Graphics vs content

User avatar
Hume2
Member
Posts: 709
Joined: Tue Jun 19, 2018 08:24
GitHub: Hume2
In-game: Hume2
Location: Czech Republic

Re: Graphics vs content

by Hume2 » Post

voxelproof wrote: I suppose that when you were writing about "expotential exploration" you meant the above pattern of cognition and our ability to categorize the scope of our knowledge. Thus we assemble species into genera, genera into families and so on. But it still doesn't mean that making new discoveries of so far unknown species doesn't bring great joy and excitement to lucky scientists or explorers. There're many examples of discoveries of this kind in recent years. I was very impressed by the discovery of a 'living fossil' plant from the dinosaur era Wollemia nobilis in Australia. It's one of the most beautiful trees nature gave birth to. I think that while systematization of our knowledge and wisdom do tend to be expotential, the core exploration and wonder accompanying subsequent discoveries are actually linear in its inherent nature.
You are right. There might be some patterns but the patterns need to be flexible enough. A generation based on random combination of given elements might be also good but each combination must be specific in a way. And also each combination should have different occurence. I think, it would be good if you could find some things you couldn't categorise when you see them first. I'd like to see this in a game.
Well, in fact in real nature we can see many very repetitive patterns (this is what makes science possible) and the above example of human faces is not an exception. This is the tiny random flaws that make each of us unique (in terms of appearance at least). And what's more the psychological properties of human psyche seem to allow even stronger categorization of certain traits constituting each (not so unique after all) personality.
Yes, categories are quite common. However, there are also many many subcategories and also many exceptions. I have seen many categories being implemented in games but I haven't seen many subcategories. I feel like I have explored the game's content when I see some instances of each category. In real life, there are really many subcategories. There are many exceptions. There are also exceptions of exceptions. This is what I'm missing in procedurally-generated games.
P.S.
Another interesting video from Extra Credits -- this one discussing procedurally generated content. I think that our ability to harness this wondrous tool is still on a very early stage and the future will surely bring amazing development in this field if only developers will be willing to risk sinking their customers into immersive immensely engaging worlds ;)

https://www.youtube.com/watch?v=TgbuWfGeG2o
Nice point. Making a procedurally generated content is also a way of human design. Sometimes I feel like knowing the way how it's generated would spoil me all the possible surprises. It would be good if the generator was obscure enough so the easiest way of spoiling these surprises was via playing the game.
P.S. II
And this one about exploration in games is really worth citing -- it's very relevant to the discussion in this thread, and very well done indeed.

https://www.youtube.com/watch?v=FE7lDFAcb4Y

(I could not find these videos without Polish subtitles. Heck, this is how information bubbles work ;))
That's also a good point. I was talking mainly about content exploration which is exponential in my opinion. If you find a few plants, you might expect many other plants to exist. However, the geographical exploration is a good aspect as well. This seems me linear. If you want to locate the plant you predicted, you have to search all places which takes you linear time.
If you lack the reality, go on a trip or find a job.

slopsbucket
Member
Posts: 199
Joined: Fri Oct 31, 2014 22:56
Location: Top End, Australia

Re: Graphics vs content

by slopsbucket » Post

How much of the content is human-designed and how much is pure generic? It looks like it's 90% human-designed but it would be cool if it was pure generic.
There were plenty of human created surprises but the real entertainment is in the Creature AI.

When you first get your creature it's a baby, you have to train it in almost every aspect. What to eat, when to eat, how much to eat. Smack it for eating villagers or pooing in their food store, or not, depending on what sort of god you want to be. The creature only shows very minimalist clues as to what's going on in it's head. Until it gets to a stage where you can begin to trust it you need to follow it everywhere and either reward ot punish it for it's actions.

An efficient way to be an evil god is to teach your creature to play Catch with you, then you start playing catch with live people, over the top of the village where the whole village can see. You become evil even more quickly if you do this with a child.

Your creature (if ignored or abused) can even get depressed and despondent and start ignoring you.


This is where Black & White II completely failed, the creature was more like a robot than a genuine living entity.

Cheers,

Andrew.

User avatar
Hume2
Member
Posts: 709
Joined: Tue Jun 19, 2018 08:24
GitHub: Hume2
In-game: Hume2
Location: Czech Republic

Re: Graphics vs content

by Hume2 » Post

slopsbucket wrote: There were plenty of human created surprises but the real entertainment is in the Creature AI.

