What is the best tile improvement in the game?

[Aurelesk]

Once upon a time, the Mine was the best tile improvement in the entire game. Then, the Gathering Storm expansion happened. All the regular TI got additional bonus in the middle game in order to compete with Mines, and that additional bonus was mostly Production (what a coincidence!):

• Lumber Mill: is unlocked earlier at Construction (Classical) instead of Machinery (Medieval), starts at +2 Production, and can be put on Rainforests with Mercantilism (Renaissance).
• Plantation: +1 Food with Feudalism (Medieval)
• Camp: +1 Food and +1 Production with Mercantilism (Renaissance)
• Quarry: +1 Production with Gunpowder (Renaissance)
• Fishing Boats: +1 Production with Colonialism (Industrial)
• Pasture: +1 Production with Replaceable Parts (Modern)

Therefore, I always wondered: is Mine still the best tile improvement in the game?

I believe it is, for multiples reasons: it doesn’t need a resource or a feature to exist, is as strong as other regular tile improvement, can be put on at least 25% of the land tiles in the game, it is straightforward, scale better with era and is linked to 2 really important technologies: Apprenticeship and Industrialization, gives minor adjacency to Industrial Zone, improve almost all strategic resources in the game, and is able to give Tourism at Flight.



We could include tile improvements from City-States and the unique from Civilizations. Fortunately, this forum is fond of elimination threads, so we can know what were considered as the best TI back them:

City-States:

1. Cahokia Mound (Cahokia)
2. Moai (Rapa Nui)
3. Batey (Caguana)
4. Colossal Head (La Venta), 5. Alcázar (Granada), 6. Nazca Lines (Nazca), 7. Monastery (Armagh)

Mahavihara and Trading Dome were not available yet.


Civilizations:

1. Kampung (Indonesia)
2. Pairidaeza (Persia), at +2 Appeal
3. Terrace Farm, at 0.5 Housing and no adjacency
4. Mekewap (Cree)
5. Ziggurat (Sumeria)
6. Outback Station (Australia)
7. Mission (Spain)
8. Sphinx (Egypt) when spammable, 9. Open-Air Museum (Sweden), 10. Chemamull (Mapuches) with no Production, 11. Great Wall (China), 12. Polder (Netherlands), Ice Hockey Rink (Canada), 14. Golf Course (Scotland), 15. Stepwell (India), 16. Nubian Pyramids (Nubia) before the 2 Food/Faith, 17. Kurgan (Scythia) before the 3 Gold and increased Faith at Feudalism, 18. Château (France) before the overhaul.

Hacienda and Rock-Hewn Church were not available yet.


This is where I think I understood: the usefulness of a tile improvement is not about really how strong it is but more about either opportunity cost or enjoying yields while not working them:

• Kampung has average or even below average strength. Meanwhile, it is incredible: there is no opportunity cost: Offshore Wind Farms or Seasteads are too late to matter, while Fisheries are underwhelming.
  
• Cahokia Mound allows to have 1~2 Housing and 1 Amenity for just 1 tile, with no set-up and no need to work the tile.
  
• Moais are a spammable and easy to use TI that allows you to do some incredible clusters, like near Coast or Volcanic Soil. You do not need to work the tile for its Tourism. Bateys are the same thing when you don’t have that much space but relies more on adjacency.
  
• Pairidaeza was insanely good to set up National Parks when it had 2 Appeal back then. It is odd that the Sphinx was only 8th at the time. Appeal and Culture are translated to Tourism, something you don’t need to work.

I started to wonder: is it possible to estimate the strength of a tile improvement accurately?

No. But I tried! So I made a spreadsheet. Here the link (Google Sheets)
You can download it, and do whatever you want. If you change some yields or value, the whole sheet is supposed to update.

According to this, the Mission on foreign continents is the TI that has the most potential, excluding the TI with 1 per city restriction. Something that didn't change from a long time, but somehow the Mission scored badly (7th) while being spammable. Accurately is a strong word, and there is no way to achieve this. But I think it is possible to roughly determine an overall power of each tile improvement. The method is simple: each Yield (Food, Production…) has a value, and sum up all of them to determine the overall Strength of it. Yes, this is a silly way to do it, because it assumes:

• A Yield has the same value though all the game (false!)
  • It is known that Culture and Faith are very valuable in the early game in order to unlock the first governments and Pantheon, then fade out to a more “common” value.
  • Some Yields have “all or nothing” value: valuable when lacking but worthless if overshoot. For example: Housing, Appeal, Amenity, Power even Food with no Housing. All those yields are wasted in abundance.
• Some Yields have more of an “Empire” value like Science, Gold or Faith than “Local” values like Food, Production or even Culture for border expansion.
• Some Yields do not need to be worked to be efficient: Power, Tourism, Appeal, Housing...
• It doesn't take into account the tile restrictions. Nazca Lines are extremely potent, but way too difficult to fully enjoy it.

