---------- Turn 9 ----------
Crippler used Protect!
Opponent used Attack but Crippler protected itself!
Crippler now has 85 HP.
---------- Turn 10 ----------
Crippler used Sand-Attack!
Opponent's Accuracy fell to stage -3
Opponent used Attack and missed!
Crippler now has 85 HP.
---------- Turn 11 ----------
Crippler used Protect!
Opponent is fully paralyzed!
Crippler now has 85 HP.
---------- Turn 12 ----------
Crippler used Sand-Attack!
Opponent's Accuracy fell to stage -4
Opponent used Attack and missed!
Crippler now has 85 HP.
---------- Turn 13 ----------
Crippler used Protect!
Opponent is fully paralyzed!
Crippler now has 85 HP.
---------- Turn 14 ----------
Crippler used Sand-Attack!
Opponent's Accuracy fell to stage -5
Opponent used Attack and missed!
Crippler now has 85 HP.
---------- Turn 15 ----------
Crippler used Protect!
Opponent used Attack but Crippler protected itself!
Crippler now has 85 HP.
---------- Turn 16 ----------
Crippler used Sand-Attack!
Opponent's Accuracy fell to stage -6
Opponent used Attack and missed!
Crippler now has 85 HP.
========== END OF CRIPPLING ==========
Crippler is still alive and has 85 HP left.
The opponent has 4 / 15 Attacking PP left.
It's Accuracy was reduced to stage -6 .
Crippler has 8 Protect PP and 14 Substitute PP left.

Now the interesting results after busting out 10^7 simulations in every case:

SubTect with Brightpowder against infinite PP paralyzed opponent

---------------------------------------------------
The probability distribution of the Accuracy stage:
Stage -6 : 0.592
Stage -5 : 0.055
Stage -4 : 0.068
Stage -3 : 0.082
Stage -2 : 0.098
Stage -1 : 0.105
---------------------------------------------------
Crippler has a 0.544 chance of surviving.
If it faints, this is the distribution of PP used by opponent:
5 PP: 0.0
6 PP: 0.004
7 PP: 0.026
8 PP: 0.087
9 PP: 0.163
10 PP: 0.176
11 PP: 0.136
12 PP: 0.116
13 PP: 0.096
14 PP: 0.077
15 PP: 0.056
16 PP: 0.036
17 PP: 0.019
18 PP: 0.007
19 PP: 0.002
20 PP: 0.0
On average, 11.2 PP is used before Crippler faints.

Note that the Crippler sometimes faints while using its last Sand-Attack, which explains the discrepancy between "Stage -6" and "chance of surviving".

SubTect with Brightpowder against 15 PP paralyzed opponent

---------------------------------------------------
The probability distribution of the Accuracy stage:
Stage -6 : 0.616
Stage -5 : 0.036
Stage -4 : 0.064
Stage -3 : 0.082
Stage -2 : 0.098
Stage -1 : 0.105
---------------------------------------------------
Crippler has a 0.574 chance of surviving.
If it faints, this is the distribution of PP used by opponent:
5 PP: 0.0
6 PP: 0.004
7 PP: 0.028
8 PP: 0.093
9 PP: 0.174
10 PP: 0.188
11 PP: 0.145
12 PP: 0.124
13 PP: 0.102
14 PP: 0.082
15 PP: 0.06
On average, 10.9 PP is used before Crippler faints.

As you can see, the finite PP doesn't matter that much with Brightpowder. Also, some opponents have way more attacking PP that would KO non-bulky Skarmory('s Substitute), for example, such as Starmie. With Leftovers and optimal HP (1 modulo 16) the program logically takes some longer to finish ten million simulations, on mine it takes around 15 minutes. Leftovers is a really big advantage though, even if the opponent might be paralyzed on Protect turns:

SubTect with Leftovers against infinite PP paralyzed opponent

---------------------------------------------------
The probability distribution of the Accuracy stage:
Stage -6 : 0.859
Stage -5 : 0.004
Stage -4 : 0.008
Stage -3 : 0.017
Stage -2 : 0.037
Stage -1 : 0.075
---------------------------------------------------
Crippler has a 0.853 chance of surviving.
If it faints, this is the distribution of PP used by opponent:
8 PP: 0.0
9 PP: 0.0
10 PP: 0.002
11 PP: 0.012
12 PP: 0.045
13 PP: 0.11
14 PP: 0.172
15 PP: 0.17
16 PP: 0.112
17 PP: 0.077
18 PP: 0.068
19 PP: 0.047
20 PP: 0.035
21 PP: 0.031
22 PP: 0.023
23 PP: 0.019
24 PP: 0.017
25 PP: 0.016
26 PP: 0.015
27 PP: 0.012
28 PP: 0.009
29 PP: 0.005
30 PP: 0.002
31 PP: 0.001
32 PP: 0.0
33 PP: 0.0
On average, 16.6 PP is used before Crippler faints.

These are just some examples. There's much more you can calculate, for example without an item, with Accuracy drops beforehand, or with non-Perfect Accuracy moves like Thunder or Meteor Mash. Or even without paralysis and -1 Accuracy (think of Ninjask which outspeeds anyway).

I think this gives good insight on how likely it is you'll send out your setup-sweeper in a good environment.
Please let me know if you enjoyed this, and if you would like to see more of these kind of calculations.

EDIT: New program and updated logs/examples as of 29/12/2020

 

[Actaeon]

Improved the program above (most important with Leftovers).

 

[Actaeon]

This post is about the new Monotype Steel team. The theme is probably clear by now: Crippler/Glue/Sweeper with some tastes of "Deutsche Wissenschaft", taken quite literally because both my lead crippler Registeel and my glue Skarmory are optimized by numerical/statistical methods.

If you don't remember: the old team used Aggron as a crippler (because it is not weak to Fire, but has many other problems), Skarmory as a crippler or endgame Toxicstaller and "German Technology" Registeel as popularized by Kommo-o. Here's the new team; the entire strategy changed and I think I'm getting quite close to the theoretical best team for Steel. Still no Metagross lol, as it can't cripple aside from Icy Wind and Flash, and has no dependable boosting moves, not even starting about moveslot syndrome problems and the inability to sweep well with only one move.

The New Mono Steel Team: "The Steel Shore"

(named after the song "Rusty Cage" performed by Johnny Cash)

Registeel @ Sitrus Berry
Ability: Clear Body
EVs: 164 HP / 12 Atk / 20 Def / 228 SpD / 84 Spe
Careful Nature (+SpD, -SpA)
- Protect
- Thunder Wave
- Mud-Slap
- Rock Tomb

Registeel could be one of the best cripplers in the game, especially on teams where its other teammates wouldn't be weak to Fire. Still, this is better than Aggron (who can run the same set or even better with Icy Wind > Rock Tomb) because it has no 4x weakness and therefore lives every non-crit, non-Focus Punch or Marowak STAB in the game. Another thing it has over Aggron is that it actually cripples many Pokemon without even needing Skarmory; a lot of stuff simply can't beat this without getting to -6 Accuracy in the process.

The moveset is simple: Protect is both for scouting out sets in order to decide whether Thunder Wave or straight Mud-Slap is best. It also aids in defeating big problems in tandem with Skarmory's Toxic (and Protect), or can help stalling out important PP, such as Overheat or Fire Blast (Lum Berry Houndoom for example). Thunder Wave is standard; lets everything on my team outspeed threats and scouts for full paralysis while I have a Substitute up. If that happens anywhere in the game, I can create an evasion-snowball effect which probably results into the opponent Struggling which is my main objective.

Mud-Slap speaks for itself, and Rock Tomb is for Ground-types I want to outspeed with Skarmory like Flygon. Another use is for DD Resters or even Anabel's Raikou, for example. Once they Rest, Skarmory still outspeeds (and Scizor as well with two Rock Tombs to be sure). As a bonus, it can bring bulky Fire-types, Crobat or Skarmory in guaranteed +6 Swarm HP[Bug] KO range. Another option over Rock Tomb would be Counter, as that would deal with dangerous Struggle users like CurseSkarm. Usually I can beat that with Scizor if I'm setup properly, and even if I can't, it usually wastes its last Rest PP before its last Curse or Toxic PP against this team. The real reason to run Counter is probably against Clear Body setuppers: Regirock and Registeel.

