SwagPlay, evaluating potential bans (basic definition of "uncompetitive" in OP)

[Haruno]

I did some iterated analysis on SwagPlay using a couple of Pokemon. I'm going to generalize the formula to eventually allow for arbitrary Pokemon based on usage data, but here's my initial findings. I ran 10000 simulations of Klefki vs. <Pokemon> for this data. This is a simplified analysis and doesn't perfectly model an actual battle, but is relatively close.

Assumptions are that neither side switches out, ever, and that they fight until the bitter end. PP is not taken into account.

Pokemon     W        L
Blissey        10000     0
Rotom-W        6255     3745
Garchomp    7343     2657
Greninja    3772     6228
Heatran        9101     899
Gliscor        8907     1093
Aegislash    1486     8514
Slowbro        7128     2872

Leftovers IS taken into account.
Klefki uses the following formula:

def swagplay(me, enemy):
if not enemy.confused:
swagger(me, enemy)
elif not me.substitute and me.hp > me.maxhp / 4:
substitute(me, enemy)
else:
foulplay(me, enemy)

In the final iteration of my program, I will improve this formula to the Swaggerificc formula.

I believe this data will aid us in determining the competitiveness of the strategy, by determining how much each match-up is due to luck.

You forgot about twave which would increase the prankster's users chance of success against most mons.

 

[Rafael J. Feliciano]

Thanks SanjiWatsuki, for bringing in good data to this discussion.

Can you report standard-error as part of the results? Standard Error for a proportion is simply sqrt(p * (1-p) / n), where n is the number of trials and p is the estimated proportion.

For example, Slowbro's proportion is 0.7128. The standard error is sqrt(0.7128 * 0.2872 / 10000) == 0.0045.

So the standard error for Slowbro is 71.28% +/- 0.45% chance to win (68% confidence interval), or +/- 0.9% (95% confidence interval). It helps remind the fact that these are statistically derived values and have some degree of unreliability to them.

I did some iterated analysis on SwagPlay using a couple of Pokemon. I'm going to generalize the formula to eventually allow for arbitrary Pokemon based on usage data, but here's my initial findings. I ran 10000 simulations of Klefki vs. <Pokemon> for this data. This is a simplified analysis and doesn't perfectly model an actual battle, but is relatively close.

Assumptions are that neither side switches out, ever, and that they fight until the bitter end. PP is not taken into account.

Pokemon     W        L
Blissey        10000     0
Rotom-W        6255     3745
Garchomp    7343     2657
Greninja    3772     6228
Heatran        9101     899
Gliscor        8907     1093
Aegislash    1486     8514
Slowbro        7128     2872

Leftovers IS taken into account.
Klefki uses the following formula:

def swagplay(me, enemy):
if not enemy.confused:
swagger(me, enemy)
elif not me.substitute and me.hp > me.maxhp / 4:
substitute(me, enemy)
else:
foulplay(me, enemy)

In the final iteration of my program, I will improve this formula to the Swaggerificc formula.

I believe this data will aid us in determining the competitiveness of the strategy, by determining how much each match-up is due to luck.

Aw damn... I knew I should've paid more attention in AP Statistics... XD

Naw but seriously, good stuff Dragontamer & SanjiWatsuki. Now that we can get some quanitative data to this debate we might be able to prove that even if some pokémon; like Slowbro/Slowking and Blissey/Chansey; can be excellent obstacles for SwagPlay users... that doesn't mean all of us should resort to incorporating one of these guys into our team on the off chance that I might encounter a "luck box" on the server. (Or as I like to call them, Jonouchi fans for those of you whoever played/watched Japanese Yu-Gi-Oh! :-)

 

[TheFirstPixel]

I see some players saying that Chansey, Slowbro, Rotom-W, etc...happen to be the best SwagPlay checks. Now note that the multiple Leppa Berry complex bans came about due to the "creative dickishness" of players. In this case, Swagger players may soon resort to this:

Whimisicott @ Leftovers
Ability: Prankster
Nature: Bold / Calm
EVs: 252 HP / 252 Def or SpD / 4 SpD or Def (whichever, really)
- Swagger
- Substitute
- Leech Seed
- Taunt

...while leaving their general game plan intact. It can easily put the pressure on those bulky sponges and deny their recovery, sapping their health away without relying on Foul Play, then proceed to Swagger all over the place anyway. It also isn't farfetched for Thundurus-I to run Thunderbolt to fry Slowbro.

