Tech Prototype Issues

[LDiCesare]

We have managed to hook up the econ/tech/military together.
However, tech output to military leads to terrific results, like raising tech level from 10 to 70 in a few 20-30 turns. Is that what is expected? Considering the scenario is Punic wars, I feel that getting from 10 to 70 is a bit too much. I believe 100 is current-day level. 70 would be a bit high for antiquity.
Mark, do you need the level you get from the food activity? If so, do we have to make research of military tactics more expensive in terms of RPs? Or introduce upkeep costs? Or can we generate less RPs from the economy?
We are going to need a lots of figure-splitting in the next days, I believe.

[Mark_Everson]

Originally posted by LDiCesare
Mark, do you need the level you get from the food activity?

Sorry, don't understand exactly what you mean. Let me guess...
I now have it set up so it takes about 20 turns to increase 1 level in farming tech when I use just the labor-generated RPs. This seems a vaguely sensible rate to me. It at least will take a long time to get to a value that's too large for classical antiquity. I define "too large" as giving a PCI of more than 2x subsistence. I may redo how the food works as more info comes in, but as a rule of thumb:

I suggest that in a 100-turn game no basic tech should more than double in effectiveness. Otherwise it will just get silly IMO. Of course new applications or techs can come in that more than double in effectiveness over a short timespan. But the overall progress of the civ shouldn't grow by leaps and bounds. In terms of tech levels then, that means that most techs should grow by 10 levels or less over 100 turns.

If so, do we have to make research of military tactics more expensive in terms of RPs?

That is the way that seems most sensible to me, and also what I've done on the econ end.

Or introduce upkeep costs? Or can we generate less RPs from the economy?

I don't think the upkeep cost concept really works to balance the system as I've described to Richard using Dawn (5k population) vs the Carthage scenario (> 1M population) as an example. Even if the two scenarios start with different tech levels, (say 0 for Dawn and 10 for Carthage) any amount of upkeep that keeps the Carthage tech levels from increasing rapidly, causes Dawn tech to go negative and stay there for a long time. (Potentially this can be offset when the Dawn civ grows to be quite large in population over maybe 100 turns.) But Richard was going to take a shot at balancing the system his way, and hopefully he'll show me to be wrong

We are going to need a lots of figure-splitting in the next days, I believe.

Yes indeed... Here are some specifics on what I've done and some other thoughts:

I cut down the tech increase rate (actually RP generation) so it is merely rapid in delenda. This makes it pretty much non-existent in Dawn.

I don't like how techs can increase without limit even if most other technologies for the civ remain at stone-age level. I have already posted on that "feature" of the current system in the D6 Tech thread.

Here are some details on how many RPs per labor, and per CC when buying RPs. They are described below:

number of RPs a unit of labor (1000 people) generates per turn
RPPerLabor = 0.0001

number of RPs an investment of one CC generates
RPPerCC = 0.001 (This is the one used when you buy RPs in the econ gui)

In the Dawn scenario I arbitrarily increase the RPPerCC to 0.05 so that there is a way to improve tech over the scale of the game. Richard will presumably also make the changes in the global variables in technology.xml to change the rate of tech progression. I haven't experimented with that yet.

Cya,

Mark

[LDiCesare]

I have said things which were not quite correct. The change of tech level we experienced was due to different levels of code, not a problem in the model. Even by putting lots of points in Combat and crushing Carthage, now I have updated the code, I can barely rise from 10 to 20 in military tactics, which is much better.
Agreed, these values may still be too high for the full game, but for the demo, I think it is important that the tech effect shows over a small period of time (with the ability to build supposedly stronger armies at some point).

[Richard Bruns]

I'm glad that military tech growth seems to be set correctly. Good job.

I believe that all of the problems that Mark mentioned are due to the fact that Rome starts with a knowledge level of 1. This doesn't make sense; Rome should start with much higher tech levels than a stone age tribe. The rapid rise in tech level is nothing more insidioius than Roman tech rising to where it should have been if the player had been developing the empire over a full game.

So what I suggest is to run the scenario, allowing Roman tech to grow until it is stabilized. Then save the game at that point and call it the scenario. People will open the savegame, and Rome will have the proper tech level and a proper amount of tech growth. That is of course a quick fix; what we need to do in teeh long run is allow the civ's starting knowledge level to be set in a text file as I mentioned above.

[Mark_Everson]

Originally posted by Richard Bruns
I believe that all of the problems that Mark mentioned are due to the fact that Rome starts with a knowledge level of 1. This doesn't make sense; Rome should start with much higher tech levels than a stone age tribe. The rapid rise in tech level is nothing more insidioius than Roman tech rising to where it should have been if the player had been developing the empire over a full game.

Hey Richard:

Unfortunately, that's not the issue, at least not for farming tech. Once the farming tech level reaches 10, it just keeps growing without bound. If your argument were correct, once farming tech reached some intermediate level its growth would slow appreciably. I have yet to do much testing with things other than farming. The farming technology feedback loop is such that a higher farming tech level increases farm productivity which increases population which puts more RPs into farming tech. With the exact specifications I have put in now (and using a traditional economy) the rate of change of farming tech level actually accelerates as the game goes on. Basic farming productivity is currently set up as 5+ farming tech level. So you see as the farming tech level grows there is more food, which in a traditional economy leads to more people, who mostly go into farming. At tech level 10 productivity is three times what it was at tech level zero, giving crudely three times the population, and three times the RP's. IIRC, as its currently set up, you only need twice the RP's to gain a level when your tech is at 10 than when it is at one. I guess if I were to divide the farming tech level in the productivity equation by a factor larger than 2 diminishing returns would eventually set in. Does that sound right to you?

It's easy to see mathematically why the results are what they are. The question is what we do to protect the model against such things happening. I can change the basic farm productivity equation so that the rate of increase of productivity per farming tech level is much lower. But since I think this problem could be much more general than just farming, I was reluctant to put a Band-Aid on it too early. The one idea I had is the thing about each technology having a helper that is all technologies combined, which would tend to slow growth in a technology that is starting to bound ahead. I'm glad that idea didn't sound too bad to you, and I agree with your post in the other thread that we should see what we can do without that approach first.

With the current RP generation settings in the testbed, it will take you a long time to be able to see the effect I am describing above. Perhaps if you edit the technologies file to allow farming tech to grow more easily, you will be able to see what I'm talking about.

[Richard Bruns]

The main cause of the problem seems to be the way the population model works. Apparently people in Clash can work the instant they are created. If we used a more accurate population model, I don't think we'd have such a problem. Population growth would be limited, and the new population wouldn't be able to do anything productive for at least ten years. This would prevent the population explosion that leads to the greater technology gain.

But that kind of modeling comes later, if at all. Perhaps we can base food RP's on the acreage of land under cultivation. That way the tech would be constant, as long as the empire is the same size. Its a kludge, but it should work for now. We can correct the population model problem later.

[Mark_Everson]

Hey Richard:

Acreage? All I can say is yuck! I'm inclinded to just divide the farming tech by 4 or something for now, and see if that doesn't control the increasing-returns aspects enough. When farm machinery and other modern stuff comes along, the rate of effectiveness growth could increase dramatically. But we're not gonna be there for a bunch of demos, so I won't worry about that for now. And hopefully by then we'll have other fixes for "tech runaway".