Jump to content

Procedurally Generated Maps... Work


236 replies to this topic

#81 Vandul

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,342 posts
  • LocationYork, New

Posted 05 August 2015 - 03:14 AM

7 Days 2 Die has been using a procedural system for awhile now. Your world can literally be infinite. INFINITE. Of course, thats on the Unity engine, not Cryengine.

Theres no reason idle servers couldnt be churning out procedural maps, dumping them into a pool, and the MM going in and fishing one out randomly. That helps the 'guy with the weak machine'.

Matches might take a bit longer to set up, but I am sure there are plenty of workarounds.

#82 Will HellFire

    Member

  • PipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 203 posts

Posted 05 August 2015 - 03:31 AM

The way I see it, PGI could create a dozen maps and upload them every two weeks to our computers when the patch is applied.

Sure, longer patches, but no wait for computationally heavy procedurally generated worlds.

And hey, new maps every two weeks.

#83 Raggedyman

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,278 posts
  • LocationFreedonia Institute of Mech Husbandry

Posted 05 August 2015 - 04:13 AM

There is a massive difference between being able to generate a map that can be played on and being able to generate one that makes for a great player experience in a specific game, especially something like a multiplayer round based FPS.

Don't get me wrong, I would love to see more maps and I love the idea of PGC, but the simple fact is that whilst the technology to make them happen is out there no-one has cracked the "how to make it work on a 'fun' level" art of it in a game like MWO. A couple of developers have tried block-based randomised environments, but I'm not aware of them satisfying both the variety requirement of those who want always changing maps and the playability demands of those who don't want to get killed because of bad luck on the map gen.

#84 Gideon Grey

    Member

  • PipPipPipPipPipPip
  • The Partisan
  • The Partisan
  • 208 posts
  • LocationMaine

Posted 05 August 2015 - 04:25 AM

Haven't read through all responses yet, so I may be redundant...

I agree more variety would be great, but the dirty little not so secret you are overlooking is that PGI effectively uses map design as part of its "balance". Terrain and structures laced to allow for cover, etc. can you imagine the lurmaggedon that would occur with randomly generated maps with little cover? Relatively flat map with no buildings? Team with range advantage: GGCLOSE.

Procedurally generated terrain is easy. Structures and buildings within the terrain is harder. Impossible? No. I wish they'd move more in that direction to improve variety, but it would require some dedicated design and testing and probably a need for a different kind of weapons balancing.

#85 Artgathan

    Member

  • PipPipPipPipPipPipPipPip
  • Knight Errant
  • Knight Errant
  • 1,764 posts

Posted 05 August 2015 - 04:29 AM

PGI can't figure out the code to handle dynamic ammo switching. You want them to write a procedural map generation system? Good luck.

#86 Vandul

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,342 posts
  • LocationYork, New

Posted 05 August 2015 - 04:30 AM

View PostArtgathan, on 05 August 2015 - 04:29 AM, said:

PGI can't figure out the code to handle dynamic ammo switching. You want them to write a procedural map generation system? Good luck.

Well, theres that

#87 Lily from animove

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • The Devoted
  • The Devoted
  • 13,891 posts
  • LocationOn a dropship to Terra

Posted 05 August 2015 - 04:32 AM

It's great for sandbox games, but for shooters I doubt it.

Edited by Lily from animove, 05 August 2015 - 04:32 AM.


#88 Mechteric

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Overlord
  • Overlord
  • 7,308 posts
  • LocationRTP, NC

Posted 05 August 2015 - 04:49 AM

Procedural does not mean random, by the way. It just means they used a predictable set of numbers to come to the same outcome when it is rendered.

Of course if done right you can get an amazingly large world, but random, it isn't.

#89 TamerSA

    Member

  • PipPipPipPipPip
  • The Merciless
  • The Merciless
  • 144 posts
  • LocationSouth Africa

Posted 05 August 2015 - 05:42 AM

I think this is actually a very interesting concept on paper. I am a fan of procedural map generation and I think it solves a niche problem that some games like survival genre had.

However I do think it is a bit late to adapt MWO for this. The game was designed to be more of an arena style PvP game, with maps designed with tactics and combat hotspots in mind.

I could see this working for a single player MechWarrior title, provided it was built with procedural generation at the core. All things considered in this topic, it's still a nice idea to maul around conceptually.

#90 Elizander

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 7,540 posts
  • LocationPhilippines

Posted 05 August 2015 - 08:20 AM

View PostArtgathan, on 05 August 2015 - 04:29 AM, said:

PGI can't figure out the code to handle dynamic ammo switching. You want them to write a procedural map generation system? Good luck.


Hey, Pixel Dungeon code is open source. It's a start. :lol: :ph34r:

#91 GroovYChickeN

    Member

  • PipPipPipPipPipPip
  • Overlord
  • Overlord
  • 209 posts

Posted 05 August 2015 - 09:01 AM

View PostJohnny Z, on 04 August 2015 - 09:40 PM, said:

https://www.youtube....e&v=4WUG55EC4v0

This shows how nav mesh is done for competent ai so it doesnt get stuck on invisible walls, or walk through walls in other situations, or terrain and can use cover and other intelligent behavior.

