View Issue Details

IDProjectCategoryLast Update
0004918AI War 1 / ClassicSuggestion - Campaign Management And SetupOct 3, 2012 10:18 am
Reporterzoutzakje Assigned To 
Status strongly consideringResolutionopen 
Summary0004918: Random amount of knowledge on planets
DescriptionHowdy folks. I'm not exactly sure under what category this belongs, so I just selected Campaign Management And Setup.
I have a lobby option idea for knowledge. I'm a big fan of randomness (or the feeling of randomness) and so I thought it'd be great to have the amount of knowledge one can gather from a single planet to be 'random'. Let's say any given number from 1000 to 5000, never exceeding the normal knowledge limit in a galaxy. A 100 planet galaxy would normally have 300k knowledge available. This should not change with random knowledge enabled. The default would be so that every player can see how much knowlegde he/she can gather from a planet once that planet has been scouted. Of course there should also be an option to turn that off, so you'll never know how much knowledge you can gather from one planet (for die-hard players).
This should add a whole new strategic element to the game and could make it significantly more difficult (or easier).

I like this idea a lot, so I figured I'd post it right here on Mantis as well
TagsNo tags attached.
Internal WeightFeature Suggestion

Activities

tigersfan

Oct 19, 2011 10:13 am

reporter   ~0016801

I definitely like this idea, so long as it's optional, of course. :-)

TechSY730

Oct 19, 2011 11:48 am

reporter   ~0016810

Last edited: Oct 19, 2011 11:48 am

Hmm, if we are going to make a variable amount of knowledge, could you make it that the AI Homes have a greater amount of knowledge on average (like, almost always near the 5k level)? That would seem to fit thematically. But then again, the Spire archives or whatever they are called seem to serve that purpose.

In any case, I would not mind this as long as the suggestion as a whole and the max-knowledge available hidden were optional.

One question though, when the max-knowledge on a planet is hidden, would the amount of knowledge available be revealed on capturing the planet, or only once the science labs report the planet is exhausted?

zoutzakje

Oct 19, 2011 12:40 pm

reporter   ~0016813

Last edited: Oct 19, 2011 12:42 pm

it should all be optional. If people want to play with knowledge the way it is right now, they can do so.

as to asnwer your question Techsy, it should be revealed at intel once planet is captured. This way players can already make plans based on the amount of knowledge, instead of having to wait for the science labs to finish which is just time-consuming and annoying.

but that's up for Chris to decide of course, if he likes my idea. Perhaps he would prefer to let it stay hidden until science labs are done.

Hearteater

Oct 19, 2011 1:15 pm

reporter   ~0016814

Last edited: Oct 19, 2011 1:20 pm

It would probably be easiest to go with something like:
10% of systems @ 5k
20% of systems @ 4k
40% of systems @ 3k
20% of systems @ 2k
10% of systems @ 1k
Much finer degrees of knowledge per system aren't really all that significant. Player systems should always be taken from the 3k system pool.

Although it may make sense for the AI home systems to always be 5k, that would reduce the chance of finding a 5k system earlier in the game. If you normally clear both AI home systems very close together you will actually end up with less knowledge on average as a result. I'd probably say making one AI world always 5k, and the other always 1k would be best. When those numbers are visible, this suggests to the player a good order to take the AIs so as to take advantage of some bonus knowledge while at the same time it doesn't reduce the overall knowledge available elsewhere.

Hearteater

Oct 19, 2011 1:25 pm

reporter   ~0016815

Also, copying and expanding on my forum post on this topic:

"Hiding knowledge values would actually prevent strategic planning since you wouldn't know how valuable planets are (knowledge-wise) until after you've claimed it. Having it be visible on all systems would add another factor for players to consider when determining which systems to take and which to skip."

It may even be nice to make is available without requiring scouting. This would give you a chance at game start to pick out some high value 5k systems to start scouting towards. This would be distinctly different from other valuable resources which are found only through scouting.

