Ret’s WAR is way too high

[ICECLIMBERS]

BMG is unlikely to remove any role. Their last resort is an overhaul like with Disguiser. The community is also resistant to change. For example, the poll for removing spy’s ability to read night chat was overwhelmingly in favor of no. Ret is a strong role but some of the ways to nerf it would anger players (such as dying as a result). A temporary solution would be refraining from reviving unique roles, or possibly bodyguards who died defending their target. Or vanillaize the revived person, but some people would leave if revived if that were the case I fear.

It’s hardly the most powerful non-killing town role so long as Transporter exists.

I think 45/45/10 is a roughly good ideal.

[Brilliand]

I would like to point out that when I introduced the "proportional winrates" idea, I didn't actually know of anyone who genuinely wanted it. I just saw it as a sufficiently coherent idea that the "60% town winrate" deserved a certain amount of respect (that is, to be treated as a feature of the game that can be improved upon, rather than a bug). As far as I can recall, I came up with the idea myself.

On the other hand, this:

I think 45/45/10 is a roughly good ideal.

...is a bit harder to justify. Apparently you think the Neutral Killing isn't "real scum"? Or you think proportional winrates are good enough for the smallest faction but not for the two largest factions?

[MysticMismagius]

Would you rather have a million kids in cages or half a million kids in cages?

What does this have to do with anything?

[MysticMismagius]

I’d rather have 45/45/10 than 67/25/8

I think Brilliand is trying to say it should be 33/33/33.

[Brilliand]

Yeah, what MysticMismagius said. 60/32/8 is a Schelling point - not a particularly good one perhaps, but it has a clear logic to it. 33/33/33 is also a Schelling point, and is also the best way to design the game, so far as I can tell. 45/45/10 is not a Schelling point, or at least I don't see the logic behind it - it just seems like an arbitrary set of numbers that has the same problems as 60/32/8 to a lesser degree.

Also since ICECLIMBERS said this:

lol imagine saying that mafia wasn’t designed for equal village/scum winrates and expecting for people to take you seriously

...I'm inclined to hold him to a higher standard than the rest of us. If he's going to espouse something even slightly proportional, he can have his own insult back.

[SantanaTheSmall]

I don’t see how it can be 33/33/33 without seriously reworking the whole game.

It’s only something achievable on custom lists with serious players.

[MysticMismagius]

Yeah, what MysticMismagius said. 60/32/8 is a Schelling point - not a particularly good one perhaps, but it has a clear logic to it. 33/33/33 is also a Schelling point, and is also the best way to design the game, so far as I can tell. 45/45/10 is not a Schelling point, or at least I don't see the logic behind it - it just seems like an arbitrary set of numbers that has the same problems as 60/32/8 to a lesser degree.

Also since ICECLIMBERS said this:

lol imagine saying that mafia wasn’t designed for equal village/scum winrates and expecting for people to take you seriously

...I'm inclined to hold him to a higher standard than the rest of us. If he's going to espouse something even slightly proportional, he can have his own insult back.

His point is 45/45/10 is achievable, and 33/33/33 may not be. It's still a VAST improvement.

From NK's perspective, it really isn't. 45/45/10 only gives NK a 2% higher win rate over 60/32/8. Any improvement for NK is a blessing, but to call this a "vast" improvement is a bit of an overstatement. NKs would still get shafted, and we'd see the same leavers ruin our day every time they roll NK.

It is a significant improvement for Mafia though.

[MysticMismagius]

Didn't specify NK. I said significant improvement generally.

I'm just bringing it to light that the winrate suggested doesn't really change anything when looking from that point of view. NKs are important too, ya know.

[ICECLIMBERS]

In a longform, very vanilla game of mafia the setup should be balanced in a way that all factions have a chance of winning. This includes 3p kill-capable scum and even additional informed minorities. Town of Salem is rapidly paced and role madness. A great number of changes would be needed to achieve that balance in a game that appears to have been designed with the idea of NK being the underdog.

And, to be frank, people would complain if town won only 1/3 of the time.

[MysticMismagius]

In a longform, very vanilla game of mafia the setup should be balanced in a way that all factions have a chance of winning. This includes 3p kill-capable scum and even additional informed minorities. Town of Salem is rapidly paced and role madness. A great number of changes would be needed to achieve that balance in a game that appears to have been designed with the idea of NK being the underdog.

