Total Posts:8|Showing Posts:1-8
Jump to topic:

Mestari's Modding Manifesto

Lucky_Luciano
Posts: 4,350
Add as Friend
Challenge to a Debate
Send a Message
6/2/2013 8:05:54 PM
Posted: 3 years ago
I'll explain my design philosophy here because I hope other mods consider it in the future. The game in which I tested this philosophy as a thought-experiment was Romance of the Three Kingdoms Mafia (http://www.debate.org...).

1. Conceptual Coherency

Role interactions need to make sense. Don't give the mafia a watcher (Royal) unless they can (A) use it as a fake claim or (B) use the results in some manner. If the mafia has a watcher there must either be no town watchers, they must be paired with an even/odd night town watcher, or they must have some sort of role that manipulates night actions to utilize the knowledge they gained. That's like making the mafia consist of a Godfather, a watcher, and a role cop. None of it works together. At best the role cop can investigate visiting roles that the watcher sees. Even then all they can do is kill the roles they investigate, which would not be significant because their knowledge would always be delayed a NP. If you are going to have these roles, the mafia has to be able to use that knowledge. They have to have a role stealer, or a recruiter, or even a redirector that acts AFTER they receive the watcher result. Roles have to make sense with one another. That is how TUF's role came into fruition. Basic conceptual analysis will determine that TUF's role only makes sense if Logic's role works in the manner that he claims it does. Make the game design make sense. Stop saying that the town needs X amount of investigative roles, Y amount of protective roles, and Z amount of negative utility roles. Start saying, does X role make sense if Y role isn't in the game? If it's not, do I want that gap of intuition to be designed into the game for structural reasons? If the answer is no, add the damn role. Would you have a godfather without a cop? Only if you are an idiot or are intentionally bastard modding the mafia, so don't do it.

Furthermore, be careful with your wording. Don't tell somebody they have an unstoppable kill if they do not. No, there cannot be an unstoppable kill and an unstoppable protection in the same game. This not only misleads the players by granting them the perception that they are free to use their actions as they please when there are consequences that they were explicitly told that do not exist, but it will frustrate them, making the game unenjoyable. (http://www.debate.org...)

Moreover, make sure the culmination of your roles are coherent. I don't care how many investigative roles there are, everybody can be an investigator or there can be none. However, if there are none then don't add framers or godfathers unless your intention is to be misleading. If there is an absurdly high amount, add framers and godfathers unless your intention is for the mafia to lose. This isn't about calculating +1s or -1s or deciding how many mislynches the town should have. This isn't about running the statistical odds of either side winning. It's about gut-checking your design with common sense.

Finally, gut-check for intuitiveness. I have learned never to use a PGO. Visiting roles intuitively are supposed to visit. They see this as beneficial. It's okay for their visits to fail, but to have them die for doing something that is only natural, visiting with a visiting role, is simply frustrating because it requires players to adopt a counter-intuitive paranoia when making their night actions. The existence of jesters also do this. They make lynching scummy players counter-intuitive. Jesters have never made a game more fun, and neither have PGOs. Thank goodness DDO doesn't seem to use jesters anymore.

2. Structural Stability

Structural stability is the ability of players to use their actions freely. A game should not be designed to where a player must have an A-grade competency to play a given role. That means the role is simply designed to be played in a manner you see as efficient. If they cannot take multiple courses of action and do well, get rid of the role, it has no place in your game. Now this doesn't mean that they have to have more than 2 options. I think TUF's role in Romance of the Three Kingdoms was a nice compliment to Logic's as a counterbalance to mis-killing. He basically had a yay or nay to blocking the kill. This is enough decision. However, if a player will be punished for not acting in a way YOU see fit, then the role is bad. Take for example a 1x vig and a jester. You may think that the 1x vig will allow the town to have an answer to the jester role, but you also just pigeonholed the vig into one possible night action. They are punished for not killing the jester because either (A) the mafia has to kill him, or (B) the town loses.

Another example of a bad role would be Yraelz's fake-claim in Alexandra Quick Mafia. Let's pretend that it was real (http://www.debate.org...).

The claim was basically,

My win condition is to survive the game, I can win along with anyone else who survives. My role is invincibility, I am unlynchable and I can't be night killed. Doing either will waste the initiators actions. [The Mod's] specification on the issue indicates that the NK will just do nothing, and that the mislynch will just become a no lynch.