The EV's are where it gets really interesting IMO. Since this is an "anti-Fire" lead, I figured I'd try Sitrus Berry as we learned from the Christmas team. Registeel can NEVER be sure to live two turns against 177 SpA Lum Berry Fire-types that use Flamethrower. Even if you would use Rock Tomb, that has 80% chance to miss so it would still be important to get that second chance. So what HP/SpD combination maximizes the probability of being 3HKOed by their Flamethrowers? I wrote a simulator for this exact problem:

leadprob.py

import numpy as np
import sys, os
from tqdm import tqdm

MonteCarlo = 1000000

Level = 50

DefenderMaxHP = 176
DefenderDefense = 218
AttackerOffense = 177

Effect = 2
STAB = 1.5

MovePower = 95
CritRate = 0.0625
MoveAccuracy = 1

LeftoversToggle = False
SitrusBerryToggle = True

#############################
LeftoversHP = int(DefenderMaxHP/16)


# DISABLE PRINT
def blockPrint():
sys.stdout = open(os.devnull, 'w')

# RESTORE PRINT
def enablePrint():
sys.stdout = sys.__stdout__

def damageRange():
DRange = []
for x in range(85,101):
D = np.floor(np.floor((np.floor(np.floor(np.floor(2*Level/5+2)*AttackerOffense*MovePower/DefenderDefense)/50)+2)*STAB)*Effect*x/100)
DRange.append(D)
return DRange


def executeMove(accuracy, level, attack, defense, power, critrate, effect, stab):
global DefenderHP
print("Attacker used Move!")
damage = 0
hit = np.random.rand() < accuracy
crit = np.random.rand() < critrate
if hit and crit:
damage = np.floor(np.floor((np.floor(np.floor(np.floor(2*level/5+2)*attack*power/defense)/50)+2)*stab)*effect*2*np.random.randint(85,101)/100)
print("It's a critical hit and it dealt", min(damage, DefenderHP), "damage!")
elif hit and crit == False:
damage = np.floor(np.floor((np.floor(np.floor(np.floor(2*level/5+2)*attack*power/defense)/50)+2)*stab)*effect*np.random.randint(85,101)/100)
print("It dealt", min(damage, DefenderHP), "damage!")
else:
print("The attack missed!")
DefenderHP = max(DefenderHP - damage, 0)


print(damageRange())

blockPrint()

NHKOVector = []

for i in tqdm(range(MonteCarlo)):

DefenderHP = DefenderMaxHP
SitrusToggle = SitrusBerryToggle

turn = 1

while DefenderHP > 0:
print("=== TURN", turn, "===")
executeMove(MoveAccuracy, Level, AttackerOffense, DefenderDefense, MovePower, CritRate, Effect, STAB)
if DefenderHP == 0:
print("Defender has fainted.")
break
if LeftoversToggle:
print("Defender restored", LeftoversHP, "HP with its Leftovers!")
DefenderHP = DefenderHP + LeftoversHP
if SitrusToggle and DefenderHP <= DefenderMaxHP/2:
print("Defender restored 30 HP with its Sitrus Berry!")
DefenderHP = DefenderHP + 30
SitrusToggle = False
print("Defender has", DefenderHP, "HP left.")
turn = turn + 1

NHKOVector.append(turn)

print(NHKOVector)

enablePrint()

for n in range(1,max(NHKOVector)+1):
print(n,"HKO:",NHKOVector.count(n)/MonteCarlo)

Then I tried all relevant SpD stats (from 170 to 222, skipping unimportant SpD values with respect to rounding, i.e. 218 is the "same" as 222 as far as damage is concerned) combined with all relevant possible HP stats from 155 to 187. You see, it gets quite subtle because if you have too much bulk, Sitrus Berry will be more likely not to activate and then you can get 2HKOed quite easily. Out of the entire table, running 176 HP and max SpD (i.e. 218) is best, as it minimizes the length of the damage range and ensures activation of the Sitrus Berry if I don't get crit. Factoring in crits, Registeel has 86.9% chance to live through two Flamethrowers. This doesn't take into account the chance of getting burned yet, but I think this would still be optimal. Will update when I know more.

First I was worried running only 176 HP instead of 187 would be scary on the physical side, but Registeel still easily survives non-crit Machamp's Cross Chop and Rhydon's EQ if necessary.

Skarmory (M) @ Leftovers
Ability: Sturdy
EVs: 164 HP / 4 Atk / 76 Def / 12 SpD / 252 Spe
Jolly Nature (+Spe, -Atk)
- Protect
- Substitute
- Sand-Attack
- Toxic

As showed by the program a few posts ago, this Skarmory has over 85% chance to reduce any non-Clear Body, non Keen Eye Pokemon's accuracy to -6 if they are paralyzed (even if it would be a Level 100 Charizard, lol). Toxic is a "filler" but sometimes a very important move if Scizor gets haxed or if I face a setup sweeper that's too dangerous, like Marowak, Salamence or Regirock.

EV-wise, Skarmory has 161 HP: the highest perfect Leftovers number it can get. It outspeeds Lati@s and a bunch of other stuff it shouldn't even be outspeeding (Moltres-1 lol), and doesn't need that much bulk as it expects its Substitutes to break anyway 76 Defense hits a jump point in defensive bulk. 12 SpD is filler, and 4 Atk EV's is not to waste any, in the not-so-very-rare case I have to Struggle (which is also why I'm running Jolly instead of Timid). Skarmory is a really good Struggler with Leftovers. Even with little investment, Skarmory is still really bulky considering all of its resistances. Also, it still survives non-STAB Thunderbolt from the likes of Lati@s if QC activates.

Skarmory is not only for the setup, it can actually be a game saver with Toxic, Protect and Substitute. I've learned a lot from the Christmas challenge.

Scizor (F) @ Salac Berry
Ability: Swarm
EVs: 4 HP / 252 Atk / 4 Def / 248 Spe
Adamant Nature (+Atk, -SpA)
IVs: 30 SpD / 30 Spe
- Swords Dance
- Substitute
- Morning Sun
- Hidden Power [Bug]

The new sweeper! It's very subtle to use, but surprisingly effective. We want to have a sweeper that's fast enough (unlike Curse Registeel, who is always screwed by crit+kill, next Pokemon crits). Agility Metagross can't boost Attack without killing the crippled lead. Scizor can easily SD up on Struggling opponents and has plenty of SD's to waste while they miss Struggles until it hits the sweet Salac Berry spot at 2HP. If its Sub then breaks somehow, simply use Morning Sun and make sure you get into a range again between 1/4 and 1/3 of your HP (i.e. anything from 37 to 48 is fine). Then you have the best position to sweep, because you're in Swarm range (which lets you KO stuff like even Salamence and Gyarados guaranteed), and can still make one more Substitute against threats like Charizard and Moltres, should you have to face them as a third Pokemon when your Substitute got down by their second. If they use Overheat on Scizor twice, then they won't beat Skarmory if you left it at full HP.

This Scizor hasn't got its HP divisibly by four, since that allows you to make one more Sub in a pinch, and you can use Morning Sun to get your HP where you want anyway.

Morning Sun adds flexibility to this team, but having only one Attack forces you to play very patiently and carefully. Another thing you can do if you're too impatient while they Struggle against Scizor is to switch out and use Substitute again to lower your HP. Be careful though: you still have to use Swords Dance three times.

Swarm + 6 HP Bug is crazy though, when you manage to pull it off. And this team often lets you pull it off.

Here's a sample streak with a lot of Fighting-types and Fire-types, in which I didn't play patiently enough to get into perfect range every time but you'll get the idea:

 

[Kommo-o]

So they call these spoilers, right?

Gamecube Pog

 

[Purple Kecleon]

Is this thread still active? It looks like Golden Blissey hasn't been on here in a while. I just recently got the urge to play the Battle Frontier again and reached 280 floors in the Battle Pyramid. I made it to 140 using a few different teams and movesets, but I thought it would be cool to clear 20 rounds using the same team and movesets all the way through.


