Process Guide Workshop for Create-A-Pokemon.

Status
Not open for further replies.

X-Act

np: Biffy Clyro - Shock Shock
is a Site Content Manager Alumnusis a Programmer Alumnusis a Smogon Discord Contributor Alumnusis a Top Researcher Alumnusis a Top CAP Contributor Alumnusis a Top Tiering Contributor Alumnusis a Top Contributor Alumnusis a Smogon Media Contributor Alumnusis an Administrator Alumnus
i see the same problem with using very good,excellent or whatever. People always naturally vote for the better ratings or better BST it happened here and happened with scylant. Then we end up with something that has so much BST or is aimed to be so good that we end up with overkill stat spreads.
It not really challenging to make a good competitive pokemon when your allowed to use 545+ stat totals every time.
I would like to see what people can do with 515 or even 505 for once.
What if I tell you that my system ranks Alakazam as 'Very Good' despite having only 490 BST, while Dusknoir is deemed only as 'Good' even though it has 525 BST. As you say, you COULD do good things with 515 or even 505 BST. My system would actually allow you to use any BST you want, as long as they are 'Very Good', 'Quite Good', 'Excellent', or whatever we want.

Yeah Gothic Togekiss, I wished I could do that earlier as well, but recently I was going through a busy patch, and actually I'm still not through it completely.
 
it amazes me how you come up with this stuff, this is some outstanding work.

Kinda worrying that Garchomp rating is twice as good as gyarados.

How do we implement this in the future CAP are we just gonna vote between "very good","excellent" etc and then just see what spreads people bring out. If not how do we narrow down options.
 
I say change the relevant polls to deciding whether it's:
1. Excellent, Very Good, etc
2. Biased towards offense, defense or no bias
2.1 How biased, if so (very, hardly, etc)
3. Biased towards physical, special or no bias
3.1 How biased, if so (very, hardly, etc)
 

DougJustDoug

Knows the great enthusiasms
is a Site Content Manageris a Top Artistis a Programmeris a Forum Moderatoris a Top CAP Contributoris a Battle Simulator Admin Alumnusis a Smogon Discord Contributor Alumnusis a Top Tiering Contributor Alumnusis an Administrator Alumnus
Moderator
I quite with everything that the current method for deciding on if it's offensive/defensive/or balanced isn't as good as we hope. I somewhat wished X-Act had came up with this rating system helluva sooner so we could gave it a good.
What? I don't know how many words are missing from this post, but I can't make sense of any of it.


As for using X-Act's formula -- I think it's a great idea. Stat totals don't mean a damn thing, and every competitive player knows it. We've been using totals because we did not have an alternative. Now that we have an alternative, I think we should give it try on the next creation.

I also like the educational aspect of using X-Act's formula -- and I'm not talking about the math. The development of this formula is a great lesson in what is important in base stats. We could probably use a good non-math-geek summary of the formulas, for inclusion/linking in the OP of the first Stat Rating Poll.
 
also how far along did the creating a new evo thing get, are you definetly running one.
I was wondering this too. Now that the initial art thread is over in the CaP, it could be a good time to try a CaPE (or EaP, which one looks different enough to not confuse people?).
 
What? I don't know how many words are missing from this post, but I can't make sense of any of it.
Funny you should say that because it seems that X-Act understood my last post enough to know what I was talking about?

How is CaPE more confusing than CaP? If you know what CaP stand for then it shouldn't be that hard to figure out what CaPE stand for. It's not like the "E" in CaPE stands for "Equipment".
 

X-Act

np: Biffy Clyro - Shock Shock
is a Site Content Manager Alumnusis a Programmer Alumnusis a Smogon Discord Contributor Alumnusis a Top Researcher Alumnusis a Top CAP Contributor Alumnusis a Top Tiering Contributor Alumnusis a Top Contributor Alumnusis a Smogon Media Contributor Alumnusis an Administrator Alumnus
I completely agree with Doug that I have learned massively from the development of this system too. What I learned the most is that it's not important that you have sucky stats when you have one stat out of PT/PS/ST/SS that is incredibly awesome. Alakazam is one such Pokemon, having incredible Special Sweepiness but crap everything else.