The reason I'm unlynchable and unkillable is because I'm a third party target. Someone, still in this game (based on the flips) has a spell which will eliminate me from the game if used 2x in a row. I can't win with the other TP because their win condition is my death.


The problem with this role is two-fold. First it is conceptually incoherent. A gut-check shows that there is not enough responsiveness available among the other players. It would be like having 5 normal cops with no godfather, no framer, and no way to mess with their results. There is nothing the other team can conceptually do to respond to that. But even if that weren't an issue, the player who had to kill Yraelz would have had their creative freedom structurally impeded upon. Their only purpose was to kill Yraelz. On top of that, every other party had their freedom structurally impeded upon as well. They had no way to kill Yraelz, and their only options were sharing a win with Yraelz, i.e. no solos except for his killer, or take second if the killer succeeded. There was nothing they could do to win the game on their own, the game defined their faith. This is an example where the structure of the game does not provide a stable foundation for creative freedom. Thank God this role wasn't real.

You should be able to take a set of equally skilled players, with undefined playstyles and consistency ratings, and assign them roles/affiliations randomly and have the game play out relatively well each time. Relatively well does not mean that one side one smash the other, but the game should always feel coherent and should allow the players to feel free with their actions. Landslides happen, so be it. That's a part of mafia. Don't restrict players because you want a 'close game.'

3. Volatility

I didn't test volatility in my Romance of the Three Kingdoms game, but it's something I would like to further study. Volatility is the ability for a small set of actions to landslide the game, and how many of these sets exist in a given game. I think I would place my Shakugan no Shana game as a 10 for volatility and my Slaves of Time game as a 1. I think an ideally designed game would be around a 7. I will look into this in the future.

Disclaimer

I do not claim to have found the perfect framework for designing games. I don't think any of my games have been 'ideal.' I do, however, believe that I put more thought into game design than most people and have a good understanding after my last experiment, I mean game, to put forth a solid framework. With that being said, I want to hear
"Age is not important" - Airmax 2014
"Australia... is that a place?" - Airmax 2014
Lucky_Luciano
Posts: 4,350
Add as Friend
Challenge to a Debate
Send a Message
6/2/2013 8:06:38 PM
Posted: 3 years ago
your critiques as mafia is an evolving game and as such our meta-discussions ought to evolve as well.

*Cut off in the prior post due to word count.*
"Age is not important" - Airmax 2014
"Australia... is that a place?" - Airmax 2014
FourTrouble
Posts: 12,757
Add as Friend
Challenge to a Debate
Send a Message
6/2/2013 8:40:19 PM
Posted: 3 years ago
I completely disagree about the volatility of Slaves of Time being 1. I have a lot to say, obviously, some good some bad. I'll get to it later.
Lucky_Luciano
Posts: 4,350
Add as Friend
Challenge to a Debate
Send a Message
6/2/2013 8:41:47 PM
Posted: 3 years ago
At 6/2/2013 8:40:19 PM, FourTrouble wrote:
I completely disagree about the volatility of Slaves of Time being 1. I have a lot to say, obviously, some good some bad. I'll get to it later.

If not a 1 then a 2, maybe a 2.5 if you want to push it. With two mafiosos after 4 and a cult dead in effectively 2 DPs, the game should have been over much sooner than it was.
"Age is not important" - Airmax 2014
"Australia... is that a place?" - Airmax 2014
FourTrouble
Posts: 12,757
Add as Friend
Challenge to a Debate
Send a Message
6/2/2013 8:48:10 PM
Posted: 3 years ago
At 6/2/2013 8:41:47 PM, Lucky_Luciano wrote:
At 6/2/2013 8:40:19 PM, FourTrouble wrote:
I completely disagree about the volatility of Slaves of Time being 1. I have a lot to say, obviously, some good some bad. I'll get to it later.

If not a 1 then a 2, maybe a 2.5 if you want to push it. With two mafiosos after 4 and a cult dead in effectively 2 DPs, the game should have been over much sooner than it was.

The game dragged on but the outcome was clear given the design of the game. I think Logic's death was far more volatile in effect than you imagine. I sincerely believe one of the mafia's most important weapons is the night-kill, and robbing them of that in a game where the town effectively got 2 lynches per Day was totally imbalanced. I know others will disagree, but I personally think F-16's play there should have got him the win, and the only reason it didn't was because the outcome of the game was decided after N1. Yes, the game dragged on for a long time. It didn't matter because the game's outcome was already set.