TechSY730

Oct 19, 2011 1:44 pm

reporter   ~0016816

Is it safe to assume that when playing with full-visbility, aka, no fog of war (though why one would want to outside of bug testing, I don't know), the knowledge available will be visible no matter what?

zoutzakje

Oct 19, 2011 3:01 pm

reporter   ~0016821

Last edited: Oct 19, 2011 3:04 pm

yes to Techsy. And to Hearteater... I agree it would probably be easier to have 10% of systems at 5k, 20% of systems at 4k, etc... But I would enjoy it much more if the numbers were really anything between 1k and 5k. I'd like seeing planets with 1472 or 3892 knowledge for example. Like randomly generated numbers between 1k and 5k (and of course always the exact same amount of knowledge in the galaxy as it normally would be). This would make a 50% 5k and 50% 1k possible and all sorts of different combinations. But since I know absolutely nothing about coding, I don't know how difficult this will be.
I agree with you on one homeworld always 1k and the other always 5k. That should make it even more interesting to decide which one to take first.
As for the knowledge always being visible without scouting... That doesn't really matter for me since I never play with 'show unexplored wormholes' on :P
I have no choice but to scout in order to see anything. But I still believe it shouldn't be visible until you scouted it. after all, you can't see enemy ships and structures either until you have scouted the planet (unless you play with full visibility of course :P)

EDIT: player homeworld(s) should start with 3k always, to avoid getting screwed over from the start (or extremely lucky)

TechSY730

Oct 19, 2011 3:12 pm

reporter   ~0016822

@zoutzkje
The only issue with truly randomized values is that part of the balance of knowledge costs are that all costs are a multiple of 250. Truly randomized values could play havok with that, like constantly finding yourself some odd, small value short of an unlock you want but not enough surplus to afford unlocking something different and still get what you want by taking another planet.

And once again, how about the idea to make AI homeworlds always have like 4k or 5k or something knowledge when this mode is on? That certainly would tempt me into trying to colonize the AI homeworlds, despite how risky that exo-wormhole makes it.

Ranakastrasz

Oct 19, 2011 3:42 pm

reporter   ~0016827

I think evenly distributing values of knowledge between all of the plansts, but with values all being multiples of 250, and the total adding up to 3k*planet count, would work fine. I would say 2k-4k however as the limits, unless you do not fear the RNG's wrath. MUAHAHAHAHA

Hearteater

Oct 19, 2011 3:45 pm

reporter   ~0016828

@zoutzakje
There isn't really any reason to have fully random K values because getting 1471 or 1378 knowledge from two different planets is about as interesting as which background you see. It also creates "mental noise" and makes it harder for people to judge the value of a system in much the same way stores price things at 4.99 instead of 5.00. If two systems are worth 1999 and 2011, they will feel very different to people even though they aren't. So it is just randomness for the sake of being random, and I think the noise aspect would actually deter more people from using the option than it would bring to it. And I think the option has enough value that I'd hate to see implemented in such a "muddy" fashion.

@TechSy730
You can only ever colonize one AI home world really, so what really matters is the difference between the two. Is it great enough that it will influence someones decision on which to capture? 1K vs 5K is the maximum swing and so creates the most weight for a player to factor. And remember, if both are 5K, you've actually reduced the possible Knowledge in the galaxy because only one can be captured and the other high-K system is effectively un-capturable.

A work around would be to pull the second AI home system from a 3k world and add a bonus +2k to it so although the galaxy total would be higher by 2k, no one could actually make use of the bonus.

zoutzakje

Oct 19, 2011 5:15 pm

reporter   ~0016843

ah yes, of course. Hadn't thought of that. Then we should have values of 1000, 1250, 1500, etc

And I have to agree with Hearteater about the enemy homeworlds. I believe having one homeworld with 1k knowledge available and the other with 5k would make for an interesting choice.

Hearteater

Oct 20, 2011 10:49 am

reporter   ~0016886

I don't know that blocks of 250 are really necessary. Zharmad also pointed out on the forum a problem of clumping which could create situations that make it easy for players to exploit for high knowledge from few systems. I think based on that I would probably only go with 2k, 3k and 4k values and make distribution non-random so that you don't get large clumps of 4k (or 2k) systems near each other.

zoutzakje

Oct 20, 2011 11:13 am

reporter   ~0016887

I'm fine with large clumps of high (or low) values. If everything is random there are odds of that happening yes. Like it's also possible to find good targets early game (3-4 ARS 2-3 hops from homeworld) or absolutely no good target early game (1 zenith power generator in a 6-7 hop radius). I've experienced both situations and I'm absolutely fine with it. that's what makes AI war such a good game, you can expect anything to happen.