Alakazam @ Leftovers
Ability: Synchronize
EVs: 6 HP / 252 SpA / 252 Spe
Nature: Modest
IVs: 21/1/15/24/11/29
Psychic
Thunder Punch
Ice Punch
Teleport

Honestly, after clearing another 20 rounds, I think that Alakazam is maybe not as good as I once thought it was for the Pyramid. Teleport is indeed very nice against a handful of wild Pokemon that can't be OHKOed easily, with Forretress being the most annoying of them. And the ability to escape any wild battle makes it easy to conserve resources. However, Alakazam's frailty means that it's not particularly good in trainer battles, and those are what really threaten your streak. So I'm not sure if the benefit of Teleport is actually worth it over something like Latios which fares much better in trainer matchups and has Levitate for double battles which would be good alongside Slaking's Earthquake.


Slaking @ Choice Band
Ability: Truant
EVs: 6 HP/ 252 Atk / 252 Spe
Nature: Adamant
IVs: 28/24/30/8/17/28
Return
Earthquake
Shadow Ball
Hyper Beam

Best 1v1 battler in the game pretty much and can OHKO most things, although Truant can be really nasty during double battles or versus a bulky setup sweeper. A good amount of wild Pokemon do carry Protect, but it wasn't that big of a deal since they aren't even guaranteed to use it. Hyper Beam's 90% accuracy can be unreliable, but the extra power is needed against things like Snorlax. I found myself leading with Slaking on 80-90% of the floors, and even sometimes when Alakazam fared a bit better against the wild Pokemon, just to decrease my chances of losing to a trainer.


Blissey @ Leftovers
Ability: Natural Cure
EVs: 252 HP / 252 Def / 6 Spe
Nature: Calm
IVs: 19/23/31/15/30/30
Aromatherapy
Softboiled
Seismic Toss
Flamethrower

I actually found Flamethrower to be quite handy in dealing with some of the wild Pokemon. Pretty nice that she can solo Brandon's entire team.


Threats include double battles and bulkier setup sweepers (especially ChestoResters) such as DD/Curse Tyranitar, Curselax, DD Dragonite, and DDMence (Intimidate is also nasty in double battles). Some of these pokemon carry things like Focus Band and Brightpowder which can get really scary in combination with Truant and Hyper Beam's 90% accuracy. I had a few close calls with the aforementioned 4 pokemon. Endure+Salac+Reversal Heracross would sweep this entire team, but luckily I never encountered it.

Thought I'd also include pictures of my bag at the end of all this. Pyramid is definitely really cool and one of my favorite facilities.

 

[Actaeon]

Updated my large Monoteam post with some big changes and minor changes:

• The Mono Normal Trick team with Linoone/Porygon2/Vigoroth has been replaced by a standard CGS Stantler/Blissey/Snorlax team because it's definitely a bit more sturdy in the long run;
• The Mono Steel team with Aggron/Skarmory/Registeel has been replaced by the Steel Shore, a better version that attempts a sweep with +6 Swarm Scizor I posted about earlier.
• On Mono Bug, Seismic Toss on Volbeat was replaced by Thunder Wave and Silver Wind on Ninjask was replaced by Double Team. This gives a little more win chance against Steel Wing and Aerial Ace users, such as Flygon and Skarmory. Not being able to finish off Claydol by dropping Silver Wind is balanced by Armaldo having evasion in many scenarios.
• Scary Face was replaced by Torment and Roar was replaced by Protect on the Mono Dark Mightyena. Yawn is still useful against physical setuppers, and Sand-Attack is a no-brainer.

I'm still trying to improve Mono Flying, Mono Ground, Mono Fighting and Mono Poison because these teams are the main reason I keep losing the ultimate challenge. Feel free to think along about these teams! If you're interested, all current teams can be found in the link above and much of my thought processes about the older, original teams can be found on pages 11-12-13 of this thread.

 

[Actaeon]

While attempting some non-live 17*7 streaks with the monotype teams, I decided to try and improve the mono Ground team a little bit. The conclusion was that even though Flygon + Quagsire provided decent setup most of the time, some threats like Starmie, Latios, Clear Body pokes that win the 1v1 against Flygon, Flygon having to choose an offensive route and then losing the next matchup, among many other things, proved the old mono Ground just wasn't meeting my monotype standards.

Through my recent experience with using Torment on other "bad" teams (Flying, Poison and Ghost) that lack a dependable crippler, I decided to give Torment Steelix a chance in the lead spot, make Quagsire less speedy but more bulky, and increase the Attack of Sandslash so it can 2HKO Flygon and OHKO many more things at +4 if I can't fully setup, while still having enough physical bulk so its Substitute can take unboosted Struggles.

Steelix is not quite the defensive behemoth it ought to be, as it easily falls to (critical hit) Water-type or Fire-type moves, but those are weaknesses Quagsire manages perfectly (unless it gets burnt). Most importantly, Torment lets Quagsire cripple their lead way more often, and with the defensive improvements it's very hard to beat if the opponent can only use their "second best move" while getting crippled and stalled by Yawn, Flash, Protect, Substitute and Leftovers. The main reason it had so much Speed to begin with was OHKO haxers, but this is somewhat mitigated by Sturdy Steelix now, and some smart manoevering with Torment and Protect on both Pokemon. Here's the new team, and I really think it's much better as I've just beaten 5 streaks in a row with it:

Steelix @ Sitrus Berry
Ability: Sturdy
EVs: 252 HP / 4 Atk / 252 SpD
Careful Nature
- Torment
- Protect
- Mud-Slap
- Earthquake / Explosion / Rock Tomb / Rest / Roar

4 Attack EVs greatly improves the odds of 2HKOing Metagross, and with Sitrus Berry it probably won't 3HKO me. I don't care about not outspeeding the other 50 Speed Pokemon anyway, so that explains the spare EVs. Otherwise, full Special defense gives the best odds against Fire Punches from Alakazam, for example, or Flamethrowers in general. On the physically defensive side, it can't really get OHKOed except by Focus Punch, which I can handle perfectly.

The other options all have their uses, but I haven't really been in situations yet where I would desperately need the other fillers.

Quagsire @ Leftovers
Ability: Water Absorb
EVs: 252 HP / 212 SpD / 44 Spe
Bold Nature
IVs: 0 Atk
- Yawn
- Protect
- Substitute
- Flash

Now with improved bulk: on the physical side it can't be 2HKOed by Rhydon's Earthquake, which it still outspeed with 44 Spe EVs, after two Leftovers turns, and the same holds for the most powerful Psychics. Generally, it has plenty of time to replenish HP with its first three moves, while slowly attempting to use Flash in between. Torment really improves this Pokemon's effectiveness.

Sandslash @ Salac Berry
Ability: Sand Veil
EVs: 20 HP / 252 Atk / 16 Def / 216 Spe
Jolly Nature
- Swords Dance
- Substitute
- Earthquake
- Hidden Power [Rock]

Now made way stronger on the attacking side, since it doesn't really have to survive (Levitate-obtained-through-Trace) Gardevoir Ice Punches anymore.

 

[JJ456]

Hey All, long-time reader, first-time poster. Just wanted to share my appreciation for all the helpful info for tackling the infamous battle tower. I really enjoyed reading and hypothesising different teams, which at times, seemed to be more fun than the actual battling. I also wanted to share the team I used if others were interested or wanted to try out. The team is a mix of recommended pokes/sets as well as my own input. I don’t think anyone has used/posted this Swampert on a high streak before, but I could be wrong. Without being special compared to top streaks, I thought this team could be helpful for the avid emerald player who can't seem to get the gold symbol but also doesn't use/want to use/know how to use techniques like RNG, breeding, trading or cloning. My pokemon are not shiny or don’t have perfect IV's (old man says they're "quite impressive"), they are all caught in Emerald, and the movesets can be taught from available TM's/tutors, so you don't need to double up or breed a specific hidden power. NB: I restarted my Emerald game multiple times because I already used the TM earthquake, toxic, rest etc. This 'no-frills' team can obviously be improved by previously mentioned techniques (as well as having it’s own flaws), but it was enough for me to FINALLY get the gold in battle tower (Level 50 - emulator) and reach a streak of 128 wins.

