Corruption
I have no problems with the aggregate corruption levels proposed, but with the execution of the model. Here's what I mean:
Right now, you have corruption in two forms (loss of shields and loss of coins) centered around twin pillars of corruption-causes. Distance and Raw Number of production centers.
The latter (number of production centers) is valid, IMO, and specifically counters (or attempts to counter) ICS. That's all well and good, but the distance factor makes no sense.
My recommendation would be to remove it from consideration entirely and replace it with a corruption factor based on government type, with distance as a remedial factor.
Examples:
Despotism: Central Control is key here, thus, for far-flung cities IN a despotism, you could expect a higher rate of corruption (which could be undone in large part with no builds at the arrival of Radio).
Monarchy: A bit more advanced. True, you still have bothersome levels of corruption, but presumdely one of the reasons you tolerate your Peers of the Realm is cos they are at least nominally loyal to you. Loyalty = Less corruption. In this case, it'd be good if there was a government-specific specialist citizen, the Justicar. Each worker promoted to Justicar decreases corruption in those cities (and obviously, the specialist is unavailable with other government types).
Republic: Similarly, a Governor's mansion/office could be built under a Republican government which would further reduce corruption.
Courthouse: Double the effectiveness, and make them Dem specific.
Communism: Labor Camp: Addresses corruption by making more productivity to counter it. Comes at the price of decreasing aggregate culture (which can be offset by using the greater productive capacity to make more cultural builds....still won't have as much as a Democracy, but you'd be a contender).
Again, none of that should be too tough to add in, and it would dramatically and decisively address the corruption problem due to distance without impacting corruption due to number of production centers to the point that the game is a walkover.
Also, each government type should come with an efficiency rating, which will further bleed productivity (borrowing from SMAC here). And, to that end, you could have some city builds that help alleviate effie problems (local precincts, lower courts, hell, perhaps even libraries could help with that!).
-Just a few more odd thoughts.
Further improving the AI
First, let me say again that Soren did a MASTERFUL job with the AI! In the ancient age, it rocks!
Trouble is, as the game advances, the AI seems to know less and less about handling the increasing flexibility of modern units. Partly, this can be addressed in the editor, by tweaking the numbers and the AI's priority flags for handling the units. Testing is currently being done with the mod proposal for that very thing.
Another thing though, that would help immensely, is to give the AI a "basic playbook" of strats to persue when fighting.
Feint, Flank, Overrun, Scortched Earth, etc.
As with a football team's playbook, the basic recepie could be spelled out for the AI, with instructions to improvise based on prevailing terrain and city layouts. As it stands now, the AI is good at massing for an attack, but it begins to fall down in terms of *executing* the attack....so the idea occurred to me, why not kinna give it a basic idea of how:
IE - The AI decides it's gonna go to war with you. It sees that it has 30 swordsmen to your 10, and 20 horsemen to your 8. Since it has the numerical edge, the decision tree points to a flank/overrun style of attack.
8 Galleys are recalled and filled to the gills with swordsmen.
These, the AI holds out of site of your units until the attack begins (with the rest of his swordsmen charging down one front, the horseman horde charging from another direction, with an eye toward laying waste to your frontier towns and pillaging roads to prevent your speedy reinforcement, and capturing workers when there's no threat from enemy troops).
Three turns after the attack begins (in order to give your own forces time to begin responding), the AI lands his flanking force of 16 swordsmen--taking advantage of the fact that most players will respond overwhelmingly to the first threat.
BOOM! The AI takes another step forward in kicking some human butt.
And playbooks like that can be devised for all the major situations the computer will be fighting in, described in broad terms to the AI so that it doesn't HAVE to cover every possible case.
-=Vel=-
I have no problems with the aggregate corruption levels proposed, but with the execution of the model. Here's what I mean:
Right now, you have corruption in two forms (loss of shields and loss of coins) centered around twin pillars of corruption-causes. Distance and Raw Number of production centers.
The latter (number of production centers) is valid, IMO, and specifically counters (or attempts to counter) ICS. That's all well and good, but the distance factor makes no sense.
My recommendation would be to remove it from consideration entirely and replace it with a corruption factor based on government type, with distance as a remedial factor.
Examples:
Despotism: Central Control is key here, thus, for far-flung cities IN a despotism, you could expect a higher rate of corruption (which could be undone in large part with no builds at the arrival of Radio).
Monarchy: A bit more advanced. True, you still have bothersome levels of corruption, but presumdely one of the reasons you tolerate your Peers of the Realm is cos they are at least nominally loyal to you. Loyalty = Less corruption. In this case, it'd be good if there was a government-specific specialist citizen, the Justicar. Each worker promoted to Justicar decreases corruption in those cities (and obviously, the specialist is unavailable with other government types).
Republic: Similarly, a Governor's mansion/office could be built under a Republican government which would further reduce corruption.
Courthouse: Double the effectiveness, and make them Dem specific.
Communism: Labor Camp: Addresses corruption by making more productivity to counter it. Comes at the price of decreasing aggregate culture (which can be offset by using the greater productive capacity to make more cultural builds....still won't have as much as a Democracy, but you'd be a contender).
Again, none of that should be too tough to add in, and it would dramatically and decisively address the corruption problem due to distance without impacting corruption due to number of production centers to the point that the game is a walkover.
Also, each government type should come with an efficiency rating, which will further bleed productivity (borrowing from SMAC here). And, to that end, you could have some city builds that help alleviate effie problems (local precincts, lower courts, hell, perhaps even libraries could help with that!).
-Just a few more odd thoughts.
Further improving the AI
First, let me say again that Soren did a MASTERFUL job with the AI! In the ancient age, it rocks!
Trouble is, as the game advances, the AI seems to know less and less about handling the increasing flexibility of modern units. Partly, this can be addressed in the editor, by tweaking the numbers and the AI's priority flags for handling the units. Testing is currently being done with the mod proposal for that very thing.
Another thing though, that would help immensely, is to give the AI a "basic playbook" of strats to persue when fighting.
Feint, Flank, Overrun, Scortched Earth, etc.
As with a football team's playbook, the basic recepie could be spelled out for the AI, with instructions to improvise based on prevailing terrain and city layouts. As it stands now, the AI is good at massing for an attack, but it begins to fall down in terms of *executing* the attack....so the idea occurred to me, why not kinna give it a basic idea of how:
IE - The AI decides it's gonna go to war with you. It sees that it has 30 swordsmen to your 10, and 20 horsemen to your 8. Since it has the numerical edge, the decision tree points to a flank/overrun style of attack.
8 Galleys are recalled and filled to the gills with swordsmen.
These, the AI holds out of site of your units until the attack begins (with the rest of his swordsmen charging down one front, the horseman horde charging from another direction, with an eye toward laying waste to your frontier towns and pillaging roads to prevent your speedy reinforcement, and capturing workers when there's no threat from enemy troops).
Three turns after the attack begins (in order to give your own forces time to begin responding), the AI lands his flanking force of 16 swordsmen--taking advantage of the fact that most players will respond overwhelmingly to the first threat.
BOOM! The AI takes another step forward in kicking some human butt.
And playbooks like that can be devised for all the major situations the computer will be fighting in, described in broad terms to the AI so that it doesn't HAVE to cover every possible case.
-=Vel=-
Comment