Of course, this method is improvable. But what do you think?

1. Is the Mine still (one of) the best regular tile improvement in the game?
2. What do you think are the best ones?
   • From the regular one.
   • From City-States.
   • From uniques.
3. What do you think about the method used to determine the overall Strength of a tile improvement in the spreadsheet?

 

[UWHabs]

I think the biggest issues with comparing tile improvements are the following:
-One per city vs "cannot be placed next to each other" vs spammable have very different valuations that really, those 3 would need to be compared separately.
-Checking the yields of a tile improvement directly doesn't factor what "else" might go on the tile. For example, Colossal Heads as a pure yield are probably one of the worse improvements out there, but they can be placed on otherwise unimprovable tiles so that definitely increases their value.
-Also, the additional bonuses of tile improvements are hard to quantify. +1 or +2 appeal is certainly a non-negligeable effect. And obviously no ranking can really perfectly quantify improvements like Nazca Lines which affect neighbour tiles. Tourism is also hard to quantify - virtually useless in domination games, but can single-handedly win you the game in a culture victory.
-And finally, there are many tile improvements which change value drastically depending on map and circumstances are very hard to quantify. Nazca Lines are fantastic if you have large desert sections, but obviously useless otherwise. That's where improvements like the Mission faltered - fantastic improvement in the right circumstance, but to get the stars to align was hard. Same for ones like the Sphinx or Chateaux in their previous incarnations - had trickier placements, so were hard to optimize.


As for the tiles overall, I think the biggest advantage for mines is that they can be placed after harvesting. So, for example, on raw yields, a lumbermill is probably roughly the same as a mine, but the power of the harvest is much much much stronger than one builder charge plus some appeal differences, so why not chop+mine? Similar for something like a deer camp - sure, it's pretty solid on its own. But why not chop+lumbermill instead? Farms are clearly the weakest individual tile, but they are very spammable, and gain their power when spammed. Triangles, diamonds, etc... create tiles that are unmatched for food, and especially if you plan then wisely and can share them, they can be doubly-effective.

Otherwise, the sheet is fairly decent in aggregating. I would generally agree that Cahokia Mounds and Nazca Lines are my #1 CS improvements when I see them, because amenities are great, and nazca are fantastic at making useless tiles into workable ones. Most of the others I have made use from, and in the right cases they can be fantastic - get the right stretch of land near water and volcanoes and your Moai zone there is absolutely bonkers. But unless if you plan for them, the others are less useful most of the time.

 

[iammaxhailme]

Vampire Castle, but that's a cheat answer.

Amojng more regular unique improvements I would say Polder or Terrace Farm if you're lucky, or Mekewap if you aren't (Mekewap still requires some luck but much less)

 

[Menocchio]

I think the core problem with your spreadsheet are that it doesn't quite take into account the E3 scale of improvements in Civilisation 6: how Early/Effectively/Easily you can build it and gain its bonuses. The mine, for example, scores well across the board: it's Early (can be built with mining), it's Effective (strong scaling production throughout the game), and it's Easy (can be built on any hills tile or suitable resource). The quarry, meanwhile, is equally Early, but not as Effective (it gains production at Renaissance Gunpowder and Atomic Rocketry as opposed to Medieval Apprenticeship and Industrial Industrialisation) and far less easy (has to be built on specific resources).

 

[KingPiggyXXI]

Strictly comparing the yields of the two, I'd say that Lumber Mills would be better than Mines. They both have similar amounts of Production (Lumber mills get +2 Production an Era earlier, but get +3 Production an Era later). However, Lumber Mills does have +1 Appeal from the Woods, and can be placed on flat land, which I'd say will push them to be slightly better than a Mine.

