Game Design: What works and what doesn't?
Moderator: Mafia Mods
Game Design: What works and what doesn't?
I'll post some of what I've learned about game design (and what others have shared with me) in this OP, but this thread is meant to be a discussion about what you think works and what you think detracts from your enjoyment of mafia as a player.
Different players like different things in a game, so it's impossible to please everyone. Still, there are some commonalities.
The thing that bugs people most is an unbalanced game where one side doesn't have a fair shot at winning. As A GM, I feel that I've succeeded in making a fair game when the final day of the game involves 2 or 3 town and 1 scum, or 2 town, 1 Indie, 1 scum. But, no matter how well balanced a game is on paper, some players will make mistakes, some will play brilliantly, etc.
Games with named characters present a challenge, especially if there are strongly identified good guys and bad guys in the story. If you do this, you should save a couple of major characters as alternate names for scum, and you should give some minor character names to some of the town, including specials. You don't want your game to be broken by a mass name-claim or role-claim.
Adding a serial killer can be a balancing role, because the SK will try to weaken whichever faction appears to be winning. An SK is basically a 3rd faction. The last man standing win condition is hard to acheive. Usually the SK will be either immune to views (will appear to be town to the seer) or immune to night kill (must be lynched). Sometimes the SK is recruitable to one side or the other, and the win condition changes to the recruiting side.
Common game design/GM mistakes:
1. Too many specials. Too many investigative specials can throw the game widely in favor of town. (I've been putting extra investigative roles in my ratz games to help you guys get used to no PM games. I'm slowly weaning you to a more common configuration). But the other problem with specials is that they tend to become confirmed town if they stay in the game. Toward the end, town may simply lynch the unconfirmed players and win if they have sufficent numbers.
2. Not doing run-throughs. Test your game design at least a couple of times, with random moves and with a couple of early lucky moves (seer NKed on night 1, scum don viewed on night 1, etc.). Does your game crater? Is it even possible for each side to win?
3. Adding new players after the role PMs have gone out. If your game was designed to be balanced for 18 players, then player number 19 can be a problem. Also, GMs usually make the late sign up vanilla town, and that can mess up the game balance if town players correctly assume the late addition is vanilla town.
4. Not nailing down the night move order. If you haven't decided in advance how to handle collisions (two players targeting each other for kill, etc.), there's a tendency to want to do the game-balancing thing. Decide in advance so you won't be tempted to play god.
5. Give yourself enough time to count the votes, and work through the night actions. Set a night action deadline and stick to it. Decide in advance whether unsubmitted moves are lost or are randomized. Don't risk messing up the game because you are in a hurry. You can't undo an incorrect lynch or night kill because the player's alignment is announced in most games.
6. Don't fraternize with the players. Stay out of the game thread as much as possible, don't answer questions unless they really need to be answered. Keep the game-related PMs brief. Keep in mind that GMs can drop tells, too.
Different players like different things in a game, so it's impossible to please everyone. Still, there are some commonalities.
The thing that bugs people most is an unbalanced game where one side doesn't have a fair shot at winning. As A GM, I feel that I've succeeded in making a fair game when the final day of the game involves 2 or 3 town and 1 scum, or 2 town, 1 Indie, 1 scum. But, no matter how well balanced a game is on paper, some players will make mistakes, some will play brilliantly, etc.
Games with named characters present a challenge, especially if there are strongly identified good guys and bad guys in the story. If you do this, you should save a couple of major characters as alternate names for scum, and you should give some minor character names to some of the town, including specials. You don't want your game to be broken by a mass name-claim or role-claim.
Adding a serial killer can be a balancing role, because the SK will try to weaken whichever faction appears to be winning. An SK is basically a 3rd faction. The last man standing win condition is hard to acheive. Usually the SK will be either immune to views (will appear to be town to the seer) or immune to night kill (must be lynched). Sometimes the SK is recruitable to one side or the other, and the win condition changes to the recruiting side.
Common game design/GM mistakes:
1. Too many specials. Too many investigative specials can throw the game widely in favor of town. (I've been putting extra investigative roles in my ratz games to help you guys get used to no PM games. I'm slowly weaning you to a more common configuration). But the other problem with specials is that they tend to become confirmed town if they stay in the game. Toward the end, town may simply lynch the unconfirmed players and win if they have sufficent numbers.
2. Not doing run-throughs. Test your game design at least a couple of times, with random moves and with a couple of early lucky moves (seer NKed on night 1, scum don viewed on night 1, etc.). Does your game crater? Is it even possible for each side to win?
3. Adding new players after the role PMs have gone out. If your game was designed to be balanced for 18 players, then player number 19 can be a problem. Also, GMs usually make the late sign up vanilla town, and that can mess up the game balance if town players correctly assume the late addition is vanilla town.
4. Not nailing down the night move order. If you haven't decided in advance how to handle collisions (two players targeting each other for kill, etc.), there's a tendency to want to do the game-balancing thing. Decide in advance so you won't be tempted to play god.
5. Give yourself enough time to count the votes, and work through the night actions. Set a night action deadline and stick to it. Decide in advance whether unsubmitted moves are lost or are randomized. Don't risk messing up the game because you are in a hurry. You can't undo an incorrect lynch or night kill because the player's alignment is announced in most games.
6. Don't fraternize with the players. Stay out of the game thread as much as possible, don't answer questions unless they really need to be answered. Keep the game-related PMs brief. Keep in mind that GMs can drop tells, too.
Re: Game Design: What works and what doesn't?
I'm bumping this thread, because I think now that folks have played several games, you might have more input for the ratz GMs.
I recently got hammered about a game design at TR, in particular a role that I created. Some people just plain didn't like the role, but a few folks really picked up on what turned out to be the major drawback with the role. And that drawback was the random nature of the targets.
People want to win or lose the game based on their skills, not on a roll of the dice. I know this. I feel the same way. And yet I sometimes design games with random elements anyway because it can help with a game's balance.
But, the recent reminders I've gotten have made me decide to avoid random elements. There are other ways around the game balance issues. For instance:
- instead of making a kill or protection random, make the role have night actions every other night instead of every night.
- Instead of a random target for a paranoid-style role, let the player select the target.
And whatever you do, avoid some of the design elements I used in the Tarot game, where a randomly selected player lost or gained a vote, lost or ganed a night action, etc., based on "influence cards".
Let people stand or fall based on their wits.
I recently got hammered about a game design at TR, in particular a role that I created. Some people just plain didn't like the role, but a few folks really picked up on what turned out to be the major drawback with the role. And that drawback was the random nature of the targets.
People want to win or lose the game based on their skills, not on a roll of the dice. I know this. I feel the same way. And yet I sometimes design games with random elements anyway because it can help with a game's balance.
But, the recent reminders I've gotten have made me decide to avoid random elements. There are other ways around the game balance issues. For instance:
- instead of making a kill or protection random, make the role have night actions every other night instead of every night.
- Instead of a random target for a paranoid-style role, let the player select the target.
And whatever you do, avoid some of the design elements I used in the Tarot game, where a randomly selected player lost or gained a vote, lost or ganed a night action, etc., based on "influence cards".
Let people stand or fall based on their wits.
Re: Game Design: What works and what doesn't?
This is amazing . Thanks for putting this together, Oblivion.
Badass Elf
- Uselesstwit
- Posts: 2290
- Joined: Sat Feb 27, 2010 1:32 am
- Contact:
Re: Game Design: What works and what doesn't?
I don't agree with the random thing if players have a chance to counteract it. The coin flips in ghat's game we annoying and really only served to allow more vote manipulation by the scum team. The tarot game on the other hand was probably one of the games I'm most proud of. We had all sorts of set backs for our little scum team, but we dealt with all of them. It was the hardest win I've ever had and also the one I'm happiest with. I will admit I would have been bummed if we had lost.
irretating wrote:you're a genius, UT!
Re: Game Design: What works and what doesn't?
I liked irre and sifaka's game at Ratskep .. the one where every player on both teams had a special role. The occasional one of those would be good, I think.
no fences
- Uselesstwit
- Posts: 2290
- Joined: Sat Feb 27, 2010 1:32 am
- Contact:
Re: Game Design: What works and what doesn't?
that one was a blast, even if it must have been a nightmare to GM.
irretating wrote:you're a genius, UT!
Re: Game Design: What works and what doesn't?
I don't like games where I'm lost ,no idea of how many teams there are ,no idea of who the scum are and who are the good guys and indeed even if I'm a good guy or a bad guy. The wizard of Oz game was horrible from my POV Watching a band wagon roll against a 'team' I was on but with no way to communicate with other members of my team!
If you posted you died, if you didn't you died, if you tried to explain you died
IMHO mafia works best when the set up is simple but the roles are complicated . paranoid cops ,insane seers one shot vigs etc .
One thing I learnt from thousands of hours DM'ing is that random die rolls can wreck your campaign it may be fair but fuckit we are not playing Poker for high stakes we are trying to have fun .
If you posted you died, if you didn't you died, if you tried to explain you died
IMHO mafia works best when the set up is simple but the roles are complicated . paranoid cops ,insane seers one shot vigs etc .
One thing I learnt from thousands of hours DM'ing is that random die rolls can wreck your campaign it may be fair but fuckit we are not playing Poker for high stakes we are trying to have fun .




