The research model in RotS: A step in the wrong direction (Full Version)

All Forums >> [New Releases from Matrix Games] >> Distant Worlds 1 Series



Message

Apheirox -> The research model in RotS: A step in the wrong direction (2/14/2011 6:40:36 PM)

I was unpleasantly surprised to find how the way research is done had been changed for Return of the Shakturi. Not only have I found it greatly hampers the AI (read about this here: http://www.matrixgames.com/forums/fb.asp?m=2726745 ), I have also concluded that the system in the original DW was superior - even if it wasn't particularly good, either.

The difference is all about the concept of map control. As a 4x strategy game, DW is a game about map control. It is about holding on to as many planets, facilities and resources as possible. This is the model which all 4x games are based on and it is a healthy model.

Unfortunately, the RotS research system does not comply with this rule. Here, there is no need to expand and hold on to many research stations in vulnerable locations. The best way to conduct research in RotS is to counter-intuitively not (sic!) have your research labs in your research stations, but rather in the safety of your space ports. You only need to put a single lab in a research bonus location and you gain empire-wide benefit. So, 1 lab module next to that black hole and the 50 others tucked away inside your large space ports. This is a really, really poor system.

In DW it was better because you had to put a much higher percentage of your labs out on the vulnerable open locations of the research bonus locations. No souped up space ports to defend them there! Still, DW's model was also poor because it was possible to put an unlimited number of labs around the same research location.

Ideally, what should be done with the game is to go back to DW's model BUT change it so that only a limited number of research labs gain the bonus from the special location - any more built than that gain no bonus. This would force empires to spread out to control multiple research locations, having research labs in many different locations - neutron stars, black holes, gas giants, all over the map. It would be much more difficult to defend a such large number of facilities - and that is how it should be, the 4x basic model.

Thanks for reading.



Data -> RE: The research model in RotS: A step in the wrong direction (2/14/2011 7:23:11 PM)

good observations, I really like these recomendations



aprezto -> RE: The research model in RotS: A step in the wrong direction (2/14/2011 7:31:07 PM)

Agreed, I didn't read the fine print and didn't see this change.



gmot -> RE: The research model in RotS: A step in the wrong direction (2/14/2011 9:03:04 PM)

Yes, I'd prefer to see only the research station at the special location get the bonus. Don't like the empire-wide bonus.

The way I see it, the bonus is due to some special characteristics of the location - e.g. proximity to the black hole enables some crazy research that isn't possible in a "normal" location. So it doesn't make sense that the bonus applies to all stations in the empire.



Shark7 -> RE: The research model in RotS: A step in the wrong direction (2/14/2011 9:12:12 PM)

I like the ROTS research model. Having only 3 major branches and having to pick and choose is a good thing. The only real problem with it are the bonuses...they do tend to be over the top. The bonuses should apply only to the station built at the site, and there should be far fewer of them.

So basically, I am in the 'keep the basic system, but tone down the bonus sites' camp.



Igard -> RE: The research model in RotS: A step in the wrong direction (2/14/2011 9:12:32 PM)

Agreed. I miss the days of building my huge research operation at some distant black hole. I used to cluster them together and build a couple of defence bases. Now, I just check the research location in the nav panel for any new finds every now and then. One facility and that's it. I don't even visit it to build a defence base, it's not worth it.



Wreck -> RE: The research model in RotS: A step in the wrong direction (2/14/2011 9:53:35 PM)

Good points.  Once I determined that a single research station at a black hole could perform all the research my little civ could possibly handle... I lost respect for the model, just a bit.

Here's some further ideas to develop the idea of forcing research into many locations.

First, break the link between research locations and (all) research bonuses.  I can see how being near a neutron star might help with say hyperdrive research.  But how does it help develop Desert Colonization?  Hospitals?  It should not.  Instead, on a per-tech basis, define a set of helpful research locations.   If you want to colonize Desert planets, build a station on a desert planet.  Or two.  Or N!  Make propulsion and torpedoes benefit from massive stellar objects; high energy stuff needs to be close to a star, colonization depend on the particular type of planet, gas extractors benefit from gas giants, etc.  There may be many techs -- i.e. computers -- with no location dependence at all.

2nd, make all research stations give a diminishing rate of return.  Here's how one might compute that: for each station, compute its base rate of research along with any multipliers.  Now sort them, best to least.  Then degrade the amount of research you get for the Nth best station, via exponential decay, using perhaps 0.9 as the base.  So, the best station gets its full research amount, 2nd gets 0.9x as much, 3rd gets 0.9^2, etc.  Huge empires thus can research faster than small, but it does not scale linearly with size, and a small empire with good multipliers might still beat a large one without.