The Team:

NB: I posted each Pokémon's stats because I found that to be a better reference (esp without optimal IVs) and I could only be 90% sure where the EVs exactly went

Swampert (M) @ Leftovers
IVs: Not sure, but old man says "quite impressive"
EVs: 252 HP / 208 Def / 50 Sp D
Stats: 186 HP / Att 133 / Def 144 / 91 Sp A / 115 Sp D / 80 Spe
Nature: Impish
- Protect
- Growl
- Mud Slap
- Roar

Latios (M) @ Lum Berry
IVs: "quite impressive"
EVs: 210 HP / 100 Sp A / 200 Spe
Stats: 176 HP / 91 Att / 114 Def / 150 Sp A / 132 Sp D / 160 Spe
Nature: Timid
- Dragon Claw
- Calm Mind
- Substitute
- Recover

Registeel @ Chesto Berry
IVs: "quite impressive"
EVs: 252 HP / 100 Att / 158 Def
Stats: 181 HP / 116 Att / 196 Def / 81 Sp A / 159 Sp D / 56 Spe
Nature: Impish
- Curse
- Amnesia
- Metal Claw (didn’t have HP steel)
- Rest (substitute already used)

Strategy:

The idea was to use defensively compatible pokemon who could offensively boost up to sweep/semi sweep. Latios is weak to ice, bug, and dragon (which Regi is resistant to) and also weak to ghost and dark (which Regi takes neutrally). Registeel is weak to fire and fighting (which Latios is resistant to) and also weak to ground (which Latios is immune to). Swampert also has great typing defensively being immune to electric (which Regi takes neutrally) and only being weak to grass (which both Latios and Regi resist).

Swampert is my lead and plays the crucial role of finding a suitable target for either Latios or Regi to setup on. He is my scout, crippler, and phazer, so most of the decision-making process is done when he’s on the field. If not used correctly or he gets knocked out prematurely, the match is challenging/dangerous at best, or lost at worst. When the opposing pokemon is revealed, the first decision is whether it needs to be phazed or not. Phazed pokemon include anyone who is a known threat (e.g. Metagross), has a OHKO move, has boosting moves/double team, or cannot be appropriately crippled (e.g. growl vs. special attacker/clear body or mud slap vs. flying/levitate). Second decision is to scout for information with protect. If there is a revealed sign of threat, I reconsider phazing. If I’ve decided to keep the poke for crippling/setup, I adjust the following moves depending if they are a physical or special attacker. If they are a physical attacker (usually ideal matchup), I use growl and alternate with protect for max recovery. After appropriately reducing their attack stat, I cripple further with mud slap to reduce accuracy if possible. If they are a special attacker, I swap the order of mud slap and growl, which is still relevant for unknown physical moves or PP stall struggle. NB: mud slap is also STAB ground, so it has decent incremental damage that could force opposing poke to recover/rest. Once the selected target is crippled, I switch in the appropriate sweeper/semi sweeper, which then makes the rest of the decision-making process relatively straight forward.

If Latios was the designated sweeper, I open with substitute to avoid CH or status ailments. Then boost up with calm mind and use recover/substitute liberally to stay healthy. Once Latios is +6 boosted, dragon claw rips into all/most pokemon. If they are not fully swept, substitute will take the hit anyway.

If Registeel was the designated sweeper, open with amnesia if they are a special attacker, or open with curse if they are a physical attacker. Use rest liberally to stay alive. I try to expect a CH, so using rest early within CH range has been helpful to mitigate low percentage hax situations. Once +6 boosted attack, defence, and special defence, Registeel becomes a force even against STAB earthquake or flamethrower. Again, I just need to be careful of CH or other low percentage annoyances, especially since Regi doesn’t have substitute, which was already taught to Latios. Finally, proceed to use metal claw, which should sweep/semi sweep most pokemon and win the match. NB: I understand HP steel is better than metal claw, but I didn’t have the correct IVs.

Threats:

• Pokemon that can’t be crippled (e.g. Metagross, clear body)
• OHKO moves (e.g. sheer cold, fissure, horn drill)
• Stat boosting pokemon/double teamers (e.g. Salamence, Marowak, Metagross)
• Special attackers/flyers can be ‘difficult’ to cripple (e.g. Articuno)
• Explosion users (more for CH and resets setup)
• Any string of hax or team of multiple threats

The Loss:

Not an exciting loss, but in retrospect, probably could have been mitigated. First, Golem exploded on Swampert with CH after 1 growl and 1 mud slap. Swampert surprisingly fainted, which made it dangerous for my next pokemon. Nidoking vs. Latios hit consecutive horn drills and blizzards, which forced me to constantly substitute and recover. Without missing and freezing me twice (first freeze was healed with lum berry), Latios eventually died. Registeel opened with curse anticipating STAB earthquake, but Nidoking hit with fireblast. It was all too late, and Nidoking hit with another perfect accuracy fireblast, which was enough to end my streak.

Final Remarks:

Overall, I appreciated using this team format after trialling numerous teams previously. This strategy had an interesting gameplan that was different to your usual non-setup sweepers, toxic stalling, explosion 2-for-1 trades. I liked the idea of basically finding one opposing pokemon to successfully setup on (which most of the time wins the match) rather than ‘beating’ 3 different pokemon back-to-back. Anyway, thanks again to all the people posting and having open forum discussions. I think this platform has been crucial for sparking new/great ideas across the board and it encouraged my problem-solving side to play pokemon again after all those years. I hope someone finds this team interesting, especially with this use of Swampert and not needing anything extra like RNG, breeding, trading etc. I don’t believe anyone has used/posted this Swampert on a high streak before, but I could be wrong. Let me know your thoughts.

 

[Actaeon]

Hey All, long-time reader, first-time poster. Just wanted to share my appreciation for all the helpful info for tackling the infamous battle tower. I really enjoyed reading and hypothesising different teams, which at times, seemed to be more fun than the actual battling. I also wanted to share the team I used if others were interested or wanted to try out. The team is a mix of recommended pokes/sets as well as my own input. I don’t think anyone has used/posted this Swampert on a high streak before, but I could be wrong. Without being special compared to top streaks, I thought this team could be helpful for the avid emerald player who can't seem to get the gold symbol but also doesn't use/want to use/know how to use techniques like RNG, breeding, trading or cloning. My pokemon are not shiny or don’t have perfect IV's (old man says they're "quite impressive"), they are all caught in Emerald, and the movesets can be taught from available TM's/tutors, so you don't need to double up or breed a specific hidden power. NB: I restarted my Emerald game multiple times because I already used the TM earthquake, toxic, rest etc. This 'no-frills' team can obviously be improved by previously mentioned techniques (as well as having it’s own flaws), but it was enough for me to FINALLY get the gold in battle tower (Level 50 - emulator) and reach a streak of 128 wins.

The Team:

NB: I posted each Pokémon's stats because I found that to be a better reference (esp without optimal IVs) and I could only be 90% sure where the EVs exactly went

Swampert (M) @ Leftovers
IVs: Not sure, but old man says "quite impressive"
EVs: 252 HP / 208 Def / 50 Sp D
Stats: 186 HP / Att 133 / Def 144 / 91 Sp A / 115 Sp D / 80 Spe
Nature: Impish
- Protect
- Growl
- Mud Slap
- Roar

Latios (M) @ Lum Berry
IVs: "quite impressive"
EVs: 210 HP / 100 Sp A / 200 Spe
Stats: 176 HP / 91 Att / 114 Def / 150 Sp A / 132 Sp D / 160 Spe
Nature: Timid
- Dragon Claw
- Calm Mind
- Substitute
- Recover

Registeel @ Chesto Berry
IVs: "quite impressive"
EVs: 252 HP / 100 Att / 158 Def
Stats: 181 HP / 116 Att / 196 Def / 81 Sp A / 159 Sp D / 56 Spe
Nature: Impish
- Curse
- Amnesia
- Metal Claw (didn’t have HP steel)
- Rest (substitute already used)

Strategy:

The idea was to use defensively compatible pokemon who could offensively boost up to sweep/semi sweep. Latios is weak to ice, bug, and dragon (which Regi is resistant to) and also weak to ghost and dark (which Regi takes neutrally). Registeel is weak to fire and fighting (which Latios is resistant to) and also weak to ground (which Latios is immune to). Swampert also has great typing defensively being immune to electric (which Regi takes neutrally) and only being weak to grass (which both Latios and Regi resist).