That said, strictly comparing them isn't going to give you a good answer. If you see a wooded hill, you almost never build a Lumber Mill. The value of the chop + Mine will be better than no chop + Lumber Mill. In many cases, the value of a chop alone is going to be better than a Lumber Mill. Lumber Mills might be able to narrowly beat Mines as the better improvement, but almost nothing beats chopping, which severely weakens how often you'll be building Lumber Mills. Lumber Mills have to not only compete with Mines, but with chopping the Woods as well. As such, Lumber Mills are stronger than Mines in theory, but in practice Mines are probably still the strongest and most useful normal Improvement.

 

[The Highwayman]

In regards to the Civ specific UIs, I'd rank the Great Wall near the very top. Cities can get to high pop (~10) relatively quickly with proper planning and now you're adding +6g/+4 culture per pop. Builder charges are plentiful with Qin's LA. Now you can buy more builders and buildings with the mass gold economy to make up for the lack of production in those cities. The massive culture spike in the late classical will be truly difference making - you can have the worst city in the world with no production but big pop that's feeding your empire global gold and culture. With experience the placement restrictions can be overcome, although there are some annoyances like floodplains, but most of these UIs have to deal with that. There are several reasons Qin is an S tier culture and science leader, and the Great Wall is one of them.

 

[Aurelesk]

I think the core problem with your spreadsheet are that it doesn't quite take into account the E3 scale of improvements in Civilisation 6: how Early/Effectively/Easily you can build it and gain its bonuses.

That's what I tried to replicate on the board at the far-right. For each era, you know how much Yields it gives. At the end, you have Averages:

• Avg as Averages accross all eras: the main problem is that TIs unlocked later have their averages leaded due to all the "0" they have. They end up to appear worst than they are.
• Avg no 0 as Averages accross all eras, not counting the 0s: the problem is now inverted: the TIs unlocked earlier have their averages leaded due to lower early Yields. They end up to appear worst than they are.
• Avg of Avgs as Average of both "Avg" and "Avg no 0": I guess it is the middle ground between both, so more representative to the overall power of the tile improvement. Not perfect, but I guess it is the least bad of the three?

If we need to rank the Tile Improvements with the "Avg of Avgs" value, then:
(not a real list since it is not fully well thought out: you should not take this seriously)

Regular ones:

1. Geothermal Plants and Seasteads (3.33): those are late games TIs, but they have no opportunity cost. Geothermal Fissure and empty Coast tiles had nothing, so you are willing to put them on. So I guess they are the best regular TIs after all!
2. Camps (2.83): they are gaining +1 Food and Production starting the Renaissance era, which is a huge improvement, and stays relevant from the Renaissance until the end of the game. The only "drawback" is it is a rather all-around TI (Gold + Food + Production) than a focused one (Mines and Lumber Mills is Production only).
3. Lumber Mills (2.48): Just slightly above from Mines. As said by @KingPiggyXXI: In practice, chopping for Mines is better since you gain more than 1.04 Yield per era from a Chop. (1.04 is from 1 (from feature), and 0.04 (from power difference). At Conservation, I guess we could replace all the Mines with second-growth Woods + Lumber Mills for 1 additionnal Production and 2 Appeal. But is it worth it? This need calculation (Production gains minus Builders costs and time to change all tiles).
4. Mines and Pastures (2.44): More like a 3bis, because they roughly have the same power. Mines can be put everywhere. This trait was not taking into account, so Mines are more valuable than expected in this ranking. (Mines should be higher)
5. Plantations and Fishing Boats (2.39): Not too far back. Let's say it is average power.
6. Quarries (2.11): As mentioned by @Menocchio, Quarries start to be lackluster due to gain their additionnal Yields later than other TIs.
7. Farms (1.61): Lackluster, as expected.
8. Oil-related TIs (1.63 / 1.56): They gain power mostly from the Oil itself (+3 Production) than the TIs.
9. Wind / Solar / Offshore Wind Farms (1.53 / 1.53 / 1.39): At least they give Power even if they are put outside of a workable range! The OSWF has no real opportunity cost before the Seastead, since Coast tiles have no proper TIs.
10. Seaside Resorts (1.44): With Appeal = 4, and Tourism not really counted. I didn't make them justice here. Their real strength is better than last. (Should be higher)

City-States ones:

