Let's talk about what's going on with the Law Units

[Thunderbrd]

So I keep hearing that people are finding that they and the AI are running up against problems with training Law Enforcement units at particular points. It has been said that they need to not obsolete so quickly and so on.

Thus, we should have a conversation about this because I've looked at this about 5 times to make sure we have no further problems and yet players continue to feel there are. So we need to figure out what's making them so difficult to train.

The design basic on these, up to modern at least, is to obsolete the unit when the next unit after the next upgrade becomes available. Thus, Watchers, for example, which are made available at Deception, don't obsolete when their upgrade, Enforcers come into play at Conduct but rather obsolete at Dualism, when Town Watchmen are unlocked.

Another design principle is that each unit requires a building to be trained and the building becomes available at the same tech level as the unit it unlocks. The buildings should also follow the same obsoletion rules as the units, obsoleting 2 steps away, while the next building is an upgrade of the previous one.

The building prerequisites for units were recently updated so that the previous law enforcement unit type can also accept the next level building as fulfilling the prerequisite. So the Watchers can still be trained if you have the Sentry Post (but for whatever reason still can't train Enforcers.)

In the modern, this gets a little more complicated because a couple of Law Enforcement unit types are intended to be concurrent to one another, with each concurrent LE unit being a little better at one roll or another. Police Cars are better at patrol and general crime reduction, while Riot Police are better at arrests (combat). The Police APC replaces the Riot Police while the Helicopter is best at spotting criminals but cannot make arrests itself. (Thus the Police Car, Police APC and the Police Helicopter are intended to all work together in their cities.)

Now, one thing I think I'm going to need to do, for the sake of the AI if not the players, is to remove transitional buildng prereqs that the prerequisite LE unit buildings may also require. For example, the BUILDING_GUARDHOUSE requires a Castle. The AI may not be able to make this leap in understanding the need for a castle enough to be getting Guards in their cities. Thus I might have to remove the Castle prerequisite.

Another thing that I found was screwing up the City Guard was the BONUS requirement of Ammunition to be problematic because Ammunition wasn't actually something that came into play until later (apparently early ammo is personally made so Ammunition wasn't in the game until some techs later than the City Guard was being enabled so the Ammunition prerequisite had become a proxy tech prerequisite that was much later than it should've been.) So Bonus prereqs may be causing some trouble in some spots as well.

What I need is for us to all take a look at this. A deep look at each step of the path, and figure out what might be causing further not yet noted issues. Apparently the sense that the player and AI are still unable to keep a consistent access to the most current LE types is still a problem in the game.

So to that end I have made sure to update these online charts for review (I've highlighted the LE unit lines):
1) Expanding Unit Review - Prereq Techs
2) Expanding Unit Review - Prereq Buildings
3) Expanding Unit Review - Prereq Bonuses

I wish I had the building's building prereqs and the building's Bonus prereqs charted out but I do have their techs charted:
4) Building Costs by Tech (Scroll down until you see the blue highlight line for the LE buildings.)

It may well be that the problems we're having lurk in the Building or Bonus prereqs for the prerequisite buildings being a bit too restrictive.


So aside from noting that these buildings need to not have any other building prerequisites so as not to confuse the AI, can you spot what may be causing problems for you and/or the AI?

 

[Raledon]

Tech prereq/obsolete-

The buildings should also follow the same obsoletion rules as the units, obsoleting 2 steps away, while the next building is an upgrade of the previous one.

Police cars act as the same type of LE as riot-police. The two steps rule should be applied for each chain of LE units.
we have 3 concurrent LE lines:
watcher->enforcer->patrol->guard->city guard->sherif->riot police ->swat teams ->sentinel
police car->police APC->police mech
police helicopter->police mech
With police car and riot police being the same tier, and later the helicopter joins them. Thus, the two step should be applied to their respective lines. so the sheriff should become obsolete when the swap teams comes into play and police car when we get the police mech (currently, the sheriff expects to go away when the police car comes). Otherwise, it should be a one and half steps, such as when the era ends (so sheriff obsolete @ modern era). This should give enough time to for the players to transition to the new units.

