Improvement Balancing

I was looking at the various improvements that we have in AND and I noticed that the yields from improvements vary quite wildly, and that there are a lot more bonuses than in regular BTS. I think it would be good to maybe prune down some of these bonuses, especially to make it so that certain improvements aren't overly superior to others. I'm specifically looking at yields here, and I am also limiting this to the 23 improvements that can be built without needing a resource. There are a total of 48 improvements in the Civilopedia, but 7 are aquatic, 3 are defensive, 5 can't be built normally (two City Ruins, Depleted Mine, Tribal Village, and Machu Picchu), and 10 can only be built on resources. I think we can look at the resource-tapping improvements later. I'm also not concerned with bonuses from traits or civics; I think balancing those bonuses falls under balancing the trait/civic, not the improvement.

I do not want to remove any improvements. Some time ago, we consolidated the Orchard improvement from Apple Orchard and Olive Orchard. When we did this, if I remember correctly, it broke saved games. I don't want to break games, so I think we can leave all existing improvements in place.

I do have two theory questions that I'd like to get some feedback on before looking at particular improvements. I want to be able to judge improvements in a framework and not just try to compare them as individual items.

First question: Should the different yields be considered equal? My current thought is that food and production should be considered equal, but not commerce. Going back to BTS, a Mine is worth +3 production (+2 base, +1 from Railroad) and a Workshop is worth +3 production (+1 base, +1 from Guilds, +1 from Chemistry)/-1 food. A Farm is worth +2 food (+1 from Irrigation, +1 from Biology), which is worth 1 Citizen, but since AND citizens consume 3 food each, we should treat a BTS farm as +3 food, which is an equal bonus to the Mine and Workshop. A BTS Town is worth +5 commerce (+4 base, +1 from Printing Press). This says to me that a +1 commerce is worth about 50-60% of +1 food or +1 hammer. Also in regular BTS, a Windmill gives +1 food/+1 production/+2 commerce and a Watermill gives +2 production/+2 commerce. Both of these suggest that if the best non-commerce yield is +3 total, +2 commerce is only worth +1 of the either two. So I think 50% is the right conversion factor.

Second question: What should be the maximum bonus from an improvement? There are a group of improvements that I consider the "baseline": Farm, Cottage, Mine, and Workshop (along with all their upgrades). These four improvements, fully upgraded and with all technology bonuses, top out at around +7 total yield without weighting. With weighting, all except Town drop to about +6; Town gets a significant downgrade and I think it could be beefed up a little. There are some improvements that don't show up until the very late game (like Geothermal Factory). I think these can qualify to be better than the baseline. However, improvements that do show up at the same time as these baseline improvements shouldn't be better. As an example, Watermill currently gets all the way up to +12 total yield. I don't think that is a good thing.

Let me know what you think and then I will move onto specific improvements.

I think I've never seen such a deep examination so I totally trust you. Only thing I've noticed is that watermill yield was very high.

I also trust you for that!
I agree that the weight of commerce is 50% inferior to the others bonus as there are many things in game that have influence on it (positive or negative). Plus, it isn't useful for a town to grow or produce, and thus for a Civilization to "physically" grow, so it needs to be treated differently. The idea of increasing the food production for farm to balance it is nice but you should try to theorize the impact on the balance of the game if making such a change.

About the second question, i haven't understood what you call weighting here: "top out at around +7 total yield without weighting". Sorry for the dumb answer here :x

Factors to consider:

1) Mines in BTS have a chance to discover resources. It can be treated as a long term investment with just +3 hammers compared to workshops which can give +4 with civic.
So given a choice, I may still build mines instead for that slight chance.

2) Towns cannot be built directly. It takes ages for a cottage to grow to a town. Comparing it directly to farms or mines which can be built in one turn is unfair.

3) Windmills and watermills may have slightly more total yield, but they can only be built on specific plots (rivers and hills respectively), compared to farms or cottages which can be spammed in more plots.

platyping said:

1) Mines in BTS have a chance to discover resources. It can be treated as a long term investment with just +3 hammers compared to workshops which can give +4 with civic. So given a choice, I may still build mines instead for that slight chance.