When you first get your creature it's a baby, you have to train it in almost every aspect. What to eat, when to eat, how much to eat. Smack it for eating villagers or pooing in their food store, or not, depending on what sort of god you want to be. The creature only shows very minimalist clues as to what's going on in it's head. Until it gets to a stage where you can begin to trust it you need to follow it everywhere and either reward ot punish it for it's actions.

An efficient way to be an evil god is to teach your creature to play Catch with you, then you start playing catch with live people, over the top of the village where the whole village can see. You become evil even more quickly if you do this with a child.

Your creature (if ignored or abused) can even get depressed and despondent and start ignoring you.


This is where Black & White II completely failed, the creature was more like a robot than a genuine living entity.

Cheers,

Andrew.
The Creature's AI seems me perspective. It's a shame they have ruined it in Black & White II. I'm curious how knowing the source code would spoil me the surprise. Is the AI only hidden behind the GUI or it's based on pure mathematics?
If you lack the reality, go on a trip or find a job.

User avatar
voxelproof
Member
Posts: 1087
Joined: Sat Aug 05, 2017 08:13
Location: Europe

Re: Graphics vs content

by voxelproof » Post

Hume2 wrote:There might be some patterns but the patterns need to be flexible enough. A generation based on random combination of given elements might be also good but each combination must be specific in a way. And also each combination should have different occurence. I think, it would be good if you could find some things you couldn't categorise when you see them first. I'd like to see this in a game.
The issue with this kind of organizing the game's content is that in my opinion it doesn't fully reflect how the real world patterns originate; while in games it's most often achieved through different parameters provided randomly (or quasi-randomly) for a given persistently constant algorithm, in reality there are moreover patterns which generate different modifications of the algorithms themselves. It's a matter of higher systemic level of organization and although it certainly is achievable in computing, the development of this kind of procedural meta-algorithms is definitely of a higher order of magnitude in terms of difficulty caused by its complexity. And of course we shouldn't forget that the sheer number of real world basic paticles interacting with each other creates complexity unimaginable in games.
I feel like I have explored the game's content when I see some instances of each category. In real life, there are really many subcategories. There are many exceptions. There are also exceptions of exceptions. This is what I'm missing in procedurally-generated games.
(...)
I was talking mainly about content exploration which is exponential in my opinion. If you find a few plants, you might expect many other plants to exist. However, the geographical exploration is a good aspect as well. This seems me linear. If you want to locate the plant you predicted, you have to search all places which takes you linear time.
I was thinking about it today and I think that there's one very basic psychological reason why you are after all near the truth when you say that the exploration in games seems to be expotential. Why aren't we so aroused by a new plant we see in game when we saw a hundred of others already? What's the essencial difference compared to the real phenomena observed?

It's the fact that we know that both this particular plant and the whole world are just an illusion, a new technologically enhanced kind of old smoke and mirrors tricks and we simply cannot treat this sort of experience seriously enough to attribute any meaningful existential value to it. If we know that all the plants, creatures, structures and so on are all in all designed by the game developers, we subconciously know that we simply cannot credit ourselves with our "discoveries'" for the very simple reason that some other sentient intelligent being contrived them to make our discoverer's trail believable. Things however change dramatically when we know that what we see and discover wasn't seen and what's much more important even predicted by anybody else. And when such amazing, providing sense of wonder and curious surprise content emerges from very simple fundamental rules of the virtual universe we are entitled to say that our discoveries now do become meaningful and that the virtual world where we make them starts to gain existential weight.
To miss the joy is to miss all. Robert Louis Stevenson

User avatar
Hume2
Member
Posts: 709
Joined: Tue Jun 19, 2018 08:24
GitHub: Hume2
In-game: Hume2
Location: Czech Republic