This doesnt show the nav mesh auto generation or height map generation or cover generation etc... This guy skipped auto generation because it would have been tougher to correct the mess than start from scratch by hand.

Like I said above many games use zombies because they will just head directly at the player and thats it and can use fairly simple nav meshes and ai. See a zombie get stuck on walls? Thats a nav gen problem.

Other games may create npc or animals in an open area and they wont travel far and that would work in procedural map generation.

Some systems are more difficult than this to use. This is an extremely good system.

Fallout 4 has made improvements to the above system to allow for procedural map generation on a small scale in predisgnated areas for base building and they are extremely proud of that. So a player can place a wall and an npc wont walk into it not knowing it is there for instance.

The Crytek engine Mechwarrior Online is using will be similar to this and I know nothing about it but I think it is good.




So I'm not trying to be mean or anything but you keep talking about AI and I don't think that's a reasonable expectation. I don't think we can expect any amazing AI in this game other than what we see in any MOBA if any at all. I mean it would be sweet to have intelligent AI in this game for something like tanks or something but I just don't think it will happen. I mean look we don't even have knockdowns back in yet and still have instant convergence.

Also generating a Navmesh at runtime isn't super hard. Yes it can be tricky but it is very doable. In fact there are free utilities that will help with it. A* Pathfinding is a good example of an extension for Unity that can update/edit a navmesh in real time based off of mesh geometry and raytracing.

Procedural level generation is relatively easy at it's core. The difficult part is making sure you correctly handle every outcome of what ever generation method you are using. I think a better utilization of any procedural generation would be a system to build out prototypes that they then be edited and polished in house to speed up level design. This way we may actually have planets that are different from each other.

Edited by GroovYChickeN, 05 August 2015 - 10:39 AM.


#92 topgun505

    Member

  • PipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 1,625 posts
  • Facebook: Link
  • LocationOhio

Posted 05 August 2015 - 10:00 AM

I would kill for dynamically generated maps! And scouts would FINALLY be worth a flying hoot. Heck, they'd be worth their weight in gold.

#93 Mister Blastman

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Survivor
  • Survivor
  • 8,444 posts
  • LocationIn my Mech (Atlanta, GA)

Posted 05 August 2015 - 03:17 PM

View PostRaggedyman, on 05 August 2015 - 04:13 AM, said:

There is a massive difference between being able to generate a map that can be played on and being able to generate one that makes for a great player experience in a specific game, especially something like a multiplayer round based FPS.


The maps we have now are fun the first few times. The problem is the players develop a pattern--a nonsensical one where every single game, they must move to those spots and specific ways.

Random maps eliminate the "rat knows the maze so well he can find the cheese every time" issue we have. A good game is a refreshing one.

I'd take new maps all the time versus hand crafted ones every several years. The quantity of good games will far exceed the... "Camp around the Citadel, nascar vroom vroom kill slow guys in rear, k?"

This way players never develop a comfort zone of what rocks to hide behind.

#94 Johnny Z

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Knight Errant
  • 9,942 posts
  • LocationDueling on Solaris

Posted 05 August 2015 - 05:56 PM

Well Mechwarrior Online already is technically doing procedural maps anyway. Its not on a basic level but they are making map sets that will be used for each planet. So each planet will have its own map set.

So far it only looks like 2 maps for each map set representing one planet. Which planets will get which map set is going to be interesting. The regular queue maps being one of the two and the current galaxy maps being the other. They will likely add a third fairly soon or already working on the third for these map sets. It is interesting to guess how exactly the will go about it.

As they add antire new map sets representing entirely different planets then some of the planets on the galaxy map will take that new map set on.

There is definately a blue water map set coming some time also...

This is technically procedural map generation. :) Just on a detailed scale and work intensive. And also a massive scale in the long run...

This is the smart way to go if anyone bothers to think it through because then they can make quest like planets or maps, player bases potentially or even space stations or captial ships for mechwarrios to walk around on. Even space for aerotech some day...

These guys are pretty smart even if they dont look it haha. :) Just kidding.

Cheers.

Almost forgot, they are definately working on the planet Solaris map set in some way, without a doubt.

Edited by Johnny Z, 05 August 2015 - 06:02 PM.


#95 Raggedyman

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,278 posts
  • LocationFreedonia Institute of Mech Husbandry

Posted 06 August 2015 - 03:50 AM

View PostMister Blastman, on 05 August 2015 - 03:17 PM, said:

The maps we have now are fun the first few times. The problem is the players develop a pattern--a nonsensical one where every single game, they must move to those spots and specific ways.

Random maps eliminate the "rat knows the maze so well he can find the cheese every time" issue we have. A good game is a refreshing one.

I'd take new maps all the time versus hand crafted ones every several years. The quantity of good games will far exceed the... "Camp around the Citadel, nascar vroom vroom kill slow guys in rear, k?"

This way players never develop a comfort zone of what rocks to hide behind.


