Slowing down research

[cryptc]

I've searched a bit on the forum but not found a mod yet that does what I want.

Does anyone know how to slow down research for the game? I don't mean just turning game to Marathon, since the problem I have is that I like playing a slow pace, not one where if I don't attack quickly, all my troops are really outdated. And I can't seem to get into a good "final battle" type situation until I'm well into future tech, which means I need to fear the dreaded nukes that I despise...

In semi-related question, does anyone know how to disable nukes?

 

[Decimatus]

You can change the values in Marathon to make various aspects of the game faster or slower.

 

[alpaca]

Two ways: Either create an xml that updates the costs of all techs manually, or use SQL to make a general change. The latter is much quicker, but the former allows you to adjust tech costs more to your taste.

For example, I'm using the following SQL syntax to scale tech costs depending on era, which drastically increases era "definition" in that there is a strong cost incentive to finish a tech from an older era rather than start a new one:

UPDATE Technologies
SET Cost = 1.25*Cost
WHERE Era = "ERA_ANCIENT";

UPDATE Technologies
SET Cost = 1.5*Cost
WHERE Era = "ERA_CLASSICAL";

UPDATE Technologies
SET Cost = 1.75*Cost
WHERE Era = "ERA_MEDIEVAL";

UPDATE Technologies
SET Cost = 2*Cost
WHERE Era = "ERA_RENAISSANCE";

UPDATE Technologies
SET Cost = 2.5*Cost
WHERE Era = "ERA_INDUSTRIAL";

UPDATE Technologies
SET Cost = 3*Cost
WHERE Era = "ERA_MODERN";

UPDATE Technologies
SET Cost = 3*Cost
WHERE Era = "ERA_FUTURE";

To disable nukes, the easiest way is to set their PrereqTech to TECH_FUTURE_TECH. This works because, even though you research the future tech multiple times, you never actually have it.

 

[cryptc]

Thanks for the tips, the SQL thing seems the simplest for me, and I picked up the modders guide to help me find the .db file to alter (Civ5CoreDatabase.db).

One thing I wonder about though, it seems that the db is made from the xml's, but I can't see when this happens. Will my change suddenly vanish on a game as the database resets? and how can I manually reset the values? will every same, new and old be affected?

 

[cryptc]

Also, this does affect the computer too right? <,<

Since I just started a fastest game type, but with tech * 10 as slow, King difficulty... and I had an AI that build Stonehenge in turn 28, before I even had chance to get a single tech. I guess he could been lucky and gotten two techs from ruins though...

 

[Spatzimaus]

Also, this does affect the computer too right?

Yes. Most mods like this do not, however, affect existing save games, only newly made ones.

Since I just started a fastest game type, but with tech * 10 as slow, King difficulty... and I had an AI that build Stonehenge in turn 28, before I even had chance to get a single tech. I guess he could been lucky and gotten two techs from ruins though...

On higher difficulties the AI starts with extra free techs. King and higher give Pottery, Emperor and higher give Animal Husbandry, and the highest difficulties get Mining as well. So with Pottery, you only need one tech to get to Stonehenge, and it's likely ONE of the AIs found that in a ruins. This is all spelled out in the HandicapInfos table.

There are several ways to slow down research through XML edits. For instance, you could adjust the ResearchPercent of the four game speeds, which'd slow down the tech pace without changing the pace of units. This'd throw the year indicators off, of course. Or instead, you could adjust the ResearchPercent of the various Eras. (Note that the ResearchPercent values in the Era file apply to games that START in that era, it's not something that's applied dynamically as you move your way up the tech tree.) Or ResearchPercent for the various map sizes. (Yes, those modify it too. Small is the default, Standard is +10%, Large is +20%, and so on.)