Re: Graphics vs content

by Hume2 » Post

voxelproof wrote: The issue with this kind of organizing the game's content is that in my opinion it doesn't fully reflect how the real world patterns originate; while in a game it's achieved through different parameters provided randomly (or quasi-randomly) for a given persistently constant algorithm, in reality there are moreover patterns which generate different modifications of the algorithms themselves. It's a matter of higher systemic level of organization and although it certainly is achievable in computing, the development of this kind of procedural meta-algorithms is definitely of a higher order of magnitude in terms of difficulty caused by its complexity.
I agree. We are trying to simulate the nature (or something that looks natural) and we get that the nature is far more complex than we previously thought. I think that the game realities should look in some ways natural too. The world should look like it might work in other reality.
It's the fact that we know that both this particular plant and the whole world are just an illusion, a new technologically enhanced kind of old smoke and mirrors tricks and we simply cannot treat this sort of experience seriously enough to attribute any meaningful existential value to it. If we know that all the plants, creatures, structures and so on are all in all designed by the game developers, we subconciously know that we simply cannot credit ourselves with our "discoveries'" for the very simple reason that some other sentient intelligent being contrived them to make our discoverer's trail believable. Things however change dramatically when we know that what we see and discover wasn't seen and what's much more important even predicted by anybody else. And when such amazing, providing sense of wonder and curious surprise content emerges from very simple fundamental rules of the virtual universe we are entitled to say that our discoveries now do become meaningful and that the virtual world where we make them starts to gain existential weight.
Exactly. When I see a new plant, I can't say that I've discovered the plant because it was supposed to generate this way. The developer made the generator generate that sort of plants. It might be a discovery only if the developer wasn't thinking about that exact combination when programming the generator. However, these discoveries are quite limited.

I think that there are two ways to improve the procedurally generated content. The first one is making the generator really complex, adding a lot of details those can be randomised and randomising the patterns. This takes a lot of time but can be easily adjusted. The main disadvantage is that each aspect must be invented before so you are exploring something that has been already invented.

The other way, which is not much used, is using a mathematic problem that is hard to solve exactly. For example chess use this. It is very hard to adjust the generator for your purposes and that's probably why it's not commonly used. The advantage is that the player can make true discoveries. The rules of chess doesn't say anything about the possible strategies. You can really discover your own strategies.
If you lack the reality, go on a trip or find a job.

slopsbucket
Member
Posts: 199
Joined: Fri Oct 31, 2014 22:56
Location: Top End, Australia

Re: Graphics vs content

by slopsbucket » Post

The Creature's AI seems me perspective. It's a shame they have ruined it in Black & White II. I'm curious how knowing the source code would spoil me the surprise. Is the AI only hidden behind the GUI or it's based on pure mathematics?
Hi Hume2,

It's not the sort of AI that you are imagining, it's a Neural Net type of program. It can't be programmed, it must be trained. Even the developers claimed that they were constantly surprised by their own creation.

https://en.wikipedia.org/wiki/Artificial_neural_network

If you have the time I recommend you have a bit of a play with it, there's many places where you can legally download it for free such as https://www.myabandonware.com/game/black-white-a33

Cheers,

Andrew.

EDIT: forgot to mention, most of these old games won't play or will behave strangely on modern Microsoft operating systems. They work perfectly in Linux under Wine. Black & White even works perfectly using a 64 bit prefix.

User avatar
Hume2
Member
Posts: 709
Joined: Tue Jun 19, 2018 08:24
GitHub: Hume2
In-game: Hume2
Location: Czech Republic

Re: Graphics vs content

by Hume2 » Post

slopsbucket wrote: It's not the sort of AI that you are imagining, it's a Neural Net type of program. It can't be programmed, it must be trained. Even the developers claimed that they were constantly surprised by their own creation.
That's good. I definitely need to try it myself. It's really a shame they have ruined that.
If you lack the reality, go on a trip or find a job.

Byakuren
Member
Posts: 818
Joined: Tue Apr 14, 2015 01:59
GitHub: raymoo
IRC: Hijiri
In-game: Raymoo + Clownpiece

Re: Graphics vs content

by Byakuren » Post

I can think of a trivial way to make a game unexplorable, which is to require playing 24 hours of match 3 puzzle mini games for every one minute of exploration gameplay.
Every time a mod API is left undocumented, a koala dies.

User avatar
Hume2
Member
Posts: 709
Joined: Tue Jun 19, 2018 08:24
GitHub: Hume2
In-game: Hume2
Location: Czech Republic

Re: Graphics vs content

by Hume2 » Post

Byakuren wrote:I can think of a trivial way to make a game unexplorable, which is to require playing 24 hours of match 3 puzzle mini games for every one minute of exploration gameplay.
I'm not sure if I understand it correctly. You can explore one subgame for a minute and then you need to play the other subgame for 24 hours. I don't think that this really solves the problem. I'd like to mention another video by Extra Credits: https://www.youtube.com/watch?v=NWZLB8CyPbM

EDIT: Also this might be interesting: https://www.youtube.com/watch?v=jVL4st0blGU
If you lack the reality, go on a trip or find a job.

