Jump to content

Randoms Feedback / Discussion Thread


Rache

Recommended Posts

To start with, here's a quick summary of how the format works.


How the randomization works:

Spoiler

The species you're given are mostly random, pulling from a list of all Pokemon that were fully evolved in their debut generation + Pikachu, minus a few exceptions that were rarely helpful (Aipom, Porygon, Nosepass). Currently only gen 1-4 species are available, but this will be expanded with gen 5 in future updates. Legendaries are also currently excluded, but will be added when they're obtainable.

 

The rules that restrict team composition are largely intended to improve team quality and make matches more fun. They are:

-Pokemon from the two highest levels (the worst performers) are limited to 1 per team.
-No type will appear more than twice on any team.
-Identical dual types such as Pidgeot + Fearow will not appear together.

-If a Pokemon with Pursuit, Arena Trap, or Shadow Tag is selected as a lead, the party will be shuffled until a new lead is chosen.

-Only one Pokemon can carry both Reflect + Light Screen on any team.

 

Each species has 1-10 (most 3-5) blueprints that sets are created from, and one of these is chosen at random when the Pokemon is generated. Within each blueprint, ability / held item / move1 / move2 / move3 / move4 possibilities are defined and in some cases weighted to make them more likely to be selected. Each is built around a particular strategy, and the separation ensures that the set you get always makes sense even if it's not the most optimal version of it. The more versatile species have hundreds of possible sets.

 

Hidden abilities are currently excluded, but will be added along with builds that take full advantage of them once they're obtainable.

 

All moves are PP Maxed. Hidden Power type is random unless defined in the builds, this is relevant when using Transform, Me First, etc. Gender and ball are random and can't be used to identify builds.

 

How the stats work:

Spoiler

Natures are all neutral, IVs are flat 31s, and EVs are flat 85s for most builds, but there are some exceptions. There are a few move/item combinations that reduce HP EVs to result in a stat divisible by 2 or 4 to make those strategies more effective. Pokemon 4x weak to Stealth Rock or with Defeatist are also set to never have HP divisible by 2 unless one of the move/item combination rules are broken by doing so.


Any Pokemon carrying Trick Room, Gyro Ball, or Metal Burst will have its speed EVs reduced to 0. The species that can potentially pull them will always show both possible speed stats in their speed display tooltip regardless of whether the one you're facing actually has the move.

 

Any Pokemon that doesn't make use of its attack stat at all will have its attack EVs and IVs reduced to 0 to minimize damage taken from Foul Play and hitting themselves in confusion.


The initial levels were chosen based on each Pokemon's competitive tier at the time of the format's implementation, and are now adjusted each month based on how well they perform:

  • 52%+ overall species winrate to level down
  • 48%- overall species winrate to level up

 

And to make sure extremely threatening outliers aren't overlooked due to their worse performing builds:

  • 54%+ individual build winrate to level species down

 

It will take a long time for everything to be appropriately balanced, but each month is another step towards that goal. Builds will be tweaked frequently as well.

 

You can see the current levels in-game by going to Matchmaking Signup -> Statistics -> Randoms.

 

How the rewards work:

Spoiler

Due to the lack of preparation required to play the format, overall rewards have been reduced. This is necessary to prevent randoms from completely replacing standard PvP for too many people.

 

Each season has its own Pokemon-themed hat like the other tiers/formats do. Leaderboard position crowns are still given as well. Automated tournaments won't be held for this format however, so seasonal tournament crowns will not be given.

 

The reward Pokemon from the fortnightly rewards are exclusively from the UU and NU tiers to hopefully give more incentive to play them. We'll be looking into more ways to do so as well.

 

Due to the lack of teambuilding and importance of matchup in the format, if you lose to an opponent whose ELO is significantly lower than yours, you'll suffer less of an ELO penalty than you would in other formats.

 

The current meta is experimental, and as such it's going to have problems. It's likely to go through a lot of revisions before it's reasonably balanced, and that's what this thread is for.

 

If you've encountered a build that doesn't make sense, please either report it here or via PM. Examples of what I'm looking for would be moves or items that don't seem to serve a purpose at all on the set you got or that make the Pokemon worse. Please don't report gimmicky yet cohesive sets for being non-optimal.

 

If you think you've identified a format-wide balancing issue or something that makes it less enjoyable, this is the place to bring it up and discuss it too. Concerns from players experienced with the format will be taken seriously.

 

If you think we can make the format better in any other way, feel free to make suggestions.

Edited by Rache
Updated
Link to comment
13 minutes ago, Bertolfoso said:

I want to see if anyone else feels like there can be way too many choice item users on 1 team. There are also a lot of eviolite mons. Not writing much, just want to see if anyone else has had the same experience.

Yeah I wonder if there's a way to filter out certain sets if you've already generated a specific item. I had a game with a light clay Raichu & light clay Ampharos on the same team and I was pretty limited with what I could do, essentially playing 6v5. 