But ask yourself WHY you want to slow it down. Do you hate the pace for the entire game, or do you just think that it starts off at a decent pace but then gets too fast (eventually reaching 2-3 turns per tech)? If it's the latter, then there are a few things you can do without SQL edits:
1> Wait until the next patch, which should be out within a week or two. They're repricing all of the techs to make ones later in the tree cost up to twice what they do now. (Effectively doing what Alpaca suggests.)
2> Decrease the effects of the +research buildings, since they're the main difference between late-game research and early. In my mod, for instance, I changed the Library, University, Observatory, Public School, and Research Lab from (+1 per 2), +50%, +50%, +50% (changed to +1 per 1), +100%, down to (+1 per 3), +30% and 1 culture, +20% and 3 culture, +40% and 1 culture, +50%. (And the Medical Lab adds +10% in addition to its food storage.) The result is that in the endgame, a fully-loaded city in a vanilla game would gain 7.5 research per population, while in my mod it's only 3.33, so the techs slow down as you go, without adjusting their costs at all.
3> Slow down the growth of cities. This doesn't hurt production/gold very much, since most of those are driven by tile yields and you'll work the best tiles first. But science is directly proportional to population, so it'd decrease faster than those others if you made it harder to grow.
So alter the growth formula. Instead of 15+8x+x^1.5, where x=(size-1), I made it 20+10x+x^1.8, although I then added more food storage to offset this a bit. The exponential part is the most important; in the late game, when cities are size 10+, it makes a huge difference in how fast the cities continue to grow.
4> Remove some of the Scientist specialist slots from early buildings. Most specialists aren't productive enough to be useful in the ancient era; an Engineer's +2 production is just less than what a good tile can give. Scientists aren't so bad.

 

[cryptc]

Thanks for some more advice, I'll ponder a bit more to try figure out what works well for me.

As to why, I just analyzed how I was playing/enjoying the game, and what seemed wrong to me. I had first started playing on standard, only to find that you very very quickly are in rather modern eras. So I tried the slower paces, which atleast allowed me the time to properly explore the map, and get some use out of single units... but with the equally slowed down build time, it still felt like I didn't really get to do anything... just when I'd start to get a game going well, I'd either be winning a cultural victory, reaching future tech, or getting dominated by someone having 90% of the map.

I really enjoy building cities, making them work well and optimally and level up military units. I found that I never ever got to build most buildings, which I realize is a design decision from the game developers, to try make people make hard choices what buildings they want yada yada... thing is, that the whole ICS thing is painful for me to play, even limiting growth of cities to keep happiness kind of hurts, but atleast that I can try optimize how it fits for myself...

In short, the game doesn't fit how I'd prefer to play it.

Anyways, the patch sounds promising (and the patches so far has greatly improved the game). Research buildings alterations also seems promising for what I'd prefer.

Final question... is it possible to limit technology to some era? for instance that noone can get past medieval period? likely only by doing some sort of mod to put prereq you can't get on them?

 

[Spatzimaus]

So I tried the slower paces, which atleast allowed me the time to properly explore the map, and get some use out of single units... but with the equally slowed down build time, it still felt like I didn't really get to do anything...

That IS the problem. "Marathon", "Epic", and "Quick" don't actually play very differently from Standard; you have a little more time to micromanage your units, but the rest is just like playing Standard but taking longer to get anything done. Back when there was no tech overflow this was a big deal, but now that that's fixed, I don't see much point.

It's easy enough to change, though; take Epic, where everything is x1.5 in lengths, and drop the unit build cost, building build cost, worker rate, etc. multipliers back down to 100. The techs will still take longer, but units and buildings will be like on Standard. While this'd be somewhat similar to the SQL changes, you could have it also affect things other than just tech costs if you wanted.

I found that I never ever got to build most buildings, which I realize is a design decision from the game developers, to try make people make hard choices what buildings they want

I think you're giving the developers too much credit. I don't think they intended it to be this way, as evidenced by what they're doing to the tech pace in the next patch. There's no possible way we could look at the vanilla game, see that techs come every 1-2 turns in the Industrial/Modern Era, and think that it was working as intended. So don't feel like these sorts of changes violate some sort of design decision.

that the whole ICS thing is painful for me to play, even limiting growth of cities to keep happiness kind of hurts,

ICS is an abomination in its own right. Thankfully, people have come up with many different ways to kill it, using a combination of the trade route equation (Thalassicus' mod), specialist slot changes (many), and Happiness changes.

For instance, take the Happiness. In my own mod, I changed the equations quite a bit; in the core game, a city costs 2 + 1.0*pop Happiness (5 + 1.34x for conquered cities). In my mod, it's 4 + 1.2*pop (8 + 1.6x for conquered). And +happiness buildings give less than before. But to balance this, there are several new city growth buildings (Aqueduct, Sewer System, etc.) that add food storage and +1 Happiness, many of which are restricted to certain sizes of cities. So an ICS empire couldn't possibly stay above zero on Happiness, because its cities wouldn't get big enough to unlock those extra buildings.
The vanilla game has already done some of this; the whole UnmoddedHappiness change in the last patch was a big part of it, and in the next patch they're reducing the Happiness from the Colosseum and Theater.