User avatar
voxelproof
Member
Posts: 1087
Joined: Sat Aug 05, 2017 08:13
Location: Europe

Re: Graphics vs content

by voxelproof » Post

Byakuren wrote:I can think of a trivial way to make a game unexplorable, which is to require playing 24 hours of match 3 puzzle mini games for every one minute of exploration gameplay.
This is an example of a brutal method of making game unapproachable, not unexplorable. When we consider 'explorability' (I'm not even sure if such a word is in use) we think about in-game content which remains interesting through the whole course of exploring it, providing ever changing experience of discovering truly new things (or at least perceived as such subjectively) and yet impossible to get to know it completely. Roguelikes which may serve as sort of reference to this topic are in fact primarily focused on exploration hindered by monsters and other guardians of nether realms. To make exploration in games purposeful and meaningful it's obviously necessary to align it with a sense of accomplishment of some kind, which in turn means that the gameplay must be demanding and to some degree very difficult, otherwise any discovery which doesn't cost significant effort will sooner or later start feeling trivial and boring.
To miss the joy is to miss all. Robert Louis Stevenson

User avatar
Hume2
Member
Posts: 709
Joined: Tue Jun 19, 2018 08:24
GitHub: Hume2
In-game: Hume2
Location: Czech Republic

Re: Graphics vs content

by Hume2 » Post

I think, I'm now sure that I understand. I think, it is the unpleasant design mentioned in that video by Extra Credits. This design would actually tell me something like "Please, don't play this game so often, it's not good. Better play these puzzles." I use again the approach to these polynomial and exponential functions. In this case, the limitation only multiplies the polynomial function by a constant. It doesn't even increase the rank. And it is also possible that the people would hack the game in order to enjoy it without the need of playing those puzzles.
If you lack the reality, go on a trip or find a job.

User avatar
voxelproof
Member
Posts: 1087
Joined: Sat Aug 05, 2017 08:13
Location: Europe

Re: Graphics vs content

by voxelproof » Post

Hume2 wrote:I'd like to mention another video by Extra Credits: https://www.youtube.com/watch?v=NWZLB8CyPbM

EDIT: Also this might be interesting: https://www.youtube.com/watch?v=jVL4st0blGU
There're scores of very good Extra Credits' videos about game design, game development, gaming styles and habits and so on. Most of my knowledge about these subjects comes from this remarkably professional YT channel.
To miss the joy is to miss all. Robert Louis Stevenson

User avatar
Stix
Member
Posts: 1385
Joined: Fri Aug 04, 2017 14:19
IRC: nil
In-game: Stix [+alts]
Location: USA

Re: Graphics vs content

by Stix » Post

voxelproof wrote:There're scores of very good Extra Credits' videos about game design, game development, gaming styles and habits and so on. Most of my knowledge about these subjects comes from this remarkably professional YT channel.
That makes two of us :).
Hey, what can i say? I'm the bad guy.

User avatar
voxelproof
Member
Posts: 1087
Joined: Sat Aug 05, 2017 08:13
Location: Europe

Re: Graphics vs content

by voxelproof » Post

Stix wrote:
voxelproof wrote:There're scores of very good Extra Credits' videos about game design, game development, gaming styles and habits and so on. Most of my knowledge about these subjects comes from this remarkably professional YT channel.
That makes two of us :).
Given the numbers of views there're probably even a few more :)
To miss the joy is to miss all. Robert Louis Stevenson

User avatar
Hume2
Member
Posts: 709
Joined: Tue Jun 19, 2018 08:24
GitHub: Hume2
In-game: Hume2
Location: Czech Republic

Re: Graphics vs content

by Hume2 » Post

I like Extra Credits' videos too. They have good graphics. I like how the abstract terms are visualised.
If you lack the reality, go on a trip or find a job.

User avatar
voxelproof
Member
Posts: 1087
Joined: Sat Aug 05, 2017 08:13
Location: Europe

Re: Graphics vs content

by voxelproof » Post

Hume2 wrote:I like Extra Credits' videos too. They have good graphics. I like how the abstract terms are visualised.
Yes, it's very simplified yet combined with the really brilliant narrative it makes even complex ideas to be easily understandable. BTW they also make very informative videos about different scientific problems. I wasn't very much to maths but thanks to them I understood some previously rather obscure to me concepts (like non-euclidean geometry -- they made a whole series about this).
To miss the joy is to miss all. Robert Louis Stevenson

Astrobe
Member
Posts: 570
Joined: Sun Apr 01, 2018 10:46

Re: Graphics vs content

by Astrobe » Post

The problem #1 in Minetest is that games are often in an open world which is practically infinite but it's difficult to provide infinite gameplay. More often than not your players will have reached their goal before the edge of the world (unless their goal is precisely that).