EDIT: the above lines are incorrect. It's:
watcher->enforcer->patrol->guard->city guard->sherif->riot police ->police car->police mech
police APC->swat teams->police mech
police helicopter->police mech
Considering that riot police can be skipped entirely (I've just done it accidently- researched radio, cars and THEN criminology, so I've never seen them), I assumed it's like the above.

 

[Dancing Hoskuld]

In vanilla BtS it is simple - you can build a unit until you can build all its upgrades at which point it stops showing up in the "available to build" list. Can't we just go back to that?

If the buildings that are required for a unit don't obsolete at a tech but are replaced by the newer versions of the building, you will be able to build the "old" unit until you can afford to build the upgrade building.

This will require that the upgrade buildings be seen as useful by the AI. Alternately if there is no intention of having them significantly different they could auto upgrade.

 

[JosEPh_II]

Thus, we should have a conversation about this because I've looked at this about 5 times to make sure we have no further problems and yet players continue to feel there are. So we need to figure out what's making them so difficult to train.

Your focus is on the wrong cities 1st off, look at the younger cities. And your philosophy of specializing cities is Not doing well for the AI either. They do not follow this as a rule.

Copied from the bug thread:

Thunderbrd said: ↑
What about the City Guard that should've been unlocked with the Guard Station long before Deputization? What's keeping the Guard Station from being built?

Nothing. That's Not the point. Your focus is on the wrong thing.

The point is that once it goes obsolete your younger cities lose access to LE because they do Not have the building chain up to that point built and now have no LE units to use against Crime. This is another reason you see the AI moving LE units across the map. And the reason when they want trades they want Your LE units you do have, those built in your older larger cities. LE units have to have some major overlap. This becomes even more critical from the Industrial Era on thru the later eras.

Our building chains and obsolescence is starting to show it's weaknesses. And rather badly at that.

And your Criminals at 1.5 str over LE's Is OP. Especially from Middle game on. I know you will disagree but you are wrong on this one.

 

[JosEPh_II]

The design basic on these, up to modern at least, is to obsolete the unit when the next unit after the next upgrade becomes available. Thus, Watchers, for example, which are made available at Deception, don't obsolete when their upgrade, Enforcers come into play at Conduct but rather obsolete at Dualism, when Town Watchmen are unlocked.

This stops at Ren Era and does not follow that design from Industrial On.

 

[JosEPh_II]

Thus, the two step should be applied to their respective lines. so the sheriff should become obsolete when the swap teams comes into play and police car when we get the police mech (currently, the sheriff expects to go away when the police car comes). Otherwise, it should be a one and half steps, such as when the era ends (so sheriff obsolete @ modern era). This should give enough time to for the players to transition to the new units.

The sheriff even here in this proposition would go out Too Early. Modern Era and even Information Era should still have sheriff's. It just the graphics model will look outdated. Sheriffs are still used world wide to this day. From Local to State to Federal here in the USA. I'm sure Europe has something similar still being used as well.

Considering that riot police can be skipped entirely (I've just done it accidently- researched radio, cars and THEN criminology, so I've never seen them), I assumed it's like the above.

The AI does this as well, the research selections.

 

[Raledon]

The sheriff even here in this proposition would go out Too Early. Modern Era and even Information Era should still have sheriff's. It just the graphics model will look outdated.

I think the issue is more that having one line of LE active at any point. I'd expect that riot and cars would be a split to "on foot" and "on vehicle" (possible should start from when horses are introduced), with one being stronger and faster (the "on vehicle") and the other better for finding criminals and as city defense. This means that around the time you get APC and SWAT, you get an upgrade to the "on foot" types as well, so sheriff will become obsolete at the modern era.

Sheriffs are still used world wide to this day. From Local to State to Federal here in the USA. I'm sure Europe has something similar still being used as well.

I'd put them under obsolete still. The job of Sheriff is not widely used, and as far as I'm aware not many countries still employ that system. I might be mistaken and it has been renamed, but me not being aware that the job still existed is an indication. (this is on the same level as buildings and such being active nowadays but being obsoleted in c2c)

And your Criminals at 1.5 str over LE's Is OP. Especially from Middle game on. I know you will disagree but you are wrong on this one.