Swampert is my lead and plays the crucial role of finding a suitable target for either Latios or Regi to setup on. He is my scout, crippler, and phazer, so most of the decision-making process is done when he’s on the field. If not used correctly or he gets knocked out prematurely, the match is challenging/dangerous at best, or lost at worst. When the opposing pokemon is revealed, the first decision is whether it needs to be phazed or not. Phazed pokemon include anyone who is a known threat (e.g. Metagross), has a OHKO move, has boosting moves/double team, or cannot be appropriately crippled (e.g. growl vs. special attacker/clear body or mud slap vs. flying/levitate). Second decision is to scout for information with protect. If there is a revealed sign of threat, I reconsider phazing. If I’ve decided to keep the poke for crippling/setup, I adjust the following moves depending if they are a physical or special attacker. If they are a physical attacker (usually ideal matchup), I use growl and alternate with protect for max recovery. After appropriately reducing their attack stat, I cripple further with mud slap to reduce accuracy if possible. If they are a special attacker, I swap the order of mud slap and growl, which is still relevant for unknown physical moves or PP stall struggle. NB: mud slap is also STAB ground, so it has decent incremental damage that could force opposing poke to recover/rest. Once the selected target is crippled, I switch in the appropriate sweeper/semi sweeper, which then makes the rest of the decision-making process relatively straight forward.

If Latios was the designated sweeper, I open with substitute to avoid CH or status ailments. Then boost up with calm mind and use recover/substitute liberally to stay healthy. Once Latios is +6 boosted, dragon claw rips into all/most pokemon. If they are not fully swept, substitute will take the hit anyway.

If Registeel was the designated sweeper, open with amnesia if they are a special attacker, or open with curse if they are a physical attacker. Use rest liberally to stay alive. I try to expect a CH, so using rest early within CH range has been helpful to mitigate low percentage hax situations. Once +6 boosted attack, defence, and special defence, Registeel becomes a force even against STAB earthquake or flamethrower. Again, I just need to be careful of CH or other low percentage annoyances, especially since Regi doesn’t have substitute, which was already taught to Latios. Finally, proceed to use metal claw, which should sweep/semi sweep most pokemon and win the match. NB: I understand HP steel is better than metal claw, but I didn’t have the correct IVs.

Threats:

• Pokemon that can’t be crippled (e.g. Metagross, clear body)
• OHKO moves (e.g. sheer cold, fissure, horn drill)
• Stat boosting pokemon/double teamers (e.g. Salamence, Marowak, Metagross)
• Special attackers/flyers can be ‘difficult’ to cripple (e.g. Articuno)
• Explosion users (more for CH and resets setup)
• Any string of hax or team of multiple threats

The Loss:

Not an exciting loss, but in retrospect, probably could have been mitigated. First, Golem exploded on Swampert with CH after 1 growl and 1 mud slap. Swampert surprisingly fainted, which made it dangerous for my next pokemon. Nidoking vs. Latios hit consecutive horn drills and blizzards, which forced me to constantly substitute and recover. Without missing and freezing me twice (first freeze was healed with lum berry), Latios eventually died. Registeel opened with curse anticipating STAB earthquake, but Nidoking hit with fireblast. It was all too late, and Nidoking hit with another perfect accuracy fireblast, which was enough to end my streak.

Final Remarks:

Overall, I appreciated using this team format after trialling numerous teams previously. This strategy had an interesting gameplan that was different to your usual non-setup sweepers, toxic stalling, explosion 2-for-1 trades. I liked the idea of basically finding one opposing pokemon to successfully setup on (which most of the time wins the match) rather than ‘beating’ 3 different pokemon back-to-back. Anyway, thanks again to all the people posting and having open forum discussions. I think this platform has been crucial for sparking new/great ideas across the board and it encouraged my problem-solving side to play pokemon again after all those years. I hope someone finds this team interesting, especially with this use of Swampert and not needing anything extra like RNG, breeding, trading etc. I don’t believe anyone has used/posted this Swampert on a high streak before, but I could be wrong. Let me know your thoughts.

Welcome!

Nice team, the strategy is an in-between cripple / 2-poke setup that's quite easy to obtain indeed. I considered using Swampert as a lead crippler on both Mono Ground and Mono Water when I was working on those, but on both teams I missed Thief and Torment (not necessarily both). Nevertheless, Swampert is a tanky Pokemon that doesn't get OHKOed very often, even by most critical hits. The thing is it doesn't really do anything to cripple the likes of Special Flying-types/Levitators, so I would really recommend a Speed-lowering move (probably Icy Wind) so Registeel can outspeed them while using Amnesia. I'm thinking of Moltres, for example, but also stuff like Special Gengar, (Fire Punch) Gardevoir and DD Gyarados (although I guess you Roar that out anyway). I can see a lot of Pokemon ruining the strategy immediately with a well-timed critical hit for which they'll get quite some opportunities (Salamence, Ursaring, etc.)

Using Roar on a bulky Pokemon like Swampert is OK I think, because usually it has the bulk to do something against the next unknown Pokemon.

If you can't get HP Steel on Registeel, another take is to use Registeel as the crippler with Thunder Wave, Mud-Slap, Rock Tomb and Protect while Swampert takes those annoying Fire-types (using Curse/EQ/HP[Flying]/Rest is probably ideal) and Latios handles the ones with really strong EQ's or Fighting-types. Heracross is probably really bad news (another reason to choose HP[Flying] over HP[Rock], but I guess both are difficult to obtain in-game and at least you can replace HP[Rock] with Rock Slide).

Yet another take: something like Crippler Registeel + SubTect Latios + Curse Swampert might work against a lot of leads, with fillers like Flash/Recover and Memento/Recover on Latios. This team has the unique strategy that the final sweeper is capable of setting up against all Clear-Body Pokemon (especially when they are paralyzed and PP-stalled), so you don't really have to worry about that. If you're not too worried about sweeping without a Substitute up (which you should be), you can run Leftovers on Latios and Chesto Berry on Swampert, although for this strategy it might be wise to give Swampert Sub/Curse/EQ/Filler @ Leftovers over Rest and slap Recover @ Brightpowder on Latios. As you can read a couple of posts up, a paralyzed opponent has a really good chance of getting crippled this way even by a Recover-less Pokemon like Skarmory.

Just a few things that came to my mind, let me know if you'd like more details!

 

[JoebertIII]

Hello All,

I'm here reporting my new personal best, 63 wins in the Open Level Battle Factory Singles! I totally set out to go further, but one lucky Crobat4 had other plans.

63 Factory Winstreak

Round 5: Raikou5, Milotic2, Salamence2 (dud set) ➡ Moltres2 ➡ Regirock6

Round 6: Starmie8, Registeel6 ➡ Steelix3, Articuno6 ➡ Zapdos4

Round 7: Latias7 (terrible moves, good EV's+item), Claydol4, Blastoise1 ➡ Ludicolo3 ➡ Gyarados3

Round 8: Moltres2, Metagross3, Dugtrio3 ➡ Lapras2 ➡ Starmie8 ➡ Suicune4
So this round was home to the most nail-biting battle I've had in awhile. Dugtrio3 wasn't pulling his weight before I heard the next opponent has Water types. So he's out and I take an unknown Lapras. Starmie I originally had pegged as the "Ice Beam/Psychic" variant so that was my rational switching Metagross in.

Round 9: Starmie7, Steelix1, Whiscash2 ➡ Regirock6 ➡ Zapdos5

Round 10: Dragonite10, Blissey4, Marowak2
It was really tough losing match one with such great starting picks. I heard the opponent had Flying types and 31 IV, 252 speed Marowak outspeeds the 3 IV, 0 speed legendary birds. But in the end, it was confuse ray that did me in. And my apprehension to blizzard into the brightpowder Crobat4. I thought fire blast will get him low enough for a Blissey ice beam, or it'll burn and I can softboil...

 

[Actaeon]

Hello All,

I'm here reporting my new personal best, 63 wins in the Open Level Battle Factory Singles! I totally set out to go further, but one lucky Crobat4 had other plans.
View attachment 309135

63 Factory Winstreak

Round 5: Raikou5, Milotic2, Salamence2 (dud set) ➡ Moltres2 ➡ Regirock6

Round 6: Starmie8, Registeel6 ➡ Steelix3, Articuno6 ➡ Zapdos4

Round 7: Latias7 (terrible moves, good EV's+item), Claydol4, Blastoise1 ➡ Ludicolo3 ➡ Gyarados3

Round 8: Moltres2, Metagross3, Dugtrio3 ➡ Lapras2 ➡ Starmie8 ➡ Suicune4
So this round was home to the most nail-biting battle I've had in awhile. Dugtrio3 wasn't pulling his weight before I heard the next opponent has Water types. So he's out and I take an unknown Lapras. Starmie I originally had pegged as the "Ice Beam/Psychic" variant so that was my rational switching Metagross in.

