It is very much worth noting that Dynamax Cannon is generally a bit better than our primary reliable STAB options, both because Dragon is a generally better typing (particularly in Ubers) and because of the extra 10 BP, which may not seem like a lot but does add up. If we want LO sets to have similar potency in OU as Etern LO sets do in Ubers, it isn't entirely unreasonable to suggest that CAP38 could have higher SpA than Etern. OU's bulkiest mons are not substantially less bulky than Uber's bulkiest mons, even though Ubers does have a bit more bulk across the board.
-
Follow our Instagram!
-
Curious about what we're currently doing about bots? Check out the latest updates from staff in this thread.
-
Welcome to Smogon! Take a moment to read the Introduction to Smogon for a run-down on everything Smogon, and make sure you take some time to read the global rules.
CAP 38 - Part 7 - Stat Limits
- Thread starter Spammernoob
- Start date
As the NDUbers guy I feel it is worth clearing a couple of things up. Spammer's post is well intended, but relies on a somewhat outdated dex that also doesn't provide any sort of context that ends up being pretty relevant. There are some ubers specific things that have been mentioned, but I'm not sure how relevent they are here. For example, yeah the general bulk is higher, but this mostly impacts offensive Pokémon as there are not many defensive Ubers to begin with. The flip side is that even the defensive Pokémon tend to be pretty fast compared to other tiers. As for the ND Ubers stuff and Life Orb Eternatus specifically...
What we should be be looking to capture is the 'vibe.' This doesn't just mean that Eternatus does x v y Pokémon, but also how it does it. I've specified that because
is what transforms Life Orb Eternatus from a strong Pokémon that is annoying to switch around into an proper wallbreaker. Additionally, the environment makes a difference as ND Ubers doesn't seem to be as entry hazard centric as CAP (to my understanding).
So what sort of Pokémon are Eternatus targeting? Defensive Arceus-Dark was given as an example, but yeah a special wallbreaker should be nabbing 2HKOes on a Pokémon concerned with the physical metagame. Calm Mind sets (usually 248 HP / 104 SpD) are a different story. Sure Eternatus doesn't KO, but that isn't surprising given the SpD invest is to survive Meteor Beam into +1 Dynamax Cannon, but it isn't far off and a Sludge Bomb Poison makes it dicey long term and risky to use Eternatus as setup fodder without popping Tera to ignore it - especially in a world with 8 Recover PP. Likewise, Eternatus doesn't get 2HKOes on Pokémon where their special bulk is a substantial part of their identity such as Ho-Oh and defensive Primal Kyogre, but can break them with Sludge Bomb procs and / or Tera Dragon over time with good play.
As a whole, Ubers is pretty similar to most metagames in that special tanks / walls excluding the blobs tend to be less numerous and varied relative to their physical counterparts. That is something that Life Orb Eternatus benefits from as will CAP 38, albeit to a lesser degree. I'll put a checklist below of the general stuff that LO CAP 38 should try to achieve to be a mini Eternatus. This is also very 'general' and we need to be honest that CAP 38 will have slightly lower efficacy at a minimum due to the STABs and that is fine. I'll also put some sort of 'LO Eternatus counterplay yap' in a spoiler in case it is useful to anybody.
So in an ideal world CAP 38:
Asked the other day on discord about tiering the different type and would like to expand on that. Mystical Fire is a tool that is appealing to defensive sets that offensive sets will rarely utilize. If you hate Equilibria with a passion then I can see it being a niche option on LO or Specs, but that is kind of it. Flamethrower and Fire Blast are a different beast and I have concerns about both, but mostly Fire Blast. That made a massive difference for Eternatus and I can see something similar with CAP 38 and think it the onus should be on the submitter to justify including it at all - even with a penalty. Presumably there is some threshold it hits outside of Equilibria (125 SpA LO 2HKOes with Stealth rock up), but would would take a cautious approach. It seems likely that Fire Blast will free up the tera type in a similar manner to Eternatus.
What we should be be looking to capture is the 'vibe.' This doesn't just mean that Eternatus does x v y Pokémon, but also how it does it. I've specified that because
is what transforms Life Orb Eternatus from a strong Pokémon that is annoying to switch around into an proper wallbreaker. Additionally, the environment makes a difference as ND Ubers doesn't seem to be as entry hazard centric as CAP (to my understanding). So what sort of Pokémon are Eternatus targeting? Defensive Arceus-Dark was given as an example, but yeah a special wallbreaker should be nabbing 2HKOes on a Pokémon concerned with the physical metagame. Calm Mind sets (usually 248 HP / 104 SpD) are a different story. Sure Eternatus doesn't KO, but that isn't surprising given the SpD invest is to survive Meteor Beam into +1 Dynamax Cannon, but it isn't far off and a Sludge Bomb Poison makes it dicey long term and risky to use Eternatus as setup fodder without popping Tera to ignore it - especially in a world with 8 Recover PP. Likewise, Eternatus doesn't get 2HKOes on Pokémon where their special bulk is a substantial part of their identity such as Ho-Oh and defensive Primal Kyogre, but can break them with Sludge Bomb procs and / or Tera Dragon over time with good play.
As a whole, Ubers is pretty similar to most metagames in that special tanks / walls excluding the blobs tend to be less numerous and varied relative to their physical counterparts. That is something that Life Orb Eternatus benefits from as will CAP 38, albeit to a lesser degree. I'll put a checklist below of the general stuff that LO CAP 38 should try to achieve to be a mini Eternatus. This is also very 'general' and we need to be honest that CAP 38 will have slightly lower efficacy at a minimum due to the STABs and that is fine. I'll also put some sort of 'LO Eternatus counterplay yap' in a spoiler in case it is useful to anybody.
So in an ideal world
CAP 38:- Threatens 2HKOes on even bulky offensive Pokémon so they have to think about switching in even with a type advantage (i.e. 125 SpA LO Sludge Bomb 2HKOes bulkless Kyurem) with a neutral stab
- May OHKO some frailer offensive threats with SE coverage, but doesn't have to
- A Sludge Bomb Poison can enable CAP 38 or a teammate to break through some defensive checks over the course of a game
- Requires some combination of support, chip, and tera to break bulkier structures
- LO chip adds up quickly enough that CAP 38 can't freely utilize its natural bulk to trade
- Isn't too strong v HO and balance / bo at the same time
Asked the other day on discord about tiering the different type and would like to expand on that. Mystical Fire is a tool that is appealing to defensive sets that offensive sets will rarely utilize. If you hate Equilibria with a passion then I can see it being a niche option on LO or Specs, but that is kind of it. Flamethrower and Fire Blast are a different beast and I have concerns about both, but mostly Fire Blast. That made a massive difference for Eternatus and I can see something similar with CAP 38 and think it the onus should be on the submitter to justify including it at all - even with a penalty. Presumably there is some threshold it hits outside of Equilibria (125 SpA LO 2HKOes with Stealth rock up), but would would take a cautious approach. It seems likely that Fire Blast will free up the tera type in a similar manner to Eternatus.
I haven't had a lot of time recently and haven't really contributed much to this process at all, but I figured I would give some insight into this question in particular given that Cresceidon is being used as a reference here and I was the Stats Leader for that process.
My main justification for why I set T = 2.75 back then was that I wanted to be pretty harsh on restricting user's abilities to submit spreads that were very fast while also being able to apply a decent amount of offensive pressure (I forget what the exact benchmark I used at the time was, and I'm too lazy to bother looking through older messages). This isn't something that I think really applies to this process, and in fact I think might actually lean on being fairly counter-productive given the offensive nature of many of the sets we have currently proposed. This makes me believe that, just based on principal, we should be going with a T-Value that is lower than what we currently have.
