Yet more on gangs

[LennyLeak]

Okay, first off this is not a suggestion to have gangs in the game. The reason for writing this post is that though most seems to agree that gangs should and will be implemented, Chris and Mark have been very quiet about it as of late. The last thing I've heard from them about it, was when they stated months ago, that they could not figure out a way to implement it, that would be viable and function as a good game mechanic. I just want to talk about a viable approach to implementing gangs, and what is necessary for gangs to be a viable game mechanic. Other topics have been created about this I know, and I hope I do not offend anyone by starting a new topic. The other topics have drifted low, and has different outsets. I believe that a new discussion will be more beneficial than continuing and old one. I wish to start from scratch and talk about the core concept of gangs. I apologize in advance for less than stellar spelling. I am Danish and speak English a lot better than I write it.

It seems to me that gangs are such an important aspect of prison culture around the world, that it can not be ignored. Even in Denmark, were the crime rate is lower and the prison system more progressive than most places, gangs are a fundamental fact of prison life. Some will say, that if gangs are not a major contribution to the game, then it should not be implemented. I think, that since gangs are such a major part of prisons, gangs should be implemented as long as it does not detract from the fun of the game. Having gangs will in itself be a contribution to the game. I do however think, that gangs can be implemented such, that it increases the fun and makes for a good game mechanic.

The problem that gangs present
For gangs to be a viable game mechanic, they must present a challenge for the player to solve. It seems that in most prisons, the existence of gangs makes crime and violence increase within the prison, while decreasing sense of security for other inmates. Therefor it makes sense that gangs in PA should have the same effect. Since keeping violence and crime down within you PA prison, and keeping prisoners needs met (those are of course connected), it seems straight forward that gangs in PA should do exactly that. The presence of gangs should increase the likelihood of violence while decreasing inmates’ sense of security... Often prisons try to deal with the gang problem by identifying and segregating gang members, and keeping them more controlled than regular inmates - even high security inmates. The problem is, that gang members will often not be open with staff about their gang affiliations, because that would lead them to lose privileges and freedom. The problem is thus to identify gang members, and keep them away from other inmates.

The solution that is available
Many prisons have a special segment of staff detailed to identify and classify gang members, so that they can be separated. They have different names in different prisons, but Gang Classification Unit (GCU) seems to be an intuitive and fairly common name. The GCU is staffed by a small group of people, whose sole job is to identify gang members (by looking at inmate tattoos and intelligence from inmate correspondence etc.). A suspicion is not enough to take away privileges, so confirming an inmate as a gang member is a continuing process, where several "markers" have to be identified. When enough indications are found, the inmate can be technically identified as a gang member. Once gang members have been positively identified, the inmate is often moved to another cell block specifically for gang members. Again the names of these varies, but Secure Housing Unit (SHU) seems to be a popular name. In the secure housing unit inmates will have less freedom, and the staff pr. inmate ratio tends to be higher. This is a costly process but keeps gang violence from escalating and makes other inmates feel more secure.

How this relates to Prison Architect
The problem and solution seems to be fairly straight forward, and IMO is a good example of a useful game mechanic. The presence of gangs will create a problem of increased violence and crime in the prison, which in turn also makes inmates feel less secure and thus unhappy. You can solve the problem by identifying and segregating gang members, which in turn will cost you money for staff and new facilities. To be more specific, you could give each inmate a Boolean variable like 'isGangmember', and if an inmate isGangmember he would be more likely to commit crime in your prison (violence and contraband). That way the more gang members you have, the more crime would go up. To solve this you could have a new room 'GCU' that would be an office that needed some staff. When a GCU is up and running, the office would identify those gang members so you could separate them. That could be done by another variable. If an inmate isGangmember, he would also have a hidden variable that designates how close he is to being confirmed a gang member - lets call it ConfirmationStatus. and say it has a value of 0-1. If a GCU was running then every time a gang member committed a crime in your prison his ConfirmationStatus would rise, and the more staff in the GCU the more it would rise, making classification faster. When it hits 1 the inmate is confirmed as a gang member, and flagged as a gang member in the UI. Maybe you would even get a notice from the GCU. To accommodate the segregation all you have to have is a new classification beside low, med and max security. Have another one called 'SHU' or whatever, and you would already have all the tools for segregating those inmates. That way you could keep them in a more secure wing, and make sure they would not have access to places where they could steal contraband, and they would not have a negative impact on other inmates. You thus spend money on staff and facilities to keep crime down and happiness up.

