CAP 4 (Aurumoth) Post-mortem: Which competitive step did we screw up on the most?

[zyrefredric]

Which competitive step did we screw up on the most?

I think it would be easier to answer the question "Which competitive step did we NOT screw up on?". My answer would be the first ability poll. Despite the TL and some other prominent members expressing their dislike for Weak Armour (even though it embodies the concept to a tee imho), it still won the ability polls. I was very happy, and gained confidence in the democratic CAP process. However, everything after that was so bad it was disgusting.

• The next two abilities just did not make sense for me. Illusion's and No Guard's risk lies on opponent's predictions and susceptibility to inaccurate attacks, which frankly, are not even that risky. Also, unlike Weak Armour where the reward (+Spe) can only be achieved by taking the risk (-Def), you could actually play a whole match where the disguise is never revealed in a crucial point, or got hit by a Stone Edge that would have missed. Thus, using the other two abilities gives you guaranteed reward without the risk even coming into play.
• Now although the stat distribution is something I liked (mixed offenses, good defense, bad sp.defense, except speed - wanted it to be lower), when I tried to make a stat spread submission myself, I felt that the stat limits imposed forces us to make high BST spreads, and so we ended up with a 600-BST CAP. This is not necessarily bad, if not for it having two powerful abilities to go with it.
• Now the movepool. I remembered one of the argument for No Guard was that if we don't give it inaccurate moves to abuse, then it's reward will be VERY minimal (e.g. Megahorn only) compared to it's risk (susceptibility to Fire Blast, Stone Edge, etc). But maybe it was forgotten? Because CAP4 got Thunder, Blizzard, Hydro Pump, Focus Blast, etc. Wow.
• And then there's Dragon Dance and Quiver Dance. With Illusion basically guarantees a boost in speed AND Atk/SpA/SpD. Might as well remove Weak Armour. Too bad it's the only thing I believe we did right this CAP.

[uwnim]

The step we screwed up the most was movepool. While the other steps were messed up, those were messed up by BMB. Stats had to be high due to the minimum requirements while No Guard was basically forced on us. Movepool was almost entirely on us though.
DD was almost fine, but instead of Close Combat, it should have had Superpower and Earthquake. That would have made DD sets fairly easy to wall. QD should have just been not allowed though.

[Meganium Sulfate]

Originally Posted by Fat BIRDKAL

As you said, jynx KB, we historically have not done revamps or edits. The reason for this is that we learn from our mistakes; making edits would only hide those lessons. In Krilowatt's case, we learned that a bloated movepool and a fantastic ability can wreck a concept.

But isn't this the same mistake we just made with Aurumoth? I don't think some of us are really learning.

[Rayquaza_]

Now that I think about it, even with Aurumoth's obscenely good movepool and stat distribution, a reasonable trio of abilities that would have worked perfectly in this CAP project would have been Weak Armor, Flare Boost and Unburden.
All three of them are high risk-high reward abilities that would have allowed the moth to be very versatile, potentially extremely dangrous but still fully counterable depending on the set it was running.
The best part is that none of them is clearly superior to the other (Unburden requires you not to run LO, which mean you'd miss on many important KO's so Weak Armor is definitely not outclassed here).

Did anyone even suggest those abilities during the project?

[Nyktos]

Both were definitely suggested, but I don't think either were ever slated.

[kingtrace]

I think that the movepool and abilities were the two parts that were most screwed up. I usually try to tank CAP's the whole way through (even mollux, who was ridiculous at some points), but I believe I abandoned following this cap around the point were Illusion and No Guard were being discussed after Weak Armor was already voted in. Abilities that actually make sense with the concept and the CAP, like Guts, were prematurely tossed out the window with a 'not risky in the RIGHT way' justification, and then we got Illusion, which is a fun ability, but not risky in any real way.

Movepool, obviously, was a huge wreck. I remember looking at the final two movepools out of curiosity and saying 'The only hope this thing has to not be broken is to not get the moveset with every coverage move in the game and one of the most broken setup moves as well.' I have no idea how it got voted in, but it did, and now the only way you learn anything from this mon are: 1. In the same way armies learn how not to screw up from major historical battle flubs. 2. In a roundabout way were the idea is 'Well, the opponent has risk by Aurumoth existing, so there's a risk study!' I honestly hope the next CAP allows more community contribution in some places and less in others. Yes, the movepool is bloated, but the ability pool ignored some decent ones in favor of an idea that doesn't seem to match the concept.

