After a quick chat with Eco, I think I would add the following rule to this game if it were to be run again:
"You cannot vote for a team which includes yourself."
Thoughts on the above also welcome.
I feel like maybe just on day 1, state how many wolves are in that pairing, but not who they are (assuming you want to keep the set-up vanilla). The rule you ended basically has to have the entire town vote for your pairing, or 1 "bus" vote. Strongly dislike that. Town could also use that mechanic to try and go for a double clear.
Yeah, there's a lot of easy 'fixes' but they open up a lot of abuse.
The day one only scum count reveal might be an option, but I think it's very powerful if there's *zero* scum in the pairing. Scum are effectively forced to kill within the pair.
Private Mod Note
():
Rollback Post to RevisionRollBack
Can we have Megiddo removed from the forum forever please?
i'm pretty sure i can find your ***** online within 3 minutes
The current setup allows two scum to bus each other all game and be basically impossible to pair-vote. If it had been Shadow+Pizza, for example, with Pizza's "I'm 100% on Shadow+somebody and nothing will ever change my mind" hilariously scummy play for eight pages straight, who would ever have voted those two as a pair? Would've been a free win. Hell, the way Pizza was playing, I was actually convinced it was Pizza+Shadow, right up until the end when Shadow didn't hammer.
And that's not even accounting for the fact that the baseline random-vote chances for this are an abysmal 21% or so.
On the other hand, yes, being told that there are no scum in the opening pairing actually rewards Town for poor play, because clearing two people is really powerful. It still leaves things Mafia-favored (Town baseline chance of winning off a no-scum opening guess is 25%, I think?), but it doesn't feel good.
The solutions I can see off-hand (such as voting for an individual instead of a pair on day one, but not lynching the individual even if he's scum), while they resolve the issue mathematically and help with the problems like two scum bussing each other all game, are unfortunately clunky as hell and feel dumb from a flavor perspective.
I'ma sleep on this, cuz I like the concept of pair-voting and wanna help figure out a setup for it that isn't a free scum win.
Osie, do you have a plan for your first scum game here (if it's not this one); specifically, is there a certain scum playstyle that you'd try to mimic/improve upon?
My original plan was to just treat the game as a town game. Same process and everything, which in context worked out. This plan steadily fell apart a bit over time, though. Like I had this readlist:
I kinda still think Osie might be town, though the hard buddying is off putting. Osie, I don't remember you being this buddy buddy in Avalon, but as we were tunneling each other, was there someone that you were buddying up to?
I buddied up some to Silver, but not nearly as blatantly. That was a pretty big mistake here, trying to take advantage of how comfortable you can be around me.
And that's not even accounting for the fact that the baseline random-vote chances for this are an abysmal 21% or so.
I calculated out the math for this in an earlier post. I might have made a mistake on the later rounds, but I don't think so.
Assuming the lynch was decided through a random process, then the chance of a mislynch Rd 1 is
1 - (2/7)*(2/6) = 90.5%
Continuing with random process, assuming any town player could be scum killed and eliminating the pairs that we've already looked at, we end up with (assuming my math is correct):
Rd 2 Mislynch: More complicated = 75.1%
Rd 3 Mislynch: More complicated = 61.9%
Our starting chances of winning are just under 38.1%…
I'm not sure that's correct. If you treat each pair as a single possible lynch, there are 21 pairs only one of which is the actual scum team. After one NK, with 6 alive there are 15 pairs. Day 3 with 5 players is 10 pairs.
Scum win if they avoid all three attempts, so they win at 1 - (1/21 + 1/15 + 1/10) which is a cool 78.5%, leaving the town random win rate at 21.5%.
I think the problem with your maths is you treat the lynch as two separate events, where actually the only way to avoid mislynching is to hit both scum at once.
I'm not sure that's correct. If you treat each pair as a single possible lynch, there are 21 pairs only one of which is the actual scum team. After one NK, with 6 alive there are 15 pairs. Day 3 with 5 players is 10 pairs.
Scum win if they avoid all three attempts, so they win at 1 - (1/21 + 1/15 + 1/10) which is a cool 78.5%, leaving the town random win rate at 21.5%.