Why this is not just another Max security
Of course, some similarities between gang members and regular max sec is bound to happen. But there are important differences that makes this viable and not just a redundant mechanic. Gang members does not necessarily have to be max sec in the first place, so they could be a sort of meta-max sec. But the important thing is the unknown factor. Having a GCU to classify gang members would give an entirely different feel to gang members than regular max sec. Of course you can up- or downgrade prisoner security based on behavior, but with gang members it's all about the race to identify them and keep them separate. What is really interesting about doing gangs however, is that is opens up a lot of possibilities with gang-on-gang interaction that I will discuss below.

Simple gang structure
With the system above in place you would have a solid foundation, and you could structure the gangs as simple or complex as you would like. You could for instance just stick to the simple system that I outlined, and assume that all gang members are just generic gang members. What would be interesting though, would be to differentiate gangs. Lets assume that we have 5 different gangs, gang 1-5. How those gang members relate to each other could result in some interesting scenarios and choices for the player. A very common thing with gangs is, that some gangs do not get along very well. By adding another variable to anyone who 'isGangmember', you could differentiate between gangs, and set the stage for some dramatic interaction. Lets say that anyone who isGangmember also has as variable of GangAffiliation 1-5. That would of course be associated with a string variable, and a GangAffilitation=1 would be a name of the gang. Gangs 1-5 would have a relation status with all of the other gangs, which would dictate each members interaction with members from other gangs. Different gangs might be at war, and a gang member with GangAffiliation=1 might attack any gang member with GangAffilitation=2 on sight, while being passive towards those with GangAffilitation=3. GangAffiliation would also be revealed when the GCU identified someone as a gang member, and thus the player would have to decide how best to house the different gang members. Would he just say fuck it, and house them all together, or would he spend money on a more secure SHU that separated different gangs. (An evil warden might even decide to assign a gang member who attacked staff to a cell in a housing unit with rival gang members). The amount of gang violence would differ depending on how well you identified gang members and how careful you are in housing them. That can alter how secure inmates feel, but could also affect the score and value of your prison in late-game.

More complex gang system
The above system could be expanded into an even more elaborate gang system structure. I will point out three possible ways of making the system even more interesting.

Dynamic gang relations: The relations between different gangs could be dynamic and change over time, either at random or even possibly from events. The GCU could then also chart the different relations between gangs. So gang1 might make peace with gang2, or even team up with gang2 to take on gang3. This would force the player to be constantly aware of the gang situation, and perhaps re-house different gangs as their relations grew sour. Much of this could just happen in the code through changing variables, but the gameplay effect would be huge, and IMO really interesting.

Gangception: Another possible and very realistic aspect could be gangs within gangs. In real life, prisons often have different layers of allegiance. Inmates are part of sub-gangs (sets or cliques) that are part of real gangs, that again is segregated into races. These function like different hierarchical layers of gangs, and if something kicks off at race level all gangs and sets will unite in that fight, even though the different gangs within a race might not be on good terms. In the same way different sets might not be on good terms, but will unite in a fight against a rival gang. This could be handled in the same way as the simple gang structure, just with several layers where the top layers trump any allegiance in sub layers. I think this would make a really interesting mechanic, and force the player even more to think about housing and prison layout.

