Property units cost vs effect over time

[Raledon]

We've seen to have an issue with units that deal with properties (LE, disease and to a certain extent, education units).
While I'm going to focus on LE and disease, education has a similar issue with different numbers.

Each unit has 4 attributes when it comes to property control:
0) tier (not a real attribute, but it implies the rest)
1) upkeep (I understand it's 5 as base, +additional for certain units)
2) base reduction
3) build up
4) promotions

I've put it all in https://docs.google.com/spreadsheets/d/1nM-ZCg3fp0fK0ES0QDuzkS2SYeSLnNl1IjZbo35xP_8/edit?usp=sharing

The bottom 4 lines in the spreadsheet shows the unit's property reduction per . (numbers are based on LE units)
We can see that if the units didn't have promotions/bonuses, we should've upgraded them for better efficiency.
By tier 5, units can build-up 5 times. At this point, we'd rather have watcher over any other LE unit, till we get police cars (or later units).
The more xp the units get, the better the units become. For example, at tier 4 promotion (which is relatively easy to obtain), the watcher provides about 50% more reduction per than the police car.

If we want the players not to use older units, I propose we implement one (or more) of the following ideas:
a) Rebalance the base crime reductions, and/or bonus upkeep.
b) Limit the promotions older units can get. A level 100 watcher wouldn't be able to take a level 2 LE promotion.
c) Limit the amount of xp older units can get.
d) Slow down the rate LE get promotions, and make the promotions better (per level). In addition, tier # LE will start with tier $ promotion. This means that a level 4 tier 9 LE will get the same number of crime reducing promotions, but their value will be higher (for example, 100 crime reduction instead of the watcher's 20)
e) Internal multiplier. A tier 1 LE will reduce 100 crime, a tier 2 150, tier 3 225 and so on. (at some rate). I'm not sure how feasible this solution is, though.
f) Remove/reduce building requirements, as to force new units to be the highest grade possible (without causing issues with missing building/resource)

Which one of the options do we wish to implement?

 

[Toffer90]

Or " f) " remove the building requirement the LE units currently have.

Making crime reduction per gold upkeep better for each unit tier is imo how it should be balanced, so I would look into option " a) " first.
LE units should imo have a much lower upkeep than the regular strength based units. And why should it increase so much for LE units for each unit tier? Total upkeep could just as well be 1 for each of the LE unit tiers.

 

[Raledon]

Or " f) " remove the building requirement the LE units currently have.

I'm not talking about building requirements at all.
My issue is that, given the option, the optimal play will be to have an army of watchers and a single mech police. The mech police will investigate crime/arrest criminals, while the rest crime control.
This sounds like a very strange situation, when compared to having an army of mech police to do so.
We (I?) would like to have the option of buildings watchers (or any of the other LE) as a fast & cheap option, but the later units will be better, so given the option I'd like to upgrade them.
The power curve should rise over time, not go down.

 

[Toffer90]

I'm not talking about building requirements at all.

If you remove the building requirement, then it's going to be mighty hard to be able to train watchers in late game as they obsolete when it's possible to build the upgraded unit. It would partly solve the problem. "Partly" in that you could always refuse to upgrade old LE units, or you could build a city that is never connected to the trade network dedicated at training the lowest tech LE that have no bonus requirement.

I just thought the option was worth mentioning even though i favour option " a) ".

 

[Raledon]

If you remove the building requirement, then it's going to be mighty hard to be able to train watchers in late game as they obsolete when it's possible to build the upgraded unit. It would partly solve the problem. "Partly" in that you could always refuse to upgrade old LE units, or you could build a city that is never connected to the trade network dedicated at training the lowest tech LE that have no bonus requirement.

I just thought the option was worth mentioning even though i favour option " a) ".

Oh, that is true... Edited the main post and added it as f).

 

[JosEPh_II]

Nice set of theories. But will it actually work in game. And can the AI "abuse" the system like you say? The AI is Not going to build "By tier 5, units can build-up 5 times. At this point, we'd rather have watcher over any other LE unit, till we get police cars (or later units)." And this does not happen in game now.

This is just another carry over from T-brd and My disagreement over ForceObsolete.

Raledon, do you see the AI doing any of this you say is possible for the player to do? Have you Ever seen it? No you have not. End of subject for me.

