• Follow our Instagram!
  • Curious about what we're currently doing about bots? Check out the latest updates from staff in this thread.

Announcement Gen 4 Random Battle Suspect Process: Stealth Rock (Result: Stealth Rock stays gone!)

Stealth Rock has a 50% winrate**
That's right, with only two asterisks, rocks are completely balanced.
*[1]: In tournament level play, they are in fact stronger. We can quibble over sample sizes and recency and tracking rocks being up as opposed to genned and yadda yadda, but we all know rocks are strong. Every teamchat goes "up" when they go up. However, the vast majority of players are not playing gen4 rbtt, and making the game only about them would be a mistake. In addition, it seems most of the actual rbtt gen 4 players prefer rocks. Many of the pro-ban votes in this very thread say they hate gen4 either way, so honestly who cares which version they dislike least. They won't play it anyway. (As an aside, if we're banning everything that's stronger in tour level play, we're gonna be banning a lot of things in a lot of formats.)

*[2]: The reason we know rocks have a 50% winrate is that each pokemon has a 50% winrate (in mid-high ladder), and each mon that could gen rocks always did. Therefore, having rocks was equivalent to having a mon with rocks, and therefore it was all 50%. However: this isn't quite true. It's possible to gen 2 mons with possible rocks, in which case the second mon won't have rocks, and therefore be largely dead weight. This depresses their winrate, leading to buffs, thus making rocks stronger. This is preventable. I'm not a fan of fucking around with the randomness, but surely making it impossible to gen 2 setters is a lot less drastic than banning rocks, and I think it'd go a long way of assuaging concerns.
 
Last edited:
Chains of Markov said:
In addition, it seems most of the actual rbtt gen 4 players prefer rocks.
Click to expand...
Click to collapse...
I didn't play gen 4 in RBTT in part because there WERE rocks; if they stay removed, there is a strong chance I would consider playing 4 this season. It isn't particularly surprising to me that the people who chose to play DPP in RBTT during the rocks meta tend to like rocks, and those that didn't tend to dislike them.

Chains of Markov said:
Many of the pro-ban votes in this very thread say they hate gen4 either way, so honestly who cares which version they dislike least. They won't play it anyway.
Click to expand...
Click to collapse...
Even outside of RBTT, every circuit competitor, and especially the playoffs competitors, are very much affected by the decision we make, and I see 9 of the current top 50 advocating in this thread to keep rocks banned. In fact, I believe the highest current circuit scorer who voted ban as of this post is Xceloh, who is tied for 51st and didn't even sign up for Gen 4 Open. I would love to choose Gen 4 in playoffs since I've played it a ton, but would be unlikely to choose it against anyone if rocks are added back.
 
Chains of Markov said:
Stealth Rock has a 50% winrate**
That's right, with only two asterisks, rocks are completely balanced.
*[1]: In tournament level play, they are in fact stronger. We can quibble over sample sizes and recency and tracking rocks being up as opposed to genned and yadda yadda, but we all know rocks are strong. Every teamchat goes "up" when they go up. However, the vast majority of players are not playing gen4 rbtt, and making the game only about them would be a mistake. In addition, it seems most of the actual rbtt gen 4 players prefer rocks. Many of the pro-ban votes in this very thread say they hate gen4 either way, so honestly who cares which version they dislike least. They won't play it anyway. (As an aside, if we're banning everything that's stronger in tour level play, we're gonna be banning a lot of things in a lot of formats.)

