Feedback: Maps and Terrain

[Xyth]

Maps and Terrain
Feedback and Development

 

[Xyth]

I'm going through the various mapscripts and fixing them up to work properly with the new mapsizes and some other HR changes. Fixing and testing these is a little time consuming so I'm not going to bother converting many of the more esoteric mapscripts. There are also a few proving more challenging than others. Here is the list of mapscripts that will be in 0.9.4:

• Archipelago
• Continents
• Fractal
• Highlands
• Inland Sea
• Islands
• Lakes
• Mixed (a slightly tweaked version of Big and Small)
• Pangaea
• PerfectWorld2
• Tectonics​

Please let me know if one of your favourite mapscripts is not on this list. Be quick though, unless something goes wrong, 0.9.4 should be finished tomorrow.

I should note that one noticeable omission from that list is Terra. Unfortunately this mapscript is giving me a lot of grief and won't be in 0.9.4. I'll try it again for 0.9.5 if people miss it.

Finally, I've not updated any of the real world maps for 0.9.4. They'll still work fine but they won't have any of the new resources on them. It's on my todo list but not a high priority at this stage.

 

[Lirvan]

World map type on giant size with 18 players seems to hang at the loading screen on the initialization phase. Forces me to force quit, so no error message. I'm running on a mac version.

 

[Xyth]

World map type on giant size with 18 players seems to hang at the loading screen on the initialization phase. Forces me to force quit, so no error message. I'm running on a mac version.

It does work it just takes a very long time (5 mins or more, depending on your CPU), especially for the larger map sizes. It's a very complicated mapscript. I meant to add a warning to the description but I forgot. Also be aware that 18 civs on a map that large could very well cause memory issues as the game progresses.

 

[civ_king]

World map type on giant size with 18 players seems to hang at the loading screen on the initialization phase. Forces me to force quit, so no error message. I'm running on a mac version.

On that size give it up to 15 minutes to generate

 

[Xyth]

I've found a very interesting and complex mapscript tool that, among many other functions, would allow me to add Marsh and/or Deep Ocean terrain to HR. However it has another feature that could potentially be very useful: I could use it to add an option to relevant mapscripts (like Archipelago) that allows coastal waters to extend more than one tile out from the land. This strikes me as more as a much simpler and more elegant option than adding a whole new new ocean terrain type, especially as it can be switched on or off as the player prefers. What does everyone think?

In particular, Howard, do you think this would be a good solution for the Archipelago games you enjoy? How many extra tiles of coastal water (on average) do you think would be ideal? It looks like I could set it as a consistent amount (e.g, coast is always 2 tiles thick around land) or I could make it somewhat random (e.g coast is 1-3 tiles thick around land).

Also, does anyone have any particular thoughts on whether marsh/wetlands would be a good addition?

 

[Azoth]

The "pollution" effect on mines and workshops not only reduces city health but also removes the river commerce bonus from the tile in question.
I don't think this was intended; and it further weakens mine and workshops. Could you restore the river commerce?

I've found a very interesting and complex mapscript tool that, among many other functions, would allow me to add Marsh and/or Deep Ocean terrain to HR. However it has another feature that could potentially be very useful: I could use it to add an option to relevant mapscripts (like Archipelago) that allows coastal waters to extend more than one tile out from the land. This strikes me as more as a much simpler and more elegant option than adding a whole new new ocean terrain type, especially as it can be switched on or off as the player prefers. What does everyone think?

I agree that extending coastal waters is more elegant than introducing a new Deep Ocean terrain type. I think a random setting of 1-3 coastal tiles would be best.

Also, does anyone have any particular thoughts on whether marsh/wetlands would be a good addition?

Would Marsh be a base terrain type? Or would it be a terrain feature, like Forest, that could be removed (Drained instead of Chopped)? Either way, I'm not sure players would appreciate another weak terrain type like Tundra or another annoying terrain feature like Jungle. Still, it could create interesting game dynamics: regions with Marsh might remain unsettled until medieval times.