TechSY730

Oct 20, 2011 11:31 am

reporter   ~0016889

I'm with zoutzakje. Outside of near unwinnable or truly unwinnable failsafes (like a raid engine spawning next to a player home or a core shield generator spawning on an AI home), there are no things put into place to prevent statistical clustering. Having a cluster of high knowledge planets would certainly not be a rare sight, but that is not a bad thing. Especially because that cluster of high knowledge planets may be out of the way of where you want to go, or there may be a cluster of low knowledge planets to balance it out, or there may be no such cluster.

This is one thing I do like about AI war. So many other games make their random "less random" by putting measures into place to avoid statistical clustering. AI War has the guts to use true randomness.

Hearteater

Oct 20, 2011 12:57 pm

reporter   ~0016894

Actually, AI Wars has quite a bit that isn't true random in map seeding. AI Home Worlds aren't because true random doesn't always produce a better play experience. If this option means every game you look at the nearby systems, find the cluster of 2-4 5000 knowledge systems and rush them for an easy game, that doesn't make it a fun option. Depending on map type, true random distribution could result in there almost always being a good cluster of systems to take early.

What about two simple rules like: No adjacent high value systems (4k/5k), no more than one high value (4k/5k) and one low value (1k/2k) system next to the player home world. That would prevent a start like, the four systems next to your home planet all have 1k knowledge which is probably un-winnable at 8+ difficulty.

If the range is kept from 1k to 5k, I also think fewer at the extreme would be better. Say 5% 1k/5k, maybe 20% at the 2k/4k level, leaving 50% at 3k.

TechSY730

Oct 20, 2011 1:27 pm

reporter   ~0016899

I would support a compromise of "goal percentages" for overall distribution and rules in place to keep planets around the player homes useful but too rewarding. Both of these fail-safes already have precedence in other mechanics of the game.

The only thing that concerns me that the clustering avoidance you proposed seems a bit too strict. How about bumping it up to no more that 2 high knowledge planets adjacent to each other (like 1k -- 5k -- 5k -- 3k would be valid, but 1k -- 5k -- 5k -- 4k would not)

Hearteater

Oct 20, 2011 1:47 pm

reporter   ~0016906

I don't think that would be a problem, but I would still leave it as no more than one of each (high and low) near each player starting system.

TechSY730

Oct 20, 2011 1:48 pm

reporter   ~0016907

Yea, no more than one of each extreme next to a player home seems reasonable.

Vinraith

Oct 22, 2011 11:54 am

reporter   ~0016972

As long as the total knowledge in the galaxy stays the same, and the knowledge values are visible after scouting, I think this is a really interesting idea. It adds another layer of strategic decision making, as it complicates (in a good way) how one values possible planet captures.

zoutzakje

Oct 22, 2011 3:48 pm

reporter   ~0016977

exactly what I was thinking vinraith ^^

zharmad

Oct 24, 2011 2:58 am

reporter   ~0017001

Last edited: Oct 24, 2011 3:00 am

- Granularity in K-distribution:
 Perhaps we can have a certain compromise at 500-intervals: 2000, 2500, 3000, etc. Thinking through most of the research unlocks, I remember there are enough techs that end with a 500, but only three or so techs I'd unlock that has a 250 or 750. Most games, actually, I end up with a 500 somewhere that's not easy to use.
 In terms of actually coding this, it might be neater to just ask the program to draw a distribution (binomial?) and round-off the decimals. We've got a large range of sector-sizes (10-120) to consider, and also I'd probably reckon this kind of thing is too influential in games below 30 planets.

Extremities:
 About what I said in the forums... I don't think reducing the "clumpiness" of random distributions will help that much. Reducing the total variation would be better. 5% is only 4 planets of 80 (which is managable), or only have distributions between 1.5k-4.5k or 2k-4k.