*[2]: The reason we know rocks have a 50% winrate is that each pokemon has a 50% winrate (in mid-high ladder), and each mon that could gen rocks always did. Therefore, having rocks was equivalent to having a mon with rocks, and therefore it was all 50%. However: this isn't quite true. It's possible to gen 2 mons with possible rocks, in which case the second mon won't have rocks, and therefore be largely dead weight. This depresses their winrate, leading to buffs, thus making rocks stronger. This is preventable. I'm not a fan of fucking around with the randomness, but surely making it impossible to gen 2 setters is a lot less drastic than banning rocks, and I think it'd go a long way of assuaging concerns.
Click to expand...
Click to collapse...
I feel like this is a bit disingenuous because suspects have always been this way - top players who can qualify to vote get the say in how the tier moves forward. I don't think rands should be exempt from how suspects work for every other format. Similar to a trustee model of representation in politics, it is assumed that the people who can qualify to vote also can make the choice on what will be best for the tier moving forward. I don't have any data to back this point up, but I am assuming it to be true: Most suspect tested items probably have roughly a 50% winrate on their ladder. It's ladder... Additionally, hating a tier should not be a reason that we disregard someone's opinion. I was very outspoken in my post about how I hate playing this tier, but I want to be clear: Disliking playing this tier does NOT mean I do not care about its future. I want the tier to be improved. As drifttrick mentioned, stealth rock are not an objectively more fun thing to have in this format - the player base may change as the result of this suspect to some extent. And that's fine.
I disagree with saying that this decision also is only about the people who would be playing gen 4 in high level tournaments. Anyone is welcome to vote on this if they can get requirements. Is it difficult? Very much so. But if there are people who strongly dislike the possibility of stealth rock remaining banned in this format, despite not playing in rbtt or wcor etc, they should prove that they can make an informed decision on the future of the format.
 
Lady Writer said:
Similar to a trustee model of representation in politics
Click to expand...
Click to collapse...
But this is exactly what I'm advocating! I don't care how easy or difficult the reqs are (though "anyone is welcome to vote" is obviously disingenuous), I just want those that do vote to consider more than just tournament level play. If you really believe that the tier is more fun or better or whatever for everyone with rocks removed then vote to remove them! I just don't want the rbtt winrate to have an outsized effect.

And to be clear, the reason that winrates mean nothing on tier ladders is because people get matched up to roughly equal opponents, making everything 50% by default. This is not the case in rands where teamgen is a separate effect that is very much measurable.
 
If it helps anyone, this is the graph of games played above the winrate tracking threshold per month for each format. For the non-cg ones that's the number of games where both players are above 1300 Elo:
1790283263954.png

The last datapoint on the right is last month, as this month is not yet finished (and would thus show a sharp drop in every format).
 
Last edited:
Chains of Markov said:
If it helps anyone, this is the graph of games played above the winrate tracking threshold per month for each format. For the non-cg ones that's the number of games where both players are above 1300 Elo:
View attachment 865518
The last datapoint on the right is last month, as this month is not yet finished (and would thus show a sharp drop in every format).
Click to expand...
Click to collapse...
As a colorblind person, this graph took me many minutes to decipher :fukyu:
Which one is the last data point exactly? Is it where the graph ends or the point right before that? Is seems that all rands ladders have been seeing less activity as of recent, but I'm not sure if I'm interpreting this graph correctly.
 
Just got reqs. I will admit, I've topped this ladder at 1860 before so I imagined it would be pretty easy to get ladder reqs here. But the combination of high level ladder players on low ELO alts and the relatively low playerbase of this gen made this much harder than I anticipated. Also I'm just bad.

I'm gonna be voting to ban rocks. On top of the things I've said in the past, I genuinely just think the format is more fun and more competitive this way. I thought gen 4 was decent before but now I think it's one of our best formats.
 
R1C3M4N said:
As a colorblind person, this graph took me many minutes to decipher :fukyu:
Which one is the last data point exactly? Is it where the graph ends or the point right before that? Is seems that all rands ladders have been seeing less activity as of recent, but I'm not sure if I'm interpreting this graph correctly.
Click to expand...
Click to collapse...
I've updated the graph to have more useful labels on the axis (indeed the rightmost point is August). Don't know an easy way to make it more colorblind friendly I'm afraid :l
I woudn't draw any major conclusions off of it or anything, but indeed most ladders had decreased activity with 1/3/5 a little increase in August over July.
 
