The Resistance

I'm a Mafia fan, but after playing it a few dozen times with the same people it gets a bit old. We've developed a few optimal strategies and tend to repeat them with mild variation. Plus we're so competitive that we'll spend an hour without advancing the game state because it provides more information to the majority group.

This game sounds like exactly what we need. I'll see how it works in practice, but I like the no-elimination, and greater flexibility with the number of players. My friends won't touch Mafia with less than 9 players, which isn't easy to come by. Plus a change to the old formula is a good thing for us, and we're the sort to enjoy the added crunch. Thanks for sharing, Greg!

One thing that jumps out at me is, given the 'bases' method of keeping score, you'll never have more than four rounds of plans to base your decisions on. That doesn't seem like enough to me.

This game sounds a lot closer to the classic Mastermind board game than to Mafia, except that sometimes the judge can lie and give you fewer pegs (black cards) than he should.

Only the second paragraph is about The Resistance. After that I'm just criticizing the way Costik plays Werewolf and offering suggestions to John Mc on his werewolf problems.

This looks like an interesting variant that would be suited to smaller groups than Werewolf handles well, but I don't think it will scale to Werewolf's sweet range. Playing to three points is probably good for 5-8 players, but at the 12 players specified in the rules, only four rounds won't provide much information to analyze -- less than a standard Werewolf game for that many players. However at more than three points a lopsided game could drag while many players lose interest.

Speaking of Werewolf, from your description of Werewolf it looks like you were taught some rules incorrectly. For a balanced game you want 18%-25% werewolves (minimum two), not "just under half." For example, in a 15 player game you should have three wolves. In this 15 player game there could be as many as seven lynchings, of which the village needs to get three correct. Much of Werewolf is reading people, but the voting patterns and discussions from the lynched villagers give a number cruncher plenty of data to chew on. I'm lousy at reading poker faces, but I still win Werewolf games because I can find the patterns in the voting. The seer is also a key piece of the game balance, and provides more analysis data.

I agree that player elimination is arguably a problem with Werewolf, though on the other hand, I've seen people who only want to observe, and even moderated games that were broadcast via webcam for spectators. However, when Werewolf is played correctly, your "basic problem" is only valid if the players are too drunk to think about the data the game provides -- but they probably aren't too worried about any kind of game design issues then. :)

John Mc, Two suggestions for your paralyzed Werewolf group:
1. Use a timer - on both days and nights. I normally use 1 minute or less for the wolves to choose a victim and as little as five minutes (though ten is better) for the daytime deliberations. In large groups use fewer wolves than the percentages above if the day time limits are less than 15 minutes. If nobody has been lynched when the time is up, then night falls and the werewolves kill again. I've kept games as large as 35 moving with those kind of time limits. (But honestly, if you have 30 players, two games of 15 will be more fun and much easier to moderate.)
2. Vary the roles, within reason. One or more of a witch (healing potion and poison), hunter (takes someone with him), sorcerer (may expose the seer), cupid (designates lovers -- it's complicated), or pair of masons (two villagers who know each other are villagers) are good for breaking up patterns your group may have developed. Most of those roles benefit the villagers, so you may need an extra wolf to balance them. Hidden victims with priests throws all of the conventional analysis strategies right out the window. (I would not recommend priests unless the group is VERY experienced.) Once everyone knows the rules for several variant roles, you can make things even less predictable by keeping the exact list of roles in the game a secret. Lots of other variant roles are out there, but remember to use them within reason! Any one of those roles is easy to teach even tipsy players, but I've seen people go overboard, with page after page of special rules. Even the moderator gets confused in those games.

We've discussed a timer, but it's no good. A plurality of the group doesn't want a timer.

You're absolutely right that we should try out some variations. We did early on, but agreed that all the versions we knew of sucked. Also, we don't get 15 players. We get 9. So that limits what we can do. We've always played 3 Mafia, 4 Townsfolk, a Doctor, and a Sheriff. However, this generally results in a Mafia favored game. We counter-balance it with a protracted round of lynch-talk, which often leads to slip ups by the Mafia.

If you could recommend one version of the game for 9 people, what would it be?

It's worth mentioning that my group might not be normal. I think we're better at logical deduction than lie detection, and generally better at subterfuge than investigation. I, for example, have never lost as Mafia. I'm great at manipulating the votes to our favor, picking townsfolk to kill, and disguising my behavior. As Townsfolk, I'm worthless. I never realize who's Mafia, and always try to lynch the wrong guy. :P

So yea, a variant might not work for us, but I'm willing to give it a shot. In general I'm thinking The Resistance might be exactly what we need though. :)

Thanks for your interest in The Resistance. As the designer of the game, I'll comment on some of the questions about numbers of players and effects thereof.

