Ask A Game Dev - Procedural Planets v2 Analysis

rpmcmurphy

http://askagamedev.tumblr.com/post/151797723437/weekly-video-analysis-3-star-citizens

And then the follow up after accusations of poor research and attempting to tear the game down (lol)
http://askagamedev.tumblr.com/post/151801073831/regarding-your-article-on-star-citizen-you-did

Turrican187

Talk is cheap.

Nuff said

MaxBacon

The same analysis can fall upon Elite Dangerous Procedural Planets or NMS as well.
Or any game using this large scale vs gameplay content.

The technology allows to amazing scales, details and visuals; however when it comes to gameplay and content it's not going to be the tech that is going to magically add things to do. You don't procedurally generate gameplay.

With the artist-driven dev on the procedural planets on SC they can be made logical and unique, however, at this stage, I don't feel we'll get any solid layer of gameplay on the planets without things like mining being added upon them at later updates.

The 2 points I see that writer is more out of sync with SC's dev

• That this is to be released upon an alpha, one first iteration of several others to come later not as a finished game so a complete gameplay loop is not expectable.
• And towards the artist work needed to populate the planet, being all this fully explorable PG planets considered non-populated past outposts you can find and such; and the artists tools as shown can scale to the necessary scale.

CrazKanuk

rpmcmurphy said:

http://askagamedev.tumblr.com/post/151797723437/weekly-video-analysis-3-star-citizens

And then the follow up after accusations of poor research and attempting to tear the game down (lol)
http://askagamedev.tumblr.com/post/151801073831/regarding-your-article-on-star-citizen-you-did

I tend to agree with regards to PG. I said it for NMS as well. Randomizing an infinite combination of elements from a finite number of assets. Imagine the limited possibilities  

Balmong

CrazKanuk said:

rpmcmurphy said:

http://askagamedev.tumblr.com/post/151797723437/weekly-video-analysis-3-star-citizens

And then the follow up after accusations of poor research and attempting to tear the game down (lol)
http://askagamedev.tumblr.com/post/151801073831/regarding-your-article-on-star-citizen-you-did

I tend to agree with regards to PG. I said it for NMS as well. Randomizing an infinite combination of elements from a finite number of assets. Imagine the limited possibilities  

Which I think is why they aren't putting all their faith in some mathematical equation, rather what the artist intends for the planet. 

I like the style of design they're going with, get PG to do 60% of the work, tweak and add as needed.

Octagon7711

My analysis so far, for what's it's worth is that this game will come in as average and a little over priced.  They are doing state of the art stuff but that's offset by increased trial and errors in trying to get stuff to work that's never been done before.  It will launch as a MVP and will continue be in a state of rebalance as promised systems are added producing a continued cycle of fixing things.  Alpha is an average alpha with most stuff broken which leads me to believe beta and launch will pretty much be the same.  

They're trying to do so much that the game will never run perfectly.  Just too many things that could break.  That's not taking into account a game that has so many details will produce a lot of lag when you have a big number of players online at the same time.  I know they're working on a system to handle that but it's a theory at this point which I doubt will work right out the box the way they want it to.  I think that's why each demo has no more than 10 players online at a time.  

Yes, I could be wrong but I've been playing games since pong and a lot of MMO's and I see a new team trying to do to much and please to many people.  And if I am wrong I will gladly post a formal apology so bookmark this and throw it in my face if the game launches with no problems.  I will eat humble pie if proven wrong.

Erillion

CrazKanuk said:

rpmcmurphy said:

http://askagamedev.tumblr.com/post/151797723437/weekly-video-analysis-3-star-citizens

And then the follow up after accusations of poor research and attempting to tear the game down (lol)
http://askagamedev.tumblr.com/post/151801073831/regarding-your-article-on-star-citizen-you-did

I tend to agree with regards to PG. I said it for NMS as well. Randomizing an infinite combination of elements from a finite number of assets. Imagine the limited possibilities  

http://www.mathwords.com/c/combination_formula.htm

Combination Formula

A formula for the number of possible combinations of r objects from a set of n objects. This is written in any of the ways shown below.

All forms are read aloud "n choose r."

 

Formula:
	Note: , where nPr is the formula for permutations of n objects taken r at a time.
Example:	How many different committees of 4 students can be chosen from a group of 15?
Answer:	There are possible combinations of 4 students from a set of 15.
	There are 1365 different committees.

Have fun

Vrika

Erillion said:

CrazKanuk said:

I tend to agree with regards to PG. I said it for NMS as well. Randomizing an infinite combination of elements from a finite number of assets. Imagine the limited possibilities  

http://www.mathwords.com/c/combination_formula.htm

Combination Formula

A formula for the number of possible combinations of r objects from a set of n objects. This is written in any of the ways shown below.

....

There are 1365 different committees.

That's all very well in theory, but in practice once you've seen the same student 10 times you'll likely remember him by heart, and 11th time will just feel like repeating the same.

Like No Man's Sky showed us, it's possible to procedurally generate nearly infinite variation. But human brain is very good at learning the types and elements used, and after that it feels just cosmetic.


I'm not saying that Star Citizen should try to do anything to fix it. It's not possible to create enough variation to fill that much space.

Phaserlight

Vrika said:

Erillion said:

CrazKanuk said:

I tend to agree with regards to PG. I said it for NMS as well. Randomizing an infinite combination of elements from a finite number of assets. Imagine the limited possibilities  

http://www.mathwords.com/c/combination_formula.htm

Combination Formula

A formula for the number of possible combinations of r objects from a set of n objects. This is written in any of the ways shown below.

....

There are 1365 different committees.

That's all very well in theory, but in practice once you've seen the same student 10 times you'll likely remember him by heart, and 11th time will just feel like repeating the same.

Like No Man's Sky showed us, it's possible to procedurally generate nearly infinite variation. But human brain is very good at learning the types and elements used, and after that it feels just cosmetic.


I'm not saying that Star Citizen should try to do anything to fix it. It's not possible to create enough variation to fill that much space.

I disagree with the above statement, it is possible. Combinatorics alone isn't enough to do it, you also have to use power laws. I don't understand why more developers don't grasp this. Or maybe they do, it's just that implementing it is more difficult than I'm imagining. No man's Sky certainly didn't seem to do this, and if the above equations are any indicator, Star Citizen won't either.  I look forward to being proven wrong. 

I've stated this many times here; if you want to see algorithmic procedural generation done right, have a look at Dwarf Fortress.