Policy Review Cascading CAP Processes

[Birkal]

Cascading CAP Release Processes
​

This is the final of a series of three policy review threads about the future of the CAP process. CAP moderator Dogfish44 is leading the first on Stage Adjustments and Quality Control. The second thread is about releasing our Pokemon at several instances in order to make a cohesive process. This one will conclude with a proposal that has been cooking for a long time. It's the one I'm most nervous about, and also the one with the most pre-conceived baggage. Let me bring you up to speed.

--------------------------------------------------------------------​

The Problem with One CAP

It's difficult to describe the slowing of CAP's processes succinctly, but it's certainly noticeable. One only has to look at the eleven projects we created in Gen4 and compare them to the six of the previous generation to realize that we have significantly slowed as a process. This is not an inherently bad thing. Our creations are more cohesive than those in the past, with extra stages that are specifically implemented to make sure that our Pokemon comes out resembling what was in the concept. Furthermore, we have an entire CAP metagame to worry about now, which involves a heightened sense of worry about how balanced our CAPs will be in their metagame. Yes, I prefer how we're set up currently, because we're having more fruitful discussions and more thoroughly baked designs. But I think the majority of CAP users would agree that in the back of their minds, they wish we made just a few more Pokemon. Perhaps we could cut off just a few days here and there to squeak out another design.

At the end of the day, we can't, unless we do something drastic. What I'm proposing here, it isn't drastic. Many of you who have been following CAP might come with some horrible visions about what a cascading CAP process could look like. You might look feverishly back on CAP25, when we made three Pokemon simultaneously, and groan about how difficult that process was. All I ask is that you look at the calendar I'm detailing below with an open mind. Because if we move forward with it, I estimate that we can cut at least two months out of every CAP process, leading to more discussions, without necessarily losing anything in the long run. How is that possible? Read on.

--------------------------------------------------------------------​

Proposal: Cascading CAP Processes

Let me provide you with a brief calendar of what this process would look like. The idea behind cascading processes is that we would start the next CAP during the final stages of the previous CAP. Here's what the calendar would look like:

• Start Topic Leader Nominations (14 days) after the conclusion of Art Poll 4 of the previous CAP.
  • These 14 days will coincide with Sprite Submissions, Moveset Discussion, and Name Submissions.
  • Topic Leader Nominations are traditionally some of the lowest trafficked threads in CAP. Rightfully so; while it's an important community aspect, it is of direct relevance to very few contributors.
• Concept Submissions (14 days) for the next CAP will start after the conclusion of Sprite Submissions, Moveset Discussion, and Name Submissions.
  • While this might conflict with some polling, the idea here is that the previous CAP would be entirely done competitively before the start of any competitive steps of the next CAP.
  • In tandem with Multi-Tiered Releases being discussed in the second PRC thread, we’d be hitting concept submission, voting, and discussion during Pokedex, Secondary Ability, and Movepool Submissions.
    • This is key, because all of our competitive, metagame-based contributors are NOT double dipping with two CAPs. They would be contributing to the next CAP strictly after competitive movepools are created for the previous CAP.
    • Also of note is the fact that our flavor-based contributors typically have very little to contribute during TLT nominations and concept submissions. This gives them plenty of time to still focus on the completion of flavor aspects of the previous CAP.
• Typing Submissions (14 days) for the next CAP would start at the Release 2.0 of the previous CAP.
  • At this point, the entire previous CAP will have been released to the public in an official capacity, meaning we are ready to fully devote ourselves to Typing Discussion, which is typically one of the most debated stages of CAP.
  • As an aside, this final timing has the advantage of hitting one of the most popular and accessible stages of CAP (typing) right during a Release 2.0, when our advertising would be highest.
• Stages that would be unaffected by cascading:
  • Typing Submissions and Polls
  • Ability Submissions and Polls
  • Art Submissions and Polls
  • Stat Submissions and Polls

--------------------------------------------------------------------​

Conclusion

The obvious advantage to this proposal is that we save about THREE MONTHS MONTHS in the process. I'm calculating we snip not only the 42 days we spend of actual cascading, but the additional month we spend over Policy Review, and general release waiting. You didn’t read that incorrectly about the timing, by the way. Using CAP23’s Art Poll 4 conclusion (October 15th) and CAP24’s Topic Leadership Nominations (January 30th), we would save three entire, human months by starting the process this early. I couldn’t use data reliably from CAP25 due to its celebratory nature, and CAP26 into CAP27 was unfair due to the transition of a generation and the updates we had to do. But at the end of the day, the biggest advantage here is that we're simply doing what we love to do with more frequency. I know I'm at my happiest when CAP is actually making a Pokemon, and I think many of you would feel the same way.

This also grants ALL CAP contributors more opportunities to contribute. The three months I’ve listed don’t feature any meaningful CAP contributions for competitive people. While these flavor stages are extremely important, they offer nothing to this demographic. Once we've completed our competitive movepool and gotten Release 1.0 uploaded onto Pokemon Showdown, it's just testing at that point. In this proposal, competitive contributors might do a little double dipping with playtesting and concept submissions, but neither require more attention than typing, stats, abilities, or movepool. As for the playtesting and battling, we're doing that all the time anyways. CAP has proven time and time again that if you sequester time for us specifically to study the metagame as the ONLY step in CAP, it usually results in us playing the metagame just as much as we have before with no significant increase. Tournaments, CAPTT, and other laddering efforts are what really gets people involved, not holding the calendar hostage.

Furthermore, this proposal gives flavor CAP contributors more opportunities to contribute and win. Particularly for artists, you must consider that having more CAPs released gives you more opportunities to win a poll. We have so many phenomenal creations that it's a shame we don't feature more art AND more artists. And with more art comes more opportunities to contribute names, movepools, pre-evolutions, sprites, and models. It allows us to make more Pokemon, which is inherently a good thing for flavor contributors. And as a result, this is more good optics from social media, which gets to praise about CAP a lot more frequently, AND always gives new contributors a new way to jump into the process.

So with all this good, why would we not implement this? I think the biggest preconceived notion that will trip veterans up is the concept of burnout. It's certainly something we saw with CAP25, and people want to avoid us overworking ourselves. If you feel that my proposal would give burnout, I highly encourage you to read the structure of how exactly I am proposing the cascade; no significant demographic of contributor is ever going to be handling two major steps at once. If you double dip into competitive and flavor, that could make you quite busy, but you're also not required to actively participate in every stage of every CAP. Similarly, artists are not required to draw for each CAP; if they're feeling burnt out by drawing a design every four months (as opposed to six months), then they're welcome to sit one out and take a break.