1. Nazca Lines (7.11): The ranking does not take its restrictions (flat Desert tiles). In reality, this TI is not as good. Huge potential, but cannot be fully used. (Should be lower)
2. Cahokia Mounds (5.44): Adds 1~2 Housing and 1 Amenity. For me, it is one of the best of CS-TIs, mostly because you don't need to work it to enjoy from it.
3. Mahavihara (5.14): It is one of the rare TIs that gives Science, and it is able to give quite a few. The Mahavihara is really interesting if you milk the Campuses with it for additionnal Science.
4. Monastery (4.56): This is unexpected. It gives Housing and Faith, and Housing count almost as half of its power. If you do not face any Housing problem, the Monastery value is way less. (Should be lower)
5. Alcázar (4.00): A Culture and Science tile improvement. I believe it is a good one but underestimated. Chopping Woods and putting Mines everywhere harms the Alcázar strength quite a lot, so the power of this TI depends a lot on how you play.
   
6. Moai (3.99): It could be at the 5th place. Being a spammable TIs was not taking into account for the Strength of this TI. It is way usable and abusable than the Alcázar upthere, and it is in fact one the best TIs because of this. For me, it is on par with the Cahokia Mounds and the Mahavihara. (Should be higher)
7. Colossal Heads (3.67): As the Moai, it is spammable but this trait was not taking into account. But contrary to Moais, the power of Colossal Heads come from adjacent feature and not itself. You cannot spam it everywhere, because it would reduce the power of each Colossal Heads.
8. Batey (3.56): A powerful TI that gain Culture from adjacency. You can milk each Entertainment Complexes for 2 Culture with it, ECs that tend to be put more and more because they also gives 2 Culture to Theatre Square. The Bateys is clearly an interesting TI if you care about the adjacencies.
9. Trading Dome: It is hard to say. If you dedicate a city to Trade, then I guess you can make a lot of money out of this. I don't know how to really rank it.

Unique TIs from Civilizations:

1. Open-Air Museum (maxed) (7.78): Not as good due to 1 per city requirement. But clearly good, but not first. (Should be lower)
2. Ice Hockey Rink (7.00): Not as good due to 1 per city requirement. Its restrictions were not taken into account (tundra/snow tiles), so it is clearly not as efficient since not all cities can enjoy from it. (Should be lower)
3. Mission on foreign continent (6.98): Spammable. This is, for me, the best TI in the game. But you need to fulfill the requirement. (Should be higher)
4. Golf Course (6.11): Not as good due to 1 per city requirement. Same thing as the Rink, except you can put it in almost all your cities. (Should be lower)
5. Sphinx (5.81): A really powerful TI that helps a lot toward National Parks creation. If you do not care about Tourism and Appeal management, this TI is not worth it.
6. Polder (5.46): Spammable. Too restrictive to use, it is a "all-or-nothing" TI. Great if you can put some of them, but overall, that TI is too situational. (Should be lower)
7. Pairidaeza (5.36): It lost a lot of power since it is now only 1 Appeal. It works like a Sphinx except it has quite a nice Culture output.
8. Nubian Pyramids (5.00): Spammable. Its restrictions were not takien into account (flat desert tiles). So it is clearly not as efficient since not all cities can enjoy from it. (Should be lower)
9. Great Wall and Mekewap (4.83):
   • Great Wall is spammable and can achieve a lot of Culture. But you need to be vigilant and plan ahead to make full use of it, because its restrictions can be quite hasardous: a border expansion can ruin your whole planning. (Should be higher)
   • Mekewap is fairly good at giving Housing and Production on flat tiles. It is an all-around TI that give cities a nice boost.
10. Terrace Farm (4.44): It is basicaly a Farm and a Mine at the same time. It depends a lot on the synergy with Incan abilities, but overall, it is a nice one for growth and production. (Should be higher)
11. Kurgan (4.28): It is spammable. You can end with a 3 Gold/Faith around all Pastures, which can give a nice economy boost and Tourism. The downside is that working the tile prevent the city to grow or produce something. (Should be higher)
12. Hacienda (4.10): It is spammable. You can manage to have a lot of Production out of this, so it should probably be at a better rank. (Should be higher)
13. Château (4.03), Rock-Hewn Church (4.01), Chemamull (4.00), Stepwell (4.00)
    • Château: It is better, and can work as a Pairidaeza except you need Bonus or Luxury resources. Why that additionnal restriction? It is also dependent on Wonders to build its Culture. It is a better version.
    • Rock-Hewn Church: It works mostly because of the Ethiopian abilities and bias. It can also achive really high Faith output on right circonstances, so it is a nice tile.
    • Chemamull: Spammable. It depends on Appeal, so the Strength of it will depend a lot on if you care about Appeal management. (Should be higher)
    • Stepwell: Quite an odd one, focusing on Growth and Faith. It has flat restrictions and can break some farms triangle on the process, making the use of it quite hard sometime. A good one for Housing purposes, but fails to be good if you do not need those Housing;
14. Mission on home continent (3.87): Spammable. Not as great as the foreign one, but can achieve some good Yields nonetheless. (Should be higher)
15. Outback Station (3.68): Spammable. Same as Hacienda except it makes use of Desert, increasing a lot its value. (Should be higher)
16. Kampung (3.66): Spammable. Have no real opportunity cost because it doesn't take the place of any other TIs. (Should be way higher)
17. Ziggurat (3.06): Spammable. Well, it is straightforward, and the early Science is incredible to put Sumeria on the top of all civilizations. (Should be way higher)