[akela]

Originally Posted by Fat Rayquaza_

Now that I think about it, even with Aurumoth's obscenely good movepool and stat distribution, a reasonable trio of abilities that would have worked perfectly in this CAP project would have been Weak Armor, Flare Boost and Unburden.
All three of them are high risk-high reward abilities that would have allowed the moth to be very versatile, potentially extremely dangrous but still fully counterable depending on the set it was running.
The best part is that none of them is clearly superior to the other (Unburden requires you not to run LO, which mean you'd miss on many important KO's so Weak Armor is definitely not outclassed here).

Did anyone even suggest those abilities during the project?

Yes, they were suggested. So was Toxic Boost, Guts, Analytic, Anger Point, Blaze, Torrent, Overgrow, Swarm, Color Change, Cursed Body, Defeatist, Defiant, Slow Start, Dry SKin, Harvest, Hustle, Special Hustle, Klutz, Light Metal, Heavy Metal, Marvel Scale, Moxie, Multiscale, Mummy, Quick Feet, Rattled, Simple, Stall, Synchronize, Tangled Feet, and a multitude of custom abilities. They were all nixed because they were not the kind of risk the thread leader wanted.

Having a time/turn limit to how long the pokemon would survive was the wrong kind of risk (Toxic Boost, Flare Boost, Guts, Tangled Feet, Quick Feet, Swarm, Overgrow, Torrent, Blaze). Having a decrease in accuracy was the wrong kind of risk (Hustle, Special Hustle). Depending on your opponent attacking you was the wrong kind of risk (Anger Point, Color Change, Cursed Body, Defeatist, Defiant, Dry Skin, Marvel Scale, Multiscale, Mummy, Rattled, Synchronize, Tangled Feet). Klutz was the wrong kind of risk because it depended on the move pool to have trick (or something like it). Harvest was the wrong kind of risk because it depended on the RNG respawning a berry. Light Metal and Heavy Metal were stupid because Aurumoth wasn't Steel Type (Despite that a custom ability that behaved like Light/Heavy Metal was the suggestion). Simple was the wrong kind of risk because no one uses stat decreasing moves. Moxie was the wrong kind of risk because Aurumoth was supposed to be too slow to make use of it. Stall was the wrong kind of risk because Aurumoth was supposed to be fast.

A lot of ideas about what risk was and how to make Aurumoth a risky mon were suggested. For whatever reason, they were nixed rather quick.

[Nyktos]

The rest is accurate, but Moxie was slated for both primary and secondary ability.

[MCBarrett]

Well I haven't been in the cap forum for quite sometime until this caught my eye. I was here for suspect test and did ladder with auromoth for quite a bit and before that I had left cap right around the start of the stats submissions. What I feared during stats I had learned not only came true but had been made even worse by the move pool. The community turned away from weak armor being the primary ability. Although it was still so in name it was quite evident that it was being neglected. and that is where I think cap4 went wrong.

We had a pretty risky typing and a very risky ability but for one reason or another we decided to focus on the other abilities that were chosen once we reached stats. I think we could have lived with no guard an illusion if we focused on making a set that maximized the risk vs. reward aspect of weak armor. Instead weak armor began to fade into the background and the stats began to favor the other two abilities. Then the movepool stage. This only amplified every mistake we had already made. Not only was it OP, taking away all risk on the spot, but it covered up the stats that were purposefully made subpar so weak armor would be useful by adding quiver dance. So basically, when weak armor was deemed too risky was when the project became unsalvageable. There were, as others said, already problems, but this is where it started to get out of hand.

[ganj4lF]

I agree with the most of you, problems started with the ability selection and were amplified by the inclusion of some moves, specifically QD, into the movepool.

CAP4 was the first CAP project I followed somewhat actively, and after Weak Armour was selected, I had the clear impression that selecting three abilities was considered "mandatory" or something (and found myself defending No Guard, which in hindsight was a big mistake, but I perceived that a "lesser evil" compared to Illusion, that has basically no risk involved at all). To me, selecting other abilities outside of Weak Armor was a big mistake (I thought that when CAP was being voted on, and I still think that now). CAP4 Weak Armour is very risky, thus very fitting to the concept, but any other ability would be more attractive to a competitive player, since the risk factor is not typically what you seek in a competitive mon. We also tailored good stats...for a Weak Armour pokemon, completely forgetting that said stats with Illusion / No Guard would be much more useful (no Def drops, ability to set-up more freely thanks to Illusion) / borderlining broken. We basically made the CAP around the weaker ability, blessing it with good physical durability and attacking stats, and ignored the other ones, that obviously took advantage much better of the stats we gave it to balance out the "riskiness" of Weak Armour.