Data -> RE: The research model in RotS: A step in the wrong direction (2/14/2011 10:04:18 PM)

quote:

2nd, make all research stations give a diminishing rate of return. Here's how one might compute that: for each station, compute its base rate of research along with any multipliers. Now sort them, best to least. Then degrade the amount of research you get for the Nth best station, via exponential decay, using perhaps 0.9 as the base. So, the best station gets its full research amount, 2nd gets 0.9x as much, 3rd gets 0.9^2, etc. Huge empires thus can research faster than small, but it does not scale linearly with size, and a small empire with good multipliers might still beat a large one without.


All hail this idea....not sure how realistic it is but I am sure it will do wonders for the gameplay.
Great one, Wreck, you should make a double post on the wishlist with this.



hewwo -> RE: The research model in RotS: A step in the wrong direction (2/15/2011 12:58:52 AM)

I felt there was something wrong with the research system, but couldn't figure it out exactly. Didn't try either tbh, but now I wont have to because this thread is exactly right! Good analysis.



thiosk -> RE: The research model in RotS: A step in the wrong direction (2/15/2011 1:21:58 AM)

I agree with general problems of the research system. I'm sad to hear the new expansion system is inferior for these reasons.

Many moons ago during the early iterations of the vanilla game I made a long winded post, and not particularly well-received post, about what I did and didn't like about the research system. I do not like the anomaly-centric system, and was vocal about it in this thread.

http://www.matrixgames.com/forums/tm.asp?m=2436663&mpage=1&key=�

Turning the anomalies into something to fight viciously over would be good.



Kayoz -> RE: The research model in RotS: A step in the wrong direction (2/15/2011 1:44:15 AM)


quote:

ORIGINAL: Apheirox

Ideally, what should be done with the game is to go back to DW's model BUT change it so that only a limited number of research labs gain the bonus from the special location - any more built than that gain no bonus. This would force empires to spread out to control multiple research locations, having research labs in many different locations - neutron stars, black holes, gas giants, all over the map. It would be much more difficult to defend a such large number of facilities - and that is how it should be, the 4x basic model.


I disagree with this one. Placing an artificial limit on the number of research labs at a location doesn't make sense. If it's 5 labs, then why not 6? 7? Why one number instead of another? It seems to be an arbitrary value which is placed with no other reason than to force players to spread out.

Diminishing returns on the number of labs at a certain place would make more sense. So if you're crammed into a corner with only a few research locations - you can keep building stations there - but the benefit will be less than you'd otherwise get from spreading them out. It'd make more sense. But an artificial limit? No - that doesn't jive with me.



Apheirox -> RE: The research model in RotS: A step in the wrong direction (2/15/2011 2:06:13 AM)

The point is to force players to spread out to control multiple research locations. How this is accomplished is secondary.



Kayoz -> RE: The research model in RotS: A step in the wrong direction (2/15/2011 2:17:28 AM)

That's not what he said - scroll up and read his "ideal situation".

As for goals - it's the goal of the treatment of a disease to kill off the bacteria/virus infecting the patient. But sticking someone in an autoclave - while accomplishing that goal - may be greeted less than enthusiastically by the patient.



unclean -> RE: The research model in RotS: A step in the wrong direction (2/15/2011 2:19:26 AM)

If that's the case, wouldn't it be easiest to lower the default research potential of empires and increase the anomaly bonuses?

I agree though, the AI's bad research is definitely annoying. On top of putting them at a tech disadvantage it also makes it stupidly easy to center your entire economy off of selling technology.



Shark7 -> RE: The research model in RotS: A step in the wrong direction (2/15/2011 2:21:57 AM)


quote:

ORIGINAL: unclean

If that's the case, wouldn't it be easiest to lower the default research potential of empires and increase the anomaly bonuses?

I agree though, the AI's bad research is definitely annoying. On top of putting them at a tech disadvantage it also makes it stupidly easy to center your entire economy off of selling technology.



I always start the AI empires, 2-3 tech levels ahead of me. Solves all the problems and makes the game...interesting. [;)]



Data -> RE: The research model in RotS: A step in the wrong direction (2/15/2011 7:03:25 AM)

That is one interesting setup, Shark. Why I am still surprised? [:)]



Wreck -> RE: The research model in RotS: A step in the wrong direction (2/15/2011 2:31:52 PM)

quote:

ORIGINAL: Apheirox
The point is to force players to spread out to control multiple research locations. How this is accomplished is secondary.

But it cannot be only about gameplay. I think that is an important point -- the game is after all a game first and a sim second. But really to me the "realism" is what matters most. Players should be able to intuit the mechanics. I'll tell you what my intuition is: more research is better, and there should be no artificial limits to research. So for example in my first few games I built research labs into every small starbase -- 100s of them. What a waste it must have been!