I think the problem with your maths is you treat the lynch as two separate events, where actually the only way to avoid mislynching is to hit both scum at once.
You only mislynch the second time if you mislynched the first time, though, so you're not calculating in the given factor there.
I'm not sure that's correct. If you treat each pair as a single possible lynch, there are 21 pairs only one of which is the actual scum team. After one NK, with 6 alive there are 15 pairs. Day 3 with 5 players is 10 pairs.
Scum win if they avoid all three attempts, so they win at 1 - (1/21 + 1/15 + 1/10) which is a cool 78.5%, leaving the town random win rate at 21.5%.
I think the problem with your maths is you treat the lynch as two separate events, where actually the only way to avoid mislynching is to hit both scum at once.
You only mislynch the second time if you mislynched the first time, though, so you're not calculating in the given factor there.
I don't follow, any pair other than the scum is a mislynch, not any pair which contains zero scum.
Also the problem in your maths is that the odds of a D1 mislynch is 1 - (2/7*1/6), since once you hit one mafia the odds of hitting the other is 1/6 not 2/6.
I'm not sure that's correct. If you treat each pair as a single possible lynch, there are 21 pairs only one of which is the actual scum team. After one NK, with 6 alive there are 15 pairs. Day 3 with 5 players is 10 pairs.
Scum win if they avoid all three attempts, so they win at 1 - (1/21 + 1/15 + 1/10) which is a cool 78.5%, leaving the town random win rate at 21.5%.
I think the problem with your maths is you treat the lynch as two separate events, where actually the only way to avoid mislynching is to hit both scum at once.
You only mislynch the second time if you mislynched the first time, though, so you're not calculating in the given factor there.
Yeah, Eco has the problem with Osie's math. 2/7*1/6 = 4.7% chance of a correct lynch D1, not 2/7*2/6 = 9.5%. Or just see that there are 21 pairs, so the chance of guessing the right pair is 1/21.
I'm not sure that's correct. If you treat each pair as a single possible lynch, there are 21 pairs only one of which is the actual scum team. After one NK, with 6 alive there are 15 pairs. Day 3 with 5 players is 10 pairs.
Scum win if they avoid all three attempts, so they win at 1 - (1/21 + 1/15 + 1/10) which is a cool 78.5%, leaving the town random win rate at 21.5%.
I think the problem with your maths is you treat the lynch as two separate events, where actually the only way to avoid mislynching is to hit both scum at once.
You only mislynch the second time if you mislynched the first time, though, so you're not calculating in the given factor there.
I don't follow, any pair other than the scum is a mislynch, not any pair which contains zero scum.
Also the problem in your maths is that the odds of a D1 mislynch is 1 - (2/7*1/6), since once you hit one mafia the odds of hitting the other is 1/6 not 2/6.
I don't think that's (edit post #666) correct, but it's probably easier to look at the town win percentage, which is just a flat 1/21 + 1/15 + 1/10 = ~21%.
Of course it doesn't count for on-pair kills, it's pure random lynch.
I don't think that's correct, but it's probably easier to look at the town win percentage, which is just a flat 1/21 + 1/15 + 1/10 = ~21%.
No... Because if you win the first round, the second round doesn't happen. (20/21) of the time, the second round happens. Assuming random scumkills, some of that time you would remove an extra pair from the picture and have 1/14 instead of 1/15.
And a similar principle applies to the third round.
I don't think that's correct, but it's probably easier to look at the town win percentage, which is just a flat 1/21 + 1/15 + 1/10 = ~21%.
No... Because if you win the first round, the second round doesn't happen. (20/21) of the time, the second round happens. Assuming random scumkills, some of that time you would remove an extra pair from the picture and have 1/14 instead of 1/15.
And a similar principle applies to the third round.
Oh yes, I believe you are correct. Still, that only gives a 20% town win rate as opposed to 21%. Not a great place to start.
I don't think that's correct, but it's probably easier to look at the town win percentage, which is just a flat 1/21 + 1/15 + 1/10 = ~21%.
No... Because if you win the first round, the second round doesn't happen. (20/21) of the time, the second round happens. Assuming random scumkills, some of that time you would remove an extra pair from the picture and have 1/14 instead of 1/15.