Give me the wine , I don't need the bread
Re: Game Design: What works and what doesn't?
I get what you are saying. I design a game like the Oz one once in a great while. I actually find a game where I know there are multiple factions (even though I don't know how many) to be easier to figure out than games where a faction or two are hidden to start with and you think you're playing a town/scum game.Feck wrote:I don't like games where I'm lost ,no idea of how many teams there are ,no idea of who the scum are and who are the good guys and indeed even if I'm a good guy or a bad guy. The wizard of Oz game was horrible from my POV Watching a band wagon roll against a 'team' I was on but with no way to communicate with other members of my team!
If you posted you died, if you didn't you died, if you tried to explain you died
IMHO mafia works best when the set up is simple but the roles are complicated . paranoid cops ,insane seers one shot vigs etc .
One thing I learnt from thousands of hours DM'ing is that random die rolls can wreck your campaign it may be fair but fuckit we are not playing Poker for high stakes we are trying to have fun .
- Magicziggy
- Posts: 4847
- Joined: Wed Mar 17, 2010 8:56 am
- Contact:
Re: Game Design: What works and what doesn't?
Seconded Feck. Oz was a nightmare for me too.Feck wrote:I don't like games where I'm lost ,no idea of how many teams there are ,no idea of who the scum are and who are the good guys and indeed even if I'm a good guy or a bad guy. The wizard of Oz game was horrible from my POV Watching a band wagon roll against a 'team' I was on but with no way to communicate with other members of my team!
If you posted you died, if you didn't you died, if you tried to explain you died
IMHO mafia works best when the set up is simple but the roles are complicated . paranoid cops ,insane seers one shot vigs etc .
One thing I learnt from thousands of hours DM'ing is that random die rolls can wreck your campaign it may be fair but fuckit we are not playing Poker for high stakes we are trying to have fun .
Re: Game Design: What works and what doesn't?
It was good ,but I think we needed more clues as to what was actually happening .