Hi! I've been playing a lot of DPP Randbats on the suspect ladder, and I fully believe Rocks should be a part of the metagame. A couple of takeaways I've had:

-Flying-types and normally Stealth Rock-weak Pokémon are dominating with the absence of Rocks in the game. Just see when you get an Articuno how hard it is for either you or your opponent to take it down.

-Speed is immensely important, and a lack of Speed control on your team means you almost certainly are going to lose. You cannot manage a lack of speed without hazard pressure.

I strongly urge those voting to return Stealth Rock to DPP Randbats. The lack of them is causing the metagame to become more and more unbalanced. I will not be playing this Gen of Randbats should Rocks continue to be banned. The metagame is dramatically worse for it.
 
Iguana said:
-Speed is immensely important, and a lack of Speed control on your team means you almost certainly are going to lose. You cannot manage a lack of speed without hazard pressure.
Click to expand...
Click to collapse...
These really aren’t correlated at all. If anything, the presence of hazards makes fast/offensive teams more overwhelming, since it’s easier for them to net KOs. The +2 Minun that outspeeds your team usually doesn’t care about the fact it took 12.5% on entry; it cares a lot more about only needing to do 87.5% damage (or less, due to Rocks weakness and/or cumulative hazard chip over the course of a game) to each opposing Pokémon in order to sweep, instead of the full 100%.

As for your first point, this has been litigated extensively throughout the thread. Rocks in the format made Articuno either very bad, or very good, depending on team-gen. Level balancing + Rocks removal will allow it to be balanced in a more consistent manner across games.
 
JustOut459 said:
Level balancing + Rocks removal will allow it to be balanced in a more consistent manner across games.
Click to expand...
Click to collapse...
drowsy69 said:
Level balancing can actually keep them in check now, which it literally couldn't in a rockfull meta if you didn't get rocks.
Click to expand...
Click to collapse...
drifttrick said:
All of a sudden, level balancing goes out the window
Click to expand...
Click to collapse...
Amaranth said:
without rocks, if level balancing is doing its job, this is easily fixed.
Click to expand...
Click to collapse...
I do not understand this near Bible thumping level of faith in level balancing when, per the data in Markov's post, level balancing demonstrably did not work on Stealth Rock. It's balanced on ladder, yet (allegedly) completely broken in tournament play.

One of the chief complaints from pro-banners is that Stealth Rock is so easy to use, but apparently not for the ladderers that level balancing is based off of. So, why should I believe that this issue will not repeat itself for gameplay elements that are harder to manage than Stealth Rock?

Vespiquen is a good example, with all the slow U-Turns it pulls off to enable breakers. That's something a ladder player is less likely to be able to capitalize on than RBTT caliber players, which leads them to lose more often using Vespiquen -> Vespiquen's level goes up when, if a better player had been using it, its level would actually go down.

If that sounds overly hypothetical, recall that the Articuno polarization (The Livid Question) is similarly based on differential estimates of Articuno's winrates in the presence of Rocks vs no Rocks, and the impact of those on its level, which isn't even calculable anymore according to Markov?
1790320811750.png
 
Wigglytuff said:
I do not understand this near Bible thumping level of faith in level balancing when, per the data in Markov's post, level balancing demonstrably did not work on Stealth Rock. It's balanced on ladder, yet (allegedly) completely broken in tournament play.

One of the chief complaints from pro-banners is that Stealth Rock is so easy to use, but apparently not for the ladderers that level balancing is based off of. So, why should I believe that this issue will not repeat itself for gameplay elements that are harder to manage than Stealth Rock?