You also might be concerned about how we will staff such an endeavor; this calls for essentially twice the amount of TLs and TLTs. That's not entirely true for the reasons I listed above relating to no single stage overlapping itself, but general leadership is a concern. The only rebuttal I can provide is anecdotal: CAP was scoffed at by some when we switched to the TLT model, foreseeing that we would never be able to scrounge together five contributors to lead a project. As it turns out, we were able to every single CAP since the TLT's implementation, and I feel that we've had great success with the volunteers we've gotten from our veteran community. The CAP staff will work hard to make sure the forum stays clean and focused, so that newcomers know exactly what's going on with a CAP, particularly when a cascade is actively happening.

Please let me know your thoughts below. I am eager to hear what you think about this proposal, and if the advantages can outweigh the potential pitfalls.

 

[Mova]

I'm all for cascading CAP processes to make them go significantly faster, but I don't have much else to say about the matter. Pretty sure it will go quite well for the most part. The previous CAP's flavor can overlap with the next CAP's concept stage and it won't hurt anything.

 

[MrDollSteak]

I would like to echo support for this proposal. I agree that it's not without some potential pitfalls in regards to burnout, but I personally find that this will be less damaging in the long run than periods of inactivity. The various long gaps that I've seen in CAP since the start of my participation in CAP 25 have seen various contributors leave due to a lack of projects to work on, and low traffic for the metagame at large. I agree that the more CAPs we can work on the better. Flashcaps are fun, but are in my opinion a pale imitation for the real CAP process, especially in periods of low activity.

I'd like to pass on a brief point that Zephyr2007 made, which was that a new user arriving in the middle of two processes may be overwhelmed. I think that this is a valid consideration but would like to refute it, at least in so far as my own personal experience. I joined CAP in the middle of the CAP 25 process, and due to the sheer volume of threads that were happening, sprites, names, Pokedex entries for three different CAPs, I actually found it less daunting to post submissions, because I was able to selectively cater to ones where I was feeling inspired to contribute, without feeling guilty about missing others, for example I submitted Snaelstrom's winning sprite, and attempted a sprite for Smokomodo, but ignored Caribolt in this instance.

 

[Quanyails]

I'm interested in hearing where the following parts of CAP fit into this proposed timescale if we start the next CAP after the proposed beta release and without a real break between cycles:

• CAP metagame updates
• Tournaments
• Pre-evos
• PRC cycles/policy updates
• 3D modeling (tagging QxC4eva)
• CAP generational updates

Do you have a Gantt chart or a timetable showing where these would fit in the timeline of things? I'd have fewer concerns about how the timing would work if I could see how these fit in. Also tagging Dogfish44 in case this is relevant to him.

--

If the intent is to shorten time between CAP cycles, we could also consider timeboxing. We have a lot of data in the CAP archive, and we could examine where the most dead time is in past forum threads. We're currently timeboxing CAP 27 prevo to keep it on schedule with CAPPL and to prevent it from potentially overlapping with the next CAP. This is a topic for another thread, though. (Personal note: coordinate how much time we schedule for each CAP based on weeks, not days, as well as the school year cycle.)

 

[Bughouse]

I agree with Quanyails' concerns, though I would particularly focus on impacts to metagame updates (including the separate thread about 1.0 vs 2.0 releases, since that might include significant movepool changes) and PRC/policy. The lack of downtime between projects seems to impact these two most heavily.

I think the other points are relevant, but I'm less concerned about them for various reasons:

• generational updates occur less frequently (and I think we'd aim for that to be a break anyway where all active projects have to finish and then we restart the cascading from 0 in the new gen).
• tournaments in other areas of the site frequently need to change banlists for new rounds due to metagame shifts.
• prevos and models are inherently flavor and can have their own scheduling updates as necessary, but should not drive the decision making here if other factors strongly suggest overlapping/cascading would be beneficial.

 

[GMars]

I’m proxyposting for Jordy:

After taking part in building up the CAP metagame to something worth taking seriously and reading this proposal, I find it important to ask two questions. How will this affect the CAP metagame? Will it make it worth it to invest any time into the CAP metagame?

The first question is rather simple; it will overwhelm the CAP metagame by churning out Pokemon at a very quick rate. I don't think there's any way to deny this. The second question is definitely an opinionated one, but if CAP is going to churn out Pokemon so quickly, I very frankly do not think it is ever worth investing time into the CAP metagame and its resources.

This may be a "good thing" for CAP's creation process, but I can only see it as a negative thing for the CAP metagame that so many people have put a lot of effort in over the past two years. It really pains me to see the CAP metagame getting such a short end of this stick. I would urge everyone to think about this and its side effects thoroughly.

 

[jas61292]

So, let me preface this by saying that I'm one of the few insane people out there who actually kinda enjoys the so called Policy Review stage we have between projects. Yes, I'm a policy nerd and and actually have fun with threads just like this one. But I have also enjoyed that phase precisely because it was a bit of downtime. I love CAP, and I love participating in CAP projects, but when you are a leader, and not necessarily in the sense of holding an actual leadership role, but just a leader in general in the CAP community, it can get a bit exhausting. Having some time where your participation in the project can be limited to looking at cool art or reading funny Dex entries is nice. So the idea of cutting out all that time and pretty much starting up the next project the moment competitive stuff is done with the last project feels like a bit much to me.

However, while I enjoy a bit of downtime, we do have way, way too much of it. While I'm not sure I could support cascading projects as proposed, I would be all for some measures to shorten, or even eliminate the gap between projects. I personally don't think I would like two projects overlapping in any sense, but I think it would be great for the project, and especially for attracting newcomers, if there was always some part of a project ongoing. Just starting one as soon as the previous one is finished would be good. The combination of flavor steps at the end and nominations at the beginning are more than enough downtime for me.

And as for my beloved PR stage... well, as long as people actually post in PR when it is needed, there is no need for a dedicated stage for that. Though I will fully admit I have a slight worry from past experience that people won't post as much without such a stage. But that is a lesser worry.