Wait, here's an idea. The superiority of Forest to Jungle has always struck me as yet another Euro-centric concept. Jungles (Rainforest) aren't particularly poorer in resources as compared to temperate Forests; and any number of classical civilizations - Mayan, Tamil, Javanese - thrived in the jungle. If you introduce Marsh, you could redesign Jungle as a Forest-replacement for tropical regions, and retain Marsh as a truly uninhabitable Jungle-replacement, as follows:

Terrain Type: Food/Hammers/Commerce

Grass/Forest: 2/1/0, +0.5 health
Plains/Forest: 1/2/0, +0.5 health
Grass/Jungle: 2/0/1, +1 commerce with River
Plains/Jungle: 1/1/1, +1 commerce with River
Grass/Marsh: 1/0/0, -0.4 health
Plains/Marsh: 0/1/0, -0.4 health

In other words,
Forests would add +1 hammer and +0.5 health to the base terrain (as they do now);
Jungles would add +1 commerce and a further +1 commerce on Rivers; and
Marshes would subtract -1 food and -0.4 health (as Jungles do now).

 

[Xyth]

The "pollution" effect on mines and workshops not only reduces city health but also removes the river commerce bonus from the tile in question.
I don't think this was intended; and it further weakens mine and workshops. Could you restore the river commerce?

Okay I can fix that, but there is a slight quirk. If you build a pollution causing improvement on a tile it will get the +1 commerce for being adjacent to a river, even if the original terrain type didn't give that. I think this only affects snow tiles though so it's probably not a big deal. Can you think of any other situations though?

I agree that extending coastal waters is more elegant than introducing a new Deep Ocean terrain type. I think a random setting of 1-3 coastal tiles would be best.

I think I might even be able to code several different Coastal Water options too. For example: Normal (1 tile), Extra (1-3 tiles), Extensive (2-5 tiles). Or variations thereof.

Would Marsh be a base terrain type? Or would it be a terrain feature, like Forest, that could be removed (Drained instead of Chopped)? Either way, I'm not sure players would appreciate another weak terrain type like Tundra or another annoying terrain feature like Jungle. Still, it could create interesting game dynamics: regions with Marsh might remain unsettled until medieval times.

Looks like it would have to be a terrain type and not a feature placed on a terrain type. If I put it in I think it would possibly be best to make it quite rare and scattered except for the occasional large Pantanal-like area. I think I can use this tool to make that possible.

I'd probably prefer to call it 'Wetlands', it's more inclusive and appropriate for the area that a tile represents. One advantage of adding Wetlands is that it could make for more interesting distribution of some resources. Suggestions?

Grass/Forest: 2/1/0, +0.5 health
Plains/Forest: 1/2/0, +0.5 health
Grass/Jungle: 2/0/1, +1 commerce with River
Plains/Jungle: 1/1/1, +1 commerce with River
Grass/Marsh: 1/0/0, -0.4 health
Plains/Marsh: 0/1/0, -0.4 health

In other words,
Forests would add +1 hammer and +0.5 health to the base terrain (as they do now);
Jungles would add +1 commerce and a further +1 commerce on Rivers; and
Marshes would subtract -1 food and -0.4 health (as Jungles do now).

The way it works for Forests and Jungles is that they alter the yield of the tile they are on. Forests add +1 production to either Grasslands or Plains and Jungles subtract 1 food from Grasslands (they don't appear on Plains). I agree that Jungles being so weak compared to Forests is unrealistic. I think there are a few options or combination of options here:

1) Make Forests also cause -1 food
2) Remove the food penalty from Jungles
3) Give Jungles +1 production like the forest
4) Give Jungles +1 commerce to distinguish them from Forests
5) Give Jungles (and Forest?) +1 commerce on Rivers

As for the (un)healthiness I'd be more tempted to lower the health bonus from Forest a bit. The unhealthiness from Jungle makes sense as it represents the danger of tropical diseases and the like. Unhealthiness and -1 food is pretty harsh though.

My initial feelings are that the food penalty is important because it simply isn't possible to provide as much food from forest/jungle as you can from cleared agricultural land - at least by default. Now that camps provide food and can be built on any Forest/Jungle tile this makes me very tempted to choose option 1 (and leave the Forest's health bonus as is). It would mean that Plains Forests would give no food by default but I'm not sure that's a bad thing. With this we could remove the GPP penalty from Redistribution.