Vespiquen is a good example, with all the slow U-Turns it pulls off to enable breakers. That's something a ladder player is less likely to be able to capitalize on than RBTT caliber players, which leads them to lose more often using Vespiquen -> Vespiquen's level goes up when, if a better player had been using it, its level would actually go down.
Click to expand...
Click to collapse...
Win-rate based level balancing can never be perfect for everybody, given that Pokémon often perform differently in the hands of differently-skilled players. This is, more generally, an issue with any sort of balancing in any sort of game; there will always be easy characters and difficult characters, easy strategies and difficult strategies, etc. Unless these characters/strategies/Pokémon/[things] are balanced differently for different Elo brackets (unironically an interesting concept, but would exacerbate the pre-existent sample size issue), the difficult [things] will always perform worse in low Elo compared to high Elo, given that they aren't as accessible to lower skill players. This means we have to choose between such [things] being weak in low Elo, or dominant in high Elo. The same principle applies for the easy [things], but in reverse.

However, this does not mean that win-rate based level balancing is unsuitable for Rands, or any other game – indeed, it's likely the best option we have at our convenient disposal (for now at least... until the days of Jaxcalibur-based level balancing are upon us). Moreover, level balancing is something that is in effect regardless of whether or not Rocks are present in the format. The argument we are all putting forward is that it is more effective in balancing Pokémon like Articuno, Ho-oh etc. in a Rocks-less format than a Rocks format, since such Pokémon's overall win-rates across three months will be more consistently similar to their win-rates in any one given match-up. Level balancing will never be perfect, but perfection should not be the enemy of progress, especially when perfection is unobtainable.

Wigglytuff said:
If that sounds overly hypothetical, recall that the Articuno polarization (The Livid Question) is similarly based on differential estimates of Articuno's winrates in the presence of Rocks vs no Rocks, and the impact of those on its level, which isn't even calculable anymore according to Markov?
View attachment 865657
Click to expand...
Click to collapse...
This sentence is not coherently relevant to the topic at hand. We have never been able to collect data on pre-Rocks-ban Articuno's different win-rates against teams with Rocks vs without Rocks – only ever how Articuno performs as a whole. We are and always have been able to calculate the change in Pokémon's win-rate after a level change. However, Markov points out that this data isn't very meaningful, since only significant outliers get level changes, and outliers by their nature are unlikely to persist into the next balancing period, so the observed change in win-rate will largely be due to this statistical noise, rather than the true effect of the level change. Either way, this doesn't have anything to do with Articuno performing poorly vs Rocks and strongly vs no Rocks, since the data for these specific match-ups does not exist (nor has it ever). In the absence of this data, the argument we are all putting forward is based on very intuitive reasoning – Articuno will be significantly weaker in games where it takes 50% on switch-in, compared to games in which it does not. If you disagree with this premise, feel free to lay out your thoughts. If not, then I believe we are on the same page.

@jaxcalibur should we ban Rocks or nah?
 
I've laddered both with and without stealth rock and my impression is that stealth rock kinda hold the tire together. Rocks punish all mons for switching in, so you can usually chip threats down, even if you couldn't beat them otherwise.
Now if you take away stealth rock, you suddenly tend to find yourself in a lot of situations where your opponent has a mon that outspeeds and threatens most of your team. And that thing can now switch in for completely for free. Fun :))
Spikes do also server this purpose, but flying types and levitators just get to switch for free anyway and you're less likely to have any hazards on your team all now.
Maybe I shouldn't complain too much about rng in randbats, but while laddering I've had some truely one sided match ups more or less because one of us just didn't have any way of punishing a certain mon. If you can't check it offensively, tough luck!

Subjectively it just feels worse too. I don't hate long games or toxic, but having to play out these games where it's obvious from turn 5 who's gonna win because they have a toxic immunity is not fun for me, even when I'm winning.