 

[Toffer90]

Raledon, do you see the AI doing any of this you say is possible for the player to do? Have you Ever seen it? No you have not. End of subject for me.

The problem is not that the AI does it, the problem is that the human player can use this exploit while the AI doesn't, the exploit should be removed.

 

[JosEPh_II]

The problem is not that the AI does it, the problem is that the human player can use this exploit when the AI cannot.

Players will always find ways to exploit, especially the uber player, but not the avg joe player. So it's rarefied air we are throwing siege mounts against. This is the problem games will always have. The regular player will upgrade the units and buildings and proced in an orderly fashion towards keeping their Empire "up to date". Only the player that purposefully trys to "break" a game will do this. All other players play for Fun, not for seeing if they can "beat", but in reality want to "Break the game".

This is pandering to that type of player at the expense of the player who plays for Fun.

 

[Raledon]

Raledon, do you see the AI doing any of this you say is possible for the player to do? Have you Ever seen it? No you have not. End of subject for me.

The AI doesn't use it, as the AI lags behind player's learning curve.
It's easier for a player to discover/learn that it's better to do X over Y, than teach the AI. In the current version, the AI believes that the best way to play is to build the latest LE units possible. I think that this SHOULD be the correct decision.
If there's a clearly better alternative, either the alternative shouldn't be possible, or the AI should use it too. It doesn't matter that much.

I think I saw you (or one of the other modders?) saying that it's a good thing that the game is getting harder, and people drop in difficulty levels (this was in regards to your recent changes to disease).
Disease had a similar issue- the AI cared about it a lot, and worked to reduce it, while players knew that it doesn't do much and simply ignored it.
With your changes, players were forced to care about it. This didn't affect the AI, or players that already cared about it at all. Only players like me that decided it's not something worth caring about, had a need to change their playstyle.
With this topic, the normal players that keep the empire "up to date", and the AI, will generally see nothing different. On the other hand, we fix an exploit, allowing certain issues to be fixed. For example, the recent obsolete/force obsolete/no obsolete debate between you and T-brd wouldn't have been such an issue, since it has a downside. The same with the general "I suddenly can't build units" that sometimes pops up.

 

[Toffer90]

Players will always find ways to exploit, especially the uber player, but not the avg joe player. So it's rarefied air we are throwing siege mounts against. This is the problem games will always have. The regular player will upgrade the units and buildings and proced in an orderly fashion towards keeping their Empire "up to date". Only the player that purposefully trys to "break" a game will do this. All other players play for Fun, not for seeing if they can "beat", but in reality want to "Break the game".

This is pandering to that type of player at the expense of the player who plays for Fun.

It is so easy to change it so that later units reduce crime more per gold upkeep than earlier units.
Why is it so important to you that newer LE units must be less effective than older ones as it currently is?
Give us an argument for why not to change it; tell us why the prehistoric "Watcher" unit should be the best crime reduction unit in the game.

@Raledon: I accidentally changed part of the spreadsheet, didn't know that any changes I made was automatically saved.
Changed it all back again, and added an easy way to adjust the base cost while at it.

 

[JosEPh_II]

It is so easy to change it so that later units reduce crime more per gold upkeep than earlier units.
Why is it so important to you that newer LE units must be less effective than older ones as it currently is?
Give us an argument for why not to change it; tell us why the prehistoric "Watcher" unit should be the best crime reduction unit in the game.

I have Never said that Ever! And the Watcher Stinks! Costs too much in the Era available. As does the Enforcer and the Town Watch man.

I get tired of ppl and fellow modders putting words in my mouth I Never Said.

I'm tired of all this. As I replied to T-brd, revert Everything I've posted about, and I mean Everything over this topic. We can go back to having 2 very necessary LE units made Unavailable when you need them the worst. We can continue to have one of them go ForceObsolete Before you can even build the Damn Unit! Cause we have the Same Obsolete tech on these 2 units and then later on 3 more (FO on same tech) of the units you can upgrade to If you can build the 2 you really can't. Killing the AI and Players game. Go ahead do it! And then tell me I caused a player exploit. smh

I Am Done. Going to step away for awhile now. Bye.

 

[Toffer90]

I have Never said that Ever!

Oh, I'm sorry. I got the impression that you were opposed to the the things proposed in this thread from your two previous posts.

And the Watcher Stinks!