Click to expand...

I agree. If I have nothing better to build in a tile, I might well put a mine on the (very) off-chance that it spawns something later on.

platyping said:

2) Towns cannot be built directly. It takes ages for a cottage to grow to a town. Comparing it directly to farms or mines which can be built in one turn is unfair.

Click to expand...

Well, that could be solved by allowing the player to build hamlets or villages directly later in the game (say, Mediaeval and Industrial). Otherwise, yes, cottages are a very slow-burn investment.

Arakhor said:

Well, that could be solved by allowing the player to build hamlets or villages directly later in the game (say, Mediaeval and Industrial). Otherwise, yes, cottages are a very slow-burn investment.

Click to expand...

But in BTS, they were a bit overpowered. At some time, you can just build cottages everywhere! (and buy a Ferrari for every citizen).

Buy a Ferrari for every citizen.

I agree that cottage economy is less useful in AND, but towns can still be destroyed and then you have to wait nearly forever to get them back, unlike mines, farms etc.

That's right

AND Mines get much stronger than BTS Mines, although I'm actually happy where those are (although Mine itself is too front-loaded for my taste). It's several other improvements that need to be cut down. Mines and Workshops are also buildable on different tiles; Mine requires Peak or Hill or bonus, while Workshop requires flatlands.

Watermills and Windmills are effectively the "hybrid" improvements. The real baselines of Farm, Mine, Workshop, and Cottage increase one yield primarily and then dip into the others at some points, mostly commerce. Town dips into both, but late. Watermill and Windmill, on the other hand, are spread out from the beginning. So I agree Watermill/Windmill could be slightly higher in total yield as long as they don't completely outclass the baselines.

I'm not quite sure how to account for the time factor involved for Cottages to grow. I think I'm more concerned about the tech limits. In AND, Hamlets don't show up until mid-Classical (City Planning), Villages are early Medieval (Civil Service), and Towns are mid-Renaissance (Social Contract). So that takes precedence.

Something that AND does that BTS does not is unhealth from improvements. Several improvements generate unhealth (Mine, Town, Industry) and I think that needs to be considered.

1 point of unhealth is worth somewhere between 0 and -1 yield. If your city is unhealthy, then each point of unhealth is worth -1 food. If your city is healthy, then there is no yield penalty. So I think the best way to count this is to split the difference and count each full point of unhealth as -0.5 yield. Regular Mines with 1/2 unhealth to start can count as -0.25. I will round off at the end.

Here is my first example, with the most basic improvement: Farm.

Farms
Farms are my baseline of baselines, since it's so difficult to grow a civilization without them. I understand that we do not want to add any extra food production early on to avoid cities getting too big too fast. In fact, I'm reluctant overall to mess with farms any more, except to remove the +1 food from Biology because it's too close to Agricultural Engineering. Biology can survive without the +1 food.
So this is the yield curve over time for Farms:

• Default: +1 food (if irrigated; but you aren't building farms without either irrigation or a resource, and irrigation spreads at Canals, which is mid-Classical Era)
• Medieval Era: +1 food from Crop Rotation
  • Net: +2 food (total yield +2)
• Industrial Era: +1 food from Agricultural Engineering
  • Net: +3 food (total yield +3)
• Modern Era: +1 food from Gene Manipulation, +1 commerce from Biomaterials
  • Net: +4 food, +1 commerce (total yield +5; +4.5 with commerce weighted at 50%)
• Transhuman Era: +1 food from Artificial Evolution, +1 commerce from Vertical Farming
  • Net: +5 food, +2 commerce (total yield +7; +6 with commerce weighted at 50%)

I think overall a +6 yield is sufficient.

Does this also explain how weighting commerce works? It means that we can give out small commerce bonuses without issues, but improvements focused on commerce need to have higher totals to be significant.

45* is on good track here! I trust Vokarya's decisions because I am very impressed with his streamlining of the game!