 

[Mx]

I think this is a very interesting proposal, but I'm still rather skeptical about it. Cascading CAP Processes would undoubtedly help with activity problems that sometimes plague the whole project, but I think this proposal underestimates how demanding a project really is. In my experience, every CAP discussion requires a lot of investment, as many times one needs to calculate stats, test various sets, investigate past metagames, weight the pros and cons of very different proposals, and many contributors have to handle that on top of tournaments, art submissions, and real life obligations, so trying to follow two processes simultaneously would be very difficult for many people and outright impossible for some.

We might not have previous experience with this exact method, but I believe the starters process show what happens when we divide our attention too much: Snaelstrom always struggled due to its mediocre typing and lack of recovery, Smokomodo is viable but it never uses the ability its concept was supposed to be based around; Caribolt is probably the only one I would call a "success", but even then, it always faced a lot of competition from other Grass-types like Necturna due to how little defensive utility it provides.

The cascading CAPs shouldn't be as demanding as the starters process was, but to be honest, I still believe with the current proposal it will lead to a noticeable drop in quality. The overlap between Moveset discussion and Concept submission would be the biggest offender here, as both are key steps in any CAP that require a lot of attention, the former being the last step competitive step, where a lot of testing is required to see what works and what doesn't, while the latter is the foundation for the rest of the project. Frankly, as someone who would like to contribute to both discussions, having these occur simultaneously sounds like a nightmare.

The metagame is another worrying issue about this proposal. Even at our current pace, the DLC releases and bans (both from OU and from CAP) have made Gen 8 a very unstable metagame. This kind of meta might be fun for a while, but it really makes it hard to develop and maintains resources, we still haven't updated the Viability Rankings after the DLC, and it was agreed that we still need a few more weeks of development before we can come up with something stable. Work on analyses is also almost impossible, as by the time one is finished, the metagame usually changes completely. This proposal would only make things worse here, as it adds yet another layer of complications to the process. That being said, I believe that this could be somewhat mitigated if we tried to establish a stronger tournament schedule. Even though this doesn't help at all with maintaining resources, a larger number of games would help paint a better picture of the meta and make it more stable.

Overall, while I can see the benefits of it, I really can't support the current proposal, as I believe it would be too the very exhausting to keep up with both projects, which will ultimately lead to inferior final products and also it destabilizes the metagame.

I think that if we instead scaled this back a bit, making TLT applications run while discussing full movepool, and Pokedex Submission, then trying to make the Concept Submissions begins with the final release of the previous CAPs. This should create much less strain on contributors. With this change, I think it might be worth giving this a try during CAP 28/29, and then make another PRC thread to reflect on how it worked and how we should act in the future. The problems with the metagame would unfortunately still persist, but as I said before, I think more active tournaments can help this to an extent.

 

[Jho]

While many other people bring up how this would negatively effect the CAP metagame, as someone who came to CAP from the metagame side and has been working to get the meta more integrated into the CAP process, I cannot get behind this proposal. While it would undoubtedly give more opportunities for our contributors to, well, contribute, it completely invalidates the metagame as a competitive tier which kind of undermines everything myself and a lot of others have been working towards for the past 2-3 years which is to improve the competitive integrity of the metagame, and to try and make it a part of the CAP process (still a long way to go, but things like testing the CAP out in matches during the moves stage is a good example of how this can be done, and I think we can all agree that has been a very successful addition to the process).

Since CAPs are made to be viable in the metagame, it is impossible for a CAP release to not have an effect on the metagame, which means that every CAP release we have to learn a new metagame and update our resources to reflect the new metagame. If we take a look at the Contributions and Corrections forum for this generation, we can see that there has been a very large upheaval and change there as it adapts to taking on incremental releasing Pokemon and changing metagames. Even in the metagames that have the most contributors, such as OU, not every Pokemon had an up to date analysis before the DLC came along and invalidated pretty much everything and I can say with fairly high confidence that not all analyses will be done before the metagame shifts again later this year. While you can argue that OU got a lot of new Pokemon, National Dex is a good example of a metagame that essentially only got 1 new addition and this has lead to the National Dex resources being put on hold for almost a month now and still going as the meta changed so drastically and has ended up with 5 Pokemon banned. These are both metagames with vastly superior numbers of players to CAP and they are still developing and adapting to changes that happened almost a month ago.

With the timeframe given in this proposal, it seems to suggest that the metagame can be developed to a stable point AND have metagame contributors continue to output high quality resources, such as analyses, sample teams, and the Viability Rankings, in a 4~ week span after a CAPs release. Even if this is possible, you are asking the metagame contributors to push out work very quickly, only for it to be invalidated a week or 2 later at the best, if not have it be immediately irrelevant upon the release of the next CAP in the cascade. This means that the metagame will never be allowed to develop.

On top of all of this, there is also no concrete metagame to talk about when developing any CAP after the cascade starts, as accurate and up to date resources that reflect the position of the metagame will very likely not be available for the vast majority of the CAP which makes it very hard to know what to do in every stage. If you don't know what the metagame looks like, how can you argue for a typing combination without knowing what metagame threats it needs to check/break; how can you make stat spreads without knowing what hits it needs to tank, what it needs to be able to break, what is common that it will need to outpace etc etc. This may seem like I am exaggerating the impact of 1 CAP release on the metagame but I can assure you I am not - below you can find the Viability Rankings changes that occurred after Equilibra's release.

vr changes post Libra

Rises:
Tornadus-T S- → S
Celesteela A → A+
Landorus-T A → A+
Tapu Koko A → A+
Tapu Lele A → A+
Colossoil A- → A
Jumbao A- → A+
Mega Medicham A- → A
Toxapex A- → A
Gliscor B+ → A-
Mega Latias B+ → A-
Mega Charizard X B → B+
Gyarados B → B+
Tomohawk B → B+
Victini B → B+
Mega Charizard Y B- → B+
Mega Garchomp B- → A-
Mega Lopunny B- → B
Mamoswine B- → B
Manaphy B- → B+
Kommo-o C → B-
Slowbro C → B-
Mega Slowbro C → B-

Drops:
Magearna S → B+
Aurumoth S- → B
Heatran S- → A+
Mega Alakazam A+ → A-
Arghonaut A+ → A-
Garchomp A+ → B
Volkraken A → A-
Kitsunoh A → A-
Rotom-W A → A-
Tangrowth A → A-
Mega Diancie A- → B
Mega Latios A- → B+
Tyranitar A- → B+
Mega Crucibelle B+ → B-
Naviathan B+ → B
Smokomodo B+ → D
Clefable B → B-
Moltres B- → C
Mega Camerupt C → UR
Crawdaunt C → UR

===

Equilibra UR → S
Mega Gallade UR → B+
Skarmory UR → C

As you can see from this there are a LOT of drastic changes that can come from releasing a CAP, Equilibra is a bit of an extreme example but nonetheless as all CAPs are made to be viable in the CAP metagame, their impact on how the meta plays wont be too disimilar to this.