And, to be frank, people would complain if town won only 1/3 of the time.

That is true. I still think 10% winrate for NKs isn't ideal, but 33% isn't either.

[SantanaTheSmall]

This is why I said proportional win rates are utilitarian. It’s technically the biggest benefit or “good” given to the most number of people.

I personally don’t believe one person should have an equal winrate than a 9 person faction. The larger evil faction, however should have a better winrate.

Secret Hitler is a very fun game because the two factions have about a 50/50 chance of winning. This problem is really only introduced and likely unsolvable when a one person faction requires the death of everyone else. That’s why NE is fun. Jester and Exe don’t require the death of an entire faction. And witch can moderately improve mafia’s or NK’s chances if they find each other early,

[Razbae]

Ret has always been OP but the only way to really nerf it is to get rid of it. If BMG implements actual nerfs to the already one click warrior, it'll make the role even worse to play and have less impact in casual modes. BMG just needs to get it out of ranked, however, they won't. Why? idk but it's not rocket science like OP is trying to make this seem like so BMG obviously know it's op and have known for a long time. Good luck getting ret out of ranked.

[Brilliand]

Ret has always been OP but the only way to really nerf it is to get rid of it.

Look at the Retri re-work options thread over in Role Ideas. There are a bunch of ret nerf ideas there - some of them are even good.

The variations on "light side necromancer" seem the most promising.

[SantanaTheSmall]

Ret has always been OP but the only way to really nerf it is to get rid of it. If BMG implements actual nerfs to the already one click warrior, it'll make the role even worse to play and have less impact in casual modes. BMG just needs to get it out of ranked, however, they won't. Why? idk but it's not rocket science like OP is trying to make this seem like so BMG obviously know it's op and have known for a long time. Good luck getting ret out of ranked.

We all know that ret is OP but the whole point of this post is to help quantify exactly how much ret is OP.

[James2]

Note: the ret games only included games where the ret successfully pulled off a revive.

This is a systemic methodological flaw that invalidates your entire analysis.

[MysticMismagius]

How? He didn’t throw out games where Ret existed but didn’t get their revive off. He just made the independent variable a little bit different from the other guy who did a study about Ret. Sure it’ll change the results but his point isn’t invalid because of it. The conclusion one can draw from this data, “In games where Retributionist revived someone, Town’s win rate is way too high” is a perfectly valid point that is supported by the data OP collected.

[Brilliand]

Note: the ret games only included games where the ret successfully pulled off a revive.

This is a systemic methodological flaw that invalidates your entire analysis.

We had a whole conversation about this, that ended with him agreeing to make a change to his methodology. But (as MysticMismagius said) the problem isn't as severe as you say - while his original data failed to prove that the ret frequently pulls off their revive, it's easy to believe intuitively that that is the case.

[SantanaTheSmall]

Yeah I’m gonna have to edit the first post to clarify this,

[James2]

“In games where Retributionist revived someone, Town’s win rate is way too high”

"In games where Plaguebearer becomes Pestilence, his winrate is too high"

"In games where Juggernaut advances to his final tier, his winrate is too high"

Etc.

It's an arbitrary criteria. The relevant winrate is the rate at which a Ret rolling leads to Town victory, not the rate at which a Ret rolling and the gameplay going in a certain way results in Town victory.

[SantanaTheSmall]

“In games where Retributionist revived someone, Town’s win rate is way too high”

"In games where Plaguebearer becomes Pestilence, his winrate is too high"

"In games where Juggernaut advances to his final tier, his winrate is too high"

Etc.

It's an arbitrary criteria. The relevant winrate is the rate at which a Ret rolling leads to Town victory, not the rate at which a Ret rolling and the gameplay going in a certain way results in Town victory.

There’s already been multiple analyses on that. I’m adding to the discussion by answering then question: “when ret DOES pull off a revive, what is town’s likelihood of winning?”

And honestly, if you see a small gap between ret’s existence and ret’s success in their ability’s use, that indicates that the percentage of rets that pull off their revive is high. So it’s NOT like plaguebearer or juggernaut, wherein their success is extremely low. It’s a false equivalence to compare them.