And a similar principle applies to the third round.
Oh yes, I believe you are correct. Still, that only gives a 20% town win rate as opposed to 21%. Not a great place to start.
Post 666 is assuming that scum always removes in-pair, which is not necessarily the case. I was very careful not to remove in-pair because of it coming back to bite us with WIFOM later on.
Wow, that was certainly something. Main takeaway for me is that in such a setup, days would need to be extended if any slot goes inactive. Town must be able to solve each person as town or mafia independently, and you really can't do that when the slot is not there. We had nothing from Atog all d1 and only got a replacement by d2. So effectively the game only had 2 viable phases instead of three. We got a little out of Anak d1, but it's still pretty much dart throwing at that point.
Otherwise the setup is scumsided but fair, which is what we knew going in.
I have issues with the site's page load times and whatnot, and I didn't much care for some of the things said in the game, but otherwise it was a good experience and there was fun to be had, which is all we can ask for.
Well played Gj in particular for not being scummy during your time in the game. That's all you really gotta do to win, if town is looking for scummy behavior. Thanks especially to rand and tom for being shiningly townie during play. Good effort shadow and anak, sorry I couldn't resolve your slots correctly based on what I saw. Osie, no hard feelings.
I do want to apologise to everyone, including Pizza again, for going off base and accusing Pizza of cheating. That sort of talk saps the fun out of the game fast, and I regret it wholeheartedly.
I'll likely be taking a break from mafia for a while due to Meatworld time constraints, but minus that snag, this was a great game, and I look forward to seeing all of you on-sitr in the future.
It's good. I only wanted to note that such things shouldn't come into play as a tactic, or even as a legitimate concern; it should always go to the game host if it were legitimate, reported to mods after the game.
During play it can be provocative, inaccurate, and otherwise people can read into it for alignment.
For example, suppose someone thought the team was Osie-Pizza for some reason.
A post like that almost immediately discounts such a team, because most people would read into it that Osie would not do that to partner pizza, and if he did, it would be such a bad way to win the game. People would get salty in post.
That's why I try to avoid posts which make it clear we can't be a team, because we're appealing to emotion too much and it looks too real, like two folks that legit dislike each other. On MU there's even a policy against discussing mod actions and reports during play because people read into that for alignment.
Basically, anything that can possibly impact people's reads but which don't take place in the game.
Your own moderation duties with regard to randwolf's posts, for example, may be better handled by another mod, with maybe a quick public note in an oog sense that heads up this is against the rules.
People wouldn't read too much into that for alignment, but it was hard to read a randwolf-Osie team after some of your posts as a moderator.
And that's not me looking to angle the game, it's just something I have to try to overlook and it's really hard to. When it's in your head you can't delete it, right?
This is all intended as constructive feedback. From what I understand you're relatively newish to the game, and moderation responsibilities sometimes can conflict with play. It's tough to know where to draw the line.
I don't take it personally, and I tried my best to get us back to a discussion that was alignment reads based on just our behavior as much as possible. That's what I aim for, and if that's wrong then that's wrong and it's all good.
Townies have a right to be wrong, and so do scums. That's part of it being a guessing game.
We just want to avoid things that make some guesses invalid that have little to do with the game itself. Avoid those as much as possible, in general, and that's a good idea for everyone, not just mods.
I'd love to be able to say I caught wolves because I'm good, not because I had access to angly information. I want the game to be as hard as possible, that's why I signed up to this one. I hate detectives and I hate vigs because I want to solve it, not have the game on easy mode.
As tough as a loss this game was, I'm still grateful for the opportunity to try to win it. That's where all the fun is derived.
tl;dr apology very accepted and no hard feelings, as I said. If any of this is constructive or useful, feel free to run with it. I had fun, and I don't hate anyone here.
The GJ way path to no lynching:
The GJ way path to no lynching:
Votecount 3.1
Anak - Osieorb (1): Pizza
Pizza - TCM (3): Anak, Osieorb, TCM
Pizza and TCM are....... NOT the scum team!
As the game will now inevitably be won by the scum (2v2 is a scum win here), I can declare that TCM and Osieorb have won the game!