What really disappoints me about this proposal is that it is essentially dooming the metagame to a showcase for CAPs creations and I believed we had moved away from that recently, and this proposal doesn't acknowledge this at all. If this proposal goes through you are agreeing to the CAP metagame not being a competitive format and just a place to showcase our creations, which I cannot get behind.

 

[quziel]

Going to do a minor post here; Off-time is one of the ways that the meta develops best, and we can use the Cascading proposal to help us generate that off time without needing to kill the release rate. The current ideal of 3 months per CAP is entirely fine, and overall a rate that I think we should preserve. That said, long periods of a stable meta are highly favorable to the development of resources (VR, analyses, etc), and then to the creation of the next CAP, so we should prioritize them as much as possible. Instability in the meta doesn't fully invalidate these resources, but can heavily reduce their usefulness (imagine if CAP 27 was released post-libra but still was prepping to take on Magearna as the #1 threat).

My proposal is then as such:
Cascade 2 caps together, starting CAP2 Concept submissions after the competitive elements of CAP 1 are done (aka moveset submissions), and after CAP2 is released, take 2 months off in order to fully develop the meta.

If we assume 3 months per CAP as an ideal, with 1 month of PRC work between caps, for 4 months per CAP in the current system, with 1 month being "safe" to cascade (aka concept submission, voting, and beginning of assessment), then we have the following competing timetables:

M1​	M2​	M3​	M4​	M5​	M6​	M7​	M8​
CAP1​	CAP1​	CAP1​	​	​	​	​	​
​	​	​	PRC​	​	​	​	​
​	​	​	​	CAP2​	CAP2​	CAP2​	​
​	​	​	​	​	​	​	PRC​

M1​	M2​	M3​	M4​	M5​	M6​	M7​
CAP1​	CAP1​	CAP1​	​	​	​	​
​	​	CAP2​	CAP2​	CAP2​	​	​
​	​	​	​	​	PRC​	PRC​

(Apologies to mobile users)

This is just a simple mockup with very simplistic assumptions, but if we adopt the cascading model (again, ideally with a swiss-style release tour after CAP1), we can hopefully get 2 dedicated months for PRC and meta development with a similar release schedule. I don't have a full GANTT chart at my fingertips for the CAP process, and I'd like to see one of competing schedules from someone who does have one, but I am tentatively for this if we have a dedicated cool-down period after every 2 CAPs.

Notes:

• I guarantee I got the numbers wrong, but you can sorta see my point that if we adopt limited cascading we can buy more time for a dedicated meta development period.
• Depending on when cascading begins we could have a lengthy dedicated meta development period and increase the release rates of CAPs overall, albeit with the second cap in each batch being a bit less related to the meta; perhaps a good time to tackle weirder concepts?
• Please someone post a GANTT chart comparing the ideal for the current process and a fully cascading model.
• Assuming that full moveset doesn't really require competitive insight.
• The release of CAP 1 would harm the resources needed to develop CAP 2 here; given that a lot of arguments in the competitive aspects are essentially just quoting VRs and analyses, how can we best remedy this. I would suggest a very immersive tour maximizing games played, again, swiss as I suggested, but that doesn't get the VR updated quickly.
• All CAPs are, in some way, designed to take on major meta threats, as they are designed to be viable, and you must account for the meta when designing something designed to be viable unless you wanna go full Kyurem-Black.
• I am repeating myself often rip.

 

[Bughouse]

I think quziel's approach is probably closer to right, though I question if a semi-cascading process with a dedicated 2 month break is actually substantially faster than just doing 2 caps sequentially with a break after each for it to be worth it.

I guess there's an ancillary benefit that the last chunk of a project that is not competitive stages doesn't dwindle and disengage the people who only really care about competitive stages, but I think a lot of that is resolved potentially just by doing the staged releases that would enable the CAP to be available before all of that winds up anyway, so that seems to me to be the simpler approach if this sort of disengagement is the concern.

 

[quziel]

My argument isn't necessarily the increase in speed, its the fact that we can maintain the current speed while having a long stretch entirely dedicated for meta development; 2 months straight of meta development will hopefully yield far more info about the meta than 2 one month stretches. 2 Months straight of "off time" gives time for a tournament to be run, as well as a fair few analyses to be written, and the VR to be updated multiple times which 1 month of downtime doesn't necessarily give. Again, these numbers are arbitrarily chosen, but I hope the central thrust comes through.

 

[Mova]

Wasn't sure at first, but I definitely like this compromise to allow some down-time for the CAP metagame to stabilize with the two CAPs they would receive afterward. It would cause a bigger burn-out if we cascaded more than two after one another, so I agree with the roughly two-month down-time away from the process.

 

[Voltage]

My sentiments mostly stand with offler's post on this one. From what I understand about the CAP process, "This is a project dedicated to exploring and understanding the competitive Pokemon metagame by designing, creating, and playtesting new Pokemon concepts." The CAP metagame is an entirely integral part of this entire system and this proposal seems to go against that very idea. In my mind, it's as Offler said. A new CAPless than every two months gives a metagame absolutely no time to stabilize. I would have been far more for this proposal during Gen 7 when the closest thing we had to new mons was Naganadel and Zeraora with USUM, but with the volatility of DLC drops, OU bans and seemingly random hidden ability releases like Libero Cinderace and Grassy Terrain Rillaboom, it's really hard for me to be on board with the proposal in its current form.

Quziel's proposal is a little bit better imo, since at least there's a brief cooldown period. This cooldown period is another point that I'd like to really emphasize CAP has. Like others have brought up, as we saw with the starters, making many mons at the same time can lead to discussion burn out. This obviously goes for all kinds of projects, but if you dive too quickly into a new project after just wrapping up one, you're likely to burn out. I don't want to exaggerate, but I think this proposal is very reactionary, and it feels valid right now is only because had a monumental gap between Equilibra and Astrolotl. In my honest opinion, if we hadn't taken the break between those two for the new generation drop, I doubt this proposal would even be a thing. But returning to Quziel's proposal, if we are intent on moving forward with a cascading process, then there needs to be some time for creators and contributors to cool down while still allowing the metagame to develop.