Link to comment

Ive found that leads can be a little frustrating, with sneasel instantly getting a free kill with pursuit vs some pokemon. I know this is probably a pretty niche situation but its happened a few times to me now. Perhaps the option to choose our leads would help? Or alternatively I dont know if its possible to have a list of pokemon that can be given to each team as a lead, that way we can get a better balance of hazard setters and stuff, and perhaps make sashes are distributed better. I dont know, it might not really be an issue for anyone else and ive just had bad luck

 

Link to comment
50 minutes ago, Paul said:

TLTPJSJ.png

 

Toxic Orb + Facade with Immunity

This one's already fixed for the next build update, the item was missed when converting builds with hidden abilities to no longer rely on them.

 

47 minutes ago, clayclover said:

image.png.7ef67e953aee78bf3b3e039197d799af.png focus sash Toxic Heal > toxic orb

This one's intentional, it's a pure offensive set. Poison Heal was chosen instead of Effect Spore to allow it to safely switch into and benefit from Toxic / Toxic Spikes. This build can pull other items too.

Link to comment

I also do feel that we have a lot lot lot of choiced mons (probably too many). In 2 of my matches I ended up with 5 choiced users (all of them were CB).

 

Another thing is level balancing dependant on tier. I know it's experimental, but I think we might have to take a look at each pokemon individually instead of looking at the pokemon's tier. For example, Slaking is so ridiculously powerful and fast due to the level gap (and bulky) that it can grab multiple KOs during one match while taking very little damage.

 

I'm not exactly fond of the NFE eviolited pokemons such as Yanma, Murkrow, Tangela or regular Porygon. I think we already have a lot of fully evolved mons which completely suck (I'm looking at you, Sunflora) regardless of circumstances and I don't think that expanding this any futher would be necessary. Additionally, when gen4/5 along with HA and legendaries will get implemented, we will be facing an incredibly wide pool of mons and 90% of them will be completely useless during a randbat match. Biggest fun of randbats was when you were getting lots of strong mons on your team, and your opponent was not far behind, not when fighting with tranz trash.

 

Just my thoughts, although I've been playing randbats for 8+ years on showdown, now I WILL play that in Pokemmo I do love it here already

Link to comment
6 hours ago, Mehagony said:

Idk if its just me but hazard stacking seems kinda op concidering how little rapid spin/defog ive seen so far atleast.

Hazards are about twice as common as ways to remove them, but I'm not sure that this is a bad thing as it means that when you do have hazard removal, it's likely going to be useful instead of a wasted moveslot. If it's widely considered an issue after more time has passed though, I'll look into either reducing hazard availability/weighting or adding more removal.

 

 

 

6 hours ago, Bertolfoso said:

I want to see if anyone else feels like there can be way too many choice item users on 1 team. There are also a lot of eviolite mons. Not writing much, just want to see if anyone else has had the same experience.

4 hours ago, RysPicz said:

I also do feel that we have a lot lot lot of choiced mons (probably too many). In 2 of my matches I ended up with 5 choiced users (all of them were CB).

 

I'm not exactly fond of the NFE eviolited pokemons such as Yanma, Murkrow, Tangela or regular Porygon. I think we already have a lot of fully evolved mons which completely suck (I'm looking at you, Sunflora) regardless of circumstances and I don't think that expanding this any futher would be necessary. Additionally, when gen4/5 along with HA and legendaries will get implemented, we will be facing an incredibly wide pool of mons and 90% of them will be completely useless during a randbat match. Biggest fun of randbats was when you were getting lots of strong mons on your team, and your opponent was not far behind, not when fighting with tranz trash.

There are quite a lot of Choice items, it's something a lot of Pokemon can take advantage of. If this is a common complaint over time, I'll reduce the number/weighting of them based on how severe of a problem it ends up being. Like with hazards though, I think it's too early to tell how much tweaking is appropriate.

 

With regards to Eviolite, these are the current NFEs, their winrates, and their placements if in the top 50. It's early enough that this data won't be accurate long-term, but it should give a rough idea of how much of an asset or liability they are.

Spoiler

Pikachu (47.82%)
Golbat (53.05%, currently ranked #26)
Magneton (50.26%)
Onix (48.96%)
Lickitung (46.57%)
Rhydon (50.89%)
Chansey (49.29%)
Tangela (52.24%, currently ranked #47)
Seadra (51.84%)
Scyther (57.38%, currently ranked #1)
Electabuzz (54.83%, currently ranked #8)
Magmar (51.88%)
Porygon (47.90%)
Porygon2 (54.55%, currently ranked #10)
Togetic (49.72%)
Aipom (41.13%, third lowest behind Spinda and Ditto)
Yanma (47.63%)
Murkrow (47.05%)
Misdreavus (52.55%, currently ranked #39)
Gligar (49.21%)
Sneasel (50.73%)
Piloswine (53.41%, currently ranked #23)
Nosepass (48.13%)
Roselia (52.11%)
Dusclops (49.73%)

 

 

The rules I used when selecting them were "was fully evolved in its debut generation or is named Pikachu" for the sake of including the viable ones like Piloswine and Scyther without it being especially arbitrary. Some NFEs perform well, but if the ones that perform poorly are widely agreed to make the format worse to play after more time passes, I'm not opposed to removing the ones that don't/won't have a niche over their evolutions. Aipom in particular stands out as complete garbage.