One of the goals in the design process was to create a game that offered more substantial 'data deduction' for the players to crunch before deciding who to trust, while retaining the important element of 'personality deduction.' John Mc's group will hopefully enjoy the game for this reason.

I think the game succeeded in this effort, but it's worth mentioning that the balance changes along the scale of players. From 5 to 9 players the game involves a stronger amount of data deduction. I admit that this range is the sweet spot (for me) for the game.

From 10 to 12, personality deduction plays a stronger role because it's more difficult to remember who received plans on each round and allow for bluffing. As gschildt mentioned, I could have added more bases to be destroyed, but I found that regardless of number of players, anymore was just too much.

Memory of what has happened is integral to decision-making in the game; without this element, it's just another Werewolf. However, the requirement at 10-12 players of two instead of one sabotaged plans makes it more likely the spies will be revealed, providing that needed hard data.

Junko

I've just gotten into mafia in the last few months playing (mostly) online games. I'm sure the forum I use will give this a try.

I've started doing a mafia night as an excuse for my friends to get together and they all love it. I tend to moderate 90% of the games because the others aren't experienced enough yet, but as they get more experienced it's going to become necessary to mix the game up a bit. Resistance sounds like a fun variant.

As for how to mix mafia/werewolf up, there are lots of game decks you can build that have a nice balance with only nine players.

2 mafia, 1 doctor, 6 town
2 mafia, 1 doctor, 1 cop, 5 town
3 mafia, 1 doctor, 1 cop, 4 town

(as long as there are more townspeople without powers than there are mafia then you can add lots of other roles)

2 mafia, 1 doctor, 1 cop, 1 mayor (can prevent lynches/or 1 lynch), 4 town

there's the little girl who can peek, the miller who appears guilty when he's not. etc.

Give the mafia players power cards too.

Godfather - always appears innocent
Framer - chooses one town to appear guilty each night
Mafia doctor

One thing to remember is that the roles have to work in a party game, so people have to make their nominations silently with as little movement as possible - but there are loads of roles you can spice it up with. I made a deck in about 30 minutes, by making a table in ms.word and typing the cards out. Then I printed them on some red card and cut them out.

email me if you want me to send you the .pdf of the deck: alexbutterfield-at-yahoo.co.uk

all the cards are ones you can use in table top play. but John Mc, the best way to spice up your games would be to get a few more people and then you can really go wild with the roles, survivors, cults, serial killers.

also if you are interested in some really fun online mafia games check out www.internationalleagueofskeptics.com/forum

First, a caveat: I may not be the best person to give this advice, because I normally run Werewolf for 15+ people at game conventions, and ~85% of the smaller games I've moderated have been vanilla teaching games for beginners.

I don't think there is a standard "best mix" that is right for every group; hopefully this will address the specific problems you've mentioned (a group that is settling into predictable strategies and getting bored with the repetition, but isn't worried about how long a game takes) and include enough discussion to help you tailor it. Since this is a public domain game with lots of similarly-named variations I tried to define terminology too. Sorry about the length all that adds up to, but I hope it's useful information.

Roles:
2 wolves / mafia
1 sorcerer / informer (evil seer)
1 cupid
1 seer / detective / sheriff
1 hunter / vigilante
1 martyr
2 villagers/townsfolk
Partial role reveals on death -- everyone knows if they killed a wolf/mafia or "villager" but the other special roles are not revealed when someone dies. If there are two deaths on one night, the causes are not revealed.

The wolves or mafia are pretty standard, they know each other and select a victim each night.

A sorcerer or informer is on the evil team (he wins if the wolves/mafia win whether he lives or not). Each night he chooses a person and the moderator will tell him if he found the good seer/detective/sheriff/investigator/whatever you want to call him. :) A sorcerer or informer does NOT know who the wolves/mafia are! This type of evil seer is half to 3/4 as strong as adding a wolf/mafia member.

I recommend cupid for your group because it will create a lot of variation between games without creating total random chaos. :) At the beginning of the game, cupid chooses any two players to be lovers. Cupid can choose himself. The lovers aren't told who cupid is. The lovers will not know each others' secret roles (unless they are both wolves). The lovers can never attempt to kill each other (neither voting to lynch nor night kills). If one lover is killed, the other will die of a broken heart (or commit suicide) that same night, or the night after their lover is lynched by the townsfolk.

Cupid has no powers during the game; he becomes a normal villager/townsfolk for all purposes, though of course he knows who the lovers are.