This leads into a third point, I want to make. The CAP metagame as a whole is only going to get more increasingly difficult for new players to join in. Obvioulsy we can't just NOT make any new CAPS, but with the cascading CAP process, I feel as though the already small CAP playerbase, will stagnate in growth. Personally, I get turned off trying to learn new metagames at the start of new generations just because there's so much to learn in a very short period of time. You can honestly liken CAP to the DLC drops impacting OU. `30ish new mons that all have varying levels of viability. Sure, someone who WILL WANT TO PLAY is likely to take the time and learn the ins and outs of every CAP ever made, but for the player who is curious about trying a new meta, how do we make a metagame that is approachable that allows us to grow as a community in that sense? I agree that the cascading process would absolutely develop the contributor side of things, but I'd argue that there's an equivalent exchange with the actual players of the CAP metagame. It's a symbiotic relationship, and while I have been more of a vocal advocate for not forgetting the competitive side of CAP, it's clear to me that we also can't leave the creative portion hanging. It's just that if we overemphasize the creation process, there's no way that new competitive players are likely to start playing this new metagame.

Furthermore, the current cascading process doesn't seem to allow time for the reCAP process (or whatever the thing we decided to call it), where we rexamine whether or not the cAP we made was too strong or too weak. If we move forward with a cascading process of some kind, there absolutely has to be considerations of the new CAP mon after the product has been presented. I think back to Equilibra with Toxic and DoomPhazing. With the current cascading process, would anything have actually been done about those problematic items? From what I'm reading, not really. Please correct me if I'm wrong about this.

Again, I don't want to make it seem like we need to stop making mons, or that our downtime between new CAPs should be extended. I think that, honestly, if we could, I would be open to starting CAP28 tomorrow. But it's about finding a balance. I don't think this cascading process that has been proposed is the right way to go. Like quziel, the engineer in me would love to see a GANTT chart to see the current process. With refinement, I'd be more on board with it, but we are trying to explore and understand the competitive Pokemon metagame, and the current CAP cascading process does not allow for a healthy metagame to develop in the first place.

 

[Menshay]

Having a more detailed outline/chart of the current process would be definitely useful, possibly including timeframes for both discussion and polling stages (including overlaps, if any); so far I have only found the process guide.

That said, I echo the sentiment that a permanent cascade with little to no downtime would just turn CAP into an assembly line. If anything some of the lighter stages could be made parallel, but I cannot assess which and how much without more details.

 

[QxC4eva]

I like the proposal for what it strives to do but the implementation seems to go a bit overboard. There's more to getting a faster process than to simply take every non-creation part of CAP and lumping them together with other stages. That won't make a more efficient process, but a process that is sloppy, disorganized and cutting corners, and that's the issue I have with the proposal - its focus on quantity over quality.

There's been a lot of talk about how it affects the metagame, and I'm gonna talk about how it affects 3D modeling in a similar way. For modeling, our biggest enemy has nothing to do with time constraints but the burnout after each model we make. That's why we're lucky to be blessed with the insane amounts of downtime between caps. If the downtime is ever removed or drastically shortened there's gonna be a lot of new problems with burnout. As a leader, the wellbeing of my team is the top priority and will always come before meeting any deadlines. A continuous cascade of caps means we may have to skip a cap every two cycles to keep our health in check. I suspect though, that alone would be demoralizing enough to ruin any further motivation to get them done on time, where they eventually fall straight back into reservation and get picked up at our own leisure. Just like how we did it back in the pre Naviathan days. That would be undoing 4-5 years of work I think so yeah take that Jho I just upped the ante ;P But yeah it's not ideal to release any of our creations without a model, and have them as sprites on the simulator for an indefinite amount of time. I'm sure there's ways to adapt 3D modeling to the new system but it can only happen if we take it a step at a time, rather than several steps like the OP.

I think the best way forward is to find a sweet spot between the proposal and the current process. I know a lot of us including myself are heavily invested in a particular area of CAP, and when a new proposal jeopardizes the hard work we've done for it our instinct is to reject it at all costs (and understandably so). But regardless of which part of CAP we stand for I think we should also keep in mind what's best for the project as a whole. For it to continue growing there needs to be some sort of give and take for everyone, and I think the proposal has potential if we can put forward ideas to help address some of our grievances with it. I was going to suggest rather than cascading the next cap after the art polls, to wait out some length of time before doing so - will help to alleviate 3D modeling burnouts. But quziel brought up an even better idea, one that eases some metagame issues and also happens to work well for modeling - I like it, seems like a good compromise for everyone. It's still somewhat far from ideal for modeling (just as it is for the metagame I assume) but at least we've got a more realistic plan to work with. So yeah I'm on board with it. It's nice that people are not only voicing their concerns but also actively seeking ideas to make the proposal work, really good stuff so far.

 

[snake]

I've taken the proper time to gather my thoughts regarding this proposal. Out of the big PRC threads open, this one must be resolved first before the others can be effectively resolved. Thus, I will be focusing on this proposal only in this post.

I believe that Birkal's point about how it can be hard to contribute to CAP with the current setup has merit. However, as many others in this thread have pointed out, the Cascading CAPs Proposal as designed in the OP is unsustainable for the CAP Metagame. The CAP Metagame is a) the basis for all competitively driven decisions in the CAP Process and b) a metagame that many users, including myself, have devoted a lot of time into developing it into what it is today. Thus, it should come to no surprise that I oppose any proposal that will cause a huge hit to the CAP Metagame. If the CAP Metagame suffers excessively, then the CAP Process suffer with it; it's as simple as that.

The fact is that no matter how many CAPs enter the metagame or how balanced we make them that they will cause a metagame shift. A balanced, A-rank threat, will cause some Pokemon to fall and some to rise: it's the game we play. Assuming users were willing to play a metagame in constant flux, this would cause CAP to be unable to create accurate threatlists and other relevant competitive decisions, simply because there's not enough time for the metagame to settle. Thus, the CAP process will lose out on how it's supposed to be grounded in competitiveness, and that's not acceptable.