Criminals are generally allright at 1.5, as the law units get bonuses against them. I agree that there is an issue, though, as it's near impossible to level LE units. This means that if a criminal gets a quality up promotion, it's going to be hell for a same tier to get it without losing an army of cops.

 

[tmv]

The job of Sheriff is not widely used

In Germany we don't vote for police officers, but I am not so sure about the rest of Europe. In fact, the only part of the executive formed by popular vote is the mayor, I think (and where I live even that is a recent development). On the state and federal level only the legislative assembly is voted for (mostly by proportional representation, with all its advantages and drawbacks), and this assembly elects the head of government (there is no distinct head of state at the state level) and the judges of the constitutional court (the latter with two-thirds-majority). This system creates very powerful party leaders.

 

[JosEPh_II]

I think the issue is more that having one line of LE active at any point. I'd expect that riot and cars would be a split to "on foot" and "on vehicle" (possible should start from when horses are introduced), with one being stronger and faster (the "on vehicle") and the other better for finding criminals and as city defense. This means that around the time you get APC and SWAT, you get an upgrade to the "on foot" types as well, so sheriff will become obsolete at the modern era.

Still disagree. Sheriffs classification also includes Marshalls, again still in use today. Also DEA officers would fall under this category as well. You can not let your impression of the USA's wild west days dictate what the sheriff line is.

In Germany we don't vote for police officers, but I am not so sure about the rest of Europe.

I guess I'm missing something here but what does "vote for police officers" have to do with this line? Or this discussion. You lost me there tmv.

 

[Raledon]

I meant that it's not widely used in the world, as there are numerous countries that do not employ the system.
According to wikipedia, there are but a few countries that use the system- all of them are/were part of the Great Britain Empire (colonies such as canada and USA are included). Other countries do not employ the system, which is why I said it's not widely used- if you are not in one of these places, you will never hear/know about sheriffs.
This is in contrast to, say, a judge which is something that every country has in similar forms.

In terms of votes, here in Israel we only vote for the Knesset (the guys who make the rules), and they in turn form the government. There is no direct voting for judges or policemen. I'm not aware of any country in the region (middle east, or the the mediterranean sea) that vote for a position similar to sheriff.

 

[tmv]

what does "vote for police officers" have to do with this line?

I was under the impression that the fact that US sheriffs are elected is an essential part of the sheriff's office. It should certainly influence the way the sheriff interacts with the people. I am not aware of any other country where someone reaches an office by an election (sorry, bad english, I know).

This is in contrast to, say, a judge which is something that every country has in similar forms.

Not quite. It depends on the legal system. In countries based on civil law (like Germany, and in fact, continental Europe) a judge is much more powerful than in countries based on common law (USA, UK, etc.). There is (usually) no jury and the judge performs the investigation at trial (inquisitorial system) instead of deciding objections. There is an overview at https://en.wikipedia.org/wiki/List_of_national_legal_systems.

 

[Thunderbrd]

Tech prereq/obsolete-

Police cars act as the same type of LE as riot-police. The two steps rule should be applied for each chain of LE units.
we have 3 concurrent LE lines:
watcher->enforcer->patrol->guard->city guard->sherif->riot police ->swat teams ->sentinel
police car->police APC->police mech
police helicopter->police mech
With police car and riot police being the same tier, and later the helicopter joins them. Thus, the two step should be applied to their respective lines. so the sheriff should become obsolete when the swap teams comes into play and police car when we get the police mech (currently, the sheriff expects to go away when the police car comes). Otherwise, it should be a one and half steps, such as when the era ends (so sheriff obsolete @ modern era). This should give enough time to for the players to transition to the new units.

EDIT: the above lines are incorrect. It's:
watcher->enforcer->patrol->guard->city guard->sherif->riot police ->police car->police mech
police APC->swat teams->police mech
police helicopter->police mech
Considering that riot police can be skipped entirely (I've just done it accidently- researched radio, cars and THEN criminology, so I've never seen them), I assumed it's like the above.