[MysticMismagius]

“In games where Retributionist revived someone, Town’s win rate is way too high”

"In games where Plaguebearer becomes Pestilence, his winrate is too high"

"In games where Juggernaut advances to his final tier, his winrate is too high"

Etc.

It's an arbitrary criteria. The relevant winrate is the rate at which a Ret rolling leads to Town victory, not the rate at which a Ret rolling and the gameplay going in a certain way results in Town victory.

There’s already been multiple analyses on that. I’m adding to the discussion by answering then question: “when ret DOES pull off a revive, what is town’s likelihood of winning?”

And honestly, if you see a small gap between ret’s existence and ret’s success in their ability’s use, that indicates that the percentage of rets that pull off their revive is high. So it’s NOT like plaguebearer or juggernaut, wherein their success is extremely low. It’s a false equivalence to compare them.

This pretty much says everything I was going to say in response to James2's comment.

[ElderSivart]

“In games where Retributionist revived someone, Town’s win rate is way too high”

"In games where Plaguebearer becomes Pestilence, his winrate is too high"

"In games where Juggernaut advances to his final tier, his winrate is too high"

Etc.

It's an arbitrary criteria. The relevant winrate is the rate at which a Ret rolling leads to Town victory, not the rate at which a Ret rolling and the gameplay going in a certain way results in Town victory.

In most games, Retributionist can revive night 2. It is generally most beneficial to the Town for them to do this.
Comparing it to Plaguebearer and Juggernaut, which usually do not achieve the conditions you mentioned until late in the game, is false equivalence in my eyes.

[Brilliand]

If you'll refer back to the conversation on the first page of this thread, I believe this topic was already adequately addressed:
viewtopic.php?p=3197586#p3197586
viewtopic.php?p=3197654#p3197654

WRT Pestilence, I'm surprised to hear that transforming into Pestilence usually results in a win. I'd have thought that it was the reverse: the Pestilence transformation is easy to achieve, but rarely results in a win.

[James2]

“In games where Retributionist revived someone, Town’s win rate is way too high”

"In games where Plaguebearer becomes Pestilence, his winrate is too high"

"In games where Juggernaut advances to his final tier, his winrate is too high"

Etc.

It's an arbitrary criteria. The relevant winrate is the rate at which a Ret rolling leads to Town victory, not the rate at which a Ret rolling and the gameplay going in a certain way results in Town victory.

There’s already been multiple analyses on that. I’m adding to the discussion by answering then question: “when ret DOES pull off a revive, what is town’s likelihood of winning?”

And honestly, if you see a small gap between ret’s existence and ret’s success in their ability’s use, that indicates that the percentage of rets that pull off their revive is high. So it’s NOT like plaguebearer or juggernaut, wherein their success is extremely low. It’s a false equivalence to compare them.

The point wasn't the specific rates. The point was that whether or not ret successfully revives is a product of player skill. Since a successful revive selects (to some extent) for skill, the value of this data is substantially reduced.

“In games where Retributionist revived someone, Town’s win rate is way too high”

"In games where Plaguebearer becomes Pestilence, his winrate is too high"

"In games where Juggernaut advances to his final tier, his winrate is too high"

Etc.

It's an arbitrary criteria. The relevant winrate is the rate at which a Ret rolling leads to Town victory, not the rate at which a Ret rolling and the gameplay going in a certain way results in Town victory.

In most games, Retributionist can revive night 2. It is generally most beneficial to the Town for them to do this.
Comparing it to Plaguebearer and Juggernaut, which usually do not achieve the conditions you mentioned until late in the game, is false equivalence in my eyes.

A lot of rets wait to revive more important roles. Saying that it's possible for a ret to revive with relative ease doesn't mean it actually happens that way.

[Brilliand]

The point wasn't the specific rates. The point was that whether or not ret successfully revives is a product of player skill. Since a successful revive selects (to some extent) for skill, the value of this data is substantially reduced.

...

A lot of rets wait to revive more important roles. Saying that it's possible for a ret to revive with relative ease doesn't mean it actually happens that way.

Skill as a ret means not waiting to revive more important roles. Just revive ASAP so no one has time to stop you.