To continue with Mekkah, we could also state explicitly that we want, for example, Rank 2 Physical Sweepiness, or Rank 3 Special Tankiness, or whatever.
 
I say change the relevant polls to deciding whether it's:
1. Excellent, Very Good, etc
2. Biased towards offense, defense or no bias
2.1 How biased, if so (very, hardly, etc)
3. Biased towards physical, special or no bias
3.1 How biased, if so (very, hardly, etc)
Yes, this is simple and direct. I don't see a problem either.


And regarding those who doesn't understand the formulas (or doesn't want to bother trying), I think a chart in the first post of each poll with the current pokémon tiered (like the ones X-Act posted in the Stats Ratings thread) would be sufficient.

In the first poll, the chart with the general rating.
In the second, the chart with the offense/defense balance.
In the third, the chart with the physical/special balance.

People can just look at that, and think "I want the pokémon to have stats as good as Magnezone, as biased to attack as Gyarados, and as biased to attack as Rampardos", and vote to the corresponding ranks.

The only ones who will have to worry about the actual formulas will be those who want to propose a stat spread. And if among those there are some that don't understand the formulas, they can always toy with X-Act's applet until they find a spread that adjusts to the requisites.
 

DougJustDoug

Knows the great enthusiasms
is a Site Content Manageris a Top Artistis a Programmeris a Forum Moderatoris a Top CAP Contributoris a Battle Simulator Admin Alumnusis a Smogon Discord Contributor Alumnusis a Top Tiering Contributor Alumnusis an Administrator Alumnus
Moderator
We will do an Evolution project. We are no longer going to poll New Species/Evolution at the start of a project. I think the next project will not be called "CAP 4", it will be "New Species 4". And when we do an evo, the project will be called "New Evolution 1". Both of these will be under the umbrella of "Create-A-Pokemon". This makes it clear that this project is always involved in creating new pokemon, but we have two different flavors of projects.

The only question is timing. I don't think we're ready to do simultaneous projects yet. There is easily enough interest. But, from a process standpoint, this current project has made many stumbles. I thought it would be much smoother, since it is our third one. It's obvious we still have some kinks to work out regarding things like stats, art, and general scheduling.

I think I probably expected too much from this one. In many ways, this is actually our "first time". Not really, but it's the first time we've tried to manage the process in a very organized process-oriented way. To think we would be a well-oiled machine right off the bat -- that was an unrealistic expectation.

But, I don't want to dive into an evo process that is assumed to be a modified version of our new species process, when the new species process is still working through issues. I know the process will always be tweaked, but we're still tinkering with it quite a bit.

I think one of the reasons this latest project has veered out-of-control at times is because the process hasn't gelled yet. If we crank up a second project now, I see it just causing more problems. I think many people are still a bit confused as to what the hell is going on. Maybe that's my fault. Maybe it's the TL. Perhaps we could have done a better job of establishing clear leadership. But, in my mind, it's hard to regulate something when the rules are ill-defined.

If we can get the process firmed up, it will be easier to manage. And we'll be able to manage multiple projects simultaneously. We should begin advertising the fact that a new evo project WILL happen in the near-future. But, not right this second.

Since Dane seems to have the best assembly of all the current process guidelines, I'd like to have a Create-a-Pokemon Process Guide sticky topic. Right now, people can't really follow the process, because they can't see it in an easily-accessible place.

BTW, Dane -- I think you should include your written timeframe of the creation process. I think DK's calendar is superior for seeing how everything runs concurrently. But, the "list format" is really nice for getting an overall sense of the phases of the process and the sequential ordering of each topic relative to the others.
 