In regards to where I think we should have it exactly, this is admittedly something I haven't given a whole lot of thought into (mostly cause I'm a bit strapped for time right now and it's a bit hard for to me really gauge this without having some idea of where the limits might be set right now; could easily see my self coming back to this later), but I do think that we want to avoid going too low, as I think this would discourage spreads that value a bit more offensive pressure for speed, which I think is an very valid route for us to take here. I think an ideal range here would be somewhere between T = 2 and T = 2.5 (Leaning more towards the lower side of the scale than the higher one). As for an exact value, I don't really have a great idea myself, but I think Guingil's idea of averaging out the T-Values of Ramnarok-Radiant and Cresceidon (1.5 and 2.75 respectively, which would result in a value of T = 2.125) which I think makes a lot of sense here given that they have a lot of parallels to the ideal offensive and defensive sets for CAP38, respectively.
This feels like the opposite of how t-value works? A higher t-value means Speed matters less, so each point of Speed costs less SpA. Conversely, it means each dropped point of Speed buys back less SpA, so high t-value discourages going slower here. Speed factor does break down around the high end of what we are looking at (125-135 Spe) so there's little incentive to move off of 135 anyways (at least, not in the sweepiness formula), but high t-value really kills all hope I think.
In practice, a fixed SS limit makes no sense outside of a particular t-value context. If we instead anchor SS to a Speed around the middle of our range, we can see how far we can go in each direction without going over the SS of this anchor value. This table shows that the higher the t-value, the less SpA we get out of lowering our speed (though realistically, no one actually thinks 140 SpA / 111 Spe is nearly equivalent to 134 SpA / 135 Spe).
| T-value | Baseline SS (135 SpA / 125 Spe) | Max SpA for 135 Spe | Max SpA for 111 Spe |
| 1.5 | 169.46 | 134 | 142 |
| 2 | 165.92 | 134 | 140 |
| 2.75 | 162.13 | 134 | 139 |
As always, feel free to check my work in the BSR calculator.
EDIT:
Also, 120SpA hits some relevant benchmarks for our LO set:
252 SpA Life Orb Venusaur Sludge Bomb vs. 252 HP / 0 SpD Hemogoblin: 398-471 (103.6 - 122.6%) -- guaranteed OHKO
252 SpA Life Orb Venusaur Sludge Bomb vs. 252 HP / 4 SpD Clefable: 398-468 (101 - 118.7%) -- guaranteed OHKO252 SpA Life Orb Venusaur Sludge Bomb vs. 252 HP / 4 SpD Clefable: 398-468 (101 - 118.7%) -- guaranteed OHKO
252 SpA Life Orb Venusaur Giga Drain vs. 252 HP / 204+ SpD Garganacl: 242-289 (59.9 - 71.5%) -- guaranteed 2HKO after Leftovers recovery
Garg calc relevant for Protect + Leftovers. I feel like we probably want 120SpA minimum in this case? It's really hard for me to see anything under that, unless it has Apple Acid (in which case we will not be on LO; the sets will look really different probably).
Finally, an interesting benchmark we discussed in Discord today was uninvested CAP38 2HKOing CM Clef w/ Leftovers. This requires the first hit to do ~64% minimum, so the second hit does ~42% minimum vs. +1 to kill. This calc is at 125SpA:
0 SpA Venusaur-Mega Sludge Bomb vs. 252 HP / 4 SpD Clefable: 254-302 (64.4 - 76.6%) -- guaranteed 2HKO after Leftovers recovery
Of course, 0SpA calcs can be bolstered w/ some EVs, but it is a useful benchmark nonetheless (especially if we want defensive CAP38 to be competitive as a Fairy answer).
Last edited:
Moving on, I have seen some general agreement over what the BSR penalties on each of the optional moves have been, as well as some of the conversations around our T=Value. I was particularly swayed by the posts of Guingil and Da Pizza Man regarding acting on a BSR in the middle ground of Ramnarok-Radiant and Creseidon, so currently I think we will go with a T-Value of 2.25. The TLT also discussed between ourselves and we came into the conclusion that a small penalty for speed that excedes a particular threshold could be in order. Similarly, Scorching Sands has been discussed as a potentially disruptive option so its being discussed for penalties as well. Right now I will post a draft regarding how our limits look like. This is not the final Draft just yet, let alone our Final Limits, But I have decided to show just what im thinking for stats right now.
Speed Cap: 111 < x < 141
*: These penalties are cumulative, meaning that should you submit a spread that has all of these 5 options as defining moves, you will induce a -40 BSR Penalty!
Lets address the big noticeable thing here: The rather strict BSR penalties on each move. The reason for this is the extremely high BSR: My objective with these limits is to give a lot of leeway for people to submit high stat spreads, while also massively restricting the options one can choose. That way, people can choose just which moves they can use for whatever purpose. Mystical Fire is slotted seperately from the rest of fire coverage since it functions slightly different that them and thus I think warrant seperate penalties. The speed penalty is there to restrict faster builds from using as much utility as they may originally want. Wisp has a hefty Penalty due to its very disruptive nature, especially for faster sets, while Apple Acid nuking your BSR by itself shouldn't really surprise anyone I feel.
100 PS limit is there to avoid getting high attack values without constraining our speed. 170 allows stuff like calcs for OHKOing Hemo with minimal investment. 140 PT is for primarily uninvested calcs such as avoiding +2 Wogerpon Cudgel 2hko, while ST is slightly lower due to lesser relevance to PT while still allowing for a fair amount of special bulk.
Leaving 72 hours for discussion.
T = 2.25 | Base Limits | Flamethrower / Fire Blast | Mystical Fire | Will'o'Wisp | Apple Acid | Speed > 125 |
|---|---|---|---|---|---|---|
PS | 100 | | | |||
SS | 170 | |||||
PT | 140 | |||||
ST | 130 | |||||
BSR | 660 | -5* | -5* | -10* | -15* | -5* |
*: These penalties are cumulative, meaning that should you submit a spread that has all of these 5 options as defining moves, you will induce a -40 BSR Penalty!
Lets address the big noticeable thing here: The rather strict BSR penalties on each move. The reason for this is the extremely high BSR: My objective with these limits is to give a lot of leeway for people to submit high stat spreads, while also massively restricting the options one can choose. That way, people can choose just which moves they can use for whatever purpose. Mystical Fire is slotted seperately from the rest of fire coverage since it functions slightly different that them and thus I think warrant seperate penalties. The speed penalty is there to restrict faster builds from using as much utility as they may originally want. Wisp has a hefty Penalty due to its very disruptive nature, especially for faster sets, while Apple Acid nuking your BSR by itself shouldn't really surprise anyone I feel.
100 PS limit is there to avoid getting high attack values without constraining our speed. 170 allows stuff like calcs for OHKOing Hemo with minimal investment. 140 PT is for primarily uninvested calcs such as avoiding +2 Wogerpon Cudgel 2hko, while ST is slightly lower due to lesser relevance to PT while still allowing for a fair amount of special bulk.
Leaving 72 hours for discussion.
In a perfect world w/o Speed floor (ignore that I broke the stat cap)
EDIT: is this the CAP for 170atk?
Last edited:
This isn't a good system. What is? Idk. But I know for sure that mystical fire isn't worth 5 BSR when a mon already has fire coverage. -5 for the first fire move, with -2 for each additional fire move is my best idea right now to make the decision for what fire moves to use still meaningful. But idk, I'm kinda lost here and any good system seems overcomplicated.
Some quick stuff that's a little odd/irritating:
I've been doing some preliminary calcs and I keep running into 101 PS and stuff with a 600 BST so I have to lose my pretty numbers. Please can this be 105 or 110? It shouldn't really impact viability in any way
.I don't think a single person wants us outspeeding zama. Also we probably need a min, likely at 111 but you could justify lower, to prevent nonsense like whatever spammer posted above. Or you can just not slate slow spreads but this seems pretty agreed upon. Also speed is MESSED UP at these high speed tiers, you can go all the way up to the max speed available, not fully invest in speed, and come out with more stats than if you ran 252+ on a lower speed set. -5 BSR for going above 130 may be valid for that reason (maybe less, but something).
Requesting that this becomes either >124 or >=125, as right now we're being encouraged to speed tie darkrai which is really yucky.
Last edited:
Okay, some quick thoughts on the limits.