As I explained, Sherrifs actually split between riot police and police cars, the riot police being the first in the line of 'best to arrest with' and 'best for crowd control' units that leads to >Police APC>SWAT Teams>Police Mech (where the lines remerge). The Police Cars are best for crime control and investigation and although they may lag in crime control alone, they maintain this role until Police Mech.

In vanilla BtS it is simple - you can build a unit until you can build all its upgrades at which point it stops showing up in the "available to build" list. Can't we just go back to that?

The strange thing about all this is I haven't changed that. I'm not even sure I specify when they forceobsolete... since there's no NEED to because they aren't limited units, they could have any forceobsolete tag usage removed and it wouldn't hurt the game. But there's still places where players aren't able to access what they should for whatever reason.

If the buildings that are required for a unit don't obsolete at a tech but are replaced by the newer versions of the building, you will be able to build the "old" unit until you can afford to build the upgrade building.

At some point it's still silly to be saying you can build a crude building though in a more advanced era, which is why the buildings don't obsolete until the second upgrade point is accessed. If you haven't upgraded the building yet then you've really neglected something.

This will require that the upgrade buildings be seen as useful by the AI. Alternately if there is no intention of having them significantly different they could auto upgrade.

That last point is certainly one being considered. I'd have to make some interesting unitprereq tags to manage that though in this case. If we start adopting this kind of approach overall, as we've been discussing, it could have a lot of benefits of untying some knots in the mod in many places.

Your focus is on the wrong cities 1st off, look at the younger cities. And your philosophy of specializing cities is Not doing well for the AI either. They do not follow this as a rule.

By, my philosophy of specializing cities, do you mean that cities should always build their own LE units? At the moment, the AI only uses one city to build its LE units, with some effort to make the cities build their own if they are on a different continent. This is how it works for the AI so to say they don't follow this can only come from a place of overlooking how they are operating. There are huge benefits in this approach, from XP to the speed of addressing the LE needs of the nation. Another thing the AI does effectively with cities on the same continent is move their LE units to the places that they are needing more help from places that are doing fine. It tends to try that first before training more if it can get away with it.

That part is working provided the AI isn't lambasting themselves with crime units and having their own criminals hanging out in their cities, which they were doing a LOT of until this fix last week.

The point is that once it goes obsolete your younger cities lose access to LE because they do Not have the building chain up to that point built and now have no LE units to use against Crime. This is another reason you see the AI moving LE units across the map. And the reason when they want trades they want Your LE units you do have, those built in your older larger cities. LE units have to have some major overlap. This becomes even more critical from the Industrial Era on thru the later eras.

Our building chains and obsolescence is starting to show it's weaknesses. And rather badly at that.

Of course the lesser cities will take a while to get updated LE unit prereq buildings. And they aren't going to be really motivated to get them updated until they start having a crime issue. You can build the update right away but again, I think some of them requiring other buildings is a big problem because the AI doesn't see them until it has had the opportunity to get the other prerequisite buildings... thus the castle makes the AI not even know that the Guardhouse can be available because it's a prereq for the guardhouse and the AI knows it must deal with crime so never gets to a point where it thinks to build the castle for the castle's merits and thus never gets the guardhouse to even reveal itself to be built and thus the Guards are never upgraded because the AI never knew they could've been. I think I do need to fix that.

This stops at Ren Era and does not follow that design from Industrial On.

Modern... and I explained why in my first post. Many LE units are meant to work together in the modern rather than overriding the need for another.

I agree that there is an issue, though, as it's near impossible to level LE units. This means that if a criminal gets a quality up promotion, it's going to be hell for a same tier to get it without losing an army of cops

Investigating units are getting an XP a round when there's a unit there to investigate. Shouldn't be that hard. Plus, all the modifiers are heavily slanted towards the LE units in the case of combat.


As for the Sheriff bit, I'm quite happy to remove force obsoletion on any of these units. That really isn't the issue. The issue is trying to figure out why AI and player aren't getting access to the best LE units for the tech level at too often a juncture. Maybe it would help to remove all forceobsolete lines for LE units though. They are needed for criminals but not LE.