While this Whimsicott is technically a pro-ban argument, it is also a classic example on how it doesn't manage to abuse Prankster Swagger to the absurd degree we are all familiar with (and loathe). This is arguable proof that Prankster Swagger itself is not necessarily the issue to address, but since it still technically leaves the game up to chance, the verdict would still highly subjective.

Prankster confusion by itself is pretty annoying at worst, but there are a good number of moves that can easily take advantage of this to put the confusion users in a somewhat unfair (again, subjective) advantage over the victims, who usually has little to no say in whether or not they want to participate in these games of pure luck. Let's say both players play a game where they may only use moves that can miss (or hell, Metronome battles). It leaves the game up to chance also, but at least both players signed up for it, which is what makes the element of chance fun. The problem starts when the element of chance is forced onto other players, like evasion boosting (this is different from lowering the opponent's accuracy, since the opponent can choose to either switch or stubbornly break through the hax). Players are not saying it is unbeatable (the fact that it has the tendency to work is another story altogether), but it removes their choices in battle. Now I know not all Prankster Pokemon can make use of Swagger as effectively as some others, but the same can be said about Double Team (the best users have recovery and a degree of bulk). Prankster Swagger will likely have to be banned on the same grounds as Double Team, even if it does hurt Pranksters that are more reliant on just Swagger to succeed with it.

I've seen that exact Whims

I see some players saying that Chansey, Slowbro, Rotom-W, etc...happen to be the best SwagPlay checks. Now note that the multiple Leppa Berry complex bans came about due to the "creative dickishness" of players. In this case, Swagger players may soon resort to this:

Whimisicott @ Leftovers
Ability: Prankster
Nature: Bold / Calm
EVs: 252 HP / 252 Def or SpD / 4 SpD or Def (whichever, really)
- Swagger
- Substitute
- Leech Seed
- Taunt

...while leaving their general game plan intact. It can easily put the pressure on those bulky sponges and deny their recovery, sapping their health away without relying on Foul Play, then proceed to Swagger all over the place anyway. It also isn't farfetched for Thundurus-I to run Thunderbolt to fry Slowbro.

While this Whimsicott is technically a pro-ban argument, it is also a classic example on how it doesn't manage to abuse Prankster Swagger to the absurd degree we are all familiar with (and loathe). This is arguable proof that Prankster Swagger itself is not necessarily the issue to address, but since it still technically leaves the game up to chance, the verdict would still highly subjective.

Prankster confusion by itself is pretty annoying at worst, but there are a good number of moves that can easily take advantage of this to put the confusion users in a somewhat unfair (again, subjective) advantage over the victims, who usually has little to no say in whether or not they want to participate in these games of pure luck. Let's say both players play a game where they may only use moves that can miss (or hell, Metronome battles). It leaves the game up to chance also, but at least both players signed up for it, which is what makes the element of chance fun. The problem starts when the element of chance is forced onto other players, like evasion boosting (this is different from lowering the opponent's accuracy, since the opponent can choose to either switch or stubbornly break through the hax). Players are not saying it is unbeatable (the fact that it has the tendency to work is another story altogether), but it removes their choices in battle. Now I know not all Prankster Pokemon can make use of Swagger as effectively as some others, but the same can be said about Double Team (the best users have recovery and a degree of bulk). Prankster Swagger will likely have to be banned on the same grounds as Double Team, even if it does hurt Pranksters that are more reliant on just Swagger to succeed with it.

I've seen this exact Whimsicott destroy my team multiple times in OU. I agree that it's is fine to include chance as a fun element in battling but forcing it onto people can ruin the game. Something should be done to nerf it, but at the moment no coding changes are going to happen so some sort of clause or ban will have to do.

 

[Petrico94]

Small note: Regenerator is actually one of the most optimal abilities to combat SwagPlay. Switching to remove confusion, attack boosts, and heal in the same turn is a really great play against it.

Do we really know that? I don't feel like anyone has done formal analysis of Pokemon outside of this scope.

My gut-feeling is that Pokemon like Rotom-W, who is everywhere, beat this strategy consistently as well. Without analysis or significant play testing, we can't know for sure. From the playtesting of SwagPlay given in this thread, for example, Rotom-W has often been cited as a Pokemon SwagPlay has troubles with.

As for the hax, there is an acceptable level of hax that removes player control in OU right now. For example, Ice Beam freeze hax. The counter-argument often made is that teams aren't made around freeze hax -- which is fair enough. But if we can prove that a large subset of Pokemon beat SwagPlay more often than Ice Beam freezes, then isn't SwagPlay more competitive than Ice Beam?

Problem, you don't want to be switching around letting everyone get paralyzed.