Round 9: Starmie7, Steelix1, Whiscash2 ➡ Regirock6 ➡ Zapdos5

Round 10: Dragonite10, Blissey4, Marowak2
It was really tough losing match one with such great starting picks. I heard the opponent had Flying types and 31 IV, 252 speed Marowak outspeeds the 3 IV, 0 speed legendary birds. But in the end, it was confuse ray that did me in. And my apprehension to blizzard into the brightpowder Crobat4. I thought fire blast will get him low enough for a Blissey ice beam, or it'll burn and I can softboil...

Aww, that's sore loss. 63 is really good though, congratulations. Maybe you could have Icy Winded over Rock Slide on turn 1 with Marowak against Mence to outspeed it and preserve some HP? Those Sludge Bombs did a lot of damage on Blissey by the way, it surprised me considering the IV advantage.

 

[Purple Kecleon]

Does anyone know if Dome Ace Tucker's Metagross is guaranteed to use Protect on the first turn if your pokemon is at high HP? From watching some YouTube videos, I've only seen it not open with Protect if your pokemon is at low HP. And if the first Protect fails (let's say you switch on that turn), is it guaranteed to go for Protect again?

I know this is kind of an obscure question, but I was wondering if anyone who has battled Tucker a bunch of times would know the answer to this.

 

[QuentinQuonce]

Does anyone know if Dome Ace Tucker's Metagross is guaranteed to use Protect on the first turn if your pokemon is at high HP? From watching some YouTube videos, I've only seen it not open with Protect if your pokemon is at low HP. And if the first Protect fails (let's say you switch on that turn), is it guaranteed to go for Protect again?

I know this is kind of an obscure question, but I was wondering if anyone who has battled Tucker a bunch of times would know the answer to this.

Take this with a grain of salt as I'm not an expert, but from what I've read/researched that would seem to be the case - going by how the AI prioritises moves, they generally opt for Protect on the first turn if there are no other moves which take priority over it.

So they will use Protect if none of the following conditions are met:
-if they have a status move and your Pokemon does not have a status condition
-if they have a healing move and are at low HP/have a status condition that can be cured
-if they have a move that can KO your current Pokemon
-if they have a boosting move like Swords Dance or Iron Defence

Tucker's Metagross has no boosting or status moves, so neither applies here. So, unless it's a cert that your Pokemon will be KO'ed by Meteor Mash, Psychic, or Earthquake (which obviously becomes more likely if they're not at full HP), I'd say it's fairly likely that it will generally go for Protect first.

The only part I'm unsure about is whether Protect failing has an impact on repeated usage. I've seen AI trainers use it twice (and have it succeed both times) and thrice (the first two successful and the third not). The AI recognises that the diminishing success rate makes it a poor choice to keep using (otherwise opponents would simply use Protect over and over) but I'm not sure whether it's predisposed to start over once the "chain" breaks.

 

[N1c69]

I've made a challenge for myself and anyone else who is interested available bellow. I will be updating regular over the next few days.

https://www.smogon.com/forums/threads/gen-3-3-game-challenge.3676755/

The challenge is to see how long a streak/how many battle points I can get with 3 tries in each facility, without RNGing or soft resseting . Unlike Werster's all gold Emerald Battle Tower Speed Run, there is no time limit or desired streak. Score is judged by the total battle points accumulated at the end of three tries in each facility or best streak.
Completing the Ruby Battle Tower Glitch is part of the challenge. Though this is a glitch, the nature of the stolen pokemon is random and you have no resets. This is the same for each starter, beldum, eevee, latios etc. The challenge can be done without the glitch by spending more time catching and breeding pokemon until can breed the desired ivs. Introducing the glitch saves time by letting you choose one pokemon of any egg group to speed up breeding your battle frontier team.

This challenge will encourage using less viable pokemon as legendaries and pokemon like tauros and chansey are rare. It will also encourage making the best of what the game's RNG gives you, creating a unique run each time. By the end of the challenge I will have raised one box full (30 pokemon) and change teams depending on the facility.
I have completed silver for all challenges twice and gold in palace and arena. I hope to get gold for each using all the strategies I've read on this thread.

 

[Kommo-o]

January 30, 2021. Make sure you save this date

 

[JJ456]

Welcome!

Nice team, the strategy is an in-between cripple / 2-poke setup that's quite easy to obtain indeed. I considered using Swampert as a lead crippler on both Mono Ground and Mono Water when I was working on those, but on both teams I missed Thief and Torment (not necessarily both). Nevertheless, Swampert is a tanky Pokemon that doesn't get OHKOed very often, even by most critical hits. The thing is it doesn't really do anything to cripple the likes of Special Flying-types/Levitators, so I would really recommend a Speed-lowering move (probably Icy Wind) so Registeel can outspeed them while using Amnesia. I'm thinking of Moltres, for example, but also stuff like Special Gengar, (Fire Punch) Gardevoir and DD Gyarados (although I guess you Roar that out anyway). I can see a lot of Pokemon ruining the strategy immediately with a well-timed critical hit for which they'll get quite some opportunities (Salamence, Ursaring, etc.)

Using Roar on a bulky Pokemon like Swampert is OK I think, because usually it has the bulk to do something against the next unknown Pokemon.

If you can't get HP Steel on Registeel, another take is to use Registeel as the crippler with Thunder Wave, Mud-Slap, Rock Tomb and Protect while Swampert takes those annoying Fire-types (using Curse/EQ/HP[Flying]/Rest is probably ideal) and Latios handles the ones with really strong EQ's or Fighting-types. Heracross is probably really bad news (another reason to choose HP[Flying] over HP[Rock], but I guess both are difficult to obtain in-game and at least you can replace HP[Rock] with Rock Slide).

Yet another take: something like Crippler Registeel + SubTect Latios + Curse Swampert might work against a lot of leads, with fillers like Flash/Recover and Memento/Recover on Latios. This team has the unique strategy that the final sweeper is capable of setting up against all Clear-Body Pokemon (especially when they are paralyzed and PP-stalled), so you don't really have to worry about that. If you're not too worried about sweeping without a Substitute up (which you should be), you can run Leftovers on Latios and Chesto Berry on Swampert, although for this strategy it might be wise to give Swampert Sub/Curse/EQ/Filler @ Leftovers over Rest and slap Recover @ Brightpowder on Latios. As you can read a couple of posts up, a paralyzed opponent has a really good chance of getting crippled this way even by a Recover-less Pokemon like Skarmory.

Just a few things that came to my mind, let me know if you'd like more details!

Thanks for the response, Actaeon. Good summary/description of the team and interesting ideas for tweaking. My Swampert is a bit of a "jack of all trades, master of none" sort of poke, which was great/enough for the first 100 wins. My team's holes did start to show as it was continuously exposed to greater numbers and increased chance of potential threats. Nevertheless, I hadn't thought about using Regi as the crippler (which could be usable) or Curse Swampert (less usable as requires egg move). That team's synergy does look interesting with complementary sets that allows the last poke to sweep easily. I'll have a play around and see how I go. Cheers.

 

[Purple Kecleon]

Just ended with a streak of 22 tournaments at the Battle Dome. The Dome is probably my favorite or 2nd favorite facility along with the Pyramid. I like how short the rounds are and the concept of planning around your opponent's team.


Slaking @ Choice Band
Ability: Truant
EVs: 6 HP/ 252 Atk / 252 Spe
Nature: Adamant
IVs: 28/24/30/8/17/28
Double-Edge
Earthquake
Shadow Ball
Hyper Beam / Frustration

This was my lead in the majority of matchups. For the most part my strategy was to just KO the first pokemon if possible and then switch in Gengar/Meta and D-Bond/Explode. Jolly definitely has its advantages here due to the 3 IV glitch, but I found Adamant to be quite helpful in scoring guaranteed OHKOs w/Double-Edge against things like Snorlax and Lapras.

Hyper Beam gets the OHKO on bulky Salamence sets after Intimidate and I had it for the first 20 rounds, but I experimented with Frustration the last 2 rounds before I lost. (Used Frustration since I had to trade my Slaking to teach it Double-Edge and Energy Roots are much faster than raising its happiness back up to max.)