It's tempting to apply the recipe used by many MMORPGs which is to release more content when they sense that the players get bored. Or to release your game with already a lot of content (at the risk of being inconsistent, prone to exploits or prone to crashes).

This problem of infinite gameplay is not a problem Minetest as a game engine can solve directly.

User avatar
Hume2
Member
Posts: 709
Joined: Tue Jun 19, 2018 08:24
GitHub: Hume2
In-game: Hume2
Location: Czech Republic

Re: Graphics vs content

by Hume2 » Post

Astrobe wrote:The problem #1 in Minetest is that games are often in an open world which is practically infinite but it's difficult to provide infinite gameplay. More often than not your players will have reached their goal before the edge of the world (unless their goal is precisely that).
I feel the same thing. I just didn't know how to express that.
It's tempting to apply the recipe used by many MMORPGs which is to release more content when they sense that the players get bored. Or to release your game with already a lot of content (at the risk of being inconsistent, prone to exploits or prone to crashes).

This problem of infinite gameplay is not a problem Minetest as a game engine can solve directly.
Yes, this happens quite often. And can happen also when the content is generic. When the content generator is too naïve, it doesn't remove the problem at all.
If you lack the reality, go on a trip or find a job.

Astrobe
Member
Posts: 570
Joined: Sun Apr 01, 2018 10:46

Re: Graphics vs content

by Astrobe » Post

I think part of the solution could be to extend the concept of code generation to mobs and items.

Hamlet did something like this recently, by programming a system that makes the stats of the mobs vary (Mobs Redo already does that for health, he have extended it to other properties). Looking at the description he may have gone too far in some areas for my taste and it probably doesn't fit all games, but I think it's the way to go. It could be a way to make each encounter more unique.

User avatar
voxelproof
Member
Posts: 1087
Joined: Sat Aug 05, 2017 08:13
Location: Europe

Re: Graphics vs content

by voxelproof » Post

Astrobe wrote:The problem #1 in Minetest is that games are often in an open world which is practically infinite but it's difficult to provide infinite gameplay. More often than not your players will have reached their goal before the edge of the world (unless their goal is precisely that).

It's tempting to apply the recipe used by many MMORPGs which is to release more content when they sense that the players get bored. Or to release your game with already a lot of content (at the risk of being inconsistent, prone to exploits or prone to crashes).

This problem of infinite gameplay is not a problem Minetest as a game engine can solve directly.
This problem is caused by misunderstanding of the psychological nature of a good playable game mechanics I suppose. Look at some pretty simple arcade games: they have a very simple mechanics, there're always the same levels to go through, yet some of them are played to this day despite their repetitive nature. It's because the repetitiveness is sometimes a key to the replayability, on the condition that each time some elements crucial for a gameplay are randomized, and that each decision/action has significant consequences for the whole rest of the particular gaming session. This is not only the case of arcade games, the same is true about old classic board games like chess. And, however it's certainly not exactly the same as "exploration", this gives those games a factorial complexity in number of possible significantly different gameplays, therefore they never stop to engage and even surprise their players.

So, in my opinion, from a player's point of view much more important than "infinite exploration" is whether a game has a high replayability. Of course it's extremely difficult to obtain it through the sheer content, and for example the developers of No Man's Sky seem to follow the path I've mentioned above. My answer to this issue is, since I play Minetest mostly as a walking simulator, creating a graphics that makes never ceasing wonder of discovering ever changing complex dreamy landscapes composed of a limited set of simple basic elements.
To miss the joy is to miss all. Robert Louis Stevenson

User avatar
Hume2
Member
Posts: 709
Joined: Tue Jun 19, 2018 08:24
GitHub: Hume2
In-game: Hume2
Location: Czech Republic

Re: Graphics vs content

by Hume2 » Post

I have another approach. Let's see how the simple obscure rules break into many consequences:

Chess rules -> all known strategies, figure values, etc.
Conway Game of Life -> all those known pre-sets with known behaviour
Mandelbrot Set (z² + c) -> repetitive patterns, numbers of arms, figure stacking, etc.
Rubik Cube -> all known algorithms to solve it

And now some bad examples. I call them naïve rules:

Most of MMORPG games -> Play it exactly the way you are told to.
Most of shooting games -> Play it exactly the way you are told to.