First and foremost, I’d suggest reducing the Mystical Fire penalty to -2 or -3 BSR. I understand the logic of having Flamethrower and Fire Blast have equal penalties: one move is weaker but more consistent, the other is stronger but can miss. Fair enough. But while Mystical Fire is still a Fire move, it fundamentally does not fill the same kind of “unga bunga” role the stronger Fire moves do. Just look at MFire vs. Flamethrower on a Life Orb 252 Timid set with 131 Special Attack (a number I’m personally using for my WIP spread, but whether you like it or not it’s on the higher end of our realistic power):
252 SpA Life Orb CAP38 (131 Special Attack) Mystical Fire vs. 252 HP / 252+ SpD Equilibra: 138-164 (33.8 - 40.1%) -- 32.5% chance to 3HKO after Leftovers recovery
252 SpA Life Orb CAP38 (131 Special Attack) Flamethrower vs. 252 HP / 252+ SpD Equilibra: 166-198 (40.6 - 48.5%) -- guaranteed 3HKO after Leftovers recovery
While this is not to say the move does nothing for CAP38 (it likely fits better on, say, a Leftovers 3A + Recovery set for example where you can just sit there, click MFire, and make Libra sad while gradually healing off damage instead of being put in range by LOrb), you really can’t use it as an interchangeable option the same way Flamethrower and Fire Blast can. Mystical Fire is a more specific tool for offensive sets than its kin, which pretty easily enable sets using both coverage moves without much thought. That obviously is a path we’ve left open, but Mystical Fire is not a move you’d use so freely, so I don’t think it should be as heavily penalized. Now, there is an argument to be made that what should happen instead is Mystical Fire stays put while Flamethrower/Fire Blast get increased to -10, but I’m not out to make that case right now. Making Mystical Fire -3 or something feels more right in my head, so I’m saying let’s do that. Open to the -10 case though.
OKAY, SO. Spammer and I were crunching the numbers related to Speed and we’ve come to realize the entirety of how Speed is being handled right now is a clusterfuck in terms of what becomes “optimal”. If I’m not beaten to it, I’ll do a write-up in the morning. Holy hell.
Secondly, I’d like to formally request a hard Speed floor of 111 be added alongside the 141 ceiling (which I think is fine as is). Not much to say here, just think it should be explicit in our limits rather than (technically) being up in the air.
Lastly, this is kind of a petty thing, but I’d suggest raising the Physical Sweep limit by exactly five points. A few people in the Discord have been putting the Attack stats of hypothetical spreads in the high 70s for aesthetic BST purposes, but doing so on faster spreads leads to the PS being technically illegal despite it realistically not mattering at all. Adding five points to the limit should fix this issue without leading to any Earthquake shenanigans later on; the highest you can take your Attack and still be faster than Torn-T (122 Speed) is 81 (104.80 PS) and against Glowking, EQ does exactly as much as Earth Power if CAP38 has 125 Special Attack:
0- Atk Life Orb CAP38 (81 Attack) Earthquake vs. 252 HP / 4 Def Slowking-Galar: 169-200 (42.8 - 50.7%) -- 2.3% chance to 2HKO
252 SpA Life Orb CAP38 (125 Special Attack) Earth Power vs. 252 HP / 252+ SpD Slowking-Galar: 169-200 (42.8 - 50.7%) -- 2.3% chance to 2HKO
Basically, while it’s pretty much just a flavor thing, having a spread that has 77 Attack and 126 Speed (for example: 101.50 PS) shouldn’t get your stats disqualified even though it’s essentially doing squat for the mon anyway. Raising it to 105 just squashes that little bug without causing any major issues for spreads worth their salt.
First and foremost, I’d suggest reducing the Mystical Fire penalty to -2 or -3 BSR. I understand the logic of having Flamethrower and Fire Blast have equal penalties: one move is weaker but more consistent, the other is stronger but can miss. Fair enough. But while Mystical Fire is still a Fire move, it fundamentally does not fill the same kind of “unga bunga” role the stronger Fire moves do. Just look at MFire vs. Flamethrower on a Life Orb 252 Timid set with 131 Special Attack (a number I’m personally using for my WIP spread, but whether you like it or not it’s on the higher end of our realistic power):
252 SpA Life Orb CAP38 (131 Special Attack) Mystical Fire vs. 252 HP / 252+ SpD Equilibra: 138-164 (33.8 - 40.1%) -- 32.5% chance to 3HKO after Leftovers recovery
252 SpA Life Orb CAP38 (131 Special Attack) Flamethrower vs. 252 HP / 252+ SpD Equilibra: 166-198 (40.6 - 48.5%) -- guaranteed 3HKO after Leftovers recovery
While this is not to say the move does nothing for CAP38 (it likely fits better on, say, a Leftovers 3A + Recovery set for example where you can just sit there, click MFire, and make Libra sad while gradually healing off damage instead of being put in range by LOrb), you really can’t use it as an interchangeable option the same way Flamethrower and Fire Blast can. Mystical Fire is a more specific tool for offensive sets than its kin, which pretty easily enable sets using both coverage moves without much thought. That obviously is a path we’ve left open, but Mystical Fire is not a move you’d use so freely, so I don’t think it should be as heavily penalized. Now, there is an argument to be made that what should happen instead is Mystical Fire stays put while Flamethrower/Fire Blast get increased to -10, but I’m not out to make that case right now. Making Mystical Fire -3 or something feels more right in my head, so I’m saying let’s do that. Open to the -10 case though.
OKAY, SO. Spammer and I were crunching the numbers related to Speed and we’ve come to realize the entirety of how Speed is being handled right now is a clusterfuck in terms of what becomes “optimal”. If I’m not beaten to it, I’ll do a write-up in the morning. Holy hell.
Lastly, this is kind of a petty thing, but I’d suggest raising the Physical Sweep limit by exactly five points. A few people in the Discord have been putting the Attack stats of hypothetical spreads in the high 70s for aesthetic BST purposes, but doing so on faster spreads leads to the PS being technically illegal despite it realistically not mattering at all. Adding five points to the limit should fix this issue without leading to any Earthquake shenanigans later on; the highest you can take your Attack and still be faster than Torn-T (122 Speed) is 81 (104.80 PS) and against Glowking, EQ does exactly as much as Earth Power if CAP38 has 125 Special Attack:
0- Atk Life Orb CAP38 (81 Attack) Earthquake vs. 252 HP / 4 Def Slowking-Galar: 169-200 (42.8 - 50.7%) -- 2.3% chance to 2HKO
252 SpA Life Orb CAP38 (125 Special Attack) Earth Power vs. 252 HP / 252+ SpD Slowking-Galar: 169-200 (42.8 - 50.7%) -- 2.3% chance to 2HKO
Basically, while it’s pretty much just a flavor thing, having a spread that has 77 Attack and 126 Speed (for example: 101.50 PS) shouldn’t get your stats disqualified even though it’s essentially doing squat for the mon anyway. Raising it to 105 just squashes that little bug without causing any major issues for spreads worth their salt.
Last edited:
I'm not commenting on the actual numbers themselves because I'm still new and lack the experience / knowledge to know what is appropriate in that regard. What I do find surprising is the ratios. I'm also not sure if there is a mistake in the numbers since this post reads a Mystical Fire having a different penalty than the other two. I'm not sure how it is considered close to the level of Flamethrower or Fire Blast either?
What exactly is checking STABS + Earth Power + Fire Blast from LO / Specs outside of the Blobs? Spammer said Dragonite which apparently is not great in CAP, but would get much better I guess? We are already designing a mon that is supposed to be a pain to switch around without some pretty specific counterplay and adding 'guess the coverage' mindgames on top of that which always favour the CAP 38 user for such a minimal penalty just feels very off.
I can recognize I do not know the metagame nearly well enough to properly understand the value of wisp, but that doesn't seem close to twice as warping as giving CAP 38 Fire Blast. Most of the chatter surrounding defensive sets has been concerns about whether or not they'll be strong enough so that is whatever on my end. Strong Fire coverage should come with a very steep price. Maybe not Apple Acid level steep, but not far off. Offensive sets can be plenty viable without the ability to blow up almost every single mon in the metagame.