X-Act said:
What I learned the most is that it's not important that you have sucky stats when you have one stat out of PT/PS/ST/SS that is incredibly awesome. Alakazam is one such Pokemon, having incredible Special Sweepiness but crap everything else.
Indeed, though Alakazam's lack of everything but Special sweeping is what brought him down from the top. I think the best Pokemon have at least two of these covered...Deoxys-A, for example, is regarded as uber, only because of his high Physical and Special sweepiness. His durability is super crap.

Doug said:
The only question is timing. I don't think we're ready to do simultaneous projects yet.
Me neither. We should work out any and all huge problems we have with the way we currently do things before we can set two trains up and running.

And yes, we need a "How we Create-A-Pokemon" sticky.
 
Updated with new information:


Process Guide Calendar by Deck Knight

List version of Calendar:
Phase I: 7 Days + Spillover
  • Main Type (2 Days)
  • Secondary Type (1 Day)
  • Style Bias (Offensive/Defensive/Mixed) (2 Days)
  • Begin Art Submission Thread
  • Build Bias (Physical/Special/Mixed) (2 Days)
Phase II: 12 Days + Spillover
  • Stat Ratings Rank Poll (2 Days)
  • Overall Stat Spread (2 Days)
  • Finalization of Art Submissions (1 Day)
  • Art Selection (2 Days)
  • Begin Sprite Submission Thread
  • Begin Main Ability Discussion Thread (1 Day discussion before polling)
  • Begin Movepool Discussion Thread
  • Begin Name Submission Thread
  • Main Ability Selection (2 Days)
  • Secondary Ability Selection (1 Day)
  • Name Poll (1 Day)
Phase III: 2 Days + Spillover
  • Begin Pokedex Entry Submission Thread
  • Movepool Specifics (1 Day)
  • Begin Analysis Thread
  • Sprite Poll Thread & Pokedex Poll Thread (1 Day Total, simultaneous threads)
  • Etc (Pre-Evolutions, Height, Weight, Finish Analysis)
  • Complete
Average time of project, not counting spillovers: 21 Days.



Specifics / FAQ:
What is the "Step [Y] Discussion Thread"?
The Discussion thread is a thread created at various points throughout the creation of a given Pokémon. It is a thread that is used to discuss any given step of a poll, before the poll is created. Multiple steps can use a discussion thread such as this, including "Overall Stats", "Ability", and "Name". These threads are used for gaining submissions, gathering ideas, and general think tanking before a poll begins.

Before, these threads would not receive their own threads, and would instead be done in the thread of the current poll. This would make things hard to follow, and ideas could be easily overlooked. Examples of how threads such as these would be used include the previously used "Art Submission", "Movepool Discussion", "Pokedex Submissions", and "Sprite Submissions" threads.



What is Spillover, and why is it included in the timetable?
Spillover is a general term for when the end of a thread's normal timetable allotment does not reach a conclusive result. A prime example of this would be if, during the main typing thread, two types end up with a total vote in the general area around 40%. Another thread would then be created to determine which result would be the final choice. This thread would be the spillover thread, and must be completed before the next step can begin.



What are the various polling choices?
There are 4 ways to poll.

First is the "standard poll", where you use the basic board options to set up a poll for up to 10 choices. This is the most basic, and can easily be used for the majority of polls.

Second is the "check poll", where you use the board options to set up a poll for up to 10 choices. The difference between the "check poll" and the "standard poll" is how many choices you can vote for. In a "standard poll", only one option can be voted on. In a "check poll", any number of choices can be voted on. This is mostly used on polls where the outcome has multiple options, or doesn't matter as much, such as with the Pokedex Poll thread.

Third is the "bold poll", where the voters must make a post with their votes cast in bold font. Make sure that, in a poll based on this method to make it clear that no one is to post the current tally of votes. Only the Topic Leader is to do this, if they wish to at any point up to the end of the vote.