Shot callers: Gangs in prisons have 'shot callers'. Those are leaders who run the day-to-day on the tiers. There are several interesting ways to incorporate this, but I have two in mind. Another Boolean variable would determine if someone isShotcaller. A small percentage of those who isGangmember would also be shot callers. For every shot caller in a wing/section of the prison, other gang members with the same GangAffilitation would be even more prone to crime. So shot callers would make the gangs an even bigger problem. Another things they could do would be to recruit other inmates in the same prison wing, who are not already in a gang - effectively turning isGangmember=false to isGangmember=true in other inmates, and giving them the same GangAffilitation. More shot callers would increase the recruit rate. This would make the shot callers very dangerous to have on your tiers, and make it very important to identify them and segregate them, so that they only come in contact with other gang members - or preferably no one at all. To identify them you could have the same system as with regular gang members. Gang members who isShotcaller could have a variable 'ShotcallerConfirmationStatus 0-1'. If you have a GCU, that value would increase whenever a new gang member is recruited or a gang member of his gang attacks another gang member. Once ShotcallerConfirmationStatus=1 they are flagged as Shot callers in the UI, and you know what they are, and can act accordingly.

The final note
At the start of the post I outlined the basics for gangs as an interesting game mechanic. Those basics are really all that is needed to make gangs a fun and interesting game mechanic IMO. Having a simple gang structure with gang-on-gang interaction would make it way more fun though. Implementing some features of the more complex gang system would make it even more interesting, and very lifelike. I think that even the basic system will be a great asset to the game, because what the system really affects is prison design. Because PA is a game about designing a prison, I think that new features that increases difficult decisions in design should be welcomed. Having a good cost/benefit decision about how much to control the gang problem is also very interesting. Gangs create violence, but dealing with them is costly. How much do you care about prison violence and other inmates welfare?

I hope to have given a bit of food for thought on this, at least I spend quite some time thinking about this. A lot of it might be intuitive, but I think it helps to get those thoughts down side by side, and a bit structured. Thanks for reading this wall of text. I really hope to see gangs in the game soon, and I think there are lots of ways it could be done in an interesting manner. Hopefully my post touches on some of those.

[snarst]

+1

Not sure about the gangs within a gang though.

[henke37]

Agreed about the subgang thing. I think that prisoners having individual opinions about others is a better solution. Gang affiliation should be nothing more than another factor when deciding the opinion.

[LennyLeak]

Thanks for your feedback. Personally I think it would be interesting to see racewars and such, but I also recognize that this is probably the hardest point on the list to implement nicely, because it requires a link between sprite selection and gang affiliation. I agree that this is perhaps not viable, as it would be a hazzle and not add all that much to the system. I think all three points from the 'complex gang system' are optional and could stand alone. I would not expect to see all three - if any - implemented into PA. I that the basic system I outlined would fit nicely, and having dynamic gang relations and shot callers would add a lot to that system. I can see how the gangception thing would be more touble than it is worth though.

[111none]

+1
My personal ideas on top of this.
Should there be some type of like informant who can tell you somethings about their recent activities? But then Have to be In a heavily secured area to avoid revenge crimes?

Also there needs to be an error ratio of the gang identification.

Lastly, not sure if this is possible coding wise, but maybe gangs can plan escapes for an important inmate, smuggle things in for money?

[LennyLeak]

+1
My personal ideas on top of this.
Should there be some type of like informant who can tell you somethings about their recent activities? But then Have to be In a heavily secured area to avoid revenge crimes?

Also there needs to be an error ratio of the gang identification.

Lastly, not sure if this is possible coding wise, but maybe gangs can plan escapes for an important inmate, smuggle things in for money?

I like the idea about informers, and have thought about som ways to do that before, but I have a had time seing it as a fresh game mechanic. Would be really cool if your GCU could debrief gang members that wanted out, and get info on other gang members. In general it would be cool to have protective custody, but there would need to be a reason for it. It has to add something more than just another level of security status, otherwise it becomes redundant. Having protectitive custody opens up the idea of having comitted certain crimes makes an inmate more likely to be attacked. I think it would be cool, but maybe it's a bit over the top. I dont know.