Option 5 makes sense, there's no reason I can think of that rivers through jungles or forests (they don't have get it in BTS) shouldn't give the commerce bonus. And option 4 is good since it makes Jungles different from Forests besides their different effects on health. So in summary:

• Grassland Forest: 1/1/0, +0.5 health, +1 commerce on river (previously 2/1/0, no river commerce)
• Plains Forest: 0/1/0, +0.5 health, +1 commerce on river (previously 1/1/0, no river commerce)
• Jungle: 1/0/1, -0.25 health, +1 commerce on river (previously 1/0/0, no river commerce)​

In all cases building a Camp would restore 1 food to the tile. In BTS deforestation was almost always the best option but since the introduction of Camps/Redistribution in HR I've wondered if we didn't swing it too far in the other direction. At first impression this proposal seems to strikes a better balance of choice. Let me know what you think (and how Wetlands as terrain might fit in as well).

 

[Howard Mahler]

Would be an improvement.
I play on huge maps.
On small maps, having extra coastal tiles might be problematic.
On huge maps having one or two extra would work.

However, a much better fix would be to have sea tiles and leave coastal alone. Sea tiles are between ocean and coastal. They worked well in earlier version of Civilization (II and/or III.)

There could be random number of sea tiles extending off of each coastal tile. Would need to try it to see how it works, but on a huge map 1 to 3 sounds about right.

Caravels would then regain usefulness.
(In Version 9.0.4 they have lost much of their usefulness.)
They could now enter sea tiles (with appropriate technology) and you could still keep the Compass for caravels going into oceans.

A separate terrain type of sea allows a progression:
1. Until Navigation (or appropriate tech) seas act as a barrier.
If there is some way to make cultural boundaries not extend into seas and oceans until the appropriate tech, so much the better.
2. With Navigation (or appropriate tech) Caravels can now enter sea squares but not oceans. As currently, Caravels can carry great people, missionaries, and spies. However, one can not use Caravels to invade or found colonies across the sea.
With Navigation (or appropriate tech) trading would presumably now be allowed across seas.
3. As currently, until Compass (or appropriate tech) oceans still represent a complete barrier.
With Compass (or appropriate tech) Caravels can now enter oceans and trading across oceans is enabled.
4. As currently, with Charter (or appropriate tech) Galleons, Frigates etc. can be built and can cross the ocean.

The above progression makes some sense both from a historical standpoint and a game standpoint.
If this is possible, I think it is worth putting in the effort to do it as above.

I think the number of sea squares should probably depend on the map size, but need not depend on the map type.

I've found a very interesting and complex mapscript tool that, among many other functions, would allow me to add Marsh and/or Deep Ocean terrain to HR. However it has another feature that could potentially be very useful: I could use it to add an option to relevant mapscripts (like Archipelago) that allows coastal waters to extend more than one tile out from the land. This strikes me as more as a much simpler and more elegant option than adding a whole new new ocean terrain type, especially as it can be switched on or off as the player prefers. What does everyone think?

In particular, Howard, do you think this would be a good solution for the Archipelago games you enjoy? How many extra tiles of coastal water (on average) do you think would be ideal? It looks like I could set it as a consistent amount (e.g, coast is always 2 tiles thick around land) or I could make it somewhat random (e.g coast is 1-3 tiles thick around land).

Also, does anyone have any particular thoughts on whether marsh/wetlands would be a good addition?

 

[Xyth]

I play on huge maps.
On small maps, having extra coastal tiles might be problematic.
On huge maps having one or two extra would work.

I might be able to make it scale with mapsize. At the very least I can make the number or range of coastal tiles player selectable when they set up the map.

There could be random number of sea tiles extending off of each coastal tile. Would need to try it to see how it works, but on a huge map 1 to 3 sounds about right.

I can definitely do a random number of coastal tiles not yet sure about sea/deep ocean.

If there is some way to make cultural boundaries not extend into seas and oceans until the appropriate tech, so much the better.

