There has been a lot of discussion on this issue. And as more people begin to realize this is probably not a bug, and some consider it a 'cheat' to overcome the computation burden. However, IMHO this is a very smart design decision, with a little more additional feature to reflect the commute time, this could strike the sweet point between realism and scalability.
And let's be frank, simcity 2013 is a real agent based game, however broken it is. This game is NOT a pure agent based simulation. The game mechanic here is more similar to Ceasar III, with the addition of traffic simulation.
Why? Because the only real agents in this game are the ambulances, garbage trucks and etc, basically all the civil service vehicles and goods transports. I call them as the REAL agent because their physical presence brings certain service to a location. Really it is the same thing as the priest, engineer, goods delivery guys in Ceasar III.
Most of the cims, i.e., the workers going to and from work, are NOT real agents. Each with a certain destination in mind, they are there just to create a traffic pattern, which may slow down the REAL agents mentioned above, thus forcing us to design a rational road network. Whether they ever reach the work place or disappear in a traffic jam has no impact on the operation of the work place at all. All the game engine needs is just to find a connection to the work place when assigning one cim to a job. (Still remember the in Ceasar III, once a citizen reached a work place, it never returns home?)
Now some may call this cheat, but I thank the developer to take this approach because I can play this game on my aging laptop. And before anyone rage-uninstall the game, think twice about whether it is really necessary to simulate every single cims as an independent REAL agent? what effect do we want/need to model anyway?
I can basically think of four, and hope others can add to it:
1. affect whether work place is operating or not depending on if the cims have arrived or left.
2. impact the customer count in commercial buildings.
3. create a realistic road traffic.
4. commute time
The first effect is really difficult to implement even with each cim as a real agent. Because then should we force a same working hour for the people? No city builder game can ever re-create the real time scale/pattern.
The second one I really would like to see, but it can be done based on the traffic amount on the road, similar to Simcity 4, which would be good enough for me at least. Otherwise more elaborated economic model needs to apply.
The third one is already achieved, that means there is working pathfinding AI, even if not optimal, but people are not always rational anyway.
The last one I think has the most visible impact on the game, as it is really easy to create a situation where people cannot reach their work place without a proper transport connection. But it is not that hard to fix. Adding a commuting time count for each cim and setting an upper limit might do. For cims living at certain place, If all the available jobs have a commuting time higher than certain threshold, then the cims cannot reach their work and will abandon the house, which takes into account the distance and the congestion along the path.
(There are two uncertain assumptions about how the game work now: first, how do cim find a job? It seems to be based on matching education level+shorter distance+random factor; second, do the pathfinding AI take into account the road congestion, i.e., if one road is congested do they change to another route in the future? this is relevant because the commuting time need to be re-caculated regularly, say every game month, to check if road congestion changes or not)
Such implementation could add considerable realism to the game without needing to revamp the game into a fully agent based simulation. I am not sure how much extra memory would be needed to store the commuting time for each cim, but possibly not much. However, any mod or update tackling this issue will need thorough test on large scale cities and see if the job-associating simulation could reach an equilibrium, because the game maybe very sensitive to the commuting time threshold and mass abandonment might occur (it might also be better to set a different value of the threshold for each cim).
All in all, if I understand the game mechanic correctly I am optimistic about seeing the teleporting issue solved. Would love to hear modders and developers’s comment on its feasibility.
And let's be frank, simcity 2013 is a real agent based game, however broken it is. This game is NOT a pure agent based simulation. The game mechanic here is more similar to Ceasar III, with the addition of traffic simulation.
Why? Because the only real agents in this game are the ambulances, garbage trucks and etc, basically all the civil service vehicles and goods transports. I call them as the REAL agent because their physical presence brings certain service to a location. Really it is the same thing as the priest, engineer, goods delivery guys in Ceasar III.
Most of the cims, i.e., the workers going to and from work, are NOT real agents. Each with a certain destination in mind, they are there just to create a traffic pattern, which may slow down the REAL agents mentioned above, thus forcing us to design a rational road network. Whether they ever reach the work place or disappear in a traffic jam has no impact on the operation of the work place at all. All the game engine needs is just to find a connection to the work place when assigning one cim to a job. (Still remember the in Ceasar III, once a citizen reached a work place, it never returns home?)
Now some may call this cheat, but I thank the developer to take this approach because I can play this game on my aging laptop. And before anyone rage-uninstall the game, think twice about whether it is really necessary to simulate every single cims as an independent REAL agent? what effect do we want/need to model anyway?
I can basically think of four, and hope others can add to it:
1. affect whether work place is operating or not depending on if the cims have arrived or left.
2. impact the customer count in commercial buildings.
3. create a realistic road traffic.
4. commute time
The first effect is really difficult to implement even with each cim as a real agent. Because then should we force a same working hour for the people? No city builder game can ever re-create the real time scale/pattern.
The second one I really would like to see, but it can be done based on the traffic amount on the road, similar to Simcity 4, which would be good enough for me at least. Otherwise more elaborated economic model needs to apply.
The third one is already achieved, that means there is working pathfinding AI, even if not optimal, but people are not always rational anyway.
The last one I think has the most visible impact on the game, as it is really easy to create a situation where people cannot reach their work place without a proper transport connection. But it is not that hard to fix. Adding a commuting time count for each cim and setting an upper limit might do. For cims living at certain place, If all the available jobs have a commuting time higher than certain threshold, then the cims cannot reach their work and will abandon the house, which takes into account the distance and the congestion along the path.
(There are two uncertain assumptions about how the game work now: first, how do cim find a job? It seems to be based on matching education level+shorter distance+random factor; second, do the pathfinding AI take into account the road congestion, i.e., if one road is congested do they change to another route in the future? this is relevant because the commuting time need to be re-caculated regularly, say every game month, to check if road congestion changes or not)
Such implementation could add considerable realism to the game without needing to revamp the game into a fully agent based simulation. I am not sure how much extra memory would be needed to store the commuting time for each cim, but possibly not much. However, any mod or update tackling this issue will need thorough test on large scale cities and see if the job-associating simulation could reach an equilibrium, because the game maybe very sensitive to the commuting time threshold and mass abandonment might occur (it might also be better to set a different value of the threshold for each cim).
All in all, if I understand the game mechanic correctly I am optimistic about seeing the teleporting issue solved. Would love to hear modders and developers’s comment on its feasibility.