 

[skies]

The problem with the Pokemon that "consistently" beat SwagPlay is that they fall into one or more of the following catergories:
1.) They're complete shit (Numel)

2.) They have better abilities to run that give them their viability in the first place (Slowbro and Slowking would rather have Regenerator)

3.) They only fit into a very specific team styles (Chansey and Blissey don't work on Hyper Offense)

They can also still be haxed by Parafusion, Toxic Stalled to death if they barely take anything from Foul Play, and Taunted.

Thundurus is a very viable Pokemon that aids a lot of teams, and can switch into everything and Taunt them. People often do forget that it is the only true viable counter 100% of the time, because:
0 Atk Klefki Foul Play vs. 0 HP / 4 Def Thundurus: 82-97 (27.4 - 32.4%) -- 71.8% chance to 4HKO after Leftovers recovery
It can also use this strategy.

 

[SergantGold]

I think that Swagger + T-Wave Should be banned

individually they are not problematic and are only marginally (at best) annoying to pokemon, but together they just make a gimmicky set that has a 50/50(guesstimation) chance of completely neutering your team and auto-winning or just outright failing. This is very bad and uncompetitive, and overall makes Klefki, who would normally be a very good strictly Supportive Dual Screener/Spike Stacker/T-wave supporter prankster mon, turn into a gimmick that nobody likes.

Swagger alone can easily be worked around and puts the user at insane risk if they swaggerd a physical mon, but it isn't banworthy
t-wave cripples a mon, but it isn't banworthy alone at all

 

[MagikaripIsOP]

stuff

In the final iteration include all s through a- (maybe through b-) rank pokemon

 

[Niftyyyy]

You forgot about twave which would increase the prankster's users chance of success against most mons.

He also has not accounted for the odds these Pokemon have if they switch in, which is considerably less in some cases.

Also all this data really shows is how few viable complete counters that klefki has. Some of the checks are shaky at best, even with a free switch.

 

[Gachikaze]

What is the point of banning an important, legit, and counterable strategy? Only because it's "not fun"? Sure, let's ban paralysis. And crits.

Also, magic bounce HARD COUNTERS it...

 

[Pikachuun]

What is the point of banning an important, legit, and counterable strategy? Only because it's "not fun"? Sure, let's ban paralysis. And crits.

Also, magic bounce HARD COUNTERS it...

Tell me what's important about this strat without saying anything related to diversity or its synonyms. If you cannot, it isn't important.

 

[TooMuchSugar]

What is the point of banning an important, legit, and counterable strategy? Only because it's "not fun"? Sure, let's ban paralysis. And crits.

Also, magic bounce HARD COUNTERS it...

Magic Bounce has been refuted so many times. It barely even checks SwagPlay, let alone counters it. Basically, all three Magic Bounce users are 3HKO'd by Foul Play, so they cannot reliably beat it (also, none can touch Klefki without running relatively niche moves).

 

[Gachikaze]

Tell me what's important about this strat without saying anything related to diversity or its synonyms. If you cannot, it isn't important.

It is important because it is perfectly viable. It is important because without it, the metagame will change. Maybe only slightly, but it will still change. If we ban confusion it will be even worse.

If we ban Swagplay, Klefki has just lost one of its main sets. Liepard will be a piece of shit in the metagame. Sableye lost a nice set it used to have. See how it is important?

 

[TooMuchSugar]

It is important because it is perfectly viable. It is important because without it, the metagame will change. Maybe only slightly, but it will still change. If we ban confusion it will be even worse.

If we ban Swagplay, Klefki has just lost one of its main sets. Liepard will be a piece of shit in the metagame. Sableye lost a nice set it used to have. See how it is important?

Evasion was viable, and its ban changed the metagame. #FreeEvasion

Klefki and Sableye have way better things to do. Liepard is already a piece of shit.

 

[SergantGold]

It is important because it is perfectly viable. It is important because without it, the metagame will change. Maybe only slightly, but it will still change. If we ban confusion it will be even worse.

If we ban Swagplay, Klefki has just lost one of its main sets. Liepard will be a piece of shit in the metagame. Sableye lost a nice set it used to have. See how it is important?

SwagPlay Klefki is unreliable and inferior to other Klefki sets. its the fact that it leaves the life of your pokemon down to a 50/50 chance on everything is absrud and as luck based as OHKO moves imo.