I am repeating myself: do not take this ranking seriously. First because the "Strength" does not take any of the tile restriction. For example, if a TI give 1000 Production but have to be put on a Snow tile next to a Mountain Desert iles would be the "best" in the ranking, while being unusable in the game. I guess we could say it is a "inner Strength" of the tile improvement, not counting its "usefulness".

 

[bengalryan9]

There is one knock on mines that I didn't see mentioned in the original post, and that is that they decrease the appeal of tiles around them. For most people I don't think this is a huge drawback, but for those of us who really like spamming national parks or playing as certain leaders (BM Teddy, John Curtin, or Lautaro for example).

For this reason (and because I really just don't personally like chopping as a mechanic) I almost always choose lumbermills over mines given a choice between the two.

 

[Archon_Wing]

You can just replant trees if you need the appeal back. By then 1 builder charge is not that big of a deal.

Funny enough if you want to use preserves, you're better off putting neither on the tile.

 

[Siddharth Venkatesh]

Rock-Hewn churches are far too low on your list. The yield may not look as impressive as other improvements, although I think you're underrating how strong early faith is, but they unlock very early and they synergize extremely well with Ethiopia's other abilities and general game goals. They combo insanely well with Monumentality golden ages (which they give 4 era score for) and with Ethiopia's bonuses from faith. They also provide a lot of tourism on flight. They used to be slightly better before the Earth Goddess nerf but they are still one of the best civ specific UI. Ethiopia isn't the fastest culture victory civ but they are incredibly reliable and Rock-Hewn Churches are one of the big reasons why.

You do lose some space for mines with the churches but I find that you can still find enough space for mines in between them and you can get a lot of production from your faith via monumentality and Grand Master's Chapel.

Terrace farms are interesting. They can be really strong on the right maps but they do compete somewhat with preserves for Pachacuti. Being able to work mountains makes preserves really good for Inca.

Mahaviras are obviously the strongest improvement if you abuse the bug where you unlock every tech instantly.

 

[Aurelesk]

Rock-Hewn churches are far too low on your list.

Rock-Hewn Churches are powerful. Its power is estimated at 4.01, more than 60% stronger than a Mine at 2.44. To make it more intelligible: 3 Rock-Hewn Churches are worth as much as 5 Mines. The problem with the formula is that it doesn't take into account:

Synergies with the Civilization's and Leader's abilities:
Rock-Hewn Churches are made 30% stronger due to Menelik II's ability that receive Science and Culture out of Faith. If we add this, the Strength climbs to 5.23, which will put it at the 7th place between the Pairidaeza and the Nubian Pyramids.
But "if" the tile improvement were available to everyone, it wouldn't be as strong. That's why I didn't took the bonuses from abilities into account.

Min-Max value and wasted potentiel:
The Strength value is an "average estimated" to say the least. Rock-Hewn Churches' Faith output vary vastly:
→ If the Rock-Hewn Church will give low Faith, therefore you will put a Mine instead.
→ If the Rock-Hewn Church will give high Faith, therefore you will put it and gain a lot from it.
The average Faith of the Rock-Hewn Churches in your empire will be, in fact, higher than the average Faith estimated.

Tourism and abuse:
Sadly, Tourism was not taken into account (how to rate it?). Meanwhile, the modifier "Faith to Tourism" is abusable, because you can cheat the system and had more Faith to the tiles and translate it to Tourism:
• From natural Wonders like Delicate Arch, Mato Tipila, Mount Everest, Mount Roraima or Uluru.
• From Nazca Lines
• From Pantheons like Earth Goddess or Fire Goddess.