Fourth is the "preferential poll", where the voters must make a post with their votes cast in preferred order, in bold font. This style of voting has the voter rank their top 25-40% of the available choices. This is then done with a weighted system. The numbers for each rank depends on the number of overall options available, and the Topic Leader chooses how many ranks the voter chooses. The voter should not have more than 4 ranks available to choose. Here's a short table for how to rank each option based on the number of choices:
2 Choices:
___Rank 1 - 3 points
___Rank 2 - 1 point
3 Choices:
___Rank 1 - 6 points
___Rank 2 - 3 points
___Rank 3 - 1 point
4 Choices:
___Rank 1 - 10 points
___Rank 2 - 6 points
___Rank 3 - 3 points
___Rank 4 - 1 point
The option with the most points at the end of the voting period is declared the winner.

The Topic Leader is allowed to use any voting option they desire for a poll, as long as it fits with the options available to vote on.



What are the Stat Ratings?
The process, created by X-Act, can be found here. Here is a temporary Java Applet for this process. This is a much more effective process than the previous BST options.



What are the Style Bias and Build Bias polls?
The Style Bias is based on the Offensive/Defensive Balance, as seen here. Offensive is an OSB rating of 5 or more, granting "Slightly towards Offense". Defensive is an OSB rating of -5 or less, granting "Slightly towards Defense". Between -5 and 5 grants a balanced style.

The Build Bias is based on the Physical/Special Balance, as seen here. Physical is an OSB rating of 5 or more, granting "Slightly towards Physical". Special is an OSB rating of -5 or less, granting "Slightly towards Special". Between -5 and 5 grants a balanced build.



What is the Stat Ratings Rank Poll?
As seen here, the Stat Ratings Rank ranges from Horrible to Exaggerated. This poll is for voting on which number range the Pokemon should fall into, and what it's overall effectiveness will then be. Exaggerated and Too Good can be left off of the polling options, as everything in Exaggerated is either Uber or is completely restricted by it's ability, and there's nothing in Too Good to compare to (but many of the most powerful Pokemon still appear in the Fantastic column). Both of those options would basically eliminate a CaP Pokemon's ability to be usable in OU, which is the overall goal.



How are the Overall Stats chosen?
The Topic Leader chooses the method in which to choose the stat spreads. The TL can either create a submission thread for anyone to submit spreads, or ask specific members to create spreads. Generally, the TL will choose members to create the spreads and have them submitted by the end of the Base Stat Total poll. The option is there, however, to have the general populace create spreads if the TL chooses to.

All spreads would need to have reasons for the given stats. At least some of the stats have to have specific reasons, such as being able to live through a specific attack, OHKO/2HKO something with a generic attack (for generic, think 80BP STAB or 70BP without STAB). The other stats can have more vague reasoning, stating which stats could change if need be to fit the Base Stat Total, which would be polled on during this time as well.

The TL then chooses which spreads to use for the Overall Stats poll, or how to adjust the submitted spreads if they don't fit.



When is the Pokemon's purpose, or concept, discussed?
This can be discussed as early as the Main typing thread, or as late as the Movepool Discussion thread. There's no set time for when people can start formulating ideas on what to do, but the purpose is cemented by the time the Movepool is complete. This means it should be discussed earlier, so that the stats, ability, and movepool can fit the theme of the Pokemon.



What do I do if the polls are close? How close is close enough for a spillover poll? How long should a spillover thread last?
This is up to the discretion of the Topic Leader. Generally, the "close enough" point would be within 10% or so of the leading vote, but this obviously cannot be the definitive statement for all polls. If it's a large optioned poll, make a spillover poll with those who were leading, and then continue until an option has at least 50% of the votes, or is a clear cut winner. A clear cut winner would need to have at least 10% more than votes than second place. Spillover threads should last a single day, as discussion should have already occured prior to the spillover thread.