SwagPlay isn't even threatening without T-wave anyway

 

[idccc]

after reading through all of this, doesnt it pretty much just come down to Smogon seeing that while statistically SwagPlay is beatable, do they want this in their metagame? the strategy does take A LOT of control away from the opposing player by forcing a stall match or risking to hit through the 50/50 without factoring twave. and by the OP's definition it fits uncompetitive, i think? im not an expert and i dont have fancy statistics but i'm pretty sure the meta would be just fine without Prankster + Swagger.

quality seems to be going down though, but who am i to tell you guys to close this lol.

 

[Chou Toshio]

Pokemon     W        L
Blissey        10000     0
Rotom-W        6255     3745
Garchomp    7343     2657
Greninja    3772     6228
Heatran        9101     899
Gliscor        8907     1093
Aegislash    1486     8514
Slowbro        7128     2872

Definitely an interesting post-- because it highlights even more to me that all of these Pokemon lose too often. Except Blissey. lol

To be fair, Rotom's "too high" level of loses will never matter because of Volt Switch. (Rotom's "true" loss rate is no where near as high as what's stated). But then on the same note-- Rotom-W will never finish off Klefki (so something ELSE has to beat the odds...) and is close to useless in beating Thundurus-I so... yeah...

Thinking about an uncompetitive strategy, Garchomp doing 7w-3l means it's losing way too often. Even gliscor's 9 wins to 1 loss makes me want to ban this strategy-- ridiculous that a poke like Gliscor will lose to Klefki that often!! (also gliscor's loses outright if it lacks Ice Fang and is fighting Thundurus-I).

Even Heatran losing 9% of the time to this shit is ridiculous-- I mean, Heatran, losing to Klefki without something as extreme as fire resist berry + HP Ground. That's terrible.

Also, ALL of these Pokemon (save Blissey) will have a much lower chance of winning if they are switching into Swagger or Substitute. (and switching in is where the "control" and "decision making" is right?)




All this is besides the point.

This data ultimately doesn't matter at all, one way or another because your argument in this post is off-basis.

I agree fully with this key! I also believe it fits in with the anti-ban argument.

I believe an insufficient level of analysis has been done to make a quick ban on SwagPlay, however! As outlined in http://www.smogon.com/forums/thread...ompetitive-in-op.3500620/page-63#post-5271166 , there is a certain level of computational analysis that should be done to determine how reliable SwagPlay counters are AND how common they are, not how reliable the strategy is. If we can determine that certain Pokemon very consistently beat SwagPlay, then how is player control being removed? A decision that the player DIRECTLY made (i.e. which Pokemon to use and what attacks they use) has resulted in them winning the match-up.

% chance of victory does not determine level of control removed. "level of control removed" is not just about luck and percentages, and is something judged by player experience.

Your simulations ultimately mean nothing, because we are not making tiering decisions to "match certain numbers." We are making tiering decisions in order to improve the level of competition in battles, and to improve player experience.


It is player experience and not game mechanics that is the variable that determines tiering/regulation decisions in all cases-- including uncompetitiveness.

It is also possible that the analysis determines that luck dominates even many of the favorable match-ups, in which case I would concede my case. The fact remains however that we lack sufficient information to make an informed decision about the ban of SwagPlay.

You're wrong. We don't lack sufficient data-- we have more than enough data on community consensus; which is the more important factor. Though if you really wanted to do analysis-- the more useful one would be based on community opinion research, not simulation. No simulation will make a player better informed than one thing: Actual battle, and the opinion formed from his own experience (because "player experience" is the target parameter, while "% chance of win" is not even a good substitute).

Because player experience is what we're looking to cater too, not raw #'s that mean nothing without human opinion attached.

In turn, it is the tiering leaders who will ultimately make a decision on this-- and dare I say they're more than well enough informed to make a good decision about this.

I believe that the decision making of the opponent to respond to SwagPlay is moved into the macro, rather than the micro turn-based scale. Furthermore, any bans should be postponed until AFTER analysis has been done.

Now I must say, it's ridiculous for one user to think that tiering decisions should wait or be determined based on his analysis. Don't make me laugh.

 

[Ununhexium]

What a lot of the people following this thread fail to understand is that this is not a discussion about whether or not it can be countered, but if it makes it noncompetitive. If you are losing to people because of luck, I would go as far as to call that noncompetitive. Ban Prankster + Swagger

 

[Sinning]

It is important because it is perfectly viable. It is important because without it, the metagame will change. Maybe only slightly, but it will still change. If we ban confusion it will be even worse.

If we ban Swagplay, Klefki has just lost one of its main sets. Liepard will be a piece of shit in the metagame. Sableye lost a nice set it used to have. See how it is important?