Emergent gameplay:
As you said, it has incredible synergy with the Monumentality golden age.

In the end, Rock-Hewn Churches end up to be stronger than it appeared. The thing is, this is also the case for some other tile improvements like the Golf Course and a gameplay around Amenity, Sphinx / Château and a gameplay around World Wonders (wait! bad example!), Kampung and a coastal gameplay... and so on.

Short: I agree with you. This is incomplete and misleading. So do not hesitate to suggest anything to improve the method.

 

[bengalryan9]

You can just replant trees if you need the appeal back. By then 1 builder charge is not that big of a deal.

Funny enough if you want to use preserves, you're better off putting neither on the tile.

One charge (to plant trees) if you're removing a mine in a tile that you want to include in a national park. I'd also always plant trees in any tile that's going to be boosted by a preserve.

Two charges if you're removing a mine in a tile that will be next to a national park - one to replant trees, and another to add a lumbermill (which you might as well do since it doesn't hurt appeal of neighboring tiles).

Old growth forests give more appeal than new growth forests, too, so I personally would rather just not chop them at all even it isn't necessarily the optimal way to play as far as production is concerned.

 

[Futumch]

Vampire Castle, but that's a cheat answer.

Hmm. My cheat answer was "a City".
Still, this looks like a good thread to read thoroughly.

 

[Tomice]

Regarding lumber mills:
It seems weird to me that they start with +2, but rather late in the game. If we can remove entire forests by chopping with our very first tech (mining), why can't we harvest them continuously? It's not even historical, as even the earliest cultures chopped wood using stone tools.

Allowing +1 lumber mills earlier would further tip the balance between mines+chopping and lumber mills, finally making the latter almost competitive.

 

[Aurelesk]

Regarding lumber mills:
It seems weird to me that they start with +2, but rather late in the game. If we can remove entire forests by chopping with our very first tech (mining), why can't we harvest them continuously? It's not even historical, as even the earliest cultures chopped wood using stone tools.

Allowing +1 lumber mills earlier would further tip the balance between mines+chopping and lumber mills, finally making the latter almost competitive.

I have no real answer, just hypothesis.

• Before Lumber Mills, we can speculate than it was mainly used as Power, either for domestic use (cooking, light...) or industrial use (smelting, ceramics...). The main leader of Deforestation around the Mediterranean was... the Roman Empire. Deforestation was also a way to make room for Farms and have enough Food to feed the Population.
• After Lumber Mills, we can speculate that Medieval Dark Age era contribute to a lesser deforestation and more used as lumber than fuel. I guess?

Before Gathering Storm, Lumber Mills were unlocked at Machinery (Medieval) for +1 Production, and +1 additional Production if next to a River. They changed it, so it is unlocked earlier at Construction (Classical), and made the Production bonuses unconditional. That's why it is +2 Production from the start.

What Ancient technology should the Lumber Mill be unlocked? Wheel?

I think the biggest issues with comparing tile improvements are the following:
-One per city vs "cannot be placed next to each other" vs spammable have very different valuations that really, those 3 would need to be compared separately.
-Checking the yields of a tile improvement directly doesn't factor what "else" might go on the tile. For example, Colossal Heads as a pure yield are probably one of the worse improvements out there, but they can be placed on otherwise unimprovable tiles so that definitely increases their value.
-Also, the additional bonuses of tile improvements are hard to quantify. +1 or +2 appeal is certainly a non-negligeable effect. And obviously no ranking can really perfectly quantify improvements like Nazca Lines which affect neighbour tiles. Tourism is also hard to quantify - virtually useless in domination games, but can single-handedly win you the game in a culture victory.
-And finally, there are many tile improvements which change value drastically depending on map and circumstances are very hard to quantify.

I am researching more ideas to add new modifier to the list, so the comparison would be less misleading. So far, I am working around 3 more: availability, workability and opportunity cost.

Availability: How many TIs can fit in a city, and how often can the TI be put. It will be distinct from "Strength". A tile improvement could be extremely strong but also being hardly available.

