Active Research Discussion

[stealth_nsk]

We have quite clear vision of the current state of the system if we look through several articles. The tech tree is more or less the same, but each technology has a prerequisite "quest", completing which the research of the tech has double speed, i.e. building quarry for speeding up researching "Masonry".

This looks quite robust to me and will definitely shape the strategy in each game differently.

What worries me is - the system is likely to deepen borders between military and peaceful paths, which could be not so good. If military techs are tied to military achievements, and vice versa, we could see actually reduced number of strategic possibilities. Hope developers will be able to balance it nicely.

 

[MIS]

For example, you could need to do 2-3 out of the 4-5 foundational techs for Medieval (chosen from Philosophy, Currency, Engineering, and Optics, say) to move on to the techs in the next era.

Civ III did this.

 

[noto2]

The system could be good for replayability. One game you start inland with horses and get faster research towards horse archers, for example, and the next game you start on the coast and are the first to build an ocean-faring vessel.

But they need to be very careful with the balance of such a system, it could easily spin out of control.

 

[Kouvb593kdnuewnd]

There is alot of possibilities and the reviwers have said all techs have an active research part, it could nerf pure science focused empires who could do very well in civilization 5 and help large, expansionist and military empires who could not focus purely on science.

 

[TheMeInTeam]

It's a much needed tuning to the tech system, but I'll be surprised if the AI can manage it decently.

 

[Ryuu Falconwing]

It's an interesting system. Functionally, you have a penalty to researching each tech until you do something that provides that beautiful boost. But there's still nothing preventing you from researching sailing from the middle of the Gobi. It'll just take longer.

I have faith the AI will handle this okayish. Hopefully it means they don't just forward settle randomly and actually choose strategic locations now.

 

[stealth_nsk]

It's a much needed tuning to the tech system, but I'll be surprised if the AI can manage it decently.

It's not a map tactics or something like this, just a system of priorities. Something AI could easily manage.

 

[SS-18 ICBM]

All that's been described is a gimmick, and one which mostly specifies your approach to expanding the way you would anyway. Try to imagine how Active Research would apply to Medeival Era technology, or Atomic Era. I can't do it.

What is the condition for accelerating Mathematics research? "Build 1 Catapult, 2 Archer, 3 Footmen AH AH AH" ?

Yeah, I do hope it's more than just bonuses for resources and tiles. I have to wonder how Active Research is implemented by the decisions your civilization makes.

 

[Kouvb593kdnuewnd]

Fighting barbarins could be seen as a decision and that will help your military to develop.

 

[dexters]

It's a much needed tuning to the tech system, but I'll be surprised if the AI can manage it decently.

I actually think this will benefit the AI. They can be taught to go after certain techs if they see certain tiles and starting locations.

Removing the single killer teching strategy will hurt the human players more. On balance.

 

[Kouvb593kdnuewnd]

In the end it comes down to the tech tree design. If it have obvious killer techs which the player will quickly go after then active research will not change the game that much.

 

[dexters]

In the end it comes down to the tech tree design. If it have obvious killer techs which the player will quickly go after then active research will not change the game that much.

Right, but with techs tied to terrain/resources it will no longer be feasible to beeline for them.

All the devs Need to do is program the AI to pick up their cheap techs by virtue of their starting location and tiles/resources available to them plus some other grand strategic considerations. I think it's going to create a much more dynamic AI because the AI sort of already foes it in civ5 , it's just that human players disregard ai considerations and always play a set teching path. Which means AI don't adapt well to human players.

I see this research mechanic as forcing human players to think like the game AI by forcing us to weigh different costs of techs. Hence my view it will even it up. AI will be in its element. Human players will have to play the cards they are dealt.

 

[Ryuu Falconwing]

I wonder how they'll use this to fix the longswordsman-musketmen-riflemen period and keep thise units meaningful

 

[TheMeInTeam]

It's not a map tactics or something like this, just a system of priorities. Something AI could easily manage.

You are assuming long-range planning from the AI, on the order of "there is this terrain so I will put a city there and then research that to confer an advantage others are unlikely to have".

As players, we will know about the possible advantages and timing to use them in advance and will made dedicated efforts to optimize for them.

Getting an AI to think in those terms is not trivial, based on what I've heard from people trying to get it to do so in the past. If the AI could anticipate benefits in the distant future, it wouldn't walk its archers in front of melee units or embark units that can be attacked. I find it hard to believe an AI that embarks units near archers would suddenly be able to "easily manage" deliberately settling for the most beneficial terrain for active research given variant circumstances.

It can probably run math to grow a city at max rate, but not so much making consistently sound decisions on planning with incomplete information I'd suspect.

 

[dexters]

You are assuming long-range planning from the AI, on the order of "there is this terrain so I will put a city there and then research that to confer an advantage others are unlikely to have".

As players, we will know about the possible advantages and timing to use them in advance and will made dedicated efforts to optimize for them.

Getting an AI to think in those terms is not trivial, based on what I've heard from people trying to get it to do so in the past. If the AI could anticipate benefits in the distant future, it wouldn't walk its archers in front of melee units or embark units that can be attacked. I find it hard to believe an AI that embarks units near archers would suddenly be able to "easily manage" deliberately settling for the most beneficial terrain for active research given variant circumstances.

It can probably run math to grow a city at max rate, but not so much making consistently sound decisions on planning with incomplete information I'd suspect.

I think you seriously under estimate the AI here. The AI steals my resources all the time in civ5 and we fight over city states because the AI wants the city states. The Civ5 AI always had a plan . It is what sets it apart from older civ games.

They can read the map and are known to want open borders to see into the fog to make precisely the kind of decisions I describe. They can certainly be programmed to see a tile or resource in their line of sight and plan accordingly.

 

[UWHabs]

I doubt we'll get the AI thinking of "oh, I don't have any horses in my capital, but if I put my second city here, and my third city there, that will have horses. And then I'll be able to put a pasture on it, and that will give me a science boost for horseback riding, so let's delay teching it until then."

AI tends to happen in small pieces. They can have overall strategies, but they work best with very small bits to work with. But the AI absolutely can look through the techs and be like, "okay, I don't have horses now, so Horseback Riding costs 200. But I have an iron mine, so Iron Working currently only costs me 100, so let's research that for our military needs."

How challenging it might be for them to do the "So I want to research Iron Working and I have an iron square in my territory. Let's make sure I have time to build a mine on the tile before researching it to get the bonus" is the real big AI question. Or even if their planning is a little more long-term like, "I want to research Iron Working right now, and I don't have Iron Working in my territory. Oh well, let's start" not realizing that they have a settler that's one turn away from founding a city with Iron, or, in the flip side, delaying it too long because they don't realize it's going to take a long time to hook up that Iron that's in ring 3 of their city.

 

[dexters]

I'm pretty sure the AI can see resources on a map not in their borders. They already agressively settle and steal tiles in civ5. Using the same Line of sight rules human players have. AFAIK.

 

[HorseshoeHermit]

We have quite clear vision of the current state of the system if we look through several articles. The tech tree is more or less the same, but each technology has a prerequisite "quest", completing which the research of the tech has double speed, i.e. building quarry for speeding up researching "Masonry".

This looks quite robust to me and will definitely shape the strategy in each game differently.
.

Oh god please no. Multipliers? The strategy will be solved statically and you'll have players picking a research path based on which of the eureka perks will come along in the most probable order to maximize every beaker.

It has to be a lump sum.

 

[Kouvb593kdnuewnd]

Both lump sum and multipliers give free beakers.

 

[MIS]

They've already said the quest gives half the tech cost.