Crime is one of the properties mechanics currently in game that tries to reflect the real life problems when it come to controlling cities. In game, it something you want to keep as low as possible, and if a city doesn't control it, it will probably end on a death spiral. I wanted to touch two topics on this topic: the first part about balance changes to balance the mechanic (and which I think most of us can agree on) and the second part about how I think the mechanic should really work as to make it more engaging and also more AI friendly
Crime balance is currently really odd. On the lower levels of crime, you get some really harsh penalties (-14 gold of mugging for 150 crime can be deadly to the economy) as soon as you hit an arbitrary level of crimes. Transition should be way smoother , with gradual levels of occurance for each crime, so a small amount of crime doesn't hurt that much and can be more easily controllable( so mugging at 50 crime would be -1 gold, at 150 -3, at 300 -8 etc). And crimes should increase the chance of discovering criminals at the higher levels (buildings like hideouts would still protect then) to try to mitigate the current tendency of badly manage cities (read:AI) to became completely unsalvageable for their owners. Also revolution modifier should be taken out of crime autobuilds. The amount of unhappiness and unhealthyness a city would produce at higher levels will probably cause a revolts for sure (normal buildings should also have their modifiers taken away- revolutions should be influenced by the civics and the situation of the city, not by buildings that break the system by making it way too stable). Finally crime per pop should probably be adjusted by era, so that while on early ones is lower, it gets higher on the latter ones (to make early crime less brutal, specifically on newly founded cities, while being higher latter when the players can control it easier)
That were some suggestions to balance the current system. However, I think that the problem with crime is in how it works, and that a more radical approach would solve death spirals and be more AI friendly while also reflecting better real life situation were crime always exists, what matters is at what amount. My suggestion is to make crime a system with several equilibrium points. What I mean with that? Basically, that crime will find an stable position,say 150 crime, and be around that level counteracting smaller changes on the level, unless a big change occurs and the system move to a new equilibrium where it becomes stable. This will be a really big change in how crime works: while in the current system a thief would be always increasing crime unless it is counter by anti-crime buildings and law enforcement units. With the equilibrium points system, the amount of damage a thief causes will be limited even without being counter, and the same for the effectiveness of law enforcement units. The system will also make the equilibrium point harder to move the higher the level of crime is, making law enforcement more effective, and vice versa for reducing it. Thus it would be really hard to reach the insane amount of crime without significant investment by foreign actors, and non-existant crime would be a very expensive option
Combine this with the balance system I propose(more specifically, making crime autobuilds and I think we would get a more interesting system for the player who could have more liberty at balancing crime, trying to find a sweet spot where the benefits of lowering the level(ex:less unhappinessl would be smaller than the benefits of keeping it, instead of the current gameplay where there is no decision to be made, always trying to keep crime down to make sure it doesn't shoot up to the sky. Law enforcement units would still be vital to counter criminals (and with most cities having some crime, spawning will be more common yet less harsh), while death spirals would be a very rare occurrence.Here is an example to make myself more clear
Example: we got a city which is hovering around 125 crime (equilibrium point number 1) and the difference between positive crime and negative crime is +1 while the amount needed to change it is +3/-15. A rogue enter the city increasing crime by 3. The equilibrium is broken and the crime level start to increase at +4 per turn as the equilibrium as been broken, reaching a new level at 250, with the amount needed to change the equilibrium now being +5/-12. If the thief is caught during the transition, the level will go back to the previous equilibrium point. If however the new equilibrium point has been reached, the player will need to get to -12 to get it back to 125
To conclude, the system would be make even more interesting if the equilibrium points were meta-stable, meaning that a city could sometimes experience spikes of crime that could move crime up to a higher equilibrium, or periods were crime moves down without player intervention, similar to a chaotic system but that's for another moment if my idea ever gets implemented
Crime balance is currently really odd. On the lower levels of crime, you get some really harsh penalties (-14 gold of mugging for 150 crime can be deadly to the economy) as soon as you hit an arbitrary level of crimes. Transition should be way smoother , with gradual levels of occurance for each crime, so a small amount of crime doesn't hurt that much and can be more easily controllable( so mugging at 50 crime would be -1 gold, at 150 -3, at 300 -8 etc). And crimes should increase the chance of discovering criminals at the higher levels (buildings like hideouts would still protect then) to try to mitigate the current tendency of badly manage cities (read:AI) to became completely unsalvageable for their owners. Also revolution modifier should be taken out of crime autobuilds. The amount of unhappiness and unhealthyness a city would produce at higher levels will probably cause a revolts for sure (normal buildings should also have their modifiers taken away- revolutions should be influenced by the civics and the situation of the city, not by buildings that break the system by making it way too stable). Finally crime per pop should probably be adjusted by era, so that while on early ones is lower, it gets higher on the latter ones (to make early crime less brutal, specifically on newly founded cities, while being higher latter when the players can control it easier)
That were some suggestions to balance the current system. However, I think that the problem with crime is in how it works, and that a more radical approach would solve death spirals and be more AI friendly while also reflecting better real life situation were crime always exists, what matters is at what amount. My suggestion is to make crime a system with several equilibrium points. What I mean with that? Basically, that crime will find an stable position,say 150 crime, and be around that level counteracting smaller changes on the level, unless a big change occurs and the system move to a new equilibrium where it becomes stable. This will be a really big change in how crime works: while in the current system a thief would be always increasing crime unless it is counter by anti-crime buildings and law enforcement units. With the equilibrium points system, the amount of damage a thief causes will be limited even without being counter, and the same for the effectiveness of law enforcement units. The system will also make the equilibrium point harder to move the higher the level of crime is, making law enforcement more effective, and vice versa for reducing it. Thus it would be really hard to reach the insane amount of crime without significant investment by foreign actors, and non-existant crime would be a very expensive option
Combine this with the balance system I propose(more specifically, making crime autobuilds and I think we would get a more interesting system for the player who could have more liberty at balancing crime, trying to find a sweet spot where the benefits of lowering the level(ex:less unhappinessl would be smaller than the benefits of keeping it, instead of the current gameplay where there is no decision to be made, always trying to keep crime down to make sure it doesn't shoot up to the sky. Law enforcement units would still be vital to counter criminals (and with most cities having some crime, spawning will be more common yet less harsh), while death spirals would be a very rare occurrence.Here is an example to make myself more clear
Example: we got a city which is hovering around 125 crime (equilibrium point number 1) and the difference between positive crime and negative crime is +1 while the amount needed to change it is +3/-15. A rogue enter the city increasing crime by 3. The equilibrium is broken and the crime level start to increase at +4 per turn as the equilibrium as been broken, reaching a new level at 250, with the amount needed to change the equilibrium now being +5/-12. If the thief is caught during the transition, the level will go back to the previous equilibrium point. If however the new equilibrium point has been reached, the player will need to get to -12 to get it back to 125
To conclude, the system would be make even more interesting if the equilibrium points were meta-stable, meaning that a city could sometimes experience spikes of crime that could move crime up to a higher equilibrium, or periods were crime moves down without player intervention, similar to a chaotic system but that's for another moment if my idea ever gets implemented