Cupid also has a special case - if one lover is in the mafia and one is a villager, then those two players have a special victory condition. They win if and only if they are the last two people alive (to enjoy their forbidden love). Normally only the wolf/mafia half of the partnership will know this. The potential of this third team existing changes the analysis a lot. A player making two [usually inappropriate] friends be lovers is often entertaining, but you can let dice choose lovers if you prefer -- the important thing is that lovers be independent of the role cards. (And yes, one out of every 36 games will be unbalanced when cupid chooses the two wolves/mafia to be lovers. A little moderator fiat to keep that game interesting may be appropriate in your group.)

The seer or detective/sheriff chooses one person each night to view, and the moderator tells him if they are part of the mafia. The detective (seer) will not detect the informer (sorcerer), only wolves/mafia.

The hunter or vigilante can take a person of his choice with him when he dies. If it happens during the day everyone knows what happened. If a vigilante dies at night don't tell anyone which kill is which in a partial reveal game.

Martyrs can be implemented several ways. For this I'll suggest they can substitute themselves for "any" night kill -- wake them last, point out each kill and have them point to who, if anyone, they want to save and give a thumbs up, or give a thumbs down to abandon those people to their fates. Do not tell the original victim they were saved by the martyr (though sometimes it will be obvious). A little tricky with the cupid, since they can't effectively save only one of the lovers. You can handle this two ways: 1. If he tries to save a dying lover, the martyr just throws his life away, and the lovers die anyway. 2. Query the martyr when the first lover dies, if he saves the first lover, neither lover knows either of them was a target. Note: if the second lover to die is the hunter, this effectively deprives the martyr of a chance to save that hunter's victim unless you "wake" them again, which reveals more information than you may want to. A martyr is weaker than a traditional healer, who can save one person per game. To fine-tune the balance you can convert the martyr to a normal villager or upgrade them to a healer/doctor.

And finally two normal villagers. You can also shift the balance slightly in favor of the townsfolk by making these two masons -- two players who know each other are both innocent villagers.

I've included so many complex roles because you said these are players that are so experienced they playing repetitively and you need to mix things up a bit. Nights will be longer than normal since you have to wake so many people, but it sounds like your group prefers to take their time thinking things through anyway.

Re: your standard mix: I'm assuming your sheriff can view one person per night to see if they are part of the mafia (a.k.a. "investigator" or "detective"). There are several ways to implement a "doctor," but unless you are playing him as a strong bodyguard (which I don't recommend) you're right about your standard mix being having an overpowered mafia; with three wolves/mafia in a nine player game you want more good specials or the third evil to be reduced strength, which fortunately is also consistent with trying to break up repetitive strategies. I prefer healers (can save a victim once per game) and martyrs (die to protect someone else) to bodyguards (survive when they protect their charge) -- especially in small games -- because a bodyguard that lives to protect again is often game-breaking. However, sometimes a bodyguard gets unlucky and is useless. That dichotomy makes it hard to avoid games that are lopsided enough to ruin much of the fun.

http://boardgamegeek.com/thread/81173 is a balance calculator discussion; it isn't perfect because a lot of roles interact with each other, the game size, and mechanic variations in ways that are more complex than the math most moderators are willing to do, but it is a good starting place. Also note the long discussions between people who disagree about the numbers. It uses Werewolf terminology, but is flexible enough to handle even weird variants. This was built for online games, which favor the villagers because the permanent record and slower pace allow for more analysis -- which more than compensates for the inability to read facial expressions or body language when playing online. You may want to round down a little when you use it for live games.

Falling into analysis-driven patterns is a common problem with smaller groups, sometimes reaching these kinds of "pre-calculated payout table" extremes: http://www.boardgamegeek.com/thread/170732 Once everybody knows the payout table for a standard set, it's mostly a game of double-think and can feel random -- especially to someone who is new to an established group.

I think Werewolf/Mafia doesn't start to hit its stride until ~11 players. I consider the good range to be 15-25 if the players are reasonably sober and taking the game seriously. However, too many players generates more data than most people are willing to remember or analyze, so it starts to feel random again around 20-30 players (depending on how serious the group is).

I am actually very anxious to try this variant out!

We have one really big problem with Mafia/Werewolf: unequal play skill.

Basically if I or one of my friends is alive, the mafia will lose every time because we are good at investigating. This means that one of us is *always* the first to die via werewolf.

Because this variant has no death I think it should solve that problem nicely!

Thanks for all the feedback guys, I'll totally bring this up with my group and see what variants work for us. I'll start with Resistance though, and see who that goes. :)