Anyway, volatility is the one thing I think your games suffer from (they're all too volatile), and it is something I personally don't like in games. I know maybe some people might like volatility, so perhaps it is personal preference, but I wouldn't like a game with a volatility rating above 5 (that itself being subjective, of course). I'd give Slaves of Time a rating of 7, at least.
drafterman
Posts: 18,870
Add as Friend
Challenge to a Debate
Send a Message
6/2/2013 9:07:52 PM
Posted: 3 years ago
At 6/2/2013 8:05:54 PM, Lucky_Luciano wrote:
I'll explain my design philosophy here because I hope other mods consider it in the future. The game in which I tested this philosophy as a thought-experiment was Romance of the Three Kingdoms Mafia (http://www.debate.org...).

1. Conceptual Coherency

Role interactions need to make sense. Don't give the mafia a watcher (Royal) unless they can (A) use it as a fake claim or (B) use the results in some manner. If the mafia has a watcher there must either be no town watchers, they must be paired with an even/odd night town watcher, or they must have some sort of role that manipulates night actions to utilize the knowledge they gained. That's like making the mafia consist of a Godfather, a watcher, and a role cop. None of it works together. At best the role cop can investigate visiting roles that the watcher sees. Even then all they can do is kill the roles they investigate, which would not be significant because their knowledge would always be delayed a NP. If you are going to have these roles, the mafia has to be able to use that knowledge. They have to have a role stealer, or a recruiter, or even a redirector that acts AFTER they receive the watcher result. Roles have to make sense with one another. That is how TUF's role came into fruition. Basic conceptual analysis will determine that TUF's role only makes sense if Logic's role works in the manner that he claims it does. Make the game design make sense. Stop saying that the town needs X amount of investigative roles, Y amount of protective roles, and Z amount of negative utility roles. Start saying, does X role make sense if Y role isn't in the game? If it's not, do I want that gap of intuition to be designed into the game for structural reasons? If the answer is no, add the damn role. Would you have a godfather without a cop? Only if you are an idiot or are intentionally bastard modding the mafia, so don't do it.

I disagree that a mafia watcher needs to have an additional role in order to be useful. Knowledge is inherently useful. Roles that visit are, on average, more powerful than roles that don't visit. If a Townie appears on your watcher report, then it allows you to better prioritize your night kill. It helps the mafia strategize and who to focus on.

I completely disagree that a mafia watcher is a free fake claim, but I 100% disagree with giving mafia free fake claims anyway. The entire point of winning as mafia is to fake your identity. It's like giving the Town a free public investigative result.

Furthermore, be careful with your wording. Don't tell somebody they have an unstoppable kill if they do not. No, there cannot be an unstoppable kill and an unstoppable protection in the same game. This not only misleads the players by granting them the perception that they are free to use their actions as they please when there are consequences that they were explicitly told that do not exist, but it will frustrate them, making the game unenjoyable. (http://www.debate.org...)

I totally agree with this. This happens, basically, from a lack of proof reading ones roles. Mods need to consider (as much as feasible) how all of the roles interact with each other. Especially roles which explicitly modify other roles (strong, redirect, BUS DRIVE).

Moreover, make sure the culmination of your roles are coherent. I don't care how many investigative roles there are, everybody can be an investigator or there can be none. However, if there are none then don't add framers or godfathers unless your intention is to be misleading. If there is an absurdly high amount, add framers and godfathers unless your intention is for the mafia to lose. This isn't about calculating +1s or -1s or deciding how many mislynches the town should have. This isn't about running the statistical odds of either side winning. It's about gut-checking your design with common sense.

Agreed. Some games have seemed like the mod just stuck their hand in a grab bag of roles and threw them together. On the other some games like that become that way through tweaking, and can happen even with experienced mods. I've had games that, as a result of a modification, made other roles nonsensical or pointless. Like having a roleblocker or bus driver in the Mafia when the Town only has 1-2 active roles that can be roleblocked or bus driven.


Finally, gut-check for intuitiveness. I have learned never to use a PGO. Visiting roles intuitively are supposed to visit. They see this as beneficial. It's okay for their visits to fail, but to have them die for doing something that is only natural, visiting with a visiting role, is simply frustrating because it requires players to adopt a counter-intuitive paranoia when making their night actions. The existence of jesters also do this. They make lynching scummy players counter-intuitive. Jesters have never made a game more fun, and neither have PGOs. Thank goodness DDO doesn't seem to use jesters anymore.