Gotta agree with bumbocat and Sunshine future on some of the numbers seeming off compared to each other.
Fire coverage seems very warping compared to something like will o wisp.
Fire blast also seems like a significantly more powerful option than the other two, clearly outdamaging our standard ground coverage against most steels. 1.2x damage difference is pretty significant and 85% accuracy has rarely been much of an obstacle.
I’m not so convinced by the calcs made to compare mystical fire and flamethrower however. Most targets can still be hit by earth power anyway, so for them it won’t make a difference since that move is 90 BP too. Those that are specifically hit my mystical fire, like equilibra and corviknight, are not strongly affected by the power difference, especially libra for which mystical fire is perfect! That move also synergises particularly strongly with flower veil and our access to recovery, given that immediate damage is the only way to make progress against us and that a fast mystical fire will reduce that damage on the special side.
I do understand some of the worries associated with will o wisp, feels like they can be expanded on a bit more (maybe it was discussed in the discord and I didn’t see)
Agree with the 124 proposed change and also there should be an additional penalty for going above zamazenta, who we agree is a relevant benchmark.
I’m not gonna give specific numbers because I don’t think my mental image of what these numbers mean is accurate enough, but you get the principle.
Fire coverage seems very warping compared to something like will o wisp.
Fire blast also seems like a significantly more powerful option than the other two, clearly outdamaging our standard ground coverage against most steels. 1.2x damage difference is pretty significant and 85% accuracy has rarely been much of an obstacle.
I’m not so convinced by the calcs made to compare mystical fire and flamethrower however. Most targets can still be hit by earth power anyway, so for them it won’t make a difference since that move is 90 BP too. Those that are specifically hit my mystical fire, like equilibra and corviknight, are not strongly affected by the power difference, especially libra for which mystical fire is perfect! That move also synergises particularly strongly with flower veil and our access to recovery, given that immediate damage is the only way to make progress against us and that a fast mystical fire will reduce that damage on the special side.
I do understand some of the worries associated with will o wisp, feels like they can be expanded on a bit more (maybe it was discussed in the discord and I didn’t see)
Agree with the 124 proposed change and also there should be an additional penalty for going above zamazenta, who we agree is a relevant benchmark.
I’m not gonna give specific numbers because I don’t think my mental image of what these numbers mean is accurate enough, but you get the principle.
Wisp should be -5. It's not moving the needle on 38s hard matchups enough to justify anything higher than that imo. Mystical Fire and Flame/Blast should have differing values but not set on which is preferable. Either choice of lowering MFire to like -3 or raising Flame/Blast to -10 works. I'm probably in the camp of -10 for Flame/Blast but would accept Mfire being cheaper too.
I would like to keep hearing thoughts regarding the BSR penalties as they stand right now. So far from both here and the discord I see some discussion regarding lowering the penalty on Wisp, while increasing the BSR penalty of non Mystical fire coverage so its most costly than it, so if more people can chime in to give their thoughts that would be ideal.
Something that has come up a lot, and in a surprisingly brazen manner in the discord, is the topic of additional BSR speed penalties. Its clear people want a speed penalty for anything faster than 124, but some people are discussing of further speed penalties for post 130 values as well as values that are faster than Zamazenta. I would like to hear your thoughts about this, but in a more respectful manner I would hope.
Something that has come up a lot, and in a surprisingly brazen manner in the discord, is the topic of additional BSR speed penalties. Its clear people want a speed penalty for anything faster than 124, but some people are discussing of further speed penalties for post 130 values as well as values that are faster than Zamazenta. I would like to hear your thoughts about this, but in a more respectful manner I would hope.
I agree w/ Zetalz re: Wisp and Fire coverage. I'm mildly leaning towards increasing Flamethrower's penalty, but not really strong feelings either way.
Regarding speed penalties, I'll propose a baseline of -5BSR for >130, and -10BSR for >138 (in addition to the current -5BSR for >125). These penalties are cumulative, so a 140Spe spread would have a total of -20BSR. This set of penalties seems reasonable as a baseline (roughly 1BSR / 1Spe penalty for each point past 120).
I know there are people who disagree and want harsher penalties, but I'll leave that argument to them.
Somebody said earlier on that Eternatus simply gets the essentials of what it needs and nothing more, and I think I agree that this approach is the best way for CAP 38 to emulate their “vibe.” Eternatus has basically every Dragon, Poison and Fire Move under the sun you could ever yearn for but it doesn’t extend too far beyond that offensively.
CAP 38 has a similar build with oodles of Grass and Poison STABs except its Fire Blast parallel is already accounted for in Meteor Beam (High risk/reward that punches through the Fire, Ice and Flying-Types which would otherwise freely sit and stare at them). Adding on damage-focused Fire-Type coverage alongside Rock would likely break the Eternatus parallels since you forfeit the setup of “2 STABS, a pocket nuke of a differing Type, and nothing more for the most part.” The restrictions are what I think will help keep sets fresh and creative.
Not to say it would be sacrilege for CAP 38 to get Skitter Smack or something weird beyond as “coverage,” but I think that with Meteor Beam being confirmed Fire’s role shouldn’t be one of raw power for 38’s kit.
Mystical Fire punches through Steels, hits somewhat harder after the Meteor Beam boost, and gives Special sweepers a slap on the wrist, that’s more interesting to me than pulling out the Gen 1 Flamethrower distribution and simply torching something until it drops dead.
Scorching Sands sounds cool too, Will-O’-Wisp is also neat on a non-Fire or Ghost.
CAP 38 has a similar build with oodles of Grass and Poison STABs except its Fire Blast parallel is already accounted for in Meteor Beam (High risk/reward that punches through the Fire, Ice and Flying-Types which would otherwise freely sit and stare at them). Adding on damage-focused Fire-Type coverage alongside Rock would likely break the Eternatus parallels since you forfeit the setup of “2 STABS, a pocket nuke of a differing Type, and nothing more for the most part.” The restrictions are what I think will help keep sets fresh and creative.
Not to say it would be sacrilege for CAP 38 to get Skitter Smack or something weird beyond as “coverage,” but I think that with Meteor Beam being confirmed Fire’s role shouldn’t be one of raw power for 38’s kit.
Mystical Fire punches through Steels, hits somewhat harder after the Meteor Beam boost, and gives Special sweepers a slap on the wrist, that’s more interesting to me than pulling out the Gen 1 Flamethrower distribution and simply torching something until it drops dead.
Scorching Sands sounds cool too, Will-O’-Wisp is also neat on a non-Fire or Ghost.
This is stupid. I hate what I'm about to suggest, and it sounds as ridiculous to me as it does to anybody, but after looking at the numbers, it seems that -20 is ultimately not even close to enough of a penalty to make fast spreads balanced comparatively to slow ones.
For example, consider the spread: 120 / 57 / 83 /134 / 68 / 122
This was a spread sunshinefuture was considering before we noticed the speed issue, and you could argue about where some stats are and it its too offensive or all sorts of things, but thats neither here nor there. The issue is that this spread is outclassed by a faster spread with a modest nature, even with a -20 BSR penalty.
120 / 57 / 83 /134 / 68 / 122 with 252 SpA / 4 Def / 252+ Spe, for a BSR of 659.90, is the baseline I'm using and yields stats of:
381 / 135 / 203 / 367 / 172 / 377
120 / 57 / 83 /130 / 68 / 141 with 252+ SpA / 4 Def / 252 Spe, for a BSR of 654.79, models the current system and yields stats of:
381 / 135 / 203 / 394 / 172 / 381
120 / 57 / 83 /123 / 68 / 141 with 252+ SpA / 4 Def / 252 Spe, for a BSR of 638.11, models -20 for outspeeding zama and yields stats of:
381 / 135 / 203 / 379 / 172 / 381
The fast spreads are able to conjure spA seemingly out of thin air. This is showing that -20 is small enough of a penalty that speed is costing less than special attack, which is a problem because speed can be converted to special attack by underinvesting. Ideally, speed should cost more than special attack, as speed comes with the added benefit of being able to run timid if desired, with any amount of investment the mon wants such that it can outspeed whatever it feels like and save as many ev's as are relevant for the team its on. Speed is more flexible than spA in this way, so it should be worth slightly more.