There is no way to change how cultural borders behave on water, I'm not sure but I think it's not possible even with the SDK. At the moment I don't yet know how they'll behave with extended coastal waters or seas/deep ocean.

The above progression makes some sense both from a historical standpoint and a game standpoint.
If this is possible, I think it is worth putting in the effort to do it as above.

It should be possible but I don't know if it's within my skill yet. It's a lot more challenging technically that extended coastal waters but I will certainly experiment and see what's possible. The hardest part is going to be finding/making suitable textures, it's already quite hard to visually distinguish coast and ocean without the land as guide. I don't want it to look too 'fake' either. I have an idea though, I'll hopefully get some time to experiment later this week.

There's is one disadvantage to going with seas/deep ocean though; while extended coastal waters can be made optional and selectable by the player, seas/deep ocean cannot.

All in all I'll try out both options and see what the technical challenges are and then we can go from there.

 

[Howard Mahler]

Sounds good.
As far distinguishing seas from oceans, check out the earlier CIV (II and/or III).
I believe that oceans were a darker blue than seas.

I do not believe we need an option. Other terrain types are not optional. Seas make sense for all map types and should not ruin play if there is some way to limit their (average) extent.

I might be able to make it scale with mapsize. At the very least I can make the number or range of coastal tiles player selectable when they set up the map.



I can definitely do a random number of coastal tiles not yet sure about sea/deep ocean.



There is no way to change how cultural borders behave on water, I'm not sure but I think it's not possible even with the SDK. At the moment I don't yet know how they'll behave with extended coastal waters or seas/deep ocean.



It should be possible but I don't know if it's within my skill yet. It's a lot more challenging technically that extended coastal waters but I will certainly experiment and see what's possible. The hardest part is going to be finding/making suitable textures, it's already quite hard to visually distinguish coast and ocean without the land as guide. I don't want it to look too 'fake' either. I have an idea though, I'll hopefully get some time to experiment later this week.

There's is one disadvantage to going with seas/deep ocean though; while extended coastal waters can be made optional and selectable by the player, seas/deep ocean cannot.

All in all I'll try out both options and see what the technical challenges are and then we can go from there.

 

[Azoth]

Some thoughts:

Pollution: I can't think of any terrain type other than Ice that doesn't naturally provide river commerce so I would go ahead with the Pollution fix. (Frankly, with the pollution health penalty, it's not worthwhile to build mines and workshops on Ice and Tundra anyway. You are better served hiring a specialist.)

Redistribution: As others have mentioned, that -25% GPP penalty is harsh. With all the other changes in 0.9.4 (pollution on Mines, no Camps on hills, improved Free Market and Bureaucracy civics), I no longer think it is necessary. No matter the other changes you make, I suggest you remove it.

Forest and Jungle:
1. A food penalty on Forests is a bad idea. Food is the lifeblood of Civilization. It is the source of everything else: research, production, gold, and culture. Jungles are chopped precisely to eliminate that -1 food penalty. If you reduce Forests to 1/1/0 tiles, then they will also be ruthlessly cut down: 2/1/1 Redistribution Camps can't compete with 3/1/0 (and later 4/1/0) Agrarian Farms.

2. Forests and Jungles don't provide river commerce in BtS so as to encourage players to clear riverside tiles first. It makes for a better model of historical agricultural practice. If you do not apply a food penalty to Forests (as I suggest), then you should leave the river commerce out. 2/1/1 River Forest tiles would be too strong.

3. Jungles are a different story. I would certainly remove the food penalty and add a commerce bonus to distinguish them from Forests. (The rationale? Forests = lumber = production. Jungles = rare metals and medicinal plants = commerce.) If you prefer to keep a health penalty, then I would increase it from -0.4 to -0.5; and add the river commerce bonus to keep it balanced. That way, players might choose to keep Jungles around, at least some of the time.

4. I also think you should allow Jungles on Plains. It would better represent the diversity of terrestrial biomes: Grassland/Jungle would correspond to tropical rainforest (the Amazon, the Congo) while Plains/Jungle would correspond to savannah and semitropical dry forest (sub-Saharan Africa outside the Congo, notably West Africa.) You'll note that most Earth maps do include Plains/Jungle, especially around the Malinese starting location.