Sometimes your strategy can vary a little but still, all strategies are known before anyone starts playing the game. Naïve rules are good in the way they are easily adjusted. However, they don't provide as much depth as the obscure rules. Look, the obscure rules actually don't tell you anything about how you should play the game or how the result should look like.
If you lack the reality, go on a trip or find a job.

User avatar
voxelproof
Member
Posts: 1087
Joined: Sat Aug 05, 2017 08:13
Location: Europe

Re: Graphics vs content

by voxelproof » Post

Hume2 wrote:I have another approach. Let's see how the simple obscure rules break into many consequences:

Chess rules -> all known strategies, figure values, etc.
Conway Game of Life -> all those known pre-sets with known behaviour
Mandelbrot Set (z² + c) -> repetitive patterns, numbers of arms, figure stacking, etc.
Rubik Cube -> all known algorithms to solve it

And now some bad examples. I call them naïve rules:

Most of MMORPG games -> Play it exactly the way you are told to.
Most of shooting games -> Play it exactly the way you are told to.

Sometimes your strategy can vary a little but still, all strategies are known before anyone starts playing the game. Naïve rules are good in the way they are easily adjusted. However, they don't provide as much depth as the obscure rules. Look, the obscure rules actually don't tell you anything about how you should play the game or how the result should look like.
I agree with everything you've written but still there're certain exceptions. I'll give you an example of a simple space shooter I'm playing for... well, long time:) It was not until after many years of playing it when I realized that the author of this game apparently implemented some algorithm which checks out the player's activity and which is adjusted to detect inexperienced type of gameplay. If such kind of behaviour is detected, the probability of weapon upgrade noticeably rises, which makes possible some non-standard tactics enhancing the chances of winning the game. Of course the simple way of thinking about such exploits is that it's a game's shortcoming, but I wouldn't say so, especially that when I realized it it gave me a real, profound joy of discovery of a hidden game mechanics (btw this kind of discoveries is discussed in the video linked in one of the previous posts). And even in this simple game there're some other not so obvious rules concerning tactics that an attentive player may figure out only after many played games(usually referring to probabilities of different power-ups).

Even the simplest game can be made engaging without putting a player into the Skinner box trap. Sadly this is not what the gaming industry is caring about: open-world (and any other) games with very high replayability value are regarded as vile monsters menacing the prospect of future gains. And this is why linear story-based role-playing games are obviously the best to keep players buying new products. It's because the only goal is to make money and to achieve it it's necessary to make a game's lifespan as short as possible. The corporate minds really don't like happy players, only those who are craving for a next release of some established franchise.
Last edited by voxelproof on Fri Aug 03, 2018 22:23, edited 3 times in total.
To miss the joy is to miss all. Robert Louis Stevenson

Astrobe
Member
Posts: 570
Joined: Sun Apr 01, 2018 10:46

Re: Graphics vs content

by Astrobe » Post

Hume2 wrote:I have another approach. Let's see how the simple obscure rules break into many consequences:

Chess rules -> all known strategies, figure values, etc.
Conway Game of Life -> all those known pre-sets with known behaviour
Mandelbrot Set (z² + c) -> repetitive patterns, numbers of arms, figure stacking, etc.
Rubik Cube -> all known algorithms to solve it
Those examples are remarkable in two ways:
- they use combinatorial explosion in some way (except Mandelbrot)
- they exclude luck or randomness

I'd like to mention that luck isn't the enemy, on the contrary. It can have two positive roles:
- Simulate complexity, as illustrated by the use of Perlin and Voronoi noise in MT. Those noise generators allow us to simulate the result of the natural processes that create landscapes; processes that would be way too expensive to simulate "by the equations".
- "buffer" the mistakes of players and offer a second chance, or on the contrary disturb the players when they are in a winning spiral (the more you have equipment or resources, the easier it becomes to get more) - perhaps even open the possibility of regression to slow down progression.