This is why I'd like to propose -15 BSR for >124 speed, -15 BSR for being >129 speed, and -10 BSR for being >137 speed. I'll look a little closer at why those specific values were chosen in a sec, but first let's try the modest vs timid thing.
120 / 57 / 83 / 115 / 68 / 141 with 252+ SpA / 4 Def / 252 Spe, for a BSR of 619.04, models -40 for outspeeding zama and yields stats of:
381 / 135 / 203 / 361 / 172 / 381
Now, compared to the original spread, we're being shorted a net 6 spA while gaining 4 speed, which can be taken out of speed and reinvested in bulk if desired. The special attack is probably(?) better, but the fast spread has the advantage of being able to, for example, run just enough speed to outspeed stratagem and put the ev's saved from the zama matchup into bulk, or run 252+ timid on a meteor beam set to outspeed zamazenta and some scarfers and aide it in sweeping. Of course, this is all casework. Different initial statlines may get different results with this method, but imo a penalty that lets any initial statline squeeze out 4 speed, 12 spA, and flexibility for pretty much nothing by cranking the speed isn't a big enough penalty.
Realistically, if someone tanks a penalty they aren't stopping right at the speed 1 point higher than the penalty because going higher is nearly free. So for example, while the suggestion is -15 BSR for >124 speed, what its really saying is "-15 BSR for 129 speed," which is 1 below the next benchmark. Thus, this table:
| Speed 1 under the next benchmark: | 124 | 129 | 137 | 141 |
| Jump in speed from the previous value above: | 5 | 8 | 4 | |
| Relevant Mons Outsped | Lotta stuff |  ,  , +1 252  |  |  |
| Proposed BSR Penalty (Cumulative) | 0 | -15 BSR | -15 BSR | -10 BSR |
| Justification | Contains the most matchups, and is the second biggest jump in speed. | A very important matchup, and the biggest jump in speed. | A less important matchup, but we can reliably outspeed almost all of the tier and flex down for ev's. |
Last edited:
I've been thinking about this a fair bit after the discussions on discord about the forumula and whatnot over the last few days. I am nowhere near experienced enough to suggest a concrete number, but think that outspeeding Zamazenta should come an incredibly steep cost and the current proposal does not accurately reflect this. Zamazenta's poor matchup with CAP 38 stems primarily from typing and its coverage not really threatening it outside of Psychic Fangs (I'm assuming is as real in CAP as other OU metagames). Consequently, I think most folks would agree that CAP 38 will end up being a soft check to Zamazenta at a minimum so why does it matter so much if CAP 38 ends up outspeeding it?
Well... CAP 38 is being designed for and in a tera metagame. Utilizing and accounting for tera adds a novel means of complexity both in game and in the teambuilder that is usually skill expressive. Most metagames have these dynamics, but tera substantially amplifies them and is important when it comes to the relationship between CAP 38 and Zamazenta. For example, if there is a Gholdengo in front of CAP 38 and it needs to Tera Ground to allow Earth Power to OHKO during a game, doing so would make CAP 38 vulnerable to Zamazenta for the remainder of the game and something to consider before popping the trigger. If CAP 38 outspeeds Zamazenta then the opportunity cost is drastically lowered since 'pinning' it by pressuring a tera and revenging with Zama is removed as an option for counterplay.
This dynamic and skill expression is lost if CAP 38 is able to outspeed Zamazenta. I consider that a bad thing, but see a world where it can be fine if it comes with substantial drawbacks - primarily to bulk and power. The penalties should be high enough that it should be very difficult to create a good stat spread that outspeeds Zamazenta, but doable if somebody is willing to put in the time and effort.
The other issue I have is that Zamazenta's Speed tier is far enough above the typical 'fast' Pokémon that CAP 38 can still outspeed them with minimal drawbacks (i.e. 140 Modest still outspeeds Tornadus-T). This frees up CAP 38's nature to an extent that seems like a stats nightmare when all of Modest, Timid, Bold, and Calm natures have to be accounted for while also making sure that offensive sets are not hitting too hard given the Speed tier and also ensuring that defensive sets hit whatever thresholds they need to.
While there has been a lot of talk about the difficulty of balancing Apple Acid, outspeeding Zamazenta falls into a similar boat imo and has largely been overlooked. I do think it will be extremely difficult to balance that Speed tier properly, but not impossible.
Well... CAP 38 is being designed for and in a tera metagame. Utilizing and accounting for tera adds a novel means of complexity both in game and in the teambuilder that is usually skill expressive. Most metagames have these dynamics, but tera substantially amplifies them and is important when it comes to the relationship between CAP 38 and Zamazenta. For example, if there is a Gholdengo in front of CAP 38 and it needs to Tera Ground to allow Earth Power to OHKO during a game, doing so would make CAP 38 vulnerable to Zamazenta for the remainder of the game and something to consider before popping the trigger. If CAP 38 outspeeds Zamazenta then the opportunity cost is drastically lowered since 'pinning' it by pressuring a tera and revenging with Zama is removed as an option for counterplay.
This dynamic and skill expression is lost if CAP 38 is able to outspeed Zamazenta. I consider that a bad thing, but see a world where it can be fine if it comes with substantial drawbacks - primarily to bulk and power. The penalties should be high enough that it should be very difficult to create a good stat spread that outspeeds Zamazenta, but doable if somebody is willing to put in the time and effort.
The other issue I have is that Zamazenta's Speed tier is far enough above the typical 'fast' Pokémon that CAP 38 can still outspeed them with minimal drawbacks (i.e. 140 Modest still outspeeds Tornadus-T). This frees up CAP 38's nature to an extent that seems like a stats nightmare when all of Modest, Timid, Bold, and Calm natures have to be accounted for while also making sure that offensive sets are not hitting too hard given the Speed tier and also ensuring that defensive sets hit whatever thresholds they need to.
While there has been a lot of talk about the difficulty of balancing Apple Acid, outspeeding Zamazenta falls into a similar boat imo and has largely been overlooked. I do think it will be extremely difficult to balance that Speed tier properly, but not impossible.
I have some things I'm still a bit confused about right now. From discussion, it feels like 120-125SpA is basically the floor for non-Apple Acid sets, due to key OHKOs and 2HKOs for LO sets. Do we have an agreed-upon ceiling? It feels like people are gravitating towards 145, but I can't tell if this is anchor bias regarding Eternatus' SpA or actually justified. Current SS limit hard-caps SpA to 145 at 111Spe (for better or for worse), but I know Pipotchi has made arguments that 150+ is a worthwhile design space to explore. Is there a reason this should be off the table? (I'm looking for specific benchmarks / calcs here, not vague vibes).
Additionally, do we have any bulk benchmarks at all? I was looking at Great Tusk calcs, and living 2 Headlong Rushes with 0HP/252Def requires around 145/80 bulk (123PT), which nukes any chance of getting to 120SpA (even with 111Spe). Is this a benchmark we care about? If not, what should we care about? (you can hit 120SpA while tanking 2 Headlongs using my patented 5HP / 245Def spread thought...)
Finally, the fact that PT limit is 10 higher than ST limit doesn't seem right to me. Granted, no one is hitting these limits anyways, but I don't see any reason this is the case, except tunnel-visioning on Ogerpon-W matchup (this Pokemon is not even that common in CAP right now).
I might update this post about my thoughts on the Speed penalties tomorrow maybe, It feels like framing them as penalties was a mistake, when in reality it's closer to "emergency adjustments / tuning to BSR formula". Also, I think tunnel-visioning on BSR formula is a lot less important than establishing key benchmarks we want to hit, want as options, or want to be off-limits (as fun as it is to enter numbers into the sheet and see what comes out).