Then, the movepool just amplified everything else. Again, as an inexperienced CAP user, I argued for an unreasonable move to be included (QD). We maybe thought that the typing alone was enough to make CAP4 risky enough, or...I don't really know why I thought QD was a good idea here. I have a feeling that this screwup is still tightly connected with the Abilities one, but that being said, giving QD to it (DD was fine IMHO) just amplified everything to huge levels.

[Rayquaza_]

About its defensive typing, I'd argue that it's not bad at all, in fact it's pretty damn good for the role Aurumoth plays the most: boosting sweeper.

The last thing a boosting sweeper wants is to be weak to common priority moves (Mach Punch, Bullet Punch and Ice Shard). Not only is Aurumoth not weak to any of those, but it also has a wonderful 4x resistance to Mach Punch, the most widespread one.

With exception of Tornadus-T's Hurricane, it also is not weak to any type common scarf users run (ice, water, dragon and electric - those are relatively "safe" attacking types to be locked into), making it very hard to revenge kill even after a single boost.

It might be SR weak, but boosting sweepers aren't as affected by SR as walls and choice users because they are supposed to switch in once, setup and sweep.
In practice one could pretend Aurumoth had 12.5% less HP while not being SR weak, which means it would still be pretty damn bulky on the physical side.

[Legend13]

Even though we never found an exact definition of risk, we accidentally found it in the Primary Ability Discussion, in the form of Weak Armor.
I remember all the hype on Weak Armor; it was a perfect example of a risky ability. However, it was overshadowed completely by illusion, and No Guard had absolutely NO justification, being an ability that eliminated misses. because of this, we had two very low risk, but very rewarding abilities, that completely overshadowed Weak Armor. The final nail in the coffin was Dragon Dance and Quiver Dance, which boosted speed, eliminating the need to use Weak Armor entirely.
Our greatest mistakes were in the secondary and tertiary ability polls, which, undid our success with Weak Armor, Which was worsened in both the Move discussions (Also the greatest mistakes of the project, next to the above), which were ridiculous; Attacking moves complementing No Guard, Boosting moves complementing Illusion - not once was Weak Armor considered.
Because of this, our risky pokemon turned into a broken set-up sweeper.
We must avoid the same mistakes we made in the Ability Discussions, namely forbidding abilities for not "following the concept in the right way".

[capefeather]

I agree with Deck Knight in pinning Concept Assessment as the stage with the biggest failure and threat discussion as the second biggest. Korski basically most of I would have wanted to (and he even pointed it out to me), while the rest I already said in the latest Smog issue. When people were saying "no counters", we should not have stopped there and we should have identified strong, reliable checks. However, I stand by my intention not to make this project about luck, because when people went into the playtest, I didn't want people to get angry because of hax, but rather, I wanted people to be forced to blame themselves for the consequences of their own decisions. I found that a much more interesting prospect than a study on hax.

I do have an idea for every stage, namely:

For typing, I initially had a very different idea from what actually happened. I felt that a typing with immunities (both offensively and defensively), priority weaknesses, and some safe aspects such as resistance to Stealth Rock would have been a good way to "ground" the project. I felt that a Pokemon that you could switch in over and over again, but whose decisions after switching in mattered every time, would have been more manageable and focused from a process standpoint than a typing with big weaknesses to U-turn and Stealth Rock whose aim was to go all in. A few weeks ago on IRC, I and a few other people massaged this idea to come up with Normal / Ground, Flygon stats, Guts, with Wild Charge as its only significant non-STAB coverage move (well, maybe less Attack because STAB Guts Facade is SCARY). I'm not convinced by bugmaniacbob's assessment that something like Guts or Flare Boost would have been boring because it doesn't really exist in OU and we'd be able to get a better idea of how it could thrive.

For the ability stage, as I said in The Smog, I think that all of the abilities had good justification, but only one should have won out. Weak Armour was great in principle (I personally disliked how I had to consider high physical bulk because of it, but other than that...). Illusion was weird, but I still think it made for a good experiment. I still think No Guard might have been cool by itself, turning threats like Politoed and Terrakion into reliable checks (I never really cared about inaccurate coverage moves). I partially blame myself for focusing so much on the principle of it instead of focusing on making reliable checks.

Stats were weird because I (and I'm sure others) felt so much pressure from so many differing views. In particular, I did not care about mixed attacking because being checked by Chansey is not exactly a big deal, yet I felt I had to up the Attack to get votes/slated. I also never really considered boosting moves; I didn't really want boosting moves at the time, and I didn't really expect them to be a big point of contention. If I were to redo the stats without any pressures, I'd probably go with something like 110 / 50 / 99 / 117 / 30 / 94.