In summary:
Grass/Forest: 2/1/0, +0.5 health, no river commerce
Plains/Forest: 1/2/0, +0.5 health, no river commerce
Grass/Jungle: 2/0/1, -0.5 health, river commerce
Plains/Jungle: 1/1/1, -0.5 health, river commerce
(To simplify the math, I would also change the Floodplains health penalty to -0.5 health. Currently, it stands at -0.4 health.)

Even with these improvements, I still think players will gradually chop down forests and jungles over the course of the game. The best late game improvements - Farms, Towns, Watermills and Windmills - are all incompatible with Forests and Jungles. So I wouldn't worry about swinging the balance too far in the direction of preservation.

Wetlands: If wetlands are a base terrain type and cannot be drained, they need something to compensate their (presumably) low yield. Civ already has its share of uneconomical (Tundra, Ocean) and useless (Desert, Ice, Sea Ice, Peak) terrain types. What improvements, if any, do you plan to allow on Wetlands tiles? Depending on that, I might suggest 1/0/0, +1.0 health for Wetlands, in recognition of their role as natural filters and reservoirs.

Sea: I'm not sold on the idea of a Sea terrain type. I remember the implementation in Civilization III; and as a practical matter, I have to say that seas do not extend from the coast into the ocean as depicted. Rather, seas are particularly large bodies of mostly shallow coastal waters: the Mediterranean, Caspian, Arabian, and Caribbean Seas are all case in point. Every one of those was navigable in the ancient and classical eras; I'm not sure you could deny that right to Galleys and Triremes. The case for Deep Ocean is stronger: ships in the Age of Sail would make voyages only on Ocean tiles, as Deep Ocean tiles would be inaccessible to all but Combustion-era warships. But I still prefer the two terrain types we have: an extended Coast, surrounded by Ocean.

 

[Xyth]

Pollution: I can't think of any terrain type other than Ice that doesn't naturally provide river commerce so I would go ahead with the Pollution fix. (Frankly, with the pollution health penalty, it's not worthwhile to build mines and workshops on Ice and Tundra anyway. You are better served hiring a specialist.)

That's what I thought too but best to get a second opinion. Added to 0.9.5. (I'm not planning to update the patch for a while, too often and it becomes tricky to support)

Redistribution: As others have mentioned, that -25% GPP penalty is harsh. With all the other changes in 0.9.4 (pollution on Mines, no Camps on hills, improved Free Market and Bureaucracy civics), I no longer think it is necessary. No matter the other changes you make, I suggest you remove it.

Yeah I agree that the other changes have made it unnecessary. I'll remove it.

Forest and Jungle:
1. A food penalty on Forests is a bad idea. Food is the lifeblood of Civilization. It is the source of everything else: research, production, gold, and culture. Jungles are chopped precisely to eliminate that -1 food penalty. If you reduce Forests to 1/1/0 tiles, then they will also be ruthlessly cut down: 2/1/1 Redistribution Camps can't compete with 3/1/0 (and later 4/1/0) Agrarian Farms.

Yeah I did some some testing and the AI went back to being axe-happy. I won't add a food penalty to Forest.

2. Forests and Jungles don't provide river commerce in BtS so as to encourage players to clear riverside tiles first. It makes for a better model of historical agricultural practice. If you do not apply a food penalty to Forests (as I suggest), then you should leave the river commerce out. 2/1/1 River Forest tiles would be too strong.

3. Jungles are a different story. I would certainly remove the food penalty and add a commerce bonus to distinguish them from Forests. (The rationale? Forests = lumber = production. Jungles = rare metals and medicinal plants = commerce.) If you prefer to keep a health penalty, then I would increase it from -0.4 to -0.5; and add the river commerce bonus to keep it balanced. That way, players might choose to keep Jungles around, at least some of the time.