Additionally, do we have any bulk benchmarks at all? I was looking at Great Tusk calcs, and living 2 Headlong Rushes with 0HP/252Def requires around 145/80 bulk (123PT), which nukes any chance of getting to 120SpA (even with 111Spe). Is this a benchmark we care about? If not, what should we care about? (you can hit 120SpA while tanking 2 Headlongs using my patented 5HP / 245Def spread thought...)
Finally, the fact that PT limit is 10 higher than ST limit doesn't seem right to me. Granted, no one is hitting these limits anyways, but I don't see any reason this is the case, except tunnel-visioning on Ogerpon-W matchup (this Pokemon is not even that common in CAP right now).
I might update this post about my thoughts on the Speed penalties tomorrow maybe, It feels like framing them as penalties was a mistake, when in reality it's closer to "emergency adjustments / tuning to BSR formula". Also, I think tunnel-visioning on BSR formula is a lot less important than establishing key benchmarks we want to hit, want as options, or want to be off-limits (as fun as it is to enter numbers into the sheet and see what comes out).
I feel like we've gotten way too into the weeds with this stage ngl. We can spend days arguing about what penalties should or shouldn't apply to correct for the imperfect BSR formula, but at the end of the day any obvious abuse of the system can be account for by the TLT and other users. This is not a new kind of problem at CAP, we've dealt with similar weirdness but got through it by using common sense to keep submitters from doing blatantly stupid shit with their subs. We can argue for days what penalties are appropriate to fix the flawed system, or we can accept the flaw and move on with the knowledge that the community aren't going to just sit around and let people take advantage of it willy nilly, like some have made it out to be. The human element is being heavily lost here.
The limits proposed in spammers second to last post (-5, -5,-10 for 126, 131, 139 Speed respectively), are serviceable as a band-aid solution the SF conundrum, with the knowledge that CAP users are made aware of the issue and scrutinize it accordingly. We do not need to radically dump massive penalties on every relevant speed tier or adjust BSR up or down to account for them imo.
Also as far as geko's original questions about offensive and bulk benchmarks, I feel like it's better to let users answer these questions via their stat submissions than in this discussion stage at this point. I think current limits are quite generous considering the moves 38 already has presumed for its kit. Better to let individual submitters speak their case for what benchmarks they value than vaguepost about it here. The stat limits stage isn't about (or at least shouldn't be about) trying to account for every possible discrepancy in the formula or laying out a strict list of benchmarks to adhere too. We still want to ensure submitters have freedom to craft different builds and visions for the mon, which I think is more than possible to do right now with the current limits and the already established threatslist we have.
The limits proposed in spammers second to last post (-5, -5,-10 for 126, 131, 139 Speed respectively), are serviceable as a band-aid solution the SF conundrum, with the knowledge that CAP users are made aware of the issue and scrutinize it accordingly. We do not need to radically dump massive penalties on every relevant speed tier or adjust BSR up or down to account for them imo.
Also as far as geko's original questions about offensive and bulk benchmarks, I feel like it's better to let users answer these questions via their stat submissions than in this discussion stage at this point. I think current limits are quite generous considering the moves 38 already has presumed for its kit. Better to let individual submitters speak their case for what benchmarks they value than vaguepost about it here. The stat limits stage isn't about (or at least shouldn't be about) trying to account for every possible discrepancy in the formula or laying out a strict list of benchmarks to adhere too. We still want to ensure submitters have freedom to craft different builds and visions for the mon, which I think is more than possible to do right now with the current limits and the already established threatslist we have.
Last edited:
While I understand the overall sentiment here with respect to the output values, I think the 141 Speed Modest = 124 Speed Timid argument downplays the opportunity cost of running Modest on the 141 spread. The overall idea that Speed value is only as worth as the matchups you outspeed is 100% something I agree with; therefore, giving up key matchups by running Modest is not trivial. This is the "human" or contextual element to this spread, seen in even Zamazenta for Iron Press running Jolly and most speedy Pokemon running a +Speed nature. And say we afford the 141 Spread this luxury of dropping Speed essentially, should we not afford the lower Speed spread the same Modest nature for context? In that case, there is also a power ceiling achievable by Modest lower Speed/higher SpA spreads that is completely unreachable by the higher Speed/lower SpA spreads which emerges as a meaningful difference.
a>b, b>c, therefore a>c.
If timid is stronger than modest, then the problem is even worse and -40 is not enough, as the fast spreads must be even stronger with timid. Also, modest on a low speed spread is junk. The speed tiers are a lot thicker around there than they are up at zamazenta range.
Its not even really BSR abuse to make a broken spread with a -20. Simply using all of the statpoints available will naturally make the spread stronger than all of the other options with no malicious intent, because you just have more points to work with due to the broken formula. I don't see why we're seeing an issue, seeing a solution, and saying "no we'll just use common sense" when just... making an unoptimized stat spread that meets the BSR cap leads to power imbalances.
Last edited:
OK I have some thoughts after discussing with Guingil this morning that I figure are worth putting in thread.
Firstly, some comments on the current BSR Limits:
I will summon MrDollSteak for this section, since I mention a decision he made in the past. Any input on this subject would be much appreciated.
After discussing with Guingil, I concluded that framing this topic as a BSR penalty was a mistake. A better framing would be an amendment to BSR, being BSR2(Spe, ...) := BSR(Spe, ...) + 2 * (Spe - 120) (with appropriate adjustments to limits).
One reason this framing is better is that it is easier to compare two spreads. During my discussion with SunshineFuture, it was very confusing to compare two spreads that were supposed to be comparable, but had vastly different BSRs. This should just make life easier for people looking at stat spreads.
Another reason is that it correctly contextualizes this as an adjustment to make BSR a more accurate measure of stat strength, rather than a subjective judgement that passing a certain threshold , which is the standard interpretation of BSR penalties. When a penalty of -20BSR for going above 138Spe was proposed, people on Discord rebutted that outspeeding Zama was not relevant enough to warrant 20BSR penalty; this is misunderstanding the proposal. What's effectively being penalized isn't passing the threshold to outspeed Zama, but the 8 points of Speed needed to go from the previous threshold (130) to this one.
One might think this is a bit overkill, but there is precedent for manual override of certain BSR calculations to adjust for certain shortcomings in the formula, though it is a bit less involved than the one proposed here ( CAP28 ). It is pretty trivial for me to add another column in the BSR calculator with the amended formula to share, if we take this route.
Anyways, Guingil expressed that I have seemed pretty frustrated with the process recently. Maybe I shouldn't force myself to engage with stages to this degree? This is my first time TLing, so I'm not sure. Also, I feel like I definitely forgot to write something, so feel free to ping me on Discord if something is missing. Also feel free to correct me if you feel I am misrepresenting anything.
Firstly, some comments on the current BSR Limits:
- I don't see a reason to have lower ST here. CAP38 will be eating a lot of Moonblasts, but more importantly, super-effective Fire / Ice coverage moves or uninvested STAB (more commonly on Special side). I think you're tunnel-visioning on Ogerpon-W (who has dropped off in usage recently) to justify Physical defense bias, when there really shouldn't be a difference at all.
- I don't understand the 170SS limit either. Pip gave a value of 149SpA for Modest CAP38 to OHKO 40HP Hemogoblin, but 170SS maxes out at 145SpA / 111Spe. Are you calcing for 0HP Hemogoblin? In my opinion, SS limit should be based on disallowing certain calcs (do we actually have an idea of what calcs indicate we've gone overboard on SpA?), or, if the goal of SS limit is to enforce a PT / ST floor, SS limit should be dropped considerably (currently, maxing SS limit leads to bulk marginally better than Darkrai's).
- Overall BSR should be raised to 670 or 675. My feeling is that CAP38 should be comparable to Roaring Moon without Knock Off or Dragon Dance. Under t=2.25, Roaring Moon has a BSR of 671.57, so I think this value should be within the range of allowable stat submissions.
- Penalty for Apple Acid can go up, assuming overall BSR is raised. I am not even opposed to fixing Apple Acid BSR at current level (645BSR) which results in a potential -25 or -30BSR penalty for Apple Acid. That is because I see Apple Acid as an X-factor tool that justifies CAP38 beyond stats.