So, what can we do about this? quziel has a fantastic post above regarding the idea of having a Cascading CAPs + Metagame Development Cycle - 2 Cascading CAPs followed by 2 months of time for the metagame to develop. If well-timed, I think that this is working towards the best compromise. A regular schedule could be created, which solves the issue of "when is the next CAP?" A period of 2 months is reliable enough to get a good tournament in. Meanwhile, CAP is rather active for a longer period of concentrated development at a time, which solves the activity issue in the best way I can see without compromising the integrity of the metagame. I really like this solution.

Now, here's my new input (with credits to -Voltage- for our discussions regarding this topic). The issue with having even 2 cascading CAPs is that once the first CAP is released, it will cause a metagame shift. And again, this will distort our decisions for the second CAP in the cycle. Thus, what I propose is that we follow quziel's proposal but with an extra twist: The release of the first CAP of the cascading cycle is delayed so that both CAPs are released simultaenously at the beginning of the 2 month metagame development period.

Of course, this is why I'm focusing on this proposal on its own; this solution doesn't mesh all too well with the Multistage CAP Releases, but we can work around this complication. Although we will have to delay the release of the first CAP, it avoids how the first CAP will skew the second CAPs creation. Furthermore, we just saw how "releasing" just two Pokemon into the metagame isn't too overwhelming. When the Gen 8 starters gained their hidden abilities, Cinderace and Rillaboom were effectively introduced into the CAP metagame as new Pokemon (sorry Inteleon), and for the most part, the metagame was accommodating them decently before the DLC. Thus, I think it's fair to say that we could adopt this similar strategy: introduce two Pokemon at the same time in the metagame, both of which have been built upon the previous metagame and remaining as independent of one another as possible. Then, we have a stable, two month period for the metagame to adjust to them, and then we can begin the next cycle of cascading CAPs - a time when the metagame can continue to exist and thrive as well.

It's hard not to think back to CAP25, when 3 CAPs were released at the same time, and it seemed a little overwhelming to have to deal with three Pokemon entering the metagame at the same time. However, I believe that three Pokemon entering the metagame has a much greater impact than just two, in the sense that there's a increasing strain on the metagame the more Pokemon that are introduced at once. Thus, assuming that the CAPs are relatively balanced, adding one new Pokemon is pretty easy to accomodate for, adding two Pokemon is still a managable task, but adding three Pokemon is when the strain becomes significant.

The downside of this modification is, of course, we can't release the first CAP when it's done, meaning we don't get as much attention as we would otherwise. However, for the sake of a cohesive Cascading CAP + Metagame Development Cycle, I think it's worth considering. To be clear, I don't dislike quziel's proposal as it is, but I think the modification improves upon it.

 

[GMars]

The release of the first CAP of the cascading cycle is delayed so that both CAPs are released simultaenously at the beginning of the 2 month metagame development period.

The downside of this modification is, of course, we can't release the first CAP when it's done, meaning we don't get as much attention as we would otherwise. However, for the sake of a cohesive Cascading CAP + Metagame Development Cycle, I think it's worth considering. To be clear, I don't dislike quziel's proposal as it is, but I think the modification improves upon it.

I appreciate what this attempts to do Snake, but in my eyes this is a terrible way to try to adjust quziel's schedule to try to "compromise" between process and metagame. This removes any possible benefits this proposal could have in trying to keep activity drawn to CAP high due to the bump CAP gets whenever we release a mon from social media and elsewhere, and I predict it would lead to a strong anticlimactic feeling for the end of half of all future caps. This modification does not improve upon quziel's suggestion, it significantly worsens it.

 

[snake]

I appreciate what this attempts to do Snake, but in my eyes this is a terrible way to try to adjust quziel's schedule to try to "compromise" between process and metagame. This removes any possible benefits this proposal could have in trying to keep activity drawn to CAP high due to the bump CAP gets whenever we release a mon from social media and elsewhere, and I predict it would lead to a strong anticlimactic feeling for the end of half of all future caps. This modification does not improve upon quziel's suggestion, it significantly worsens it.

I definitely noticed this only as I was typing out the post last night (so I ended up just finishing out the post anyway)...even if we’d be compressing down a good bit of time in a CAP cycle in some places, a simultaneous release would have released be more sparse than they are now. I don’t think any of these solutions are flawless, but I can agree that any solution that tanks social media presence would be a bad one.

In any case, I hope my post conveyed that I still support quziel’s proposal as a better implementation of cascading CAPs than that in the OP, and that I don’t think continuously cascading CAPs is a good idea.

 

[quziel]

Ok, so had a few conversations on discord and in PMs, and have been thinking through a number of things with respect to this process.

I will argue that this PR proposal is not necessarily an argument for Cascading CAPs, but instead a question of when to start the next CAP. If, hypothetically, Astrolotl was released the day before CAPPL, we could probably start TLT noms at week 3 or 4 of CAPTT; however, as we are now, with CAPPL starting 1 month after Astrolotl's release, we can put up TLT noms at week 3 or 4 of CAPTT. That is, we need the following for a CAP to be produced. We need a stable metagame and updated resources, both of which require substantial high level play to develop. Because the CAP ladder has its issues, this high level play can really only be generated through tournament play. Our resources for generating this high-level tournament play are as follows: CAP Seasonals, CAPTT, and Release Tours. CAPTT is obviously the largest source of information we have, and syncs up fairly well with Astrolotl's release quite well.

My proposal is then as such;
Aim to have a tournament shortly after the release of each CAP, running PRC threads concurrently, and start TLT noms after a few weeks of this tournament play.

This scheduling would ensure that we can get the resources needed for the next CAP updated in time, wiht significant amounts of tour play informing VR and Analysis writing before Concept Assessment has begun. Numbers aren't my strong suit here, as I don't have a full estimated schedule, but my ideal would be that Concept Voting is conducted on the week the tournament is ending (either final round or just finishing), and Concept Assessment is done the week after the tournament is over.

Edit: I've spoken before about my like of a Swiss-Style tournament for this application, as it ensures that we maximize the number of games played, and means that average or new players will get to play (and hopefully win) a number of games without just being shoved out of the tournament after round 2 or 3.

