Most of improving how we play mafia is going to come down to the players improving rather than through game design.
We need to lynch and Night-kill players who don’t participate. Vigilantes are especially useful for this.
Not just new players, but old players who approach the game differently ;). Players need to look for motivations rather than heterodox word usage and tactics.
I agree. But the problem is a lot of players cannot distinguish someone arguing with them or disbelieving them from an actual personal attack.
For my part, I always refer to others in the game as a “player”, not “person” or “people”. I hope it gives a reminder that we’re in a game.
Let me amend that: True skepticism is pro-Town. But that includes not only not assuming that a player must be Scum, but also not assuming that a player must be Town.
But yes, this is one point where my playstyle often seems to cause some friction. I’ll place an early vote for someone, and someone will ask me if I really think that player is probably Scum. To which, on early days, the answer is usually “no”: I think that that player is more likely to be Town than Scum, even though I’m voting for them. You don’t need to think that someone is more than 50% likely to be Scum to vote for them; you just need to think that they’re more likely to be Scum than anyone else.
This is of course a very important point. That’s why I exhort people to think probabilistically. If 3 out of 4 people are Townies and you think someone is only 60 percent likely to be a Townie themselves, then they’re actually “>rand scum” and a fine candidate to push.
Also, a corollary to my “realize that most people are on your team”: clearing people – that is, figuring out that some people are more likely to be Town – can be just as important as finding scum, for reasons I’ve explained on a few occasions. For one thing, it allows you to narrow your pool of prospective scum candidates, and thereby sharpens and focuses your analysis.
I don’t actually think the game ended if the scum was killed. Their items went up for a vote, and whoever received it (was voted the item by other players) became the new scum.
… which, now that I think of it, explicitly breaks a stated rule in the game, that there were no forced alignment changes.
I also think we should emphasize that “mafia” is strongly luck-based. We’re mainly playing a murder mystery story and “being right” and “being wrong” are largely just a matter of chance. I think this difference is why no-reveal games work in other venues. By accepting that the lynch is essentially random, it removes the angst of “being wrong.”
That is funny, and a little weird. I’d think my posting style and play tactics were fairly unique.
I didn’t follow that game at all–never enough time to read a game I’m not playing in.
Those are good tactics and ones I use myself. But a lot of people simply don’t think analytically (and it’s probably better for the town if there’s a diversity of approaches). I just wish players wouldn’t get attacked for analytic thinking–some players think “I can’t understand that reasoning, so that player must be scum”. Not helpful to town.
I think that’s only true if the players are about the same skill level. Good players definitely improve the chances of their team winning. But even the best players only can expect to win slightly more often than chance dictates.
~ ~ ~ ~ ~
Oh, I thought of something that really irks me: meta-gaming. By which I mean using any information from outside the current game. It feels ungamely to me.
Strongly disagree with all of this. Poker is largely luck-based too from instantiation to instantiation, but that doesn’t mean you can’t learn to play poker better than other people such that you have much better results over a larger sample. Similarly, there are “good” Mafia/werewolf players, “bad” Mafia/werewolf players, and players who are in-between, and you can get better by figuring out how to play. There’s an element of chance/variance/what-have-you, but at bottom it is intrinsically a skill game.
I agree entirely. In fact, I am not sure I have ever been lynched as scum for any kind of actual scum-slip or suspicious move that was actually motivated by me being on the scum team. I know, I know, someone will dig up one now, but the majority of time, I have just been lynched because of luck…or bad luck when I’m Town.
I enjoy the game a ton, though, but there is a ton of luck and not much we can do about it.
Again, I disagree. If you let me hand-select a Town team from the people I play with on my other site, they would win >>>>>>>>>>>rand against a randomly-selected Scum team of people with average experience.
Similarly, if you let me hand-select a Scum team from the people I play with on my other site, they would win >>>>>>>>>>>rand against a randomly-selected Town team of people with average experience.
Luck plays a part here and there, but the game is fundamentally skill-based.
I think it’s rare for a player to leave no detectable scum tells. I dont believe Mafia is as luck-based as you’re saying. If it was, the best way to play would be to put everybody’s name in a randomizer and lynch whoever “wins.” I absolutely believe careful consideration and logical deduction can lead to better-than-random lynches. In my mind, that’s the point of the game: town need to overcome the information difference, and scum need to overcome the numbers difference. Both sides employ strategies and tactics to achieve their goals.
A sufficiently-skilled Scum team can, if they choose, make the lynch effectively random, by playing exactly as they would if they were Town, and thus providing the actual Town with no real information. This requires one to be able to compartmentalize away the extra information they have, and in practice, I don’t know how common true mastery of this skill is, but it’s at least possible in principle. Of course, Scum can also decide at some point that their odds are improved by not ignoring their extra information, and move the game away from randomness to varying degrees.
I’m not going to get into an argument about this. I think players put too much stock into “being right” or “being wrong,” when a significant part of the game is luck-based.
I agree with you that “being right” and “being wrong” should not necessarily be valid bases for reads within a game.
I disagree with you that some people cannot end up “being right” far more often and “being wrong” far less often than other people over a large sample, given the role skill plays in this game.
Also: as you gain information within a given game, there should theoretically be less variance in the lynch targets. That is, to the extent that a d1 lynch will end up lynching a Townie x% of the time, theoretically a d2 lynch will end up lynching a Townie less than x% of the time, and so forth…not just because there are fewer Townies to lynch, but because the Town is given an increasing amount of information to use when making its decisions.
And sometimes the moderator makes it easier by not revealing the names of the other scum in the role PM. So unless scum actually tell each other their names on the scum board, one can play “perfectly”. (I consider a game that allows this to have a design error.)
Skill has its place, I just think skill has a significantly smaller role than we like to admit. Certainly a scum player with the skill to not accidentally post to the game thread when wanting to post to the scum thread is better prepared to win than one that does not. But the bottom line is that the way the game works out is strongly influenced by random things. If I really wanted to (and I used to actually think this) we could say that “Oh that game was won because the detective had the skill to select X as his investigation target and X was scum! Good job! Great skill!” But I don’t think that way anymore. Sometimes skill does play into it, but most of the times X was chosen, “just 'cause.”