You can do a lot of things along these lines to prevent ICS from being viable, and there are many mods out there that'll already do this for you if you don't want to do it yourself.

Final question... is it possible to limit technology to some era? for instance that noone can get past medieval period?

Easily, and there are several mods out there that already do this. But you'd really need to tie it to some other balance changes, or else the game would basically stall out just when you were getting going.

 

[cryptc]

Played around a bit now... increased techs *5 in technologies, and I halved the yield of all science buildings in building_yieldmodifiers...

Not sure if I did everything right though, since I got out teched badly in the game I tried... might just be that I got slightly bad start, and no bonus techs from ruins, but I was basicly at half my opponents points constantly, just struggling in constant warfare to keep my cities... skilled ballista's and strategies helped alot with that (and was a fun challenge for a while).

But looks like I'm waiting for the patch, and maybe tweaking something more after that.

 

[cryptc]

Thinking about it a bit more, I think what I'd really want is to have technologies increase in cost exponentially... that might even out things alot more, make for a closer game where tech while useful, doesn't mean you auto-dominate someone with less tech, and means that if you have a tech lead it doesn't mean that it just becomes a bigger and bigger lead.

I always did find it a shame that if you get a ruin upgrade for a unit, or even two, they rarely stay useful more than a handful of turns as a powerful unit... would be nice if they actually do more than help you put down a few barbarians early.

It would maybe ideally start really stalling out before industrial then... But I'll still wait for the new patch and see

 

[Pembroke]

Two ways: Either create an xml that updates the costs of all techs manually, or use SQL to make a general change. The latter is much quicker, but the former allows you to adjust tech costs more to your taste.

Thank you very much! Slowing down the research while keeping the build times as they were was just what I wanted. Your post made me download the SDK and do my very first Civ5 mod!

The SQL statement works like a charm (multiplied research costs by 5) but, just like Spatzimaus mentioned, the year counters are now off as was to be expected. Can I adjust them somewhere, too?

 

[alpaca]

I've found that science output follows a power law between 2 and 3 (t^2.5 or so, I didn't check very thoroughly) over the course of a decently successful game, with a lot of noise of course. So to keep turns researched at the same level, the same progression should be applied to techs.

If you use a function that increases a lot faster than this, like an exponential function, you will end up with extremely long research times in the late game. If you use a function that grows much more slowly, like vanilla, your research times will become shorter and shorter. They should become a bit shorter over time to give you a feeling of accomplishment, in my opinion, so if you use an exponential increase, it will feel very frustrating in gameplay and be unrealistic unless you make the time between turns decrease a lot faster.

 

[cryptc]

I've played around a bit more with the slowing of research. This time I did progressively more slowing by era, in exponential fashion, x1 for ancient, x2, x4 ... x64 ... so that it would slow down research and eventually come to a semi-stop.

I thought it worked reasonably well, but I think I've located the reason I get out-teched so badly in research slowed games... the research agreements... does anyone know how to turn them off?

It wouldn't be a major problem, except that AI's have a love of denouncing me right before a research agreement is finished and it unbalances the slowdown I'm trying to do... come to think of it, I should do something about great scientists too hmm...

 

[Spatzimaus]

with a lot of noise of course.

It's not noise. They're correlated, systematic uncertainties. You're fitting an exponential curve to a step function; sure, it might fit fairly well on average, but the deviations from that curve aren't random; they'd happen in the same place in every game you played, because they're tied to specific techs and such.

Research output is mostly a step function; you unlock Libraries, Universities, Public Schools, and Research Labs at certain techs, and see a huge increase in research output when you build them. Or you reach the Renaissance and unlock the Rationalism tree, and see a huge increase as you start unlocking those policies and then slow down again once you have them all. Or you go on a conquering spree and see a huge increase as you add new cities to your empire (although the fact that cities lose half their population, by default, mitigates this one). It's not a smooth curve, because each of the above are tied to reaching specific techs or eras, and so you can see them coming. (The only exception to this is how Golden Ages boost money output, which can be used to buy lots of research agreements. But RAs are another topic, below.)