Mines are the hill/peak production improvement. Their yield curve jumps up higher than Farms initially and then flattens out.

• Default: +2 production, -0.5 health
• Classical Era: +1 production from Iron Working, +1 production from Road (available at Monarchy)
  • Net: +4 production, -0.5 health (total yield +3.75)
• Renaissance Era: Shaft Mine is available with Metallurgy. This nets +1 commerce and 0.5 unhealth over basic Mine.
  • Net: +4 production, +1 commerce, -1 health (total yield +4.5; +4 with commerce weighting)
• Modern Era: Modern Mine is available with Modern Seismology. This nets +1 production over the Shaft Mine.
  • Net: +5 production, +1 commerce, -1 health (total yield +5.5; +5 with commerce weighting)
• Transhuman Era: +1 production from Nanomining
  • Net: +6 production, +1 commerce, -1 health (total yield +6.5; +6 with commerce weighting)

The problem that I see with Mines is that the default Mine gets bonuses very quickly. Going back to regular BTS, the only bonus that a Mine gets over the entire course of the game is +1 production once a Railroad is built on that tile. We start the +1 hammer bonus at Road instead of Railroad and give Mines additional +1 hammers at Iron Working, but then there really isn't another hammer bonus until Modern Seismology and the Modern Mine (the regular Mine with Iron Working and the Shaft Mine are both +3 hammers).

For comparison, here are the curves by era for Farms vs. Mines. They wind up equal in the end, but Mine is way ahead for the early game.

So the only change I would like to make here is to reduce Mine, and only Mine (not Shaft Mine or Modern Mine) production by 1. This would make the curve rounder and not give Mine so much of an early bonus.

platyping said:

Factors to consider:

1) Mines in BTS have a chance to discover resources. It can be treated as a long term investment with just +3 hammers compared to workshops which can give +4 with civic.
So given a choice, I may still build mines instead for that slight chance.

Click to expand...

That is the reason I prefer to build mines on hills and mountains.The only case that I may not build them is when a resource already exists so there is no possibility to discover resource.Also I don't like workshops because of -1 food.I would build them only around cities with very,very low base production.
Watermill is my favorite improvement.

Arakhor said:

Well, that could be solved by allowing the player to build hamlets or villages directly later in the game (say, Mediaeval and Industrial). Otherwise, yes, cottages are a very slow-burn investment.

Click to expand...

Totally agree. Either the option to build hamlets or villages later with tech would be nice, OR a tech/building/civic that greatly boosts the upgrade rate of the cottage/hamlet/village would be very nice. I'm thinking public works, post-scarcity, free market, etc.

Another option if the above are not that feasible: have the growth rate from cottage/hamlet/village increase if there is a road/railroad built on top of it. More advanced roads increase growth rate further. That also has the advantage of reflecting the real world historically. For example: Road: -5% time required to upgrade. Paved Roads: -10% time required. Railroads: -20% time required. Highways/Advanced Railroads: -30% time required.

Regarding the Farm: a +1 yield boost starting in the late classical era makes sense, but I wouldn't add more than that. Earlier than that and you could see some really rapid growth though.

For most of the game, the Watermill beats the farm easily - no thinking required. The +1 yield to farm will help, but we could also consider restricting the watermill so they can't be built adjacent to each other (would be pretty harsh though), or something similar.

Delekhan said:

Another option if the above are not that feasible: have the growth rate from cottage/hamlet/village increase if there is a road/railroad built on top of it. More advanced roads increase growth rate further. That also has the advantage of reflecting the real world historically. For example: Road: -5% time required to upgrade. Paved Roads: -10% time required. Railroads: -20% time required. Highways/Advanced Railroads: -30% time required.

Click to expand...

Actually, that is a really great idea! It should be feasible in C++. Let's wait the team answer.

Watermills are restricted to one side of a river. On a river that's bordered by 2 tiles, they can only be built on one side and not both.

The river has to run down a tiles edge, and even then, rarely, you can't build a water mill on a tile, due to a water mill on another seemingly unrelated tile.