[LegoDino77]

All I want is to have people whit colored sweatbands to stick togetherr in yard and build tunnels together, and fight the gangs that have different colour sweatbands ! (Also I wanna the people in forestry (when that comes) hide guns and weapons in trees )

[blacksythe]

There is already a workable solution for the shotcallers - the controlling trait, by using this no extra code is needed to determine who runs the gangs.

As for the gangs themselves I like the idea of gangs having a relationship table implemented that looks a bit like a multiplication table

Gangs 1 | 2 | 3 | 4
......1 X | F | H | N
......2 F | X | F | N
......3 H | F | X | N
......4 N | N | N | X

Key;
X = N/A
F = Friendly
H = Hostile
N = Neutral

As this is early game I say that to start with gangs have a 33% change of being hostile or friendly with another gang and a 34% chance of being neutral
Using a simple table that is dynamic means new gangs can be created;

Gang 4 are a new gang and at this early stage basically would perform like normal prisoners (but when near each other would increase the safety need of prisoners not in or near their gang)

Gang 2 are a friendly happy bunch and have banded together for protection and have no enemies (but when near each other would increase the safety need of prisoners not in or near their gang)

Gang 1 and 3 are hostile towards each other and are more likely to commit crimes against these gangs (in the future this could be set up for gang fights when to many crimes are committed against 1 by 3 or vice versa)
Again if gang 1 are together it makes other prisoners or gang members on their own uneasy and
if gang 3 are together it makes other prisoners or gang members on their own uneasy.

If a new prisoners with hasTrait="controlling" after 24 game hours he starts recruiting and thus Gang 5 is added to the table

If all the members of the gang are killed or get out of jail (escape or finish their sentence) the gang is removed from the table (cleanup just to be tidy)

[dequ]

+10
This would make the gameplay very interesting, and i think it's so relevant, and easy (since it basicly copy/paste of other functions but with other names) to implement, that it would be a deal breaker to not get gangs in PA (At least in the final product).

[Ric666]

Seems to be getting a little over complicated to be just one small part of the game.
Also, traditionally gangs in prisons are sorted by different race. Do IV really want to get into the racial stuff? Who knows....if so, the best way may be to just make up gang names completely. Names like "The snakes" or something very generic. It shouldn't be country specific either as we've seen the mistaken news topics on that recently.

I've never been too bothered about the need for gangs.

However, each to their own. It's not a bad idea. It would just need to be thought about very carefully on how they would introduce them to the game & how it would effect the general day to day running of the prisons. It almost leads to more problems.

May write more on this later when I get time.

[screaminasian]

i'd love to see riots and targeted assassinations that shift the power balance between gangs in the prison.

[Jureth]

You could have it where affiliations change so it would be more of a matter of moveing prisoners in and out of shu.

[Raaaahman]

I read the entire OP, and I'm sorry but i don't see any game mechanics that is "interesting" IMO. You seem well documented, and I won't talk about technic details you put in it because I don't know *** in game development.

But as a gamer, adding one more security level with new specialized staff members isn't very appealing. I like the simple min/med/max switchable security levels because it force me to make choices: Did I trust the security level the prisoner come in with? Am I more afraid of kleptomaniacs (which can do every kind of misconduct more efficiently), of smart evaders or of agressives brutes?

If you add the gang layer on top of that, you'll have to choose about isolating gangs (it would be broken actuammy with more than 3 gangs in fact) or putting the more dangerous members into highest security area.

Having the sole "isGangmember" system without any more complexity would be worst than not having gangs at all IMO. All you have to do, is waiting for your GCU to identify gang members for you and then segregating them into "high+ security " areas.

All I want is to have people whit colored sweatbands to stick togetherr in yard and build tunnels together, and fight the gangs that have different colour sweatbands !

That's more about what I want to see with gangs. Some crew mechanisms that made gangs being more agressive towards each other while being able to help their friends, effectively bringing up the risk level in your prison but with precise goals, like eradicating rival gang, or attempting mass evasion.

I believe it should start with some friends & foes system, with a "socialization" need and affinities/antipathy betweens your inmates that made them team up together when a fight/riot break up even if some of them have fully satisfied needs.