 

[Raledon]

Not quite. It depends on the legal system. In countries based on civil law (like Germany, and in fact, continental Europe) a judge is much more powerful than in countries based on common law (USA, UK, etc.). There is (usually) no jury and the judge performs the investigation at trial (inquisitorial system) instead of deciding objections. There is an overview at https://en.wikipedia.org/wiki/List_of_national_legal_systems.

Both systems still possess a judge. In one the judge is weaker, but still an active an important role. If I'm not mistaken, the sharia has a person that acts as a judge, and I assume most of the customary systems out there. I guess part of this question is how exactly one defines "judge", though.

As I explained, Sherrifs actually split between riot police and police cars, the riot police being the first in the line of 'best to arrest with' and 'best for crowd control' units that leads to >Police APC>SWAT Teams>Police Mech (where the lines remerge). The Police Cars are best for crime control and investigation and although they may lag in crime control alone, they maintain this role until Police Mech.

I'm sorry, that makes no sense in context of the SVN's current version. a) police car replaces riot police (or police squad. I assume they are the same). b) police car is stronger, so making the arrest with it makes more sense. These 2 issues will have to be solved for riot police to have a place. (Also, what do you mean by crowd control?)
In any case, Sherrifs are to be replaced by police cars in the said system, so going by the two-step way, we should see them till we get a new step after cars. With your proposed system, we get riot police (obsoletes city guards), and either go for cars or radio, in order to make car police. If you research radio before you go for cars, you risk being stuck being unable to train proper (crime control and investigation) LE units. Currently you only have an issue having a couple of turns you have to build a police station so you can build a LE unit (radio->criminology causes no LE to be trainable, unless you build police station).
With your proposed system, the sheriffs should stay longer, so that you will have the time to switch infrastructure to cars.

By, my philosophy of specializing cities, do you mean that cities should always build their own LE units? At the moment, the AI only uses one city to build its LE units, with some effort to make the cities build their own if they are on a different continent. This is how it works for the AI so to say they don't follow this can only come from a place of overlooking how they are operating. There are huge benefits in this approach, from XP to the speed of addressing the LE needs of the nation. Another thing the AI does effectively with cities on the same continent is move their LE units to the places that they are needing more help from places that are doing fine. It tends to try that first before training more if it can get away with it.

I'm not sure this is the actual case, instead of a result of selective looking and/or traces of the old bug with crime units:
The ai tends not to "place and forget" LE units in cities, but move them around. This seems ineffective, as it cannot build-up anti-crime promotions, and the units may not grant their full benefits while moving. I suspect it does this with health units, too.
I'd expect it to have some units to reduce the crime/disease to close to 0 or 0/turn, and move a stack to arrest criminals and/or reduce crime when needed.

Investigating units are getting an XP a round when there's a unit there to investigate. Shouldn't be that hard. Plus, all the modifiers are heavily slanted towards the LE units in the case of combat.

Could you elaborate on the XP? Does it require the unit to take the investigation buildup? I'm pretty sure I never noticed them getting any levels, regardless.
The issue I pointed out is in regards with criminals getting a vast power up with quality-up. In my case, I have cities with robbers with 52 power, while my cars have only 26. This means that I need bonuses of 100% in order to start hoping for a 50/50. Another issue on this part is that you can't see the odds when going for an arrest, so I don't even know how dangerous an attempt will be. I probably have a decent shot getting them if I promote one of my cars to arrest criminals, but I'm still not sure what will be my odds.

As for the Sheriff bit, I'm quite happy to remove force obsoletion on any of these units. That really isn't the issue. The issue is trying to figure out why AI and player aren't getting access to the best LE units for the tech level at too often a juncture. Maybe it would help to remove all forceobsolete lines for LE units though. They are needed for criminals but not LE.