I am adding this because I realized Apple Acid w/ potentially 655-660BSR is pretty crazy.
I've seen a sentiment (mainly from Zetalz and DaddyBuzzwole) that we are trying to make an Uber instead of a Mini-Uber. If we are talking about Pokemon that have been banned, most have been due to explosive boosting potential (too many to count); the banned Pokemon that did not have boosting (Iron Bundle and Chi Yu) can't be compared to us at all (this is one difficulty with this project; we have so little reference as to what a fast tank looks like). I feel like the best analogue for CAP38 among Pokemon banned to Ubers is Mega Metagross? Unfortunately, I am not convinced it is possible to make a BSR argument based on that Pokemon.
If by "Uber" we are referring to box legends populating the top of the Ubers VR (which is the connotation), 660 is still significantly lower than unoptimized Pokemon like Groudon and Kyogre (it's better than Deoxys-A at least), not to mention Eternatus, who is a solid 70BSR above these. If we benchmark to 70BSR above CAP's premium-statted, mid-speed mons like Great Tusk (629.78), Chromera (611.35), Gholdengo (584.07), Ogerpon-W (589.98), we can see that 660 is actually a bit low (not a lot, and this methodology may be a bit flawed).
I've seen a sentiment that we have put off giving CAP38 certain high-impact tools (especially in Ability) with the promise of stats, and current limits are too restrictive to fulfill that promise. I would like to remind everyone that there is no more kicking the can down the road (unless you want to try to salvage during Secondary, which I don't recommend). At the end of the day, Solgaleo was good, but not overbearing when it was tested in OU. We effectively do what Solgaleo does, but trade bulk for Speed; I won't pretend this isn't a good trade, but we shouldn't overstate what we're doing either (Solgaleo has 721.20 BSR, by the way).
If by "Uber" we are referring to box legends populating the top of the Ubers VR (which is the connotation), 660 is still significantly lower than unoptimized Pokemon like Groudon and Kyogre (it's better than Deoxys-A at least), not to mention Eternatus, who is a solid 70BSR above these. If we benchmark to 70BSR above CAP's premium-statted, mid-speed mons like Great Tusk (629.78), Chromera (611.35), Gholdengo (584.07), Ogerpon-W (589.98), we can see that 660 is actually a bit low (not a lot, and this methodology may be a bit flawed).
I've seen a sentiment that we have put off giving CAP38 certain high-impact tools (especially in Ability) with the promise of stats, and current limits are too restrictive to fulfill that promise. I would like to remind everyone that there is no more kicking the can down the road (unless you want to try to salvage during Secondary, which I don't recommend). At the end of the day, Solgaleo was good, but not overbearing when it was tested in OU. We effectively do what Solgaleo does, but trade bulk for Speed; I won't pretend this isn't a good trade, but we shouldn't overstate what we're doing either (Solgaleo has 721.20 BSR, by the way).
I will summon MrDollSteak for this section, since I mention a decision he made in the past. Any input on this subject would be much appreciated.
The fundamental problem we are aiming to solve is that BSR is very insensitive to increases in Speed at the upper end of the range we are looking at. For example, a spread going from 130SpA / 125 Spe to 130SpA / 141 Spe leads to BSR increasing around 2 points, from 685.47 to
687.55 (100 in all other stats, for reproducibility). Adjusting t-value does not fix this in an appreciable way; changing to t=1 leads to an increase of 4 points, from 718.74 to 722.67. In contrast, increasing SpA to 131 results in a BSR of 687.84. No sane person thinks that 131 SpA / 125 Spe is a better spread than 130 SpA / 141 Spe.
A simple solution is to manually factor in Speed to the BSR calculation. That is:
BSR2(Spe, ...) := BSR(Spe, ...) + C * (Spe - 120)
If we believe that 1SpA is roughly equivalent to 1Spe at the stat ranges we are interested in, C=2 is a reasonable choice; this also led to the happy coincidence of 131SpA / 122Spe being rated very similarly to 115SpA / 139Spe; these two spreads lead to the same stats at 252/252 Timid and 252/252 Modest respectively. Rating them the same places the same value on the ability to trade Speed for Power via nature; tracey will argue that the latter spread is stronger, but litigating the exact impact is outside the scope of this proposal, so I will settle for them being equivalent.
In a previous post, I noted that BSR started flattening off at around 125 Speed; I am a bit too lazy to readjust the value of C to fit that case, but maybe C = 2 is fine enough for starting at 125 anyways.
What are the ramifications for this proposal? Realistically, we would rebalance BSR limits to allow for our fastest spreads to hit our SpA floor of 120SpA (maybe a bit lower?); this would incentivize trading speed for power, as SpA can actually increase at a reasonable rate relative as we decrease Speed; contrast this with the current state of affairs. Zetalz said that this proposal would restrict the range of viable spreads, but in practice, it would give people a reason to go higher than around 125SpA, increasing diversity.
687.55 (100 in all other stats, for reproducibility). Adjusting t-value does not fix this in an appreciable way; changing to t=1 leads to an increase of 4 points, from 718.74 to 722.67. In contrast, increasing SpA to 131 results in a BSR of 687.84. No sane person thinks that 131 SpA / 125 Spe is a better spread than 130 SpA / 141 Spe.
A simple solution is to manually factor in Speed to the BSR calculation. That is:
BSR2(Spe, ...) := BSR(Spe, ...) + C * (Spe - 120)
If we believe that 1SpA is roughly equivalent to 1Spe at the stat ranges we are interested in, C=2 is a reasonable choice; this also led to the happy coincidence of 131SpA / 122Spe being rated very similarly to 115SpA / 139Spe; these two spreads lead to the same stats at 252/252 Timid and 252/252 Modest respectively. Rating them the same places the same value on the ability to trade Speed for Power via nature; tracey will argue that the latter spread is stronger, but litigating the exact impact is outside the scope of this proposal, so I will settle for them being equivalent.
In a previous post, I noted that BSR started flattening off at around 125 Speed; I am a bit too lazy to readjust the value of C to fit that case, but maybe C = 2 is fine enough for starting at 125 anyways.
What are the ramifications for this proposal? Realistically, we would rebalance BSR limits to allow for our fastest spreads to hit our SpA floor of 120SpA (maybe a bit lower?); this would incentivize trading speed for power, as SpA can actually increase at a reasonable rate relative as we decrease Speed; contrast this with the current state of affairs. Zetalz said that this proposal would restrict the range of viable spreads, but in practice, it would give people a reason to go higher than around 125SpA, increasing diversity.
After discussing with Guingil, I concluded that framing this topic as a BSR penalty was a mistake. A better framing would be an amendment to BSR, being BSR2(Spe, ...) := BSR(Spe, ...) + 2 * (Spe - 120) (with appropriate adjustments to limits).
One reason this framing is better is that it is easier to compare two spreads. During my discussion with SunshineFuture, it was very confusing to compare two spreads that were supposed to be comparable, but had vastly different BSRs. This should just make life easier for people looking at stat spreads.
BSR (-4 penalty vs. -38 penalty):
120 76 83 131 66 122 598 99.67 159.64 112.06 93.71 651.82
120 50 83 115 66 139 573 73.82 145.05 112.06 93.71 617.83
BSR2:
120 76 83 131 66 122 598 99.67 159.64 112.06 93.71 655.82
120 50 83 115 66 139 573 73.82 145.05 112.06 93.71 655.83
120 76 83 131 66 122 598 99.67 159.64 112.06 93.71 651.82
120 50 83 115 66 139 573 73.82 145.05 112.06 93.71 617.83
BSR2:
120 76 83 131 66 122 598 99.67 159.64 112.06 93.71 655.82
120 50 83 115 66 139 573 73.82 145.05 112.06 93.71 655.83
Another reason is that it correctly contextualizes this as an adjustment to make BSR a more accurate measure of stat strength, rather than a subjective judgement that passing a certain threshold , which is the standard interpretation of BSR penalties. When a penalty of -20BSR for going above 138Spe was proposed, people on Discord rebutted that outspeeding Zama was not relevant enough to warrant 20BSR penalty; this is misunderstanding the proposal. What's effectively being penalized isn't passing the threshold to outspeed Zama, but the 8 points of Speed needed to go from the previous threshold (130) to this one.