Thanks everybody for playing and putting in the effort for a silly and most likely very scumsided setup! Thoughts welcome of course.
"You cannot vote for a team which includes yourself."
Thoughts on the above also welcome.
GG scum!
I feel like maybe just on day 1, state how many wolves are in that pairing, but not who they are (assuming you want to keep the set-up vanilla). The rule you ended basically has to have the entire town vote for your pairing, or 1 "bus" vote. Strongly dislike that. Town could also use that mechanic to try and go for a double clear.
The GJ way path to no lynching:
The day one only scum count reveal might be an option, but I think it's very powerful if there's *zero* scum in the pairing. Scum are effectively forced to kill within the pair.
And that's not even accounting for the fact that the baseline random-vote chances for this are an abysmal 21% or so.
On the other hand, yes, being told that there are no scum in the opening pairing actually rewards Town for poor play, because clearing two people is really powerful. It still leaves things Mafia-favored (Town baseline chance of winning off a no-scum opening guess is 25%, I think?), but it doesn't feel good.
The solutions I can see off-hand (such as voting for an individual instead of a pair on day one, but not lynching the individual even if he's scum), while they resolve the issue mathematically and help with the problems like two scum bussing each other all game, are unfortunately clunky as hell and feel dumb from a flavor perspective.
I'ma sleep on this, cuz I like the concept of pair-voting and wanna help figure out a setup for it that isn't a free scum win.
My original plan was to just treat the game as a town game. Same process and everything, which in context worked out. This plan steadily fell apart a bit over time, though. Like I had this readlist:
+ Confirmed Town
+ ~ Townread
~ + Townlean
~ Null
~ - Scumlean
- ~ Scumread
- Hard Scumread
— Confirmed Scum
+ Osieorb18
X + Tom
X + Randwolf
+ ~ Shadow
~ Anak
~ - GJ
- ~ Pizza
Non-Lurking Teams:
Anak + Pizza
GJ + Pizza
Anak + GJ
~~~~~~~~~~~~~
I buddied up some to Silver, but not nearly as blatantly. That was a pretty big mistake here, trying to take advantage of how comfortable you can be around me.
I calculated out the math for this in an earlier post. I might have made a mistake on the later rounds, but I don't think so.
Scum win if they avoid all three attempts, so they win at 1 - (1/21 + 1/15 + 1/10) which is a cool 78.5%, leaving the town random win rate at 21.5%.
I think the problem with your maths is you treat the lynch as two separate events, where actually the only way to avoid mislynching is to hit both scum at once.
Anak, sorry, I was pretty sure you were scum by the end there.
GJ, well done, short game and you managed to fool me just enough that I was soft town reading you.
You only mislynch the second time if you mislynched the first time, though, so you're not calculating in the given factor there.
I don't follow, any pair other than the scum is a mislynch, not any pair which contains zero scum.
Also the problem in your maths is that the odds of a D1 mislynch is 1 - (2/7*1/6), since once you hit one mafia the odds of hitting the other is 1/6 not 2/6.
1-(1/21)
.95238095238095238096
(1-(1/21))*(1-(1/15))
.88888888888888888890
(1-(1/21))*(1-(1/15))*(1-(1/10))
.80000000000000000001
Except this doesn't account for whether or not scum kills someone on-pair.
That is true.
Of course it doesn't count for on-pair kills, it's pure random lynch.
No... Because if you win the first round, the second round doesn't happen. (20/21) of the time, the second round happens. Assuming random scumkills, some of that time you would remove an extra pair from the picture and have 1/14 instead of 1/15.
And a similar principle applies to the third round.
To be fair, the moment that you town-cleared me for was completely legitimate.
Oh yes, I believe you are correct. Still, that only gives a 20% town win rate as opposed to 21%. Not a great place to start.
Post 666 is assuming that scum always removes in-pair, which is not necessarily the case. I was very careful not to remove in-pair because of it coming back to bite us with WIFOM later on.
Scumchat part 2 is attached as a screenshot.