As for movepool, I think that the coverage moves were excessive on most of the submissions. Out of the significant coverage moves, any one of them would have been sufficient. I probably should have voted for jas's movepool in the last poll, heh... Not to toot my own horn, but I still like my submission for this.

Also I love Rhys DeAnno's assessment of Gengar, which managed to discuss Substitute of all things in a way that never really happened during CAP 4's discussions.

[Wolgfang]

How about we implement beta testing? I think that auromoth was a great mon that scrapped simply because the CAP community got a little happy. It's not obvious enough that the mon is broke after a quick glance at its stats/abilities/movepool? Let a playtest drive that point home. With a few minor revisions, I'm sure auromoth can be made into a viable Pokemon that in some way realizes its concept.

It took sixty days of hard work to hammer out this failure. Let's not let that time go to waste.

[Eol]

Originally Posted by Fat Wolgfang

How about we implement beta testing? I think that auromoth was a great mon that scrapped simply because the CAP community got a little happy. It's not obvious enough that the mon is broke after a quick glance at its stats/abilities/movepool? Let a playtest drive that point home. With a few minor revisions, I'm sure auromoth can be made into a viable Pokemon that in some way realizes its concept.

It took sixty days of hard work to hammer out this failure. Let's not let that time go to waste.

This, so much.

I know my low postcount betrays me, and while it is true I didn't actively took part of the project, I sure lurked a lot during CAP4. To me Aurumoth is a pokémon with great potential, but as Wolgfang said, I think people went somewhat crazy in the end, because hey wanted Aurumoth to be good so badly.

And it is not all BMB's fault - while it is true he may have taken the wrong path, so did a majority of the people involved in the project. Instead of blaming, we should try and see how we can correct those mistakes.

EDIT: The elders have spoken.

To answer the initial question, I think the stats wouldn't be that much of a deal if the abilities (Illusion and No Guard, of course) as well as the movepool were reasonable. (Flare Boost seems more than appropriate if you're in dire need of examples).

[Pwnemon]

We undid the revamps of pyroak and we are never going down that road again. Please stop mentioning them

Also, we do do a playtest, Wolgfang

[CiteAndPrune]

Revamps aside, this topic exists so that the errors of Aurumoth can teach us something useful for later CAP Projects, so we (hopefully) avoid making those mistakes again.


But since those revamps were so long ago - what was so bad about making changes to an already completed CAP mon? What was the worst problem with the revamp, so much that Pyroak was reverted to its previous condition? If problems with it were identified and addressed, then what... or did the revamp bring with it brand new kinds of problems? Or did it just fail to fix what needed fixing? Or....

A little background on those times would be nice. Looking back might refresh another useful lesson, especially for the newer CAP members.

[Deck Knight]

To this day I still give Umbreon Dan a ration of shit because of how revisions turned out. Essentially the original revisions were as close to mindless power buffs and in some cases completely overhauled how the Pokemon functioned.

Pyroak was the worst offender - it got a lot more offensive but kept its bulk, and essentially started slaughtering the metagame with Dragon Dance Rock Head Flare Blitz/Wood Hammer along with EQ or Stone Edge. It hit like a truck, had few exploitable weaknesses, and excellent bulk. Revenankh basically got a straight stat boost. There were other changes here and there, but those are the major ones I remember. Basically the problem was we couldn't retroactively apply a concept to those CAP creations.

Even with a concept on Aurumoth, I think it sets a bad precedent, even if or when we completely change the leadership structure. Aurumoth is a discussion we've already had (however poor), and there is no way to get the interest in fixing it that there was when we initially built it. CAP Revisions have always been a rabbit hole that we have invariably regretted going down, so we killed off the idea for good.

[srk1214]

Notable among the revisions that have actually remained until today:

Giving Syclant Earthquake. Megahorn was also given, but was later taken back, as you can even see in the Swords Dance Syclant on-site: "X-Scissor is the primary STAB move of choice, as unfortunately, Syclant no longer learns Megahorn." Even now, the lone remaining meaningful addition Earthquake singlehandedly makes SD Syclant viable.

Stratagem got Trick. I don't think anyone would argue that wasn't helpful.

But really, most of the revisions were very minor. And anything that was deemed overpowered was later removed again anyway. It was largely a waste of time. And I'm very much of the opinion that CAPs exist in the meta they were created for. Updating Gen IV CAPs for Gen V with new TMs and Tutor moves etc is not a good idea.