What is the difference between a Offensive/Defensive/Balanced Pokemon? What about the difference between a Physical/Special/Mixed Pokemon?
Using the Stat Rank System, this is easily calculated. The applet will tell whether a spread is offensive or defensive, physical or special, and by how much any bias might be.



If a Pokemon's style is voted as being balanced, how would I go about polling it's build?
If a Pokemon is voted to be of balanced style, then during the Build poll you would have voters pick a build for both it's offensive and defensive sides. This would be done with a bold voting method, with voters casting one vote for either Physical or Special on the Offensive side, and one vote for either Physical or Special on the Defensive side. The possible voting choices are each on seperate lines, and the possible outcomes are:
Offensive: Physical
Defensive: Physical

Offensive: Special
Defensive: Physical

Offensive: Mixed
Defensive: Physical

Offensive: Physical
Defensive: Special

Offensive: Special
Defensive: Special

Offensive: Mixed
Defensive: Special

Offensive: Physical
Defensive: Mixed

Offensive: Special
Defensive: Mixed

Offensive: Mixed
Defensive: Mixed


Questions, comments, concerns, suggestions?
 

DougJustDoug

Knows the great enthusiasms
is a Site Content Manageris a Top Artistis a Programmeris a Forum Moderatoris a Top CAP Contributoris a Battle Simulator Admin Alumnusis a Smogon Discord Contributor Alumnusis a Top Tiering Contributor Alumnusis an Administrator Alumnus
Moderator
And yes, we need a "How we Create-A-Pokemon" sticky.
I like that name for the thread. I'm working on assembling information from this thread into an indexed sticky. As we make changes here, we can update the sticky. Dane, I plan to use your post as the core of the thread. I'll break it up into "chapters", somewhat like the CAP Pokedex sticky with an index in the OP.
 
Dane, I plan to use your post as the core of the thread. I'll break it up into "chapters", somewhat like the CAP Pokedex sticky with an index in the OP.
Sounds good. By breaking it up, that also allows for direct showing of X-Act's process rather than merely linking to the seperate thread.
 

eric the espeon

maybe I just misunderstood
is a Forum Moderator Alumnusis a Researcher Alumnusis a Top CAP Contributor Alumnusis a Tiering Contributor Alumnusis a Top Contributor Alumnus
Dane said:
lots of good stuff
looks very good, but what about forced spillover to save time?

I say change the relevant polls to deciding whether it's:
1. Excellent, Very Good, etc
2. Biased towards offense, defense or no bias
2.1 How biased, if so (very, hardly, etc)
3. Biased towards physical, special or no bias
3.1 How biased, if so (very, hardly, etc)
agreeing with this
 
Spillovers haven't been needed this project, so I figured forcing them wasn't necessary to be added to the timeline.
 

DougJustDoug

Knows the great enthusiasms
is a Site Content Manageris a Top Artistis a Programmeris a Forum Moderatoris a Top CAP Contributoris a Battle Simulator Admin Alumnusis a Smogon Discord Contributor Alumnusis a Top Tiering Contributor Alumnusis an Administrator Alumnus
Moderator
I made a sticky titled "How we Create-A-Pokemon". I didn't break out the stat stuff. Now that the sticky exists, we can change the content or format. If someone wants to write-up new sections, I'll be happy to add them to the thread and the chapter index.

One new thing I would like to see come out of this thread is a chapter on how we go about revising a pokemon. We say that we can change pokemon, but honestly I don't know how we would do it. Personally, I'd love to see Syclant get changed. But we don't really have a process for initiating and carrying out change. I'd like to see a concrete process for it.
 
Seeing as a change should be for the better, the first step would be analyzing what is wrong with the Pokemon? If it's something as simple as "it OHKOs Tentacruel jesus christ" then perhaps a slight nerf to its attacking stat should be enough so that it doesn't anymore.

However, in the case of something like Revenankh where Shedrest is deemed broken, there are more options to discuss. You could make it less bulky so it has less of a guarantee to pull it off, or you could remove either Rest or Shed Skin (the latter more likely).
 