Jesters don't exist.
drafterman
Posts: 18,870
Add as Friend
Challenge to a Debate
Send a Message
6/2/2013 9:13:18 PM
Posted: 3 years ago
At 6/2/2013 8:05:54 PM, Lucky_Luciano wrote:

2. Structural Stability

Structural stability is the ability of players to use their actions freely. A game should not be designed to where a player must have an A-grade competency to play a given role. That means the role is simply designed to be played in a manner you see as efficient. If they cannot take multiple courses of action and do well, get rid of the role, it has no place in your game. Now this doesn't mean that they have to have more than 2 options. I think TUF's role in Romance of the Three Kingdoms was a nice compliment to Logic's as a counterbalance to mis-killing. He basically had a yay or nay to blocking the kill. This is enough decision. However, if a player will be punished for not acting in a way YOU see fit, then the role is bad. Take for example a 1x vig and a jester. You may think that the 1x vig will allow the town to have an answer to the jester role, but you also just pigeonholed the vig into one possible night action. They are punished for not killing the jester because either (A) the mafia has to kill him, or (B) the town loses.

Another example of a bad role would be Yraelz's fake-claim in Alexandra Quick Mafia. Let's pretend that it was real (http://www.debate.org...).

The claim was basically,

My win condition is to survive the game, I can win along with anyone else who survives. My role is invincibility, I am unlynchable and I can't be night killed. Doing either will waste the initiators actions. [The Mod's] specification on the issue indicates that the NK will just do nothing, and that the mislynch will just become a no lynch.

The reason I'm unlynchable and unkillable is because I'm a third party target. Someone, still in this game (based on the flips) has a spell which will eliminate me from the game if used 2x in a row. I can't win with the other TP because their win condition is my death.


The problem with this role is two-fold. First it is conceptually incoherent. A gut-check shows that there is not enough responsiveness available among the other players. It would be like having 5 normal cops with no godfather, no framer, and no way to mess with their results. There is nothing the other team can conceptually do to respond to that. But even if that weren't an issue, the player who had to kill Yraelz would have had their creative freedom structurally impeded upon. Their only purpose was to kill Yraelz. On top of that, every other party had their freedom structurally impeded upon as well. They had no way to kill Yraelz, and their only options were sharing a win with Yraelz, i.e. no solos except for his killer, or take second if the killer succeeded. There was nothing they could do to win the game on their own, the game defined their faith. This is an example where the structure of the game does not provide a stable foundation for creative freedom. Thank God this role wasn't real.

You should be able to take a set of equally skilled players, with undefined playstyles and consistency ratings, and assign them roles/affiliations randomly and have the game play out relatively well each time. Relatively well does not mean that one side one smash the other, but the game should always feel coherent and should allow the players to feel free with their actions. Landslides happen, so be it. That's a part of mafia. Don't restrict players because you want a 'close game.'

Totally, 100% agree. As a mod, I understand the desire to have construct roles with a certain interaction in mind, but all mods need to understand that players will use the roles as they see fit! This is no more evident than in blackhawk's LOTR mafia where Logic broke the game through innovative use of his role. You have to consider how the role be used outside of your intentions.

One thing to consider is this: If every townie full claimed, could they construct a gamebreaking strategy?
drafterman
Posts: 18,870
Add as Friend
Challenge to a Debate
Send a Message
6/2/2013 9:15:25 PM
Posted: 3 years ago
At 6/2/2013 8:05:54 PM, Lucky_Luciano wrote:

3. Volatility


I didn't test volatility in my Romance of the Three Kingdoms game, but it's something I would like to further study. Volatility is the ability for a small set of actions to landslide the game, and how many of these sets exist in a given game. I think I would place my Shakugan no Shana game as a 10 for volatility and my Slaves of Time game as a 1. I think an ideally designed game would be around a 7. I will look into this in the future.

It's called swinginess. In line with the previous mention of having in mind certain role interactions, don't make any role a lynch-pin, such that it's loss ruins the game.

However, there are plenty of roles (enablers, disablers, back-ups, JOATs) that inherently reduce swinginess

As a rule, more numerous powerful roles, more complex roles, tends to increase swinginess. Go simple.