But Kayoz is right about diminishing returns. Another part of my intuition (and also a part of the real world realism) is that there are diminishing returns in research. Maybe there are experiments you can do near a black hole that would help hyperdrive research, but there are not infinitely many such experiments at any one time, and in fact probably the number of really useful ones that could run in parallel is pretty low. The diminishing returns, once you have any lab at all near the hole, should be rather sharp IMO. Just to throw some numbers out, I'd propound a function as follows: 
rate(N) = 1 + (bonus - 1)*0.7^N
where N is the number of labs you have at that particular bonus-location, and bonus is its research multiplier. The 0.7 in the base of the exponent gives a sharp decay in research potential for later bases, but 1 is the minimum. (Do this computation before the one with the 0.9 base I propose above.)



Zenra -> RE: The research model in RotS: A step in the wrong direction (2/15/2011 2:46:59 PM)


quote:

ORIGINAL: Wreck
<snip>

I can see how being near a neutron star might help with say hyperdrive research.  But how does it help develop Desert Colonization?  Hospitals?  It should not.  Instead, on a per-tech basis, define a set of helpful research locations.   If you want to colonize Desert planets, build a station on a desert planet.  Or two.  Or N!  Make propulsion and torpedoes benefit from massive stellar objects; high energy stuff needs to be close to a star, colonization depend on the particular type of planet, gas extractors benefit from gas giants, etc.  There may be many techs -- i.e. computers -- with no location dependence at all.

2nd, make all research stations give a diminishing rate of return.  Here's how one might compute that: for each station, compute its base rate of research along with any multipliers.  Now sort them, best to least.  Then degrade the amount of research you get for the Nth best station, via exponential decay, using perhaps 0.9 as the base.  So, the best station gets its full research amount, 2nd gets 0.9x as much, 3rd gets 0.9^2, etc.  Huge empires thus can research faster than small, but it does not scale linearly with size, and a small empire with good multipliers might still beat a large one without.


+1 on this. I was already thinking a diminishing return formula was in order, but I really like Wreck's addition of setting appropriate research locations to realize a bonus.

I definitely agree with the OP - forcing research locations out into other systems would make DW an even better game due to the premium it would place on exploring, expanding into and exploiting the value of those locations with respect to the research bonuses provided.



Apheirox -> RE: The research model in RotS: A step in the wrong direction (2/15/2011 3:40:44 PM)

quote:

ORIGINAL: Mitchell
I definitely agree with the OP - forcing research locations out into other systems would make DW an even better game due to the premium it would place on exploring, expanding into and exploiting the value of those locations with respect to the research bonuses provided.


Absolutely. This is a 4x game - that is Xplore, Xpand, Xploit, Xterminate. The current research model does not reflect this at all - but with a few simple changes, it will, and the game will be much better for it.



J HG T -> RE: The research model in RotS: A step in the wrong direction (2/16/2011 5:07:34 PM)

Good points here. Agreed with the suggestion of "outing" the research to other systems/anomalies.
My main things currently are fixing the bonus system (bonuses only for the station, not the entire empire) and fixing the AIs research habits. Make AI build more research stations on their planets when possible.



Larsenex -> RE: The research model in RotS: A step in the wrong direction (2/16/2011 5:55:49 PM)

If you remove the empire wide research and put it only on a 'per station' value, how is over all research then calculated? Would it not then make the speed different on each of the main trees vary greatly? As an example if I had 3 stations at black holes and each only was for say weapons (8%, 15% & 33%), I should in theory research the weapons branch of tech (faster) then energy and high tech. I for this change. It would add great flavor to an otherwise already great game. I also like some of the 'directional' weapons that are mentioned in another thread.



hewwo -> RE: The research model in RotS: A step in the wrong direction (2/17/2011 11:10:09 AM)

Also, it would be cool to surgically strike and destroy certain research stations of a rival empire. Actually, if you make it a bit difficult to figure out where these stations are it gives the spying part of the game a big boost. Would be totally awesome! (and terrible for the ai)



Carewolf -> RE: The research model in RotS: A step in the wrong direction (2/17/2011 1:42:46 PM)

Put diminishing returns on everything. Well at least research and mining.



Bingeling -> RE: The research model in RotS: A step in the wrong direction (2/17/2011 1:43:29 PM)

One easy fix could indeed to lower the starting potential.

As for having huge bonus for research stations in special locations... Research is done in labs and offices. Preferably at nice location with good night life and food (even for a lot of researchers). Data is collected in special places. Of course, some research benefit more from location, like colonization tech. Notice that in the real world, researchers go into the field to collect data, then they sit at "home" for two years to examine the data collected. So, require one research module to be placed on a station at a tech bonus location as a data collector, but like today it should not matter where the rest are. Stuffing all your researchers at a black hole makes no sense.

Also, notice that this is a game with instant message flow across the galaxy, so don't argue about distances and delay for result sharing :-)



Page: [1]

Valid CSS!




Forum Software © ASPPlayground.NET Advanced Edition 2.4.5 ANSI
0.65625