Well, after adequate testing period (which Syclant has more than had), I'd say one of the Mini-Tournament Organizers or Mods from the Shoddy server begins a petition thread.

Some arbitrary number of "signatures" later (15-20 perhaps), the thread is locked (and deleted probably, since it really serves no purpose afterwards), and a new thread is created. This thread would then discuss possible changes to be made, why they need to be made, and any other possible issues that need addressing.

After debate within the thread is adequate, a vote is conducted on the issues with enough worthy reasoning, deciding whether it will be changed or not.


How's that sound?
 
I'd prefer something in the line Mekkah was going. Once enough people say there's something wrong with a pokémon, a thread is started. First, the goal i to identify the problem. If everyone says a CAP pokémon is broken/is too weak, but there is no agreement on what's wrong, then you can't really do anything. Not only that, maybe then there isn't anything really wrong, but just a collection of anecdotes or particular problems.

Once the problem(s) has been identified, we should focus on fixing that problem. Similarly to the movepool and ability threads, people start discussing about the possible changes that would fix the problem. The ideal situation would be to find a general agreement with the discussion alone. However, if there are a number of solutions that are heavily supported by several people, then maybe they could go to a poll. And I say maybe because balance is a delicate issue, and not everyone has good enough knowledge and competitive experience to make their opinion valid.

Once a solution is decided, it is implemented in the server, and a new testing period begins.
 

DougJustDoug

Knows the great enthusiasms
is a Site Content Manageris a Top Artistis a Programmeris a Forum Moderatoris a Top CAP Contributoris a Battle Simulator Admin Alumnusis a Smogon Discord Contributor Alumnusis a Top Tiering Contributor Alumnusis an Administrator Alumnus
Moderator
What if each time we start a playtesting thread, we begin compiling a list of "Certified Playtesters" for that pokemon. An individual's playtesting certification would be sponsored by a Server Moderator. Mods should only sponsor someone that they know has spent an acceptable amount of time playtesting the pokemon. Only certified testers would be allowed to submit ideas for changes and vote on them. The changes would be bold polls, and only certified votes would be counted. This give people incentive to actually play the pokemon and not just sit around a theory-bitch about it. If you can't get a mod to vouch for you, then you really aren't trying very hard. Certified testers will be reset every pokemon. So if you played Syclant and got certified, that doesn't mean you are certified for Revenankh. It also gives more "server only" users a reason to make Smogon accounts and get involved in the forum side of the project.
 

eric the espeon

maybe I just misunderstood
is a Forum Moderator Alumnusis a Researcher Alumnusis a Top CAP Contributor Alumnusis a Tiering Contributor Alumnusis a Top Contributor Alumnus
One new thing I would like to see come out of this thread is a chapter on how we go about revising a pokemon. We say that we can change pokemon, but honestly I don't know how we would do it. Personally, I'd love to see Syclant get changed. But we don't really have a process for initiating and carrying out change. I'd like to see a concrete process for it.
I have a few ideas about how this could work, unfortunately I do not have the time right now to write up a full guide, however I would like to speak to whoever is writing that guide and tell them my ideas.

Basically what I would suggest is have a thread for people to say things that could be changed (for Syclant it would be stuff like "to high stat total", "too high speed", "tail glow" that kinda thing) the most popular ideas would be put in a bold text poll with the options for each idea being "Change/Leave". Those that are selected as Change would be polled (for too high speed we would how much to reduce its speed) we would do the most popular changes first, nearer the end we may have nerfed it enough so the less popular changes could be re polled to see if we should implement them.


Another thing is what we should do with pre-evos, I am of the opinion that we should make a single thread for each pre-evo and give it UU competive movepool and stats.
 
Status
Not open for further replies.

Users Who Are Viewing This Thread (Users: 1, Guests: 0)

Top