I do see the point that rocks can give players a huge advantage with little cognitive effort, but... is it too much to ask to give every team a stealth rocker? It's really not an uncommon move and I feel like it would go much further towards actually creating a balanced game.
I see the noble intention of those looking at a 70/30 matchup spread and going: "wow! these evil rocks must be the only culprit! we shall get rid of them posthaste!" but I feel like that's missing the point. Rocks aren't unique in that aspect. Spikes, tspikes and even toxic immunes can make for very boring and uncompetitive games when only one side has them.
Now I don't have any statistics to back that up, so you'll either have to believe my accounts or go ladder and experience it yourself.
Point is: if we're banning stealth rock, might as well ban spikes, tspikes and toxic too. And if that's what you want, gen2randbats is right there. (But nobody plays that tier because it's hot garbage.)

If stealth rocks seriously get banned I won't touch this tier unless I really have to for a tour or something.

PS: Since this is technically a gen4 suspect, I am legally obliged to say the following:
anything but jirachi
 
A Cake Wearing A Hat said:
As stated in the OP, we are not going to do this.
Click to expand...
Click to collapse...
I knew this would come, but I feel like you should actually engage with the people suggesting things other than ban rocks yes or no.
Maybe this isn't the thread for it, but we should have a discussion outside of the arbitrary dichotomy of rocks or no rocks. Let's clearly identify issues with the metagame, outline where we would like to take it and discuss options from there. Then we can vote on those options.

I don't have much hope that I'll get more than "that's not our policy" but
1790404948864.png
 
ohko by scald said:
I knew this would come, but I feel like you should actually engage with the people suggesting things other than ban rocks yes or no.
Maybe this isn't the thread for it, but we should have a discussion outside of the arbitrary dichotomy of rocks or no rocks. Let's clearly identify issues with the metagame, outline where we would like to take it and discuss options from there. Then we can vote on those options.

I don't have much hope that I'll get more than "that's not our policy" but View attachment 865806
Click to expand...
Click to collapse...
Hi,

Enforcing Stealth Rock on every team is more complicated under the hood and has more knock-on effects than would be desirable or reasonable in most circumstances. To do this we would likely have to:
-create one catch-all role solely for and including all stealth rock setters with no exceptions (with fully generalized code for move enforcement and item generation for ALL rockers all at once), or otherwise declare a new global variable in the data to mark pokemon as stealth rock setters
-make a new method in the code that specifically picks one of the pokemon marked as a stealth rock setter for slot 1
-outright reject all possible stealth rock setters in all other slots, for the sake of not needing to worry about stealth rock setters being weaker when they aren't running rocks (so you will NEVER get hippowdon and clefable together ever again, for example. This isn't a negotiable part of this.)
-ensure that level balancing remains stable for the stealth rock setters, given they're all leads now

I dont have the time to do all the logistical math for this scenario, but I can confirm it'd be a nightmare on the backend. This is why we aren't going to do this and why it isn't ever going to happen.


And if you'd like to also force removal on every team (which people WILL demand if we do this), that is even more problematic due to there only being 17 of those; that means these spinners will each appear on 5.9% of teams, or in 1/8 of games. I don't think you want one eighth of all gen 4 games to have a delibird in it, but I could be wrong.

I hope this is adequate engagement on the topic, despite still not allowing it to be possible.
 
Last edited:
It's time to start voting!

These are the qualified voters:
billy zubat
O1V7O2X9O
Shadows4
Mikon
Rowlet69420
Chaos23333
Tom Bus
Lady Writer
Critsomnia
edgy king
hammy3581
JustOut459
drowsy69
paysa
drifttrick
Xceloh
soTsoT
xzeroxman
anankaios
Ghostilex
Drud
Drifting
mind gaming
MichaelderBeste2
Wigglytuff
RainListener
Diamonds_realm
R1C3M4N
blittius
35Q71N
livid washed
Gom
Texas Cloverleaf

20 votes will be required to keep Stealth Rock banned; 14 votes will be required to unban Stealth Rock.

IF YOU HAVE BEEN TAGGED BY THIS POST, PLEASE VOTE AT https://www.smogon.com/tools/suspects/view/254
 
You must log in or register to reply here.
Back
Top