Edit2: Electric Boogaloo; A lot of the need for this thread is driven by the large gap between Equilibra and Astrolotl, and now Astrolotl and CAP28, which are largely driven by the generation change (noone is gonna play cap when there's 50 new mons to mess with), and DLCs. We can help ourselves to overcome these issues by rigorously planning out the schedule for both the CAPs and the subsequent tournaments needed to have up to date resources and a developed meta.

 

[Birkal]

I want to start by thanking everyone who has participated in the conversation thus far. Seriously, I know firsthand how much of a pain it can be to go out of your way to post in CAP PRC threads. I wanted to address some of the concerns here and put forward a new proposal at the end of this post.

General

It can be a common reaction to consider your own specialty when it comes to CAP, particularly in response to this post. Sure, many users double dip in many facets of CAP, but there are also many who stick to one part of CAP, sometimes even one specific stage of CAP. Whether you come at it from a metagame perspective, modeling perspective, flavor perspective, and more, you are hearing the proposal in this thread of “less time.” And that idea of less time is inherently unflattering. Models take an extreme amount of time to build, rig, animate, color, and critique. Metagames take a long time to settle, as people develop new sets, checks, and counters. Everything we do in CAP takes time, and it takes time to do it well. I hear that concern and want to make sure that CAP continues to have a high standard of quality.

What I want you all to hear is that the cuts I’m proposing are not really that drastic in terms of time. Yes, we are saving a lot of process time, but I’m going to detail below how much time is actually being spent, saved, and how much down time you will actually have with my revised proposal.

--------------------------------------------------------------------​

Revised Proposal

This wasn't built exclusively by me, but from people on Discord. It's taking my proposed process in the OP, and moving the cascade back one step. That means we'd start TLT nominations during the final flavor aspects of a CAP, which include full movepool, Pokedex, and models. That would mean that the Release 2.0 would coincide with the introduction of Concept Submissions. It's more lenient on timing overall, and still gives a significant process part (concept subs) some good optics.

Please feel free to look at the visualization of this on this here Google doc, if you please.

--------------------------------------------------------------------​

Effect on the CAP Metagame

As per the above Google doc, it took us 144 days to make CAP27. That’s the addition of one new Pokemon to the metagame. I’d now like to demonstrate what shifts look like in other metagames that have taken place recently. I know that a removal is different than an addition, but it’s still a similar principle that the metagame needs to settle once there is a change. Take a look:

• SS OU: 22 days between Arena Trap and Melmetal
• SS OU: 40 days between Melmetal and Dracovish
• SS UU: 54 days between Mamoswine and Haxorus
• SS LC: 55 days between Rufflet and Chlorophyll
• SS Monotype: 58 days between Melmetal and Kyurem-B

Yes, they are not a direct comparison, because CAP has a lot more going on than a single metagame does. But I feel safe in saying that roughly double the amount (100 days) would give us enough time to settle down our metagame. That's 100 days to edit threatlists, update VR, and host at least a tournament. I personally think that's doable, and that's based on past data. We've been building for the CAP metagame since Kerfluffle and Pajantom, and both hit under the 100 day mark, and yet managed to have ample support from the growing metagame community at the time. We're much bigger now (believe it or not), and I feel that this is a good ballpark for us to work in. I do think we could do all three of those things within 100 days, but correct me if I'm wrong.

Finally, I'd want to highlight these two sequential posts by CAP admin DougJustDoug that he made five years ago. This is the PRC thread that changed CAP from making Pokemon for OU and into the CAP metagame. It was a massive discussion, and Doug's posts ring so true today. We had a CAP veteran back then named ginganinja, who was an OU mod alongside me during BW. He was obsessed with balancing metagames, and while he approved of the change in direction, he was deeply concerned about the balance of this new metagame. I'd argue we've done a pretty good job so far of trying to maintain a balanced CAP metagame. We've made several nerfs, and even tiered a Pokemon (Clefable) separately from OU.

I don't say any of this to diminish the work of our competitive CAP contributors. I truly think the world of everyone working on the Viability Rankings, doing analyses, and running and playing in CAP tournaments. These are all things I have supported in the past, and will continue to support into the future. My argument is that chasing a balanced metagame is a golden goose: it's something we should strive for, but will probably never achieve. Think about our process already. We accept OU's bans in the CAP process, which could change the entire creation of a Pokemon midstream. We've had that happen several times, and those Pokemon still turn out quite balanced (Necturna and Volkraken come to mind, for their respective metagames). We've had the opposite happen, where the metagame hardly shifted at all, and the CAP turned out needing nerfs; Equilibra is a very recent example of this. No, Equilibra was not a failure, because we learned a ton during its process, but it just shows how fickle balance can really be. It's a target to shoot for, not something we discredit entirely if we don't hit the bullseye.

None of this is giving the CAP metagame the "short end of the stick," as Jordy put it in his post above. Shortening the down time would have two results for the CAP metagame, and both I listed above: 1) we'd have to update VR, threatlist, and host at least a tourney in roughly 100 days, and 2) we'd have to be content with understanding that balance inherently unachievable, but is something we should absolutely strive for as we continue to play and test the metagame. I'm sorry if anything I said here is offensive, and comes off as me not caring about the CAP metagame. I care very deeply about both the CAP process and the CAP metagame, and a change like this won't be implemented unless both communities are in favor of it (or at least neutral).

quziel's Proposal

I like what this proposal seeks to do: provide a common ground with what we have today with what I'm proposing. I think it's great on paper, but that's the largest criticism I have with it: the stars would need to absolutely align for us to time it correctly. Think of us implementing it right now. We wouldn't be able to cascade CAP29 efficiently because of the DLC. We'd ideally need to start CAP28 after the release of DLC in November in order to make it work, and who even knows if something will come up or not between now and then. GameFreak has truly messed up Gen8, not just for CAP, but Smogon as a whole. I'm just not convinced we could use this pattern without tripping over ourselves. I'd rather pick a cascading policy, and then make exceptions as needed based on DLC and game releases by GameFreak.

Effect on CAP Modeling

I appreciate your post QxC4eva, and it sounds like you're generally in favor with some form of lessened cascading. I am hoping this revised proposal will suit us from a modeling perspective, but I'd like to hear your thoughts on this again. And also, please consider that in tandem with releasing CAP multiple times, we would have much more time and less pressure to finish a model each time. Mental health is a big concern, so I think if the modelers ever said "this is too much for us right now," we could happily play with a sprite until a model could exist someday, very similar to our past-gen creations now.

--------------------------------------------------------------------​

When Will CAP28 Start?