The building thing is especially pronounced. Libraries multiply your research by x1.5. Universities multiply again by x1.5. Public Schools multiply by another x1.667. Research Labs multiply by another x1.667. (And those are all the multiplicative numbers, not the usual +50%+50%=+100% thing.) Add in an Observatory or National College, and you're looking at 7.5 research per population point in the endgame instead of the 1 that a city without any buildings would have. Since nearly all research is population-based, your research rate jumps by a large amount, N turns after reaching the techs that unlocks those buildings, where N is how long it takes to build the new buildings in your larger cities.
So if you reprice the techs to have a much steeper or exponential cost curve, then it cripples anyone who doesn't take those +research techs at the first possible opportunity, and/or gives too big of an advantage to the people who do.

This is why, in my mod, I basically halved the research output of the research buildings instead of raising the costs of the techs. Too much hinges on whether your cities have these structures, and I wanted them to be more comparable to the military buildings, where you'd build them in your larger cities when you had the time but didn't feel obligated to build them as soon as they unlocked. So in my games it peaks at 3.333 resarch per population in the modern era, although my future eras take it a bit higher.

the research agreements... does anyone know how to turn them off?

In the Eras file, you can just set their prices to something prohibitively large. Each era has a ResearchAgreementCost entry that sets the price for a civ in that era (modified by pacing, of course.) In my own mod I've basically doubled the costs, and it makes a HUGE difference in how often the AI enters into them. It's not just that they're more expensive, it's how much more expensive they are than the other things the AI can spend on. The AI, each turn makes a value decision based on the cash on hand. Does it:
A) Save the money
B) Spend the money on a research agreement
C) Spend the money bribing a city-state
or D) Spend the money rushing a unit or building?

The problem is that in the later eras, RAs cost less than rushing a unit or building, and less than a good bribe. So there'll be plenty of turns where, depending on the AI's cash-on-hand, D isn't an option at all and C is questionable, where it's basically A-vs-B; while it might decide to save on one turn, it'll re-decide the next turn, and if it's going to be 10 more turns before you have enough cash to rush a unit, there are plenty of chances to pull the trigger. So the AI spends the cash on an RA before he has the chance to save up enough to use it for a unit or building.
But when you double the cost of an RA, the situation reverses; the AI will rarely make a research agreement, because he'll almost never have enough cash to do so, having used it several turns earlier to rush something in one of his cities. (Or even if he does have enough money, the chances that anyone else will have enough money any time soon are slim, so he'll spend his before someone else gets up there.) It makes the AI even more dangerous on the mundane level, with more units, better cities, and more city-state alliances, but less of a problem in the tech race.

It's definitely made a huge difference in the play of my own mod on higher difficulties (where the AI gets much more cash than you do).

 

[cryptc]

Ok, I'm resurrecting this old thread since recently starting to play again I'm having a problem.

I've slowed down research massively, and it's lead to a very enjoyable slow paced wargame exactly how I wanted. However, apparantly the research wraps around and becomes negative numbers, since I now suddenly get alot of -60 000 research points 1 turn research. Checking the technologies table it all looks fine, but is there some sort of maximum value there before it wraps around? here is the sql I use in my mod:

UPDATE Technologies
SET Cost = 2*Cost
WHERE Era = "ERA_ANCIENT";

UPDATE Technologies
SET Cost = 8*Cost
WHERE Era = "ERA_CLASSICAL";

UPDATE Technologies
SET Cost = 32*Cost
WHERE Era = "ERA_MEDIEVAL";

UPDATE Technologies
SET Cost = 128*Cost
WHERE Era = "ERA_RENAISSANCE";

UPDATE Technologies
SET Cost = 512*Cost
WHERE Era = "ERA_INDUSTRIAL";

UPDATE Technologies
SET Cost = 2048*Cost
WHERE Era = "ERA_MODERN";

UPDATE Technologies
SET Cost = 8192*Cost
WHERE Era = "ERA_FUTURE";

update Eras set ResearchAgreementCost = 10*ResearchAgreementCost;

 

[Thorn]

Check out the Civ5Eras.XML file if you haven't already. It controls the research time for each era. The line <ResearchPercent>n</ResearchPercent> is the one you want. If you set n to a value of 1 you will research techs in 1 to 2 turns. Set it something like 999 and it will take a very, very long time indeed! Experiment with the number to get what you are looking for. I use this if I want to spend a lot of time in a specific era. The great thing about it is you can change the values during a game. It doesn't affect saves.

 

[cryptc]

Ah thank you will give it a try

 

[cryptc]

Having tested it a bit now, it seems this has the exact same problem as when changing in technologies, around when it hits 200 000 beakers it wraps around to cost -200 000. So I'll have to be careful to avoid those limits