 

 

 

4 hours ago, RysPicz said:

Another thing is level balancing dependant on tier. I know it's experimental, but I think we might have to take a look at each pokemon individually instead of looking at the pokemon's tier. For example, Slaking is so ridiculously powerful and fast due to the level gap (and bulky) that it can grab multiple KOs during one match while taking very little damage.

We started with tier-based levels to keep it simple, but if manual adjustments would make it better, it's something we can potentially look into in the future. Doing this manually would be extremely complicated though given the huge amount of variety in the builds. If we were to seriously pursue this sort of thing, I think we'd probably take the approach of automating it based on species winrate at the end of each season instead, increasing level if winrate is too low and decreasing it if it's too high. In combination with tweaking or removing problematic builds, the format would theoretically balance itself over time. While I'm aware that winrates don't tell the full story of a Pokemon's success, I think it's a bit more relevant in randoms than in standard tiers as you can't build around abusing or facing specific things.  I think I'd rather wait and see how things go with the current system before committing to a huge change like that regardless though.

 

Slaking was #2 when I checked the usage for the NFEs above but is now #1 with a 57.02% win rate (Scyther's fell to 56.77%). Slaking seems to be a common complaint, so for now it's been dropped to lv81 (from 83) as the speed change becomes relevant at this level. If this isn't enough, its position in Untiered allows it to go as low as 80.  Feedback from matches after the level drop would be appreciated.

 

 

3 hours ago, Kamowanthere said:

I don't know why no one mentioned /randbats and /randdubs (soontm?) commands. Seeing possible moves on opponent pokemon is important unless you want to learn it hard way and memorize every set. I would appreciate /randbats for showing all possible moves on Randoms pokemon.

Somewhere we intentionally deviated from Showdown was with how the builds are constructed - there are far more of them and you're not supposed to be able to memorize or map them out (without an obscene amount of effort, a few have over 1000 sets). While experience can make it a bit easier to guess what you're facing, you're going to guess wrong a lot too. Surprise value and concealing your gimmicks is part of the design here and we'd rather keep it that way, it's also a lot simpler to pick up and play when you're not expected to be issuing commands mid-battle if you want to win often. While the format is derivative of Showdown's, the goal isn't to make a clone, but to use it as a base for something a little different.

 

Dubs-specific builds is something I might look into putting together in the future, but the current builds were many months of work and there are still two generations left to do, so it'll be a while. I don't feel like I'm familiar enough with doubles to build a good meta for it yet either.

 

_________

 

I've also been made aware that Showdown does a small amount of team rigging, and it sounds pretty good, so we'll be adding this in a future update:

 

-Limit of 2 Pokemon from the lowest tier (lv83).

-Limit of 2 Pokemon sharing a type (if you pull Pidgeot and Raticate, you won't get another normal type).

-Limit of 1 Pokemon with any specific dual type (you can't pull Pidgeot and Fearow on the same team, but you can pull Pidgeot and Salamence).

Edited by Rache
Link to comment
1 hour ago, Leviatharian said:

Just exclude it then, instead of giving people an instant 5v6?

It's against the spirit of the format as it's a fully evolved Pokemon, just like Unown. Limber Ditto should hopefully be less obnoxious after gen 4-5 species are added and it's less common as a result.

 

1 hour ago, OrangeManiac said:

I think Defog/Rapid Spin should become more common with the Pokemon that can learn it. Entry hazards are a bit too strong. And I'm aware that's just the nature of random battles in general but compared to Showdown entry hazard removal feels a bit more common there and thus more pleasant to play.

 

I'm revising the builds at the moment based on the feedback I've received, both here and elsewhere. The next set of build updates should include more hazard removal and/or fewer hazards, fewer Choice items, fewer low quality items, and some tweaks to individual builds that were poorly received. I'm not sure when these changes will be implemented, but hopefully it ends up being an improvement.

 

For the current state of the format, some Untiered Pokemon have received some level adjustments based on their performance compared to the Pokemon that share their level, they should be a bit more manageable or less bad now.

 

Spoiler

Omastar from lv83 to lv80
Poliwrath from lv81 to lv80
Scyther from lv81 to lv80

Miltank from lv81 to lv80
Slaking from lv81 to lv80

Misdreavus from lv83 to lv81

Jynx from lv83 to lv81
Pinsir from lv83 to lv81

Electabuzz from lv83 to lv82

Grumpig from lv83 to lv82

Kingler from lv83 to lv82
Ninjask from lv83 to lv82

-
Slowking from lv80 to lv81
Ninetales from lv81 to lv82
Crawdaunt from lv81 to lv82
Cradily from lv81 to lv82
Huntail from lv81 to lv82

Marowak from lv82 to lv83
Wobbuffet from lv82 to lv83

 

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use and Privacy Policy.