Ghost Pizza
Lynchbait Pizza
Summon Pizza
Love Pizza
Life Pizza
Excellent entrance Pizza
Shadow Pizza
Tom Pizza
HOW YOU LIKE ME NOW PIZZA
Town Pizza
Giffy Pizza
Iffy Pizza
Rotten tomato Pizza
Tom (ato) Pizza
Pizza Wolf
Hey Pizza
Lock Wolf Pizza
Oh snap Pizza
Vote on Pizza
Daaaamn Pizza
Wall of Pizza
Thoughts on Pizza
Null Pizza
Lock Pizza
Pizza for Breakfast
Reread Pizza
Not Pizza
Overestimating Pizza
Randwolf Pizza
Townlean Pizza
Ticking Pizza
Anak Pizza
Good Pizza Analysis
Lynching Pizza
Doubt yourself Pizza
Don't read Pizza
Killed Pizza
Right Pizza
Slight tinfoil on Pizza
Follows Pizza
Overwhelmingly townie townie Pizza
Reversing on Pizza
Inside Pizza
GJ Pizza
Garbage Pizza
Sorry Pizza
Wow, that was certainly something. Main takeaway for me is that in such a setup, days would need to be extended if any slot goes inactive. Town must be able to solve each person as town or mafia independently, and you really can't do that when the slot is not there. We had nothing from Atog all d1 and only got a replacement by d2. So effectively the game only had 2 viable phases instead of three. We got a little out of Anak d1, but it's still pretty much dart throwing at that point.
Otherwise the setup is scumsided but fair, which is what we knew going in.
I have issues with the site's page load times and whatnot, and I didn't much care for some of the things said in the game, but otherwise it was a good experience and there was fun to be had, which is all we can ask for.
Well played Gj in particular for not being scummy during your time in the game. That's all you really gotta do to win, if town is looking for scummy behavior. Thanks especially to rand and tom for being shiningly townie during play. Good effort shadow and anak, sorry I couldn't resolve your slots correctly based on what I saw. Osie, no hard feelings.
Thanks Megs for hosting. See you guys around!
sorry for getting sick
ty for mod
I'll likely be taking a break from mafia for a while due to Meatworld time constraints, but minus that snag, this was a great game, and I look forward to seeing all of you on-sitr in the future.
During play it can be provocative, inaccurate, and otherwise people can read into it for alignment.
For example, suppose someone thought the team was Osie-Pizza for some reason.
A post like that almost immediately discounts such a team, because most people would read into it that Osie would not do that to partner pizza, and if he did, it would be such a bad way to win the game. People would get salty in post.
That's why I try to avoid posts which make it clear we can't be a team, because we're appealing to emotion too much and it looks too real, like two folks that legit dislike each other. On MU there's even a policy against discussing mod actions and reports during play because people read into that for alignment.
Basically, anything that can possibly impact people's reads but which don't take place in the game.
Your own moderation duties with regard to randwolf's posts, for example, may be better handled by another mod, with maybe a quick public note in an oog sense that heads up this is against the rules.
People wouldn't read too much into that for alignment, but it was hard to read a randwolf-Osie team after some of your posts as a moderator.
And that's not me looking to angle the game, it's just something I have to try to overlook and it's really hard to. When it's in your head you can't delete it, right?
This is all intended as constructive feedback. From what I understand you're relatively newish to the game, and moderation responsibilities sometimes can conflict with play. It's tough to know where to draw the line.
I don't take it personally, and I tried my best to get us back to a discussion that was alignment reads based on just our behavior as much as possible. That's what I aim for, and if that's wrong then that's wrong and it's all good.
Townies have a right to be wrong, and so do scums. That's part of it being a guessing game.
We just want to avoid things that make some guesses invalid that have little to do with the game itself. Avoid those as much as possible, in general, and that's a good idea for everyone, not just mods.
I'd love to be able to say I caught wolves because I'm good, not because I had access to angly information. I want the game to be as hard as possible, that's why I signed up to this one. I hate detectives and I hate vigs because I want to solve it, not have the game on easy mode.
As tough as a loss this game was, I'm still grateful for the opportunity to try to win it. That's where all the fun is derived.
tl;dr apology very accepted and no hard feelings, as I said. If any of this is constructive or useful, feel free to run with it. I had fun, and I don't hate anyone here.