Jungle unhealthiness has never been 0.4, it's 0.25. Floodplains are 0.4. I think river commerce in Jungles but not Forest doesn't make sense - if 2/1/1 Forest tiles are too strong then so are 2/0/2 Jungle tiles. Plus it's not realistic to have one and not the other. I'd prefer to leave the health penalty as is and not add river commerce to either Forest or Jungle. I'm fine with removing the food penalty from Jungles and adding a commerce though.

4. I also think you should allow Jungles on Plains. It would better represent the diversity of terrestrial biomes: Grassland/Jungle would correspond to tropical rainforest (the Amazon, the Congo) while Plains/Jungle would correspond to savannah and semitropical dry forest (sub-Saharan Africa outside the Congo, notably West Africa.) You'll note that most Earth maps do include Plains/Jungle, especially around the Malinese starting location.

Someone recently made some Savannah-style tree graphics that I've been considering using. There's also some Scrub graphics that could possibly be used in Deserts. I'll see what's possible.

(To simplify the math, I would also change the Floodplains health penalty to -0.5 health. Currently, it stands at -0.4 health.)

I think the 0.4 is meant to represent 1/3, so you get unhealthiness for every 3 floodplains but 1 unhealthiness for every 4 Jungles. I suspect 1 for every 2 might be too much for either feature. I wonder if it's worth lowering the Forest health bonus though, to 1/3 or 1/4.

Wetlands: If wetlands are a base terrain type and cannot be drained, they need something to compensate their (presumably) low yield. Civ already has its share of uneconomical (Tundra, Ocean) and useless (Desert, Ice, Sea Ice, Peak) terrain types. What improvements, if any, do you plan to allow on Wetlands tiles? Depending on that, I might suggest 1/0/0, +1.0 health for Wetlands, in recognition of their role as natural filters and reservoirs.

I'm not sure yet. It will depend how rare/regional I'm able to make them. We could possibly give them 1 commerce for similar reasons to Jungles.

Sea: I'm not sold on the idea of a Sea terrain type. I remember the implementation in Civilization III; and as a practical matter, I have to say that seas do not extend from the coast into the ocean as depicted. Rather, seas are particularly large bodies of mostly shallow coastal waters: the Mediterranean, Caspian, Arabian, and Caribbean Seas are all case in point. Every one of those was navigable in the ancient and classical eras; I'm not sure you could deny that right to Galleys and Triremes.

Yeah I wouldn't call it Sea. If I put it in it would be Coast -> Ocean -> Deep Ocean. Coast would 1 tile like now and there would need to several tiles of Ocean before Deep Ocean began, preferably scaling with mapsize.

 

[Howard Mahler]

I disagree.
Redistribution is not broke.

Seems to be an unfortunate tendency to jump to conclusions, prior getting feedback from several people who have played the new version.

The fact that mines are unhealthy and camps can not be built on hills is largely irrelevant. The point is the bonus you get from Redistribution versus running nothing or some other civic.
The bonus is still very useful and the only thing you can run early.

Free Markets is not noticably stronger than it was and is available late.
Bur. is good but only comparable to Redistribution if you build a lot of lumbermills.
Professionalism is weaker than Redistribution.

In the game I am playing eventually switched out of Redistribution because of the 25% hit to GPP (and to try new Bur. civic.)
Without that I would probably still be in Redistribution.

Please wait for feedback from some other players.
Look at saved games.

For example, in my current game in 1708 AD, 4 AI civs in Redist. and 5 AI civs in Bureaucracy.
Nobody in Professionalism. Free Market largely not available yet and Environ. not available yet.

You have just changed the civics again.
Give it much more than a week to see how all the many changes interact.

 

[Xyth]

Actually Howard is right. I was running some test games at the time to see how the AI reacted to the removal of the GPP penalty on Redistribution and it seemed like there wasn't a problem. However, I've just realized that I hadn't reset the food penalty on Forests that I was trialling and that was obviously skewing the results significantly. This morning I fixed that and tried again.

The GPP penalty will need to stay, because although the AI knows how to deal with the unhealthiness from Mines it isn't aware that building more will cause more (but it will switch to more food production to counter it and thus the unhealthiness never gets out of hand). The relevant bit though is that unhealthiness from Mines won't factor at all into the AI's decision about whether to stay in Redistribution or not.