While I want to say "lets go for that!" I find there are a couple of issues with the current balance:
1) It seem like older units are more efficient at fighting crime/disease/education than newer ones. Taking arrest and investigation aside (LE are much too weak in size matters to act as real defense, and investigation wants only one real investigator), 5 watchers with crime control build ups will have similar upkeep to a car, but a much higher reduction in crime due to utilizing promotions and buildup multiple times. (If this statement is incorrect, it's really not clear that it's the case.)
2) For the production cost of one car, you could build about 50 watchers.
3) Cultural units, which could be built for the free promotions and upgraded with cash, thus getting better LE units for relatively a cheap price while ignoring the national unit cap.
4) Build old LE and upgrade with cash, as a cheap way to convert gold to hammers (should be extremely efficient, too)

An alternative option is to have every era give a free building that allows construction of the latest unit. So researching modern era, for example, will allow you to build cars (or sheriff, so the player still wants to build police station) even without building their required building. This means that in terms of buildings, you should always be able to build some units, assuming you have the required resources. The units will obsolete as normal one era later (so in the information era till police mechs, you will always be able to build cars)

 

[Thunderbrd]

I'm sorry, that makes no sense in context of the SVN's current version. a) police car replaces riot police (or police squad. I assume they are the same). b) police car is stronger, so making the arrest with it makes more sense. These 2 issues will have to be solved for riot police to have a place. (Also, what do you mean by crowd control?)

Let me review the charts more closely. I think I intended the riot police to at least be a lot cheaper and I don't think they should be replaced by police cars so I may have goofed on that to set police cars up as an upgrade for them. Crowd control is the anti-revolution factor LE units have. Again, let me review how they were setup... there could be some inconsistency of intent there.

In any case, Sherrifs are to be replaced by police cars in the said system, so going by the two-step way, we should see them till we get a new step after cars.

Ok, that makes sense.

With your proposed system, we get riot police (obsoletes city guards), and either go for cars or radio, in order to make car police. If you research radio before you go for cars, you risk being stuck being unable to train proper (crime control and investigation) LE units. Currently you only have an issue having a couple of turns you have to build a police station so you can build a LE unit (radio->criminology causes no LE to be trainable, unless you build police station).
With your proposed system, the sheriffs should stay longer, so that you will have the time to switch infrastructure to cars.

Ok, so that's one of the key loopholes I'm looking to solve for. Thanks for explaining.

I'm not sure this is the actual case, instead of a result of selective looking and/or traces of the old bug with crime units:
The ai tends not to "place and forget" LE units in cities, but move them around. This seems ineffective, as it cannot build-up anti-crime promotions, and the units may not grant their full benefits while moving. I suspect it does this with health units, too.
I'd expect it to have some units to reduce the crime/disease to close to 0 or 0/turn, and move a stack to arrest criminals and/or reduce crime when needed.

They do have some resistance to being moved so you shouldn't see cities sending out everything in them, just shaving some off the top of their stack when they have more than they need and other areas have need. They probably do move a lot more than human players would though.

Also, I believe the brokerage system does take distance and production time and other items in the queue into account so it's probably a little overstretching to say that the units are all made out of the same city. It's a complex evaluation process and Koshling actually built a lot of that out in a generic system I'm tapping into.

Could you elaborate on the XP? Does it require the unit to take the investigation buildup? I'm pretty sure I never noticed them getting any levels, regardless.

No... the BEST investigating unit, when there is a unit to investigate, is getting a little XP every round for the effort and gets a bit more whenever a criminal is successfully made 'wanted'.

Future development will include buildups for ongoing training specifically. Not just for LE units but for most types.

The issue I pointed out is in regards with criminals getting a vast power up with quality-up. In my case, I have cities with robbers with 52 power, while my cars have only 26. This means that I need bonuses of 100% in order to start hoping for a 50/50. Another issue on this part is that you can't see the odds when going for an arrest, so I don't even know how dangerous an attempt will be. I probably have a decent shot getting them if I promote one of my cars to arrest criminals, but I'm still not sure what will be my odds.

I agree that there's a factor here but if you're playing with LE units getting XP from specialists (including citizen specialists), it's not that hard to get a quality up for them either.