One might think this is a bit overkill, but there is precedent for manual override of certain BSR calculations to adjust for certain shortcomings in the formula, though it is a bit less involved than the one proposed here ( CAP28 ). It is pretty trivial for me to add another column in the BSR calculator with the amended formula to share, if we take this route.
As a TL, I am a bit conflicted as to what to do here. The bottom line is that I do not want to drag this stage of the process out; freezing the process to fix BSR is off the table. One part of me wants to leave this to Geko and let it ride, while another thinks I should use TL override (I think this is the principled / responsible option). I guess we'll find out how I feel about it tonight when thread closes.
Regarding the human element that Zetalz and spoo brought up, Guingil and I discussed this as well. Guingil expressed that me that framing this as "subjective" is strawmanning, but I don't really think so. The way I see this stage (and the next) in the framing of subjective / objective is as follows.
Obviously, things are not that clean, and we are making subjective decisions with BSR, but the goal of BSR is to be an objective measure of a stat spread's strength; any subjective decisions we make regarding it are directed towards this goal. Thus, it should be a no-brainer that any adjustment bringing BSR closer to this goal should be accepted (though it seems that it isn't?).
Adjudicating BSR abuse should be reserved for edge-cases; leaving too much to SL discretion makes it difficult for stat submitters to, as they have to constantly question whether the SL will accept a spread or not. I had this experience submitting to Pet Mods, where the parameters were extremely loose, but constantly under scrutiny from mod leader; this experience was pretty miserable. I don't mean to say that we should restrict the range of acceptable submissions, but rather that we should clearly demarcate to submitters what is and isn't acceptable from the outset.
BSR abuse is also extremely fuzzy. For example, Brambane's submission to Chuggalong's stat stage had an unusually low 45HP (I remember when it was 30HP). The motivation was to encourage EV investment to bolster bulk in certain matchups, but it can just as easily be construed as an attempt to circumvent the PT and ST limits. At what point does this tip into BSR abuse? (this is relevant to this CAP as well, possibly moreso, since encouraging smart EV allocation is explicitly pro-concept)
Finally, I wonder Zetalz' dismissive response to tracey's proposal will have a negative effect on discussion going forward. If there are good arguments against tracey's proposal, shouldn't they be made here, instead of handwaving it with "it's good enough as-is"? If accepting or rejecting the proposal would both move the process along (which is the main concern Zetalz seems to have with this discussion), what's wrong with accepting the harsher penalties and moving on? Personally, I was very unsatisfied during Obliteryx's Stat submission stage, where it felt my arguments fell on deaf ears and were being dismissed for very subjective reasons, so I would rather not subject someone else to a similar experience.
Regarding the human element that Zetalz and spoo brought up, Guingil and I discussed this as well. Guingil expressed that me that framing this as "subjective" is strawmanning, but I don't really think so. The way I see this stage (and the next) in the framing of subjective / objective is as follows.
- Objective: Measure the strength of a stat spread: BSR
- Subjective: Setting BSR Limits: decide what stat thresholds must be met, which cannot be broken, and which should be optional with tradeoffs (BSR penalties belong here)
- Subjective: (emergency) Adjudicating BSR Abuse
Obviously, things are not that clean, and we are making subjective decisions with BSR, but the goal of BSR is to be an objective measure of a stat spread's strength; any subjective decisions we make regarding it are directed towards this goal. Thus, it should be a no-brainer that any adjustment bringing BSR closer to this goal should be accepted (though it seems that it isn't?).
Adjudicating BSR abuse should be reserved for edge-cases; leaving too much to SL discretion makes it difficult for stat submitters to, as they have to constantly question whether the SL will accept a spread or not. I had this experience submitting to Pet Mods, where the parameters were extremely loose, but constantly under scrutiny from mod leader; this experience was pretty miserable. I don't mean to say that we should restrict the range of acceptable submissions, but rather that we should clearly demarcate to submitters what is and isn't acceptable from the outset.
BSR abuse is also extremely fuzzy. For example, Brambane's submission to Chuggalong's stat stage had an unusually low 45HP (I remember when it was 30HP). The motivation was to encourage EV investment to bolster bulk in certain matchups, but it can just as easily be construed as an attempt to circumvent the PT and ST limits. At what point does this tip into BSR abuse? (this is relevant to this CAP as well, possibly moreso, since encouraging smart EV allocation is explicitly pro-concept)
Finally, I wonder Zetalz' dismissive response to tracey's proposal will have a negative effect on discussion going forward. If there are good arguments against tracey's proposal, shouldn't they be made here, instead of handwaving it with "it's good enough as-is"? If accepting or rejecting the proposal would both move the process along (which is the main concern Zetalz seems to have with this discussion), what's wrong with accepting the harsher penalties and moving on? Personally, I was very unsatisfied during Obliteryx's Stat submission stage, where it felt my arguments fell on deaf ears and were being dismissed for very subjective reasons, so I would rather not subject someone else to a similar experience.
Anyways, Guingil expressed that I have seemed pretty frustrated with the process recently. Maybe I shouldn't force myself to engage with stages to this degree? This is my first time TLing, so I'm not sure. Also, I feel like I definitely forgot to write something, so feel free to ping me on Discord if something is missing. Also feel free to correct me if you feel I am misrepresenting anything.
Last edited:
Have only been loosely following this thread. BSR is broken and I'll fix it after this CAP. Yeah it doesn't account properly for speed and how it varies. Yeah the -5 -5 -10 probably fixes our issues. Yeah that allows for gaming of stat spreads and could probably be interpolated. But equally -5 at 125, -10 at 130, -10 at 140 works well enough for our issues with the CAP. There's probably some modest fuckery that can be done where high speed spreads technically outdo low speed spreads with timid, but equally a low speed spread can run modest too. See how like 145 SpA/111 speed can shove on modest, outrun Kyurem, and cleanly 2hko GKing with LO Epower in a way that other spreads can't do nearly as easily.
If you wanna be more punitive to high speed spreads, change the -5 at 130 to -10 and keep the cumulative nature.
Edit given the problems we're having put t lower btw. Make it 1.5 or even lower.
If you wanna be more punitive to high speed spreads, change the -5 at 130 to -10 and keep the cumulative nature.
Edit given the problems we're having put t lower btw. Make it 1.5 or even lower.
Last edited:
That link is broken
Okay, so I was gonna make a massive post, but I’ve been a combination of busy with other things and too tired to make it. Basically, as the person who went into the BSR mines with Spammer, I’m on board with the BSR2 formula they proposed. I think that framing it as an adjustment rather than a penalty is the best way to understand the adjustment. Considering we can't halt the process to fix our formulas, I feel that some sort of BSR2 is likely the best course of action. Alternatively, I would also be okay with cumulative BSR penalties at 126, 131, and 139 speed if a BSR2 is deemed too much. I'd personally want the final total penalty at 139+ Speed to be -40 BSR, given that should be sufficient to minimize potential abuse of the formula, but I suppose I'd be okay if the number was something like -30 or -25, so long as there is explicitly greater emphasis put on making balanced spreads within the limits rather than making optimized ones this time. While no CAP truly has to have an optimized stat line, mini-Eternatus in particular is a mon that should never be able to "do it all" on one set, even if on paper it has the potential to do many things across sets. Regardless of where the final stat limits land, I do think we as a community should try to keep this in mind.
Supporting:
-spoo
- Speed penalties at relevant benchmarks (-5 at 126, -10 at 131, -15 at 139; cumulatively 30)
- Overall bsr increase to 670
- Lower wisp penalty to 5
- Weak fire coverage penalty should be 5, strong fire coverage penalty should be 10
- Apple acid penalty should be 20
- Consider adjusting the t-value downwards
- Keep SS / ST limits
-spoo