Always good to share opinions, by the way.

[111none]

I like the idea about informers, and have thought about som ways to do that before, but I have a had time seing it as a fresh game mechanic. Would be really cool if your GCU could debrief gang members that wanted out, and get info on other gang members. In general it would be cool to have protective custody, but there would need to be a reason for it. It has to add something more than just another level of security status, otherwise it becomes redundant. Having protectitive custody opens up the idea of having comitted certain crimes makes an inmate more likely to be attacked. I think it would be cool, but maybe it's a bit over the top. I dont know.

Hm.. Maybe that they would go into a certain type of spy-warfare? Like if you had prisoners in the kitchen, they would poison the food?
I see your point though. In that case, rapists should be under protective custody as well. (I believe I watched a Tv show on a Miami prison, and how that the rapists are threatened almost daily, one guy said something like: "We are robbers, murders, pickpockets, but we ain't the ones to go around raping people; those B******'s don't deserve to live." Could someone remind me of the Tv show if possible? Thanks)

Edit, maybe this can be added into these?
http://forums.introversion.co.uk/viewtopic.php?f=43&t=47194&p=558198#p558198

(Look for the Rules of engagement part.)

[Scorpy_CFS]

I think that things like gangs should be closer to end game content but and could be handled as a risk = multiplier based on number of gang members in the prison. This means that in the early phases, even if you did indeed have gang members, they wouldn't be as big a problem as when you have a larger facility. One could argue that the larger your prison, the more incentive and prestige for gangs to want to 'own' it, and thus the increased risk of having unmanaged gang links in your prison.
I like most of the ideas suggested thus far. I also remember the devs mentioning not having our prisons in a bubble later down the road but rather have effects from the prison bleed into the hypothetical city and vice versa. I think that this would be great in that one could have random gang events (Kind of like disasters or the like in city builders) that would have happened in the city that would impact on your prison gang affiliations. Also, unmanaged gang links causing problems within your prison could lead to an increase in random negative events within the city (and thus a snowball effect) whereas well managed gang threats would reduce the chance of negative random gang events in the city.
I hate to bring up Simcity but I think it used a good description of how PA should be and that is that as you push your prison further, your prison 'pushes back.' The bigger your prison grows, the more aspects you need to account for.
I think a gang system could be very well integrated into existing structure.
ie, Before purchasing any gang unit infrastructure, your psyche could give you a general gang risk level similar to the contraband chart.
Once a gang unit officer is purchased (Let's not make them stupidly expensive like the armed guards) they are able to give you a general risk level for each gang and can give you a detailed analysis of gang relations.
Over time as gang members are identified the general risk levels are replaced with specific risk.
When your gang unit members are patrolling cells where gang members are located, they have the chance of intercepting messages for specific events that will alert you ahead of time to assassination attempts, or general disquiet, etc. Without intercepting these messages, it would just seem like random gang violence appears from nowhere but after you invest in a gang unit and they start spending some time doing their job, one can begin preempting hostilities.
Also, gang unit presence coupled with good rehabilitation systems like education could actually increase the chance of gang members dropping their affiliation during their stay or after their release, thereby decreasing negative random gang events in the city. Having said that, as a push back mechanic each consecutive gang member encouraged to leave a gang would be successful so one can never truly eliminate gangs in their prison. (Unless killed of course) But then it's only a matter of time before another affiliate is locked up and thus the situation repeats.

Anywho, I see gang systems as a valuable end game interest as I think we all can get to a point where our prisons are running flawlessly and everything becomes quite routine. At that point I know I manufacture discontent by going into lockdown. Doing unnecessary shakedowns or just leaving my front door open until someone does something stupid. I see a well done gang system as always keeping your prison teetering on the edge of something interesting happening but not unmanageably. Ie, one doesn't need a high discontent level for sudden breakout attempts or attacks on other prisoners, etc. But when it does happen it is unlikely to end in a riot unless you've really neglected that aspect of managing your prison.