Gengar @ Lum Berry
Ability: Levitate
Nature: Modest
IVs: 21/21-22/13/31/19/23
Thunderbolt
Ice Punch
Fire Punch
Destiny Bond

I definitely see how Substitute can be useful in conjunction with Destiny Bond, but Gengar can still get the job done fine without it for the most part. Fire Punch actually came in quite handy on occasion, especially against Forretress which walls my other 2 physical attackers.


Metagross @ Leftovers
Ability: Clear Body
EVs: 252 HP, 236 Atk, 12 Def, 8 Spe
Nature: Impish
IVs: 30/31/4/5-6/25/16
Meteor Mash
Earthquake
Shadow Ball
Explosion

The EVs were copied from an old Smogon set many years ago when I had no idea what I was doing and I'm too lazy to breed and train another one lol. I think running max or near max speed would have been better for the purposes of Exploding.

Upon discovering that ties in the Battle Dome actually favor the player, I thought it would be a really good idea to form a team with 2 Destiny Bonders/Exploders along with Slaking for its BST and OHKO potential. However, in practice, it's not nearly as effective as I thought, since you can't really ever be guaranteed to force a tie with Destiny Bond + Explosion. Especially when you're not sure what the AI will pick and in what order. In fact, I only won by tieing once throughout the entire run.

Also, my team is actually very vulnerable to Tucker. Tucker's Metagross has Protect which nullifies Slaking and one Quick Claw activation vs my own Metagross would spell defeat. Ended up losing to Curse + Rest Registeel. Tried fishing for crits with EQ but it never happened. Should have brought Gengar instead of Metagross, but I'm not really too disappointed since this team was never destined to make it extremely far due to its vulnerability to hax and Tucker.


I'm curious as to people's opinions on whether Hyper Beam or Return is better as the 4th move on Slaking. Looking at the Dome leaderboard, it seems like the opinions are somewhat mixed.

 

[average fella]

625 rooms in Pike

streak

pike is fun. after playing normal (non-blissey) pike, i had enough fun to want to try blissey, and my goodness, being able to plow through nostalgia rooms is really, really good.

Starmie @ Lum Berry
Ability: Natural Cure
Level: 50
EVs: 4 HP / 252 SpA / 252 Spe
Modest Nature
IVs: 10 Atk
- Surf
- Thunderbolt
- Ice Beam
- Psychic

Metagross @ Choice Band
Ability: Clear Body
Level: 50
EVs: 164 HP / 252 Atk / 92 Spe
Adamant Nature
- Meteor Mash
- Earthquake
- Shadow Ball
- Brick Break

Blissey @ Leftovers
Ability: Natural Cure
Level: 50
EVs: 172 HP / 252 Def / 84 Spe
Bold Nature
IVs: 4 Atk
- Seismic Toss
- Toxic
- Soft-Boiled
- Aromatherapy



the metagross/blissey speed creeps are definitely not original, I believe they were both directly copied from runeblade's record, who got it from jellal. blissey is for haxrein, i know this much.


basic strategy:
its kind of one of those "use the moves to kill the thing" in-battle. if blissey can easily wall something, farm hp off of it, if you have to sack something, starmie is the most valuable member of the team, then blissey, then metagross.

this team is notable for not really having a solid doubles strategy, and not having an eq immunity makes doubles maybe more threatening than singles, it definitely is more threatening if your members are statused. focus the thing that is the most threatening to the lead two. if metagross has to eq, dont worry about it, starmie will live a non-crit, meaning that blissey just has to use softboiled a couple times to top starmie off. if blissey is out there in a double battle, the battle has gone firmly off the rails, or you only had 2 mons in the first place

in terms of rooms: it goes like this:

aroma of pokemon: ALWAYS, wilds are free, singles with healing are likely wins
notalgia: ALWAYS
whispering: if healthy, take this. it's quite likely to be a no event, but if one of your leads are para'd/frozen/sleep'd, or if metagross is burned and it needs to use a coverage move to take something out, you can be in a difficult position where one or more mons needs to be sacked to get aromatherapy off, or to natural cure starmie.
is it a trainer?: i will take these rooms if it's the 7th room, and my mons aren't healthy. it's a lot easier to get a cleric off when there's only 1 enemy mon to work around as opposed to two, and when you can get natural cures off easily.

how i lost:
even the healthiest of teams can fall to ohko moves, and that's precisely what happened to me. everyone was completely healthy. Aroma room, so I only need to win to get a full heal, no preservation is needed.

vs.
Electrode3/4 (likely 3)
Meganium3
Dugtrio3/4

Starmie cannot ohko electrode, and tbolt ohkos it, so it gets out of the way. Electrode twaves blissey, blissey gets parahaxed as it tbolts. I try toxicing it as it... booms. It was about a 11/15 roll for it to kill Blissey, and it kills. I had no way of knowing that it'd have boomed precisely this turn, but it's now 2 on 2.

Starmie comes in, as Meganium3 comes in. Bad matchup, unavoidable. I get out of the way, Metagross tanks a Giga Drain. Meganium outspeeds it (telegraphing Meganium4), and it earthquakes. Meteor Mash gets a good roll to ohko, but Metagross is now in dugtrio eq kill range. Even if I switched metagross in initially, a min roll meganium eq and a min roll dugtrio eq would still kill. Here's the relevant mash roll, and the relevant eq rolls:

252+ Atk Choice Band Metagross Meteor Mash vs. 0 HP / 0 Def Meganium: 145-171 (93.5 - 110.3%) -- 62.5% chance to OHKO
255 Atk Meganium Earthquake vs. 164 HP / 0 Def Metagross: 69-82 (39.2 - 46.5%) -- guaranteed 3HKO
255+ Atk Dugtrio Earthquake vs. 164 HP / 0 Def Metagross: 112-132 (63.6 - 75%) -- guaranteed 2HKO

And the last, Dugtrio3/4. Metagross dies to eq as expected, and in comes Starmie. I have no choice but to pray that either fissure misses, or that it chooses eq, but king's rock doesn't proc...

Fissure is chosen, fissure hits. I don't know if there's a way I could have won that, I didn't make some blatantly wrong decision, the ai just booms at unpredictable times, and I didn't predict it, and ohko moves are indeed a great run-killer.




I feel like it's possible that I can get farther with this exact team, but you can only go so far before you encounter haxrein/crapras in a double battle with a mon that threatens one of my leads on the other side, or a triple freeze event after Starmie eats its lum berry, followed by a gyarados2 lead on the next aroma room, or any number of potential double leads that can screw over a paralyzed lead tandem, or Starmie gets a low roll vs. Heracross4, and Reversal either destroys my team then and there, or makes it unhealthy enough that the next room picks it off, all of which are things that have happened to me in losses. I might come back to Pike later, but 625 is a pretty good streak, so it'd be months before I would seriously think about it.

 

[Purple Kecleon]

Speaking of the Battle Pike, I was wondering what the odds of the nostalgia room are. Looking at some websites, it’s either a 45 or 50 % chance of status and either a 50 or 55 % chance of heal. But in my experience, I get the status room like 85% of the time. Have I just been unlucky?

 

[Kommo-o]

Speaking of the Battle Pike, I was wondering what the odds of the nostalgia room are. Looking at some websites, it’s either a 45 or 50 % chance of status and either a 50 or 55 % chance of heal. But in my experience, I get the status room like 85% of the time. Have I just been unlucky?

You were just unlucky

 

[Purple Kecleon]

Just finished a streak of 347 rooms in the Battle Pike! Honestly, I'm surprised my team even made it this far considering how bad my Latias and Heracross's IVs are.


Latias @ Lum Berry
Ability: Levitate
Nature: Mild
EVs: 6 HP, 252 SpA, 252 Spe
IVs: 0/4/28/18/25/12
Calm Mind
Psychic
Thunderbolt
Ice Beam

This was the Latias I caught in my Emerald many years ago before I even considered doing the Battle Frontier. I'm just glad that it at least has a pretty decent nature. I had Dragon Claw from back when I used to play the Battle Frontier, but I decided to test out Ice Beam and it seemed to do much better than Dragon Claw would have. (OHKOing Salamence/Dragonite was especially important in Double Battles and hitting ground types super effectively is nice as well)


Heracross @ Choice Band
Ability: Guts
Nature: Adamant
EVs: 6 HP, 252 Atk, 252 Spe
IVs: 2/9/21/19/5/31
Megahorn
Brick Break
Earthquake
Rock Slide

Yes, the IVs are cringeworthy I know. I believe I bred this Heracross long before I even knew about the Everstone nature passing effect, which is why I settled with such bad IVs. But yeah, Latias's Levitate + Banded Earthquake rocks in Double Battles. And Guts makes Heracross even more deadly when it gets poisoned or burned from the nostalgia room.