It is currently the best crime reduction unit in the game when it, and the others it is compared to, have four crime reduction promotions.

 

[Thunderbrd]

a) Rebalance the base crime reductions, and/or bonus upkeep.

Not really possible without inflating the crime levels by a lot to counter a stronger anti-crime improvement arc on the units or making the units a lot cheaper. Neither is a desireable solution - or has been perceived as one due to the way it was fairly balanced to begin with as a whole system. But after some tests recently with education, it looks like we could probably manage a stronger base unit arc of anti-crime if we made each era introduce an extra crime per population. So this might work. The building arcs for crime would also have to adjust and criminal and ruffian units would have to be made stronger at creating crime as well. This will intensify the 'swing' between succeeding at managing crime and failing at it.

b) Limit the promotions older units can get. A level 100 watcher wouldn't be able to take a level 2 LE promotion.

After doing some math last night, I came to the conclusion that this doesn't completely work either. Take 5 tier 4 LE units and compare that to 4 tier 5 LE units and the same amount of promos, like 3, and then the buildups and you'll see what I mean.

c) Limit the amount of xp older units can get.

Too complex and subject to arbitrary issues and then do other unit types have the same rules or are we just punishing property control?

d) Slow down the rate LE get promotions, and make the promotions better (per level). In addition, tier # LE will start with tier $ promotion. This means that a level 4 tier 9 LE will get the same number of crime reducing promotions, but their value will be higher (for example, 100 crime reduction instead of the watcher's 20)

Would likely exacerbate the problem further.

e) Internal multiplier. A tier 1 LE will reduce 100 crime, a tier 2 150, tier 3 225 and so on. (at some rate). I'm not sure how feasible this solution is, though.

The property system is very resistant to this kind of end manipulation in the way it was setup. It's... difficult to work with in this way.

f) Remove/reduce building requirements, as to force new units to be the highest grade possible (without causing issues with missing building/resource)

Leads us right back to the same 'can't train anything' problem, just without buildings being a part of it. Let's just say that it wouldn't be different to what we have now where the buildings, in the later game, now have the basic prereqs as the simplest units they unlock. And in the earlier side of the game the units can use the next level of building as a prereq as well.


Also, keep in mind if we change values for crime control, we should replicate the same value shifts for disease control across the board as I've always wanted both to be a reflection of the other for the sake of player sanity.

 

[Raledon]

Also, keep in mind if we change values for crime control, we should replicate the same value shifts for disease control across the board as I've always wanted both to be a reflection of the other for the sake of player sanity.

I agree. I do think we should add the other properties (education, currently) to work in a similar fashion for uniformity. Not a must, though

Not really possible without inflating the crime levels by a lot to counter a stronger anti-crime improvement arc on the units or making the units a lot cheaper. Neither is a desireable solution - or has been perceived as one due to the way it was fairly balanced to begin with as a whole system. But after some tests recently with education, it looks like we could probably manage a stronger base unit arc of anti-crime if we made each era introduce an extra crime per population. So this might work. The building arcs for crime would also have to adjust and criminal and ruffian units would have to be made stronger at creating crime as well. This will intensify the 'swing' between succeeding at managing crime and failing at it.

It's a rebalance project, I agree. There is no problem with greater swings, if the crime buildings are adjusted as well. If crime starts at 500 instead of 50, a swing of 50 instead of 5 will have the same effect.

After doing some math last night, I came to the conclusion that this doesn't completely work either. Take 5 tier 4 LE units and compare that to 4 tier 5 LE units and the same amount of promos, like 3, and then the buildups and you'll see what I mean.

Comparing 4 tier 5 units (A), and 5 tier 4 units(B), both have max build up, and equal tier of promotions. (tier 4 has 4, tier 5 has 5). C is 4 tier 5, with tier 4 promotion
A) upkeep of 45, reduction of 215. ratio of 4.7
B) upkeep of 40, reduction of 220 ratio of 5.5
C) upkeep of 40, reduction of 180 ratio of 4.5
In this case, assuming units has a lot of XP, we want to upgrade them. This does mean we want them to have a lot of xp, though. It does mean that in this case, with 4 promotions, we'll want tier 4 units instead of tier 1 which is a step in the right direction.

Would likely exacerbate the problem further.