Give me the wine , I don't need the bread
Re: Game Design: What works and what doesn't?
I thought Oz was great fun .. Pretty sure that was at least partly due to being picked up so quickly by my team's leader and gaining access to PM communication ... In hindsight I can see that by being completely open I probably didn't play it the best way possible for my team, but I thought it was a good way to better understand what was going on with the teams. I agree that the information given could have been less oblique and confusing ... but giving too many clues may have had too much impact on the premiss, which was for the teams to find their members while striving to meet their win condition.
I like the mix of game styles .. the different set ups and themes keeps it fresh and challenging.

I like the mix of game styles .. the different set ups and themes keeps it fresh and challenging.
no fences
Re: Game Design: What works and what doesn't?
The asylum bloodbath mafia you co GMed was tooUselesstwit wrote:that one was a blast

Looking forward to irre joining the other GM's here.
no fences
Re: Game Design: What works and what doesn't?
sorry Charlou In OZ at least you actions had a bearing on the game ,you had a team ,plans could be made . from my point of view my 'team' couldn't talk or PM the first member that died set the rest of the players hunting us because there was no apparent scum team . Even after the fact ,with the game post mortems I still have no fucking Idea what Happened ,I wouldn't mind but I was sober for most of it !




Give me the wine , I don't need the bread
Re: Game Design: What works and what doesn't?
Yeah, that's why I was lucky to have been picked up by my team leader on night one. Before that I was adrift like you and the others. Your team was particularly unlucky that your leader was lost at the start. Yeah, that was pretty unfortunate .. and perhaps unfair on your team, too.Feck wrote:sorry Charlou In OZ at least you actions had a bearing on the game ,you had a team ,plans could be made . from my point of view my 'team' couldn't talk or PM the first member that died set the rest of the players hunting us because there was no apparent scum team . Even after the fact ,with the game post mortems I still have no fucking Idea what Happened ,I wouldn't mind but I was sober for most of it !
no fences
Re: Game Design: What works and what doesn't?
At least as Miller masons we could talk ,It was like being a miller mason with a fuvcking blind fold on ! Team Wot team? the first I knew that anybody was actually meant to be on my side was when they died ,I STILL don't know what I could have done in that game apart from hide in a corner ,crash my car, get arrested ,my camera's stolen and hope nobody was heartless enough to lynch me .




Give me the wine , I don't need the bread
Who is online
Users browsing this forum: No registered users and 3 guests