A pokemon's place in OU is of no importance if we truly care about competition. We do not care that sometimes peoples' favorite pokemon are not useable. We simply care to create a fair game, and we don't care if klefki or liepard loses it's best utility by removing an unfair strategy. Perhaps this means that pokemon like liepard are simply too poor (outside of an uncompetitive strategy) to be used in the OU environment, while pokemon like klefki have legitimate uses in OU, and would actually be used in their intended purposes if we removed this unfair tactic.

No (good) player cares that Kyogre is removed from OU. Kyogre is viable, and without it metagames have shifted (we used politoed instead), however banning it did not make anything "worse" and neither would this ban. To say that keeping an unfair strategy around because it's users would lose viability is just a poor argument -- you don't see anyone arguing to keep moody in order to keep it's users viable.

 

[Shion SinX]

Losing because of bad luck is not problem. Losing too often because of of luck is.

You lost a battle because your fire blast missed twice in a row? ok
you lost 20 battles in a row because your pokemon refused to attack the enemy while under confusion? not ok

I changed my mind, I agree with Prankster + Swagger ban.

 

[Chou Toshio]

It is important because it is perfectly viable. It is important because without it, the metagame will change. Maybe only slightly, but it will still change. If we ban confusion it will be even worse.

If we ban Swagplay, Klefki has just lost one of its main sets. Liepard will be a piece of shit in the metagame. Sableye lost a nice set it used to have. See how it is important?

As mentioned elsewhere in this thread-- diversity is something we want to foster, and something that the community values; however it's not something the community values enough to sacrifice competitiveness.

Sure, it's great seeing Pokemon like Slowbro, Staraptor, and even Rhydon finding places in the metagame; but we wouldn't unban moody just so Bidoof and Octillery could do so as well.

 

[Redew]

It is important because it is perfectly viable. It is important because without it, the metagame will change. Maybe only slightly, but it will still change. If we ban confusion it will be even worse.

If we ban Swagplay, Klefki has just lost one of its main sets. Liepard will be a piece of shit in the metagame. Sableye lost a nice set it used to have. See how it is important?

1. Klefki has better sets.
2. Liepard is not meant for OU.
3. Sableye has other sets it can use. This isn't what is keeping it OU.

"Important" doesn't mean "keeps bad things in a metagame." Other "sucky" Pokemon have niches in OU, yes, but not because of SwagPlay. Having to prepare for these Pokemon restricts teambuilding. We would like a wide variety of Pokemon and sets, but if it comes to a point where you have to prepare for them JUST IN CASE, it makes it unfun and ruins the metagame.

 

[the-bumper-car]

I can't believe that after 70 pages, there are still people who think that this strategy shouldn't be actioned on with a ban or a clause. Seriously, guys?

 

[Subject 18]

What is the point of banning an important, legit, and counterable strategy? Only because it's "not fun"? Sure, let's ban paralysis. And crits.

Also, magic bounce HARD COUNTERS it...

Define important. Just because its viable doesn't mean it shouldn't be banned for being uncompetitive. Mega Lucario, Mega Gengar and Blaziken were all viable does that mean they shouldn't have been banned? No.

Banning Paralysis and crits would mean that we need to alter game mechanics. That is not going to happen.

Magic Bounce doesn't counter it. None of the Magic Bounce users appreciate taking a Foul Play. This has been discussed before, please read the thread.

 

[fenyxofshadows]

Banning Paralysis and crits would mean that we need to alter game mechanics. That is not going to happen

Then explain how they banned evasion, putting more than one monster to sleep at a time, and why many of the users in this thread wish to ban Confusion caused as a primary effect, regardless of usage?

Yes, it would change game mechanics. But so does every ban. The goal is to ban an uncompetitive strategy with as little effect on game mechanics as possible.

This isn't an anti-ban argument, just an argument against the "reasons" why you can't ban Paralysis; you easily could.

 

[Redew]

Then explain how they banned evasion, putting more than one monster to sleep at a time, and why many of the users in this thread wish to ban Confusion caused as a primary effect, regardless of usage?

Yes, it would change game mechanics. But so does every ban. The goal is to ban an uncompetitive strategy with as little effect on game mechanics as possible.

This isn't an anti-ban argument, just an argument against the "reasons" why you can't ban Paralysis; you easily could.

Evasion clause bans the moves, not what happens if you use it. The ONE TIME we messed with game mechanics was sleep clause, and that was a one-time thing. Many users would LIKE to ban confusion, but we have said, like, a thousand times that we aren't. Not sure if you saw that or not.

"Every ban" does not change the mechanics. The only one that did was sleep clause. Not sure where you see this. If you see it on a ladder then you probably aren't playing on Smogon's ladder.