With automated workers, they seem to prefer watermills over farms, so they would seem to be overpowered.

Hi,

good idea to review the improvements.

As far as improvement balancing goes: watermills are too good. I think they should lose some of the extra benefits they get with technologies, so that they eventually become obsolete (superseded by Industry and/or farms). We don't still have water-wheels these days!

If we can't eliminate improvements without breaking games, perhaps we can just make any which we want to delete no longer buildable? Perhaps there'd be scope for 'solar farm'. These seem to be more and more in the news these days, and being able to do something useful with Desert would be nice I think. I would see it as primarily a improvement, perhaps a or two.

Cheers, A.



My personal wish would be that improvements would result in a more realistic countryside. As it is, I have almost no farms by the Industrial Era, which I find unreasonable. My experience is that farms are just not needed in most cities, since watermills provide so much food, and even a lumbermill provides two food when it's on green land. Then you add in Canneries and the other food buildings, some food corporations, and even a tundra city can reach a decent size - surrounded by nothing but industry!

In reality, of course, an aerial view of almost anywhere that people live will show that farms absolutely dominate the countryside.

But I'm not sure how we'd achieve that in Civ.

What I'd really most like to see is production move from the countryside around the city to inside the city itself. By this, I mean that factories and other production-oriented buildings would open Worker slots, where citizens produce production hammers. Early in the game there are only a few of these Worker slots available - perhaps from the Carpenter, boatyard, etc... But, during the Industrial era, vast numbers of new factory jobs would become available, allowing for drastically increased production capacity. To feed all of these factory workers, the area around the city would need farms.

This would mean that a city on productive farmland could and would become a huge production centre - OR you could specialise it in other ways (more flexibility). I would find this much more realistic - we don't have factories in the Pilbara (mining area of Western Aus'); the factories are in Melbourne and Sydney. And it's probably the same everywhere, right?

I would expect that it's also easy to teach the AI how it works, since the city Governor already knows about hammers in assigning specialists.

Perhaps that's all a bit too drastic a change.

Vokarya said:

I was looking at the various improvements that we have in AND and I noticed that the yields from improvements vary quite wildly, and that there are a lot more bonuses than in regular BTS. I think it would be good to maybe prune down some of these bonuses, especially to make it so that certain improvements aren't overly superior to others.

Click to expand...

I'm glad to see this, as I think there are still a few issues with balance.

Vokarya said:

First question: Should the different yields be considered equal? My current thought is that food and production should be considered equal, but not commerce. Going back to BTS, a Mine is worth +3 production (+2 base, +1 from Railroad) and a Workshop is worth +3 production (+1 base, +1 from Guilds, +1 from Chemistry)/-1 food. A Farm is worth +2 food (+1 from Irrigation, +1 from Biology), which is worth 1 Citizen, but since AND citizens consume 3 food each, we should treat a BTS farm as +3 food, which is an equal bonus to the Mine and Workshop. A BTS Town is worth +5 commerce (+4 base, +1 from Printing Press). This says to me that a +1 commerce is worth about 50-60% of +1 food or +1 hammer. Also in regular BTS, a Windmill gives +1 food/+1 production/+2 commerce and a Watermill gives +2 production/+2 commerce. Both of these suggest that if the best non-commerce yield is +3 total, +2 commerce is only worth +1 of the either two. So I think 50% is the right conversion factor.

Click to expand...

I personally think food is even more important than production. Larger food yields can support larger cities, larger cities can work more tiles, support more specialists, etc. Whereas larger production does not directly cause larger cities.

I think the method you started using for scaling is fine though. I agree with both the farm and mine proposed changes.

Afforess said:

I personally think food is even more important than production. Larger food yields can support larger cities, larger cities can work more tiles, support more specialists, etc. Whereas larger production does not directly cause larger cities.

Click to expand...

I kind of agree, but I don't have a good way of really comparing food to production. I am trying to be the most careful about food yields. I know we have removed a lot of bonuses from Farms that we had in past versions of AND, and I'm treating the Farm as the baseline that everything else gets compared to and brought down to if necessary.