 Other notes:
 Would it be possible to have planet-level influence the abundance of knowledge? I think this makes sense intuitively (better local AI derives from better local equipment?), but am not sure it's game-play safe.

martyn_van_buren

Jul 12, 2012 4:08 pm

reporter   ~0026716

How about making both AI Homeworlds 5k but not counting them out of the pool of 5k planets available? I feel like the decision about which homeworld to attack first is complicated and important enough that I can't imagine going for one over the other for a measly 4k knowledge (which I'd have to colonize, research, spend, and then build whatever I unlocked while I'm making my rush for the second homeworld).

There would be an extra 4k knowledge in the galaxy (as in an extra 2k per homeworld), but you could only ever get 2k of it and that 2k seems like a perfectly reasonable bonus for colonizing a homeworld.

Faulty Logic

Aug 4, 2012 2:30 am

reporter   ~0027349

Last edited: Sep 1, 2012 6:42 am

There are already enough factors requiring you to pick and choose which planets to take.

I like that k, one of the most valuable resources in the game, is the same across all planets. It is one of the main incentives to even consider a high-AIP game. Adding variation to the knowledge count would reduce the number of viable strategies, if viewable. Just having it be random would be incredibly frustrating on low rolls, and make some games far easier than they should be on high ones.

Valtiel

Oct 3, 2012 10:18 am

reporter   ~0028577

The big problem I see here is that it gives an added opportunity for the RNG to screw you. If you're unlucky, all the must-capture worlds (Advanced Factory, ARS) end up with a paltry 1k knowledge.

I'd support it if ARSs were guaranteed to spawn on 3k worlds. At that point, it's just another thing encouraging or discouraging you from going for Extra Shiny Toys like golems.

Issue History

Date Modified Username Field Change
Oct 19, 2011 10:09 am zoutzakje New Issue
Oct 19, 2011 10:13 am tigersfan Note Added: 0016801
Oct 19, 2011 10:13 am tigersfan Internal Weight => Feature Suggestion
Oct 19, 2011 10:13 am tigersfan Status new => strongly considering
Oct 19, 2011 11:48 am TechSY730 Note Added: 0016810
Oct 19, 2011 11:48 am TechSY730 Note Edited: 0016810
Oct 19, 2011 12:40 pm zoutzakje Note Added: 0016813
Oct 19, 2011 12:42 pm zoutzakje Note Edited: 0016813
Oct 19, 2011 1:15 pm Hearteater Note Added: 0016814
Oct 19, 2011 1:20 pm Hearteater Note Edited: 0016814
Oct 19, 2011 1:25 pm Hearteater Note Added: 0016815
Oct 19, 2011 1:44 pm TechSY730 Note Added: 0016816
Oct 19, 2011 3:01 pm zoutzakje Note Added: 0016821
Oct 19, 2011 3:04 pm zoutzakje Note Edited: 0016821
Oct 19, 2011 3:12 pm TechSY730 Note Added: 0016822
Oct 19, 2011 3:42 pm Ranakastrasz Note Added: 0016827
Oct 19, 2011 3:45 pm Hearteater Note Added: 0016828
Oct 19, 2011 5:15 pm zoutzakje Note Added: 0016843
Oct 20, 2011 10:49 am Hearteater Note Added: 0016886
Oct 20, 2011 11:13 am zoutzakje Note Added: 0016887
Oct 20, 2011 11:31 am TechSY730 Note Added: 0016889
Oct 20, 2011 12:57 pm Hearteater Note Added: 0016894
Oct 20, 2011 1:27 pm TechSY730 Note Added: 0016899
Oct 20, 2011 1:47 pm Hearteater Note Added: 0016906
Oct 20, 2011 1:48 pm TechSY730 Note Added: 0016907
Oct 22, 2011 11:54 am Vinraith Note Added: 0016972
Oct 22, 2011 3:48 pm zoutzakje Note Added: 0016977
Oct 24, 2011 2:58 am zharmad Note Added: 0017001
Oct 24, 2011 3:00 am zharmad Note Edited: 0017001
Jul 12, 2012 4:08 pm martyn_van_buren Note Added: 0026716
Aug 4, 2012 2:30 am Faulty Logic Note Added: 0027349
Sep 1, 2012 6:42 am Faulty Logic Note Edited: 0027349
Oct 3, 2012 10:18 am Valtiel Note Added: 0028577