The era you point at is one where criminals were getting pretty lethal in relationship to the law enforcement of the time and historically, this sort of thing you're talking about was a real issue. Yes, some criminals are intended to be able to get to the point where they start being a true physical threat to any who would dare to try to arrest them. Starts requiring that you specialize some LE units to arrest (and if you do that from the beginning you can earn them quite a few XP along the way themselves, especially if you don't always try too hard to control crime so as to allow some criminal spawning.)

1) It seem like older units are more efficient at fighting crime/disease/education than newer ones. Taking arrest and investigation aside (LE are much too weak in size matters to act as real defense, and investigation wants only one real investigator), 5 watchers with crime control build ups will have similar upkeep to a car, but a much higher reduction in crime due to utilizing promotions and buildup multiple times. (If this statement is incorrect, it's really not clear that it's the case.)

Might be cheaper to make them and a little cheaper to staff them and thus would be a bit cheaper in terms of gold but they wouldn't on an individual basis be able to exceed what a more advanced one can do... the base improvements are significant. However, even making it so they don't forceobsolete doesn't mean they don't obsolete in a way... they become unavailable to train as soon as better is possible basically. The upgrade path is important to be careful with and I think at the car we have a problem as you pointed out.

2) For the production cost of one car, you could build about 50 watchers.

Should never be possible. I suppose it could be if the buildings are never obsoleting... Perhaps the buildings should skip 2 generations before obsoleting rather than just 1.

3) Cultural units, which could be built for the free promotions and upgraded with cash, thus getting better LE units for relatively a cheap price while ignoring the national unit cap.

Pretty expensive but if that's how you want to use your gold, just make sure they keep the promos right?

4) Build old LE and upgrade with cash, as a cheap way to convert gold to hammers (should be extremely efficient, too)

I do this elsewhere in the game so why not here too? I sometimes don't complete getting access to a resource before I've really built up a cheaper force I can upgrade. But the problem with doing that with LE units is that if you really need them somewhere, the place they are is going to require the buildings to upgrade the units too.

An alternative option is to have every era give a free building that allows construction of the latest unit. So researching modern era, for example, will allow you to build cars (or sheriff, so the player still wants to build police station) even without building their required building. This means that in terms of buildings, you should always be able to build some units, assuming you have the required resources. The units will obsolete as normal one era later (so in the information era till police mechs, you will always be able to build cars)

If we went this far, we might as well remove the building prerequisite entirely. And maybe that is the way to go... just make the building more of an XP generator for them rather than a prereq.

 

[JosEPh_II]

By, my philosophy of specializing cities, do you mean that cities should always build their own LE units? At the moment, the AI only uses one city to build its LE units, with some effort to make the cities build their own if they are on a different continent. This is how it works for the AI so to say they don't follow this can only come from a place of overlooking how they are operating. There are huge benefits in this approach, from XP to the speed of addressing the LE needs of the nation. Another thing the AI does effectively with cities on the same continent is move their LE units to the places that they are needing more help from places that are doing fine. It tends to try that first before training more if it can get away with it.

OMG!!! No wonder they are so screwed up and the player can waltz all over them. Of course Every AI city should be Allowed to build it's own LEs! SMH!!!! You have effectively handcuffed the AI. And you did not see this coming? Changes this and you will solve this whole issue. Well, at least for all the Older cities of the AI especially. Younger cities just don't have the buildings built up to build the top LEs anyway.

@Raledon ,

The above is Why you see the AI LEs moving around. I used to think it was for reducing Crime in adjacent tiles, Because I did Not know the the AI is Handcuffed to only having 1 city build LEs! Get rid of this ridiculous restrict and the AI will handle Crime better and Maybe stop killing itself by building Criminals it does not understand how to use.

 

[Thunderbrd]

OMG!!! No wonder they are so screwed up and the player can waltz all over them. Of course Every AI city should be Allowed to build it's own LEs! SMH!!!! You have effectively handcuffed the AI. And you did not see this coming? Changes this and you will solve this whole issue. Well, at least for all the Older cities of the AI especially. Younger cities just don't have the buildings built up to build the top LEs anyway.