Blissey @ Leftovers
Ability: Natural Cure
Nature: Calm
EVs: 252 HP, 252 Def, 6 Spe
IVs: 19/24/31/15/30/30
Aromatherapy
Softboiled
Seismic Toss
Flamethrower

Just like in the Pyramid, Blissey was also immensely helpful in the Pike. Didn't use Flamethrower except on wild Dusclops so Toxic is maybe something to consider as an alternative.


After Battle 280, wild Electrodes start appearing and I noticed my Latias sometimes failed to escape (and then got blown up on). Turns out a Level 95 Electrode with 31 IVs, 0 EVs, and a neutral nature will have exactly 300 speed, which is the same speed as my Latias. Although I'm not sure if the wild Pokemon have fixed stats here.


Latios @ Lum Berry
Ability: Levitate
Nature: Mild
EVs: 6 HP, 252 SpA, 252 Spe
IVs: 16/31/7/16/18/21
Calm Mind
Psychic
Thunderbolt
Ice Beam

To resolve this problem, I swapped in my very recently obtained Latios in place of Latias from room 308 onwards. With 309 speed, I never failed to run from Electrode up until the point I lost, but maybe it's still possible that the Electrodes could have a + Speed nature. And once again I got lucky with a Mild nature. (at least by my standards lol)

I know Latios is better anyways, but the reason I used Latias is purely because I had used it previously and I wanted to stick with the same team. But I figured that wasn't worth essentially forfeiting my streak so I ended up switching.


In terms of room picking strategy, my rule was to enter the "distinct aroma" and "whispering" rooms and avoid the "trainer" rooms. I picked the nostalgia room if it was the first 2 rooms (since he will only status 1 pokemon) or the 7th. I only deviated from the rule a couple of times when my team was in bad shape. I once had only a statused Blissey alive after the 2nd room, but somehow got insanely lucky with healing and non-battling rooms and even managed to beat Lucy at the end lol.

How I Lost:
I got into a double battle vs a Muk and Tauros with my Latios and Blissey paralyzed. Heracross OHKOs Tauros w/Brick Break but Muk KOs Latios with a critical hit Sludge Bomb. This was really bad since now my Heracross couldn't switch and was locked into Brick Break and my Psychic user was dead. If my Blissey wasn't parahaxed multiple times, I would have won the battle with a solo Blissey with 1 more room to go. But oh well, I would still say I was pretty lucky throughout the run all thing considered.

 

[wtset]

Speaking of the Battle Pike, I was wondering what the odds of the nostalgia room are. Looking at some websites, it’s either a 45 or 50 % chance of status and either a 50 or 55 % chance of heal. But in my experience, I get the status room like 85% of the time. Have I just been unlucky?

If I recall correctly, the game deletes all healing rooms from the list of rooms to randomly choose from in case your team is already fully healed. This can definitely lead to you encountering the status room far more often on average (technically the status room can also be deleted from the list if your whole team is statused already, but in my experience this is much less likely to happen than having a fully healed squad).

 

[Purple Kecleon]

If I recall correctly, the game deletes all healing rooms from the list of rooms to randomly choose from in case your team is already fully healed. This can definitely lead to you encountering the status room far more often on average (technically the status room can also be deleted from the list if your whole team is statused already, but in my experience this is much less likely to happen than having a fully healed squad).

Yes you are right, I just saw this on Werster’s FAQ recently and this must have been why I encountered the status room so frequently. According to that FAQ, you also can’t get a double battle with only 1 Pokémon alive.

 

[Actaeon]

Small but significant update on the Monotype Challenge. Recently I had quite a few losses with Mono Psychic (I thought this was the most dependable team), but apparently I either underestimated or didn't encounter too many Houndoom, Shadow Ball Gengar, Metagross, Crobat and sometimes Heracross can be a pain as well.

Heracross was the main reason I chose to use Xatu instead of Espeon; both Pokemon have access to a two-stage Attack-reducing move (Featherdance and Charm), but Xatu actually survives a CB Megahorn after Featherdance where Espeon can't say the same if it wants enough Speed to become interesting over Xatu. The latter having access to Thunder Wave as well convinced me to use Xatu back then:

The old Mono Psychic

Grumpig (F) @ Choice Band
Ability: Thick Fat
EVs: 196 HP / 220 Def / 92 Spe
Timid Nature (+Spe, -Atk)
- Trick
- Icy Wind
- Flash
- Skill Swap

Xatu (M) @ Salac Berry
Ability: Synchronize
EVs: 60 HP / 252 Def / 196 Spe
Timid Nature (+Spe, -Atk)
- Featherdance
- Thunder Wave
- Endure
- Reflect

Latias (F) @ Lum Berry
Ability: Levitate
EVs: 36 HP / 252 Def / 220 Spe
Timid Nature (+Spe, -Atk)
- Calm Mind
- Substitute
- Recover
- Dragon Claw

Xatu, however, doesn't offer dependable crippling against Houndoom (only Thunder Wave) and Gengar (it doesn't outspeed and gets obliterated by CB Shadow Ball). Of course, an untimely crit from physical Pokemon can be dangerous as well, because they often still break Latias' Substitute with CB even at -2 Atk, which means I can't setup without fearing another CH.

These problems made me have a second look at Espeon, which gets the 100%-accurate Sand-Attack as well. It's at a way better Speed tier than Xatu, and back then I probably didn't consider adding another substrategy with Substitute + Protect, ditching the advantages of Thunder Wave in favor of setting up against -2 Struggles, possibly with accuracy reductions:

Espeon (M) @ Leftovers
Ability: Synchronize
EVs: 36 HP / 252 Def / 220 Spe
Timid Nature (+Spe, -Atk)
- Protect
- Substitute
- Charm
- Sand-Attack

EVs: 173 Speed needs no explanation, 252 Def gives it the most chance of surviving a Struggle after it outstalls e.g. Meteor Mash and uses Charm. The HP EVs makes my HP one more than a multiple of 16, so that's a nice coincidence!

How does this help the strategy against the aforementioned problems?

• Against Crobat, I lead with Icy Wind, so I can Charm and hopefully beat it by outspeeding it while it has no Attack boost from Choice Band (not that different from the strategy with Xatu)
• Metagross has to be lucky enough to get 2 Attack boosts from the 5 Meteor Mashes it will get off, since Grumpig lives through a CB Meteor Mash and will use Skill Swap. Espeon can then get off at least one Charm after it outstalls Meteor Mash.
• Houndoom's Crunch and Gengar's Shadow Ball is outstalled, then Charm cripples its Struggle; this Espeon actually outstalls every 15 PP move if necessary.
• After Heracross KO's Grumpig (before or after Trick, depending whether it hits on turn 1 or not), this outstalls the Megahorns and brings Heracross' Attack to -2, so Struggle doesn't break Latias' Substitute anymore.

With this strategy, I'm going to fight more Struggle users with Latias, so an option would be to use Bold Latias with 153 Speed, full Defense investment, and the rest into HP. This makes sure its Substitute survive +0 CB Metagross Struggle over 80% of the time, and it also gives me 87.5% chance to survive a critical hit Meteor Mash, should I encounter a QC Metagross while already sweeping. But this is at the cost of not outspeeding stuff like opposing Espeon, Gengar, Starmie, Aerodactyl, etc. I'm not sure yet whether this is worth the payoff of having a bulkier spread for the setup, so I guess that's up to some experimentation.

Edit: Latias now has Bold with 76 HP / 252 Def / 180 Spe as I definitely feel it's better than a speedy set; this Latias is crazy bulky on the physical side so it's almost guaranteed to setup and most of the Pokemon can't beat it after setup, even if they hax it a little bit. For example, it lives through the hits of many Pokemon that outspeed it after 6 Calm Minds (Starmie, non-physical Gengar, Espeon).

Thanks for reading and please let me know if you have other suggestions!