Also, be aware that even if it sounds like I'm making snap decisions on the forums, I'm almost always running (or will be running) test games to verify what's being discussed. In this instance my test was faulty though.

 

[Azoth]

Didn't get a chance to respond to this earlier:

I disagree.
Redistribution is not broke. [...]

The fact that mines are unhealthy and camps can not be built on hills is largely irrelevant. The point is the bonus you get from Redistribution versus running nothing or some other civic.
The bonus is still very useful and the only thing you can run early.[...]

In the game I am playing eventually switched out of Redistribution because of the 25% hit to GPP (and to try new Bur. civic.)
Without that I would probably still be in Redistribution.[...]

For example, in my current game in 1708 AD, 4 AI civs in Redist. and 5 AI civs in Bureaucracy.
Nobody in Professionalism. Free Market largely not available yet and Environ. not available yet.

The GPP penalty will need to stay, because although the AI knows how to deal with the unhealthiness from Mines it isn't aware that building more will cause more (but it will switch to more food production to counter it and thus the unhealthiness never gets out of hand). The relevant bit though is that unhealthiness from Mines won't factor at all into the AI's decision about whether to stay in Redistribution or not.

OK, I'm confused. Howard, you seem to be suggesting that we need a penalty on Redistribution to encourage players to switch to the other Economy civics. Xyth, you argue that without a penalty, the AI will be stuck building boosted Redistribution-mines and accrue large health penalties that it cannot foresee. Am I the only one who finds that to be rather circuitous reasoning?

Let me restate my argument against the GPP penalty:
1. Howard, you're right. Redistribution does provide useful bonuses and it's the only civic in its column that's available early. I think we can all agree that Redistribution should be unambiguously stronger than the default Reciprocity civic, such that every player switches to Redistribution in the Ancient Age. In that case, the GPP penalty simply depresses global Great Person growth. Is there any advantage in that?
2. Professionalism will always be a niche civic. It is only useful if you are running lots of Workshops; unlimited Merchants are not worth switching civics for. So I'm not surprised the AI never adopts this civic; I rarely adopt it myself. (I expect I would adopt Professionalism on an Archipelago map, since I would be desperate to extract hammers from my few land tiles, and would require extra specialist slots for my burgeoning seafood cities; but I rarely play Archipelago maps.)
3. Bureaucracy is a big step up from Redistribution and Professionalism. Watermills and Lumbermills bring higher yields than Mines and Workshops; and they do not carry a health penalty, either. They do take a long time to build, however; so the transition to Bureaucracy is necessary delayed until the necessary infrastructure is in place. (The GPP penalty is irrelevant in this calculation.) Ideally, players should also move their capital to their Shrine city in preparation; but the AI will never do that. You will note, though, that 60% of leaders (Howard and 5 AIs) had made the transition to Bureaucracy by 1700 AD.
4. Free Market and Environmentalism are the best civics in the Economy branch. (In this case, the fact that mines carry a health penalty and camps cannot be built on hills is relevant; you will have fewer of those improvements around, so the other civics are better by comparison.) I agree that they are both available too late in the game; I have previously argued that they should be unlocked at Economics and Ecology, respectively.

Bottom line: I don't think the GPP penalty is the answer to the balance issues in the Economy branch. It's simply an unnecessary, punitive penalty.

 

[Xyth]

Redistribution discussion shifted to the Civics feedback thread.

 

[Howard Mahler]

It appears that the AI has somehow managed to build workshops in desert (pollution.)?!?

 

[Xyth]

It appears that the AI has somehow managed to build workshops in desert (pollution.)?!?

Workshops can be built wherever there is a freshwater source, so long as the tile is flat. I haven't changed this from BTS so I guess these Workshops are next to a river, oasis or city with an Aqueduct.

 

[Howard Mahler]

Useful to know. Thanks.

In this case, there is no river or oasis. The enemy must have had an Aqueduct in the city which was destroyed when I took the city. Mystery solved.

Workshops can be built wherever there is a freshwater source, so long as the tile is flat. I haven't changed this from BTS so I guess these Workshops are next to a river, oasis or city with an Aqueduct.