• "1 per city" would be at 1 (per city), "not next to each other" would be at 3, and "spammable" would be at 9.
• If additional restrictions, like Flat, Tundra, Rivers, next to resources... then a modifier as percentage applies. For regular Continent maps, I believe this is around:
  • Snow is ~2.5%, Tundra is ~10%, Grassland is ~37.5%, Plains is ~37.5% and Desert is ~12.5%.
  • Flat is ~66.5%, Hill is ~25% and Mountain is ~7.5%.
  • ~15% of land have rivers, and ~30% of river tiles are floodplains.
  • ~30% of passable lands have ressource (spread as 50% Luxury, 25% Bonus and 25% Strategic)
  • ~30% of passable lands have features (excluding Floodplains)
    • 10% of flat Grassland is Marsh.
    • 25% of Plains is Rainforest.
    • Woods percentage seems to vary: 15% for Plains, 20% for Tundra and Grassland?

Workability: Some Yields needs to be worked, some don't. Even those worked, some are more City-wise and some are more Empire-wise (Science, Gold, Faith). So I will split into categories:

• Need to be worked:
  • To city: Food, Production. Culture for border expansion.
  • To empire: Science, Culture for civic, Gold and Faith.
• Do not need to be worked: Housing, Amenity, Appeal, Tourism, Power, Loyalty

Opportunity cost: Mostly for non-regular tile improvements. Basically, how much more would give those special TIs. I will probably just compare to Mine on Hills, Farm on Flat lands, Fishery on Coast tiles, and nothing on Flat inhospitable tiles. But I do not like much that idea: sometimes you don't care the TI is not as good as another, because you are searching for a specific thing instead of raw strength.

 

[Tomice]

What Ancient technology should the Lumber Mill be unlocked? Wheel?

That's the best candidate, I agree!

 

[Menocchio]

Technically, the "lumber mill" in-game represents the sawmill, which was not invented until around the 200s AD (by the Romans) based off of more complex water wheel technology. Instead, I would change its production increases from Steel and Cybernetics (what on Earth do these technologies have to do with sawmills??) to Machinery and Steam Power. That would make them peak sooner than mines but still offer the question of whether to chop (immediate gain) or not (long-term gain, now scaling better). That said, all end-game tile yield improvements ought to be pulled back era-wise, I think. Getting a +1 production boost at a technology like Smart Materials, which also unlocks Exoplanet Expedition and essentially marks the very end of the game, is basically a joke.

 

[aieeegrunt]

There is one knock on mines that I didn't see mentioned in the original post, and that is that they decrease the appeal of tiles around them. For most people I don't think this is a huge drawback, but for those of us who really like spamming national parks or playing as certain leaders (BM Teddy, John Curtin, or Lautaro for example).

For this reason (and because I really just don't personally like chopping as a mechanic) I almost always choose lumbermills over mines given a choice between the two.

Chopping honestly feels gamey and broken to me, it’s too much of a boost

Regarding lumber mills:
It seems weird to me that they start with +2, but rather late in the game. If we can remove entire forests by chopping with our very first tech (mining), why can't we harvest them continuously? It's not even historical, as even the earliest cultures chopped wood using stone tools.

Allowing +1 lumber mills earlier would further tip the balance between mines+chopping and lumber mills, finally making the latter almost competitive.

Technically, the "lumber mill" in-game represents the sawmill, which was not invented until around the 200s AD (by the Romans) based off of more complex water wheel technology. Instead, I would change its production increases from Steel and Cybernetics (what on Earth do these technologies have to do with sawmills??) to Machinery and Steam Power. That would make them peak sooner than mines but still offer the question of whether to chop (immediate gain) or not (long-term gain, now scaling better). That said, all end-game tile yield improvements ought to be pulled back era-wise, I think. Getting a +1 production boost at a technology like Smart Materials, which also unlocks Exoplanet Expedition and essentially marks the very end of the game, is basically a joke.

A lot of Civ6 is like this, good concepts, awful execution

 

[lilnev!]

I like lumbermills better than mines on wooded hills. Those are generally the highest production tiles available, and will remain so for the rest of the game.
Flatland woods might get chopped or might get a lumbermill. Often depends on the value of the tile as a district or farm.
Rainforests generally get chopped early, though if they are still around or I acquire them later, I would consider a lumbermill.

 

[Mr_PeaCH]

Chopping honestly feels gamey and broken to me, it’s too much of a boost

I think about this too; here are some ideas I've come up with.

Chopping should be most useful early, in the first 2 or 3 ages. After that the boosts should taper off imo.

Old growth forest should be better than replanted forest later in the game; maybe +2 appeal vs. +1 or +1 and have replanted forest not give appeal.