Not sure if we are thinking of the same thing.
Assume we have two chains of promotions. Chain @ does nothing (or might do something, doesn't matter). Chain # requires a similar tier of @ to proceed, and reduces crime.
A tier 1 LE will start with tier 1 @. It has to spend 3 promotions to reach tier two in #. (2 in #, 1 in @).
A tier 2 LE will start with tier 2 @. This means he requires only 2 promotions to reach tier two in #.
If a tier 10 LE will start with tier 10 @ and tier 4 #, so it will require only 6 promotions to reach tier 10 #. a similar watcher will require 19 promotions to reach the same effect.
If tier 5 # is 2 times better than a tier 4 #, we will want prefer to invest in the newer models for the increased crime.

 

[Thunderbrd]

Not sure if we are thinking of the same thing.
Assume we have two chains of promotions. Chain @ does nothing (or might do something, doesn't matter). Chain # requires a similar tier of @ to proceed, and reduces crime.
A tier 1 LE will start with tier 1 @. It has to spend 3 promotions to reach tier two in #. (2 in #, 1 in @).
A tier 2 LE will start with tier 2 @. This means he requires only 2 promotions to reach tier two in #.
If a tier 10 LE will start with tier 10 @ and tier 4 #, so it will require only 6 promotions to reach tier 10 #. a similar watcher will require 19 promotions to reach the same effect.
If tier 5 # is 2 times better than a tier 4 #, we will want prefer to invest in the newer models for the increased crime.

You're talking about limiting the promotions they can get by unit type.

I think I can achieve this quicker and cleaner by taking an hour or so to correct the automatic assignment of era unitcombats and making the crime control and other similar promotions require a unit to have the era unitcombat that correlates to the era that they are unlocked in by tech. This would not be disruptive... just gotta make sure it works with a touch of testing afterwards. This gets us close to the goal without screwing up the existing game balance for this release at least.

It's a rebalance project, I agree. There is no problem with greater swings, if the crime buildings are adjusted as well. If crime starts at 500 instead of 50, a swing of 50 instead of 5 will have the same effect.

This further intensifying of the pendulum swing that takes place more and more as the game goes on has been difficult to convey the importance of observing as well so far. So such an adjustment of the balance as would be necessary here will really cause some hard to explain response needs to the rest of the team. Thus why I avoided intensifying it like the plague when restructuring the units and promos to begin with.

 

[Raledon]

You're talking about limiting the promotions they can get by unit type.

I think I can achieve this quicker and cleaner by taking an hour or so to correct the automatic assignment of era unitcombats and making the crime control and other similar promotions require a unit to have the era unitcombat that correlates to the era that they are unlocked in by tech. This would not be disruptive... just gotta make sure it works with a touch of testing afterwards. This gets us close to the goal without screwing up the existing game balance for this release at least.

Isn't this in regards to option b) rather than option d)?
One is about limiting, while the other is about slowing earlier units, compared to newer ones.

 

[Thunderbrd]

Isn't this in regards to option b) rather than option d)?
One is about limiting, while the other is about slowing earlier units, compared to newer ones.

Option d is really not good because I want the later units to have a chance, at least, of being able to get the full string of promos. Plus, I really dislike having to take middle man promos along a promotion chain. The AI doesn't work with it very well either because they don't look at what's behind the next option that might get unlocked by what they select.

But yes, now that I'm understanding D a bit more, my comment was intended more to apply to option B. Which is also imperfect, but between that and some moderate degree of obsoletion, which I guess we do still have at this point, then we should be OK for release.

 

[Raledon]

I agree that option b) will most likely be the easiest and least disruptive to implement.

Out of curiosity, what if we were to reduce the upkeep?
For example, if we reduce the upkeep to 1/2*tier or 1/3*tier (rounded up), the numbers seem to be a lot closer to the intended values. This means that while there may be an optimal play on which unit to build at each stage, it wouldn't be that effective.
Coupled with option b), you may decide to forget about units who don't have any more levels, but units will want to upgrade to use their promotions to the fullest, and upgrading other ones is not "a bad choice".

The downside is that we reduce the amount of upkeep the player will have to pay directly, for good and bad.

 

[Thunderbrd]

I agree that option b) will most likely be the easiest and least disruptive to implement.