For the same reason, we have to consider Aurumoth in the meta it was created. Genesect existed, and Aurumoth was hardly broken in its own meta. Good but not broken. We built Aurumoth to function in OU as it stood at that moment, which included Genesect. How Aurumoth functions in the CAP meta now, which doesn't have Genesect, since OU banned it, and also has all the other CAPs is entirely irrelevant.

What’s that? Ah — Revisions? Don’t talk about — Revisions? You kidding me? Revisions? I just hope we can make a CAP! Another CAP? (brownie points if you get the reference)

[erisia]

I wasn't particularly involved in the discussion for this CAP, but there was one point of contention that i'd like to address; the stat spread discussion. It can be said that giving Aurumoth exceptional physical bulk reduced the risk in using the bulky quiver dance set significantly, and even without investment, Aurumoth can tank a lot of attacks that /should/ kill it without worry (eg Scarf Terrakion's Stone Edge) In my opinion, the main fault was the overt restriction of stat spreads to rigid guidelines. Why did they confine us to making a pserudo legend automatically? Why did all of the poll options have over 100 base HP, which was a strict stipulation that has hardly any relevance in Aurumoth's use? Overall, i think that more diversity of slate options for this stage would have been particularly welcome.

[capefeather]

If nothing else, the main reason we don't do revisions is that the prospect of revisions causes people to take on the attitude that any "mistakes" we make can just be fixed later.

[TordenOfItami]

I was following the project quite closely, and I even had the feeling during Aurumoth's creation that everyone had a relatively different interpretation of the original outline. Risk is objective. And while I attempt to stay clear of criticism of bugmaniacbob, I feel that advancements forward in whose opinion makes the cut were vastly varied, and we began building on a base that was slowly straying from the original outline. This may even be nobody's fault, but it is true in building that small imperfections early in your project lead to large disproportions. Dare I say, but, perhaps increasingly throughout Aurumoth's creation, we tried to "compensate", ultimately leading to an even quicker rate of alienation. In such ways, I suppose our mistakes were in carelessness. I don't believe any major changes are necessary to the CAP system, etc. I just perhaps, suggest we all proceed with more caution.

[0urher0nik0]

Well, i don't often partake in the creation of pokemon, but i come by every now and then to see what new has came up in CAP. My first thought when i saw the new pokemon, was the idea of "risk taker," and yet the pokemon had AMAZING physical stats, and quiver dance.

With stats like that, whats the risk?
You can invest in physical bulk and hp, and then use quiver dance to cover the lack of speed, special d, and special a.
Not much of a risk.
No guard covers all chances of missing, so there's no risk in taking high powered moves over low powered.
It learns both overheat and hydropump, so it works great in both weathers.
It learns fantastic physical moves, as well as has a high amount of physical attack, so special walls, like blissy or chancy, dont stand a chance from a close combat or a megahorn.

It looks less like a risk taker pokemon, and more like you wanted to make a new hyper-threat...

I would say, lower the defensive stats quite a bit. You need to find an opportunity to send him in. At the moment, its bulk makes it quite easy. Especially with over 404 hp.
Its movepool is too vast, taking away its weather advantage moves, and limiting either its special or physical movepool would be ideal.
And less low accuracy moves.

So, from my scrounging around and short term studying...

I would say its stats were one part, and then movepool.
I wouldnt say both individually broke it, but both together did.

Maybe the opinion of an outsider may shed a little light.

[420]

This is constructive criticism not a troll: Too many people ended up getting their way who shouldn't have. And I honestly feel like bugmaniacbob and let his bias seep in which is what caused it.

Rather than have the project evolve naturally with a bunch of different people giving their opinions and having the bad ones weeded out through VOTING, bugmaniacbob essentially only listened to a vocal minority of people at important stages of design, and those people happened to be wrong. Whereas normally this wouldn't affect the process too much because the wisdom of the crowd's vote would shut them down, I believe certain biases allowed their ideas to persist longer than they should have. *cough noguard* These biases also caused certain ideas that could have been good and were popular and well liked among the majority to get shut down before even getting to the polls.

That is just my opinion anyway. This post will probably get deleted anyway because "hurrr he said bad things about a respected member of the community must be a troll!!" when I'm really just trying to help.

[Verminator]

I got involved fairly late with the project (around the movepool stage) but even at that late stage, one thing I noticed was that there was a lot of conflict between wanting to fulfill the goals of the project (create a risky Pokémon to play) and creating something that would be viable in the top echélons of OU. I think the problem with that was that the definition of what risk is is pretty broad. I think we could benefit from the next CAP having a more focused concept, or at least one that's further focused during the assessment-of-concept stage.