This is something we talked about on Discord for a big portion of today. And this is sort of what the conversation of this thread has turned into. I'm fine with that, but I hope it's made it abundantly clear how impossible it is to schedule the release of a new CAP. We could start TL nominations today, but it was brought to my attention that this would interfere with CAPTT (of which I'm a captain, and actively participating). That's a fair reason to wait -- CAPTT is a huge part of the CAP metagame culture that I personally love. A similar argument could be posted by the pre-evolution team, and that's totally fair too -- our pre-evolutions are superb and feature some of the best flavor work in our community. Another argument brought up is that the metagame has not yet settled to Astrolotl, and could use at least a month after CAPTT just for the metagame to balance around it. While I don't think balance is ever something we really can achieve, I appreciate the metagame tiering leaders for aiming at that target, so I think their request for another month to process is also fair. And then the DLC comes in November, which will completely shake up the metagame. It would be too much pressure to squeak in CAP28, and that's a fair opinion to have; no one likes a rushed CAP.

Do you see how all of these reasons are fair? It's because CAP has so many good things going on. It's part of the reason our CAPs have taken so long to make; we've got more quality community events and contributions going on than ever before. But you now also have to see that CAP28 would not begin until 2021 if we waited for every fair request. And maybe that's alright with you; if you think it's fine to wait until 2021 for us to start another CAP, please let me know. I personally find it horrifying, but maybe I'm not in the community consensus there.

This is why I think we need to establish from this thread some sort of general timeline that will dictate when we start a new CAP. I'm sick of trying to balance all of the great things we have going on with CAP with the mission of CAP: to make competitive Pokemon. It's just an impossible task, and whichever proposal we go with here should give us a general guideline on when a CAP should start. Of course concessions can always be made as necessary, but what determines what is necessary in the first place? At what point do we stop worrying and start rolling up our sleeves to play ball?

 

[cbrevan]

I have a number of concerns with cascading CAPs that have not been addressed by the revised proposal Birkal posted. A lot of this will be echoing what has already been brought up.

My chief concern is going into the competitive stages of a CAP process with just a month of testing since the latest CAP. Competitive discussion lives or dies by the general understanding of the CAP metagame. In a stable metagame where there is a consensus of what the biggest threats are, what the prominent builds are, and what sets are most effective, discussion can go deep as players and contributors are able to examine metagame trends fully with information backed by experience and evidence via tour replays. A stable metagame fosters a healthy enviroment for discussion that even those not fully up to date with the metagame can engage in. In an unstable metagame, such as the ones we'll be likely to see if we go with the revised cascading structure, discussion will be drastically limited by both a lack of resources and on hands experience with the metagame. Not everyone has the time to get acquantinted with a new metagame, which is why its important for CAP competitive resources to be in order during a CAP process. In such an enviroement, I would expect the quality of discussion to be lessened as the resources needed to back an argument are not available.

Another concern with building in a still stabilizing metagame is that there will be a discrepancy between the metagame threats and trends seen during typing and those discussed during the later stages of the CAP. We see this every CAP process as the metagame is an ever evolving thing, but the short testing period in between competitive discussions makes it more likely for trends from the previous metagame to still be seen as players are still exploring with what does or does not work. While this isn't necessarily a bad thing, I do think the CAP processes suffer in quality if the Pokemon we're targeting to beat at the beginning of a CAP differ from those we need to beat by the end of it.

As for burnout, I can guarantee we'll see it with cascading CAPs. Birkal has gone to lengths to emphasize that different facets of the CAP will be at work at any given time, and that no two groups will have to work on any two projects, but that doesn't take into account the ever present workload competitive contributors possess in keeping metagame resources updated. One month is a very short window for metagame resources to be updated in full, which means analyses, VRs, and other resources will most likely have to be updated in tandem with competitive discussion. When you factor that the teams responsible for these resources happen to include some of our most prominent and prolific competitive posters, it becomes clear that we're going to be asking our contributors to both participate in the newest CAP process as well as update the resources to ensure that process goes smoothly. That's a recipe for burnout that will be expressed either by a decrease in CAP participation or a stall in getting metagame resources updated, with either outcome working to the detriment of the overall process.

Now, I'm sure some of you have noticed that everything I've brought up relies on the idea that a single month is too short of a timeframe for the metagame to adapt to one new Pokemon. It's entirely conceivable that the new CAPs don't impact the metagame that drastically, such as what we saw after the starters dropped. However, if you look at CAP processes since CAP 21, Mega Crucibelle, you'll notice that every CAP aside from the starters have lead to significant shifts in the CAP metagame. Jho already touched upon this point with his Equilibra example, but I want to assure any skeptics that modern CAP creation has consistently released metagame shifting Pokemon. Mega Crucibelle and Kerfluffle were the two shaping factors of late gen ORAS, and all three of Jumbao, Pajantom, and Equilibra defined their release metagames. I can't speak for Astrolotl since CAPPL is still in week 1, but history has shown that the CAP community makes good Pokemon, which makes perfect sense given the nature of our process.

I believe that these issues will be present with any sort of overlap between processes, if just because of how much of the process relies upon competitive discussion. At best, we can host processes consecutively, but even then I worry about the workload we'll be putting on our contributors. Birkal spoke of compromises for the benefit of the overall process, but I truly believe that the only thing cascading CAPs will have us compromising is the quality of our project. Instead of something that can backfire as badly as this, I'd prefer if we examined alternative solutions, such as timeboxing and better scheduling, both of which have already been mentioned in this thread.

 

[Birkal]

The community consensus here is pretty clear. Cascading CAPs are not something we would like to pursue at this time as a PRC. While we are ditching this proposal, it sounds like many would still like to pursue avenues that cut back the amount of time it takes to make a CAP, such as better timeboxing and a more economical use of our time. I don't think I have what it takes to make such a proposal, but if you're interested in putting pen to paper to get some official timeboxing going, shoot me a message and I can give you the basics.

Thank you all for your participation and sharing your thoughts. Finally, I wanted to apologize to anyone who this PR thread offended. Please know that I made it with the best of intentions. I didn't seek to divide the CAP community, to downplay the importance of anyone's contributions, or discourage people from participating in CAP in the future. I was horrified to find that some people took it that way; I am sorry for the sloppy running of this PR thread in general. If you have any further beef, please contact me via PM and let's talk it out.