Out of curiosity, what if we were to reduce the upkeep?
For example, if we reduce the upkeep to 1/2*tier or 1/3*tier (rounded up), the numbers seem to be a lot closer to the intended values. This means that while there may be an optimal play on which unit to build at each stage, it wouldn't be that effective.
Coupled with option b), you may decide to forget about units who don't have any more levels, but units will want to upgrade to use their promotions to the fullest, and upgrading other ones is not "a bad choice".

The downside is that we reduce the amount of upkeep the player will have to pay directly, for good and bad.

The upkeep for these units is a big part of what's helping to control gold in the ongoing stages of the game. Not really desireable to reduce these numbers.

Free promotions that cannot be earned (like certain religious promotions). Then you have AND-requirements for the higher promotions. I don't know how efficient that is, but at least it should be pure XML.

Edit: It might be necessary to forbid giving the free promo to visiting units (a tag like FreeAtTrainingPromo).

Getting the era unitcombat autoassignments fixed will pretty much accomplish this but do so through a less graphically intrusive manner. And could benefit things elsewhere too. And would be very safe for inclusion before release.

Tell you what why don't you just revert everything I've brought up so Sheriffs and Riot police go obsolete Together on the Same Tech Before you can build much the 1 of 1st units of sheriffs. That is what it seems you Want. So let's go back and do it All over again. How about that? Would that make it "better"? Then You can add this to your plate as well. Instead of wasting your time arguing over something that was Obviously Broke. And a late game killer too. We don't really need to play into the Modern Era at all. Conquer the world players such as you will never get there anyway as they will have won the game in Classical era and never ever have to worry about the late game play. I give up.

Players will always find ways to exploit, especially the uber player, but not the avg joe player. So it's rarefied air we are throwing siege mounts against. This is the problem games will always have. The regular player will upgrade the units and buildings and proced in an orderly fashion towards keeping their Empire "up to date". Only the player that purposefully trys to "break" a game will do this. All other players play for Fun, not for seeing if they can "beat", but in reality want to "Break the game".

This is pandering to that type of player at the expense of the player who plays for Fun.

Is it so hard to understand how an exploit of this nature ruins the game for so many? It's like saying here's some chocolate but don't eat it. Just keep it there on your desk... you'll not like what happens ultimately if you eat it. No matter how bad you know you can and want to.

I'm not trying to say the points you make or have made about the problems you've pointed out are invalid. I'm trying to say you need to see the bigger picture before demanding a particular type of response that will create another problem in the process. I don't know why you have such a hard time understanding that the gold efficiency is stronger with earlier units. This is the reason that hard obsoleting was enforced.

I think we've determined a few ways to address this. One will be soft and will work enough for now. But the deeper solution is going to be a bit more painful and will require a bit more rebalancing after it throws things out of whack, which is to rework how much the core units give without promotions so that they increase in core benefit a great deal more than they do now, which is then going to require that more crime be injected into the game as the eras progress. Some of this will be a matter of trying some numbers and crossing fingers in hopes that the balance will be about right. We'll try that next cycle.


@Raledon & @Toffer90 If we increase the crime per population by one with every gateway tech earned, do you think you could find a reasonable gradiation of numbers to assign to core anti-crime values for LE units that would be a good and effective counter for that? Let's try to leave the promotions alone. They are currently nice and straight forward. What base values do you suggest?

 

[Toffer90]

@Raledon & @Toffer90 If we increase the crime per population by one with every gateway tech earned, do you think you could find a reasonable gradiation of numbers to assign to core anti-crime values for LE units that would be a good and effective counter for that? Let's try to leave the promotions alone. They are currently nice and straight forward. What base values do you suggest?

Property per population changes by era sounds like a good idea.
Shouldn't this wait until after v38 is released?
My opinion on the matter is that special units like the LE should have a much lower upkeep than the regular strength oriented units, so I would change the upkeep for LE so that it is the same for all the tiers.

The upkeep for these units is a big part of what's helping to control gold in the ongoing stages of the game. Not really desireable to reduce these numbers.

There are many ways to increasing gold expenses in the game. A squad of policemen would not imo have a higher upkeep than a Company of modern tanks, I would add upkeep modifiers to the group volume combat classes and a generally higher upkeep for all military units.

e.g.
One Clubman unit have 2 upkeep while one Rifleman unit have 60 upkeep
One Watcher unit have 1 upkeep and one Sheriff unit have 1 upkeep.