The thing is, that comfort zone/known map situation is actually a thing that quite a lot of people like in a game. If you take the Bartle model of player behavior (old, reliable, a bit imprecise these days but still useful for conversation) then the Achiever and Killer player types often like having a map that they can learn and get good at. And judging from the feedback that turns up on the forums for the first week or so everytime a new map is out ("it sucks! because I don't know it so got killed!!") I'm happy to suggest that the Explorer type of player isn't that prevalent in these parts, at least not when it comes to playing fields.

Don't get me wrong, I personally like PGC and would love to see it being used in the game. However that is very different to thinking it would go down all that well or that the cost of developing true PGC (as opposed to "randomised preset/block" maps, though even they would get a lot of flak imo) would be worth the funds raised from it as (at this current time) MWO isn't the kind of game (ie it's a round based combat game, rather than a single world, persistent explorer game) that would benefit from it.

Short version: Full/"True" PGC (like any other game content method) is great if used in the right kind of game and awful if used in the wrong one.

View Posttopgun505, on 05 August 2015 - 10:00 AM, said:

I would kill for dynamically generated maps! And scouts would FINALLY be worth a flying hoot. Heck, they'd be worth their weight in gold.


Only if the map was scoutable. Remember that if it's a generated map you can often get bad ones that don't provide the game experience you are after. Unless you are willing to accept frustration as part of the game play (some people can, so I'm not saying you can't) then randomisation can be really upsetting when you find out your toon is totally unsuitable for the round.

Edited by Raggedyman, 06 August 2015 - 04:13 AM.


#96 Raggedyman

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,278 posts
  • LocationFreedonia Institute of Mech Husbandry

Posted 06 August 2015 - 03:53 AM

View PostJohnny Z, on 05 August 2015 - 05:56 PM, said:

Well Mechwarrior Online already is technically doing procedural maps anyway. Its not on a basic level but they are making map sets that will be used for each planet. So each planet will have its own map set.


Question: do you mean randomly generated maps for that planet or each planet getting a random sequences from a set selection of prebuilt maps?

Either is cool, counts at Procedural Generated Content, and adds to the fun IMO, I just wanted to be clear on what you meant before I get all excited.

#97 Rhaythe

    Member

  • PipPipPipPipPipPipPipPipPip
  • Philanthropist
  • Philanthropist
  • 4,203 posts

Posted 06 August 2015 - 03:54 AM

Though procedural maps are cool, I think the real problem with MWO is the severe lack of unique game modes. I've said it before, and it's worth repeating. Hawken, a "dumber" mech game than MWO, has far better game modes than this one does.

#98 Raggedyman

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,278 posts
  • LocationFreedonia Institute of Mech Husbandry

Posted 06 August 2015 - 03:59 AM

View PostRhaythe, on 06 August 2015 - 03:54 AM, said:

Though procedural maps are cool, I think the real problem with MWO is the severe lack of unique game modes. I've said it before, and it's worth repeating. Hawken, a "dumber" mech game than MWO, has far better game modes than this one does.


The number of game modes has more to do with the size of the player base and the need to get people into games at an acceptable speed whilst maintaining balance than anything else. Each time you add in an extra game mode, or produce a game mode so unique that a significant portion of the player base will opt out of it, you start getting problems.

It's a tricky balancing act though, as a lack of game modes can restrict the amount of players you get into a game

#99 Rhaythe

    Member

  • PipPipPipPipPipPipPipPipPip
  • Philanthropist
  • Philanthropist
  • 4,203 posts

Posted 06 August 2015 - 04:04 AM

View PostRaggedyman, on 06 August 2015 - 03:59 AM, said:


The number of game modes has more to do with the size of the player base and the need to get people into games at an acceptable speed whilst maintaining balance than anything else. Each time you add in an extra game mode, or produce a game mode so unique that a significant portion of the player base will opt out of it, you start getting problems.

It's a tricky balancing act though, as a lack of game modes can restrict the amount of players you get into a game

Unless you're sporting around 3 game modes that are incredibly similar to begin with. These modes are stale. They involve shooting red doritos, and really, little else. This is the real reason the gameplay feels old, even after a new influx of maps. Maps make it feel new briefly, but it's not a lasting thing.

This game would be much more "fresh" if it scrapped two of its three public queue modes and created two brand-new ones that had far more depth than "hunt red".

But that's all, like, my opinion.

#100 Raggedyman

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,278 posts
  • LocationFreedonia Institute of Mech Husbandry

Posted 06 August 2015 - 04:14 AM

View PostRhaythe, on 06 August 2015 - 04:04 AM, said:

Unless you're sporting around 3 game modes that are incredibly similar to begin with. These modes are stale. They involve shooting red doritos, and really, little else. This is the real reason the gameplay feels old, even after a new influx of maps. Maps make it feel new briefly, but it's not a lasting thing.

This game would be much more "fresh" if it scrapped two of its three public queue modes and created two brand-new ones that had far more depth than "hunt red".

But that's all, like, my opinion.


Agreed that maps add a refresh to game modes, and that if the mode is stale then maps themselves won't provide a long term fix.

Out of curiosity: what two modes would you remove and what would you replace them with?





41 user(s) are reading this topic

0 members, 41 guests, 0 anonymous users