@Raledon ,

The above is Why you see the AI LEs moving around. I used to think it was for reducing Crime in adjacent tiles, Because I did Not know the the AI is Handcuffed to only having 1 city build LEs! Get rid of this ridiculous restrict and the AI will handle Crime better and Maybe stop killing itself by building Criminals it does not understand how to use.

A new city will pause to take 30 rounds to build a crime unit whereas the primary production center can build multiples of them in one round. So what it takes a little time to get them to the city? They are far better trained, and thus far more valuable than the ones the outlying cities would train and thus are more worth the upkeep they cost. This system does actually work. And it works very well.

What has not been working has everything to do with the AI selecting crime CAUSING units to staff in their cities to spot criminals with because criminals are as good at spotting other criminals as dogs are and they have more 'abilities' thus are more attractive. An attempt to fix that system failed and we still saw the collapse that it caused before the first effort to correct it. The failure was due to relying on functions that were already in place and not looking deeply enough into them to ensure they would fulfill their roles.

Seriously, if cities are given to build their own LE units, it would be a far substandard method. Again, however, cities that are not on the same landmass as the primary unit training center should be attempting to train their own and the system takes into account the distance, speed of training, and quality of the resulting units, and other current demands on that city's build queue to determine what the best city to train LE units for a particular destination will be. I'm overstating it to say ONLY one city can train LE units.

The biggest weakness in the vanilla AI was that it could never produce good units because it tried to produce them everywhere rather than out of concentrated specialized cities. It could possibly create a lot of units quickly this way but at a cost of those units all sucking AND distracting too many cities from more effective growth protocols, for both the city and the nation. It's an unwise strategy to have cities all training their own LE units - they need them long before they are setup and efficient and effective at training such units.

Furthermore, with huge globals on crime penalties, it is specifically those cities trying to train for themselves that are causing the initial seed of a breakdown because they aren't fast enough at covering other bases before getting to crime control and by then it's too late. Therefore, one big improvement in the AI planned for the future is to better setup a willingness and ability to transport property control units to needed destinations - something that is currently lacking.

But again, when the AI is doing great and then starts flooding its own cities with ruffian units there's no way it can keep up with the damage it's doing to itself. This should no longer take place and as a result, I am confident that it would be very difficult to get an AI to fail at property control. Even if you attack them with waves of your own criminals.

 

[JosEPh_II]

A new city will pause to take 30 rounds to build a crime unit whereas the primary production center can build multiples of them in one round.

This is because you've forced obsolete the older units. Stop that bad decision and this argument is dead. And your Primary production center can not keep up with demand either. The evidence is clearly seen in game for crying out loud.

 

[Thunderbrd]

This is because you've forced obsolete the older units. Stop that bad decision and this argument is dead.

I was going to but some good points have been made about overutilizing cheaper units to get the most out of their promotions and resulting cost (both in training and upkeep) to crime control ratio. They might be right that without limiting the level of promotions that a particular type of LE unit can obtain, this may present its own issues. One should never be able to build a watcher simply because they've never upgraded their building in that city. I can try to remove the force obsoletions and let the buildings handle it but I'm pretty sure I may end up with another bug report that demands it be put back in place. IIRC, this has actually happened already - I tried to remove them and it was shown to be a bad idea. So perhaps I need to just give them yet more time, go another generation in before obsoletion.

I'm working on a few solutions now.

I'm not hostile towards your evaluations, Joe. They have been helpful. There's just sometimes a bigger picture that needs to be considered before we create more problems trying to solve one. I mean, you see why/how some players could overly exploit these units if they don't obsolete at some point right? You make sense to point at what you're pointing at but there is still a limit to how far we can allow ourselves to go with that. If you're building Watchers in the Modern Era, you're going to be able to really skew the system.

 

[JosEPh_II]

See edit to above post.

 

[Thunderbrd]

This is because you've forced obsolete the older units. Stop that bad decision and this argument is dead. And your Primary production center can not keep up with demand either. The evidence is clearly seen in game for crying out loud.

When that production center is training criminals to go into its own cities right on the heels of the LE units it's building to maintain itself, you can see how it can't keep up. The stronger that center is, the worse it is for the AI when it's being used to defeat itself.