User avatar
Hume2
Member
Posts: 709
Joined: Tue Jun 19, 2018 08:24
GitHub: Hume2
In-game: Hume2
Location: Czech Republic

Re: Graphics vs content

by Hume2 » Post

voxelproof wrote:I agree with everything you've written but still there're certain exceptions. I'll give you an example of a simple space shooter I'm playing for... well, long time:) It was not until after many years of playing it when I realized that the author of this game apparently implemented some algorithm which checks out the player's activity and which is adjusted to detect inexperienced type of gameplay. If such kind of behaviour is detected, the probability of weapon upgrade noticeably rises, which makes possible some non-standard tactics enhancing the chances of winning the game. Of course the simple way of thinking about such exploits is that it's a game's shortcoming, but I wouldn't say so, especially that when I realized it it gave me a real, profound joy of discovery of a hidden game mechanics (btw this kind of discoveries is discussed in the video linked in one of the previous posts). And even in this simple game there're some other not so obvious rules concerning tactics that an attentive player may figure out only after many played games(usually referring to probabilities of different power-ups).
Yes, there are exceptions. It's interesting how some simply-looking games are much more exciting than the big games with a huge set of naïve rules. After all, games like Tetris, Snake or Arkanoid are still played. And they are very simple.
Astrobe wrote: Those examples are remarkable in two ways:
- they use combinatorial explosion in some way (except Mandelbrot)
- they exclude luck or randomness

I'd like to mention that luck isn't the enemy, on the contrary. It can have two positive roles:
- Simulate complexity, as illustrated by the use of Perlin and Voronoi noise in MT. Those noise generators allow us to simulate the result of the natural processes that create landscapes; processes that would be way too expensive to simulate "by the equations".
- "buffer" the mistakes of players and offer a second chance, or on the contrary disturb the players when they are in a winning spiral (the more you have equipment or resources, the easier it becomes to get more) - perhaps even open the possibility of regression to slow down progression.
Well, there is a thing called deterministic chaos. It means that whole future of given system is determined by its current sate. However, even a tiniest change might result in a massive change of the future. The examples I listed use deterministic chaos. For example, in chess, you might be winning but one bad move can turn the situation. Of course, these games might involve randomness too but they actually don't need it.

There are also examples those involve randomness, for example Perlin noise. Or also Tetris might be a good example because you are free to think your own strategy and you always get a random piece. There are also rules with random pre-sets, like most of card games. The cards are always mixed randomly at start but the rest of the game doesn't involve randomness at all.

And there are also bad examples those involve randomness. Imagine this simple sword generator. Each sword has a random power between 10 and 100. This actually provides no depth because none of those random swords are special. A sword with power 67 is not special at all, it's the same as sword 65, just a bit stronger. This is a naïve rule. Some designers solve it in the way, they add another aspects those can be randomised. Let's say that each sword has random attack speed between 0.5 and 2.0. The depth is increased a bit but it's still too low. A sword with power 67 and attack speed 1.2 is not much different than a sword with power 65 and attack speed 1.1. Now there are two naïve rules. Well, if there are more aspects randomised, there is a smaller probability of getting swords with close aspects. However, it's only a question of time when you get them. Adding more and more naïve rules really increase the depth but not much. Randomness might be good, but the logics of the generator shouldn't rely on randomness too much.
If you lack the reality, go on a trip or find a job.

Astrobe
Member
Posts: 570
Joined: Sun Apr 01, 2018 10:46

Re: Graphics vs content

by Astrobe » Post

The naïve, single-stat-enhancement sword generator can be better than you think depending on how you use it.

If:
- you place those swords in a treasure chest hidden in the depth of the mapgen'ed dungeons and you don't regenerate those chests,
- you don't allow to repair swords,

(in other words if there's a fixed number of those swords in a world for all its lifespan)

Then you have created a relative cheap way to have players generate quests by themselves: they will explore dungeons in search for the exceptional 95+ hit swords.

Moreover, players will quickly do the maths: they'll use the lesser swords against weak monster and reserve exceptional swords for more dangerous monsters. This creates additional minor gameplay aspects (wear both kind of swords or make sure you have your good sword with you if you go for a boss fight) and bargain opportunities in player-to-player trade. Bargains will evolve over time as more and more swords are discovered and then as more and more are broken. As a consequence, it will become more and more profitable to explore the world.

I would however make sure to use a normal (bell curve) distribution rather than the usual uniform distribution for maximum effect.

Post Reply

Who is online

Users browsing this forum: Bing [Bot] and 18 guests