Jump to content

Pseudo-random Tactical Map generation for PUG games


17 replies to this topic

#1 IxxxI

    Member

  • PipPipPipPipPipPip
  • 360 posts

Posted 02 February 2012 - 05:56 AM

There are lot of things bothered me in different MMO games, but one of them - static maps and spawn positions - is a very common sin among PvP oriented games, leading to plain step-by-step best-known-tactics, which made lot of PUG fights too unvaried (lot of team fights as well, TBH).
One solution: play in team, preferrably the best on server B), that way is also very rigid from the perspective of tactical gaming (let me skip examples, I belive of you have similar expirience).
Second solution: dinamic tactical environment resulting in different conditions for every match. My proposal is: create large detailed map for each planet/location available and randomly select warfare area for every match, giving the example: we have large scale map for campaign location and can select variants 1-4 for particular matches, note, that teams spawn points should change coordinates every time:

Posted Image

*Red circle is just a possible solution to save some level designers' time. Number of map variants is finite of course, but it's better than same map for every location every time, quiaff?

Variety of battle's tactical environment could be affected by weather/day/night/et.c. conditions, providing players with almost unique gaming expirience. The idea could be used not only for PUG games, but for MercCrops/Houses campaigns as well.

In some cases, like historical battles or siege operations, tactical environment should be more defined: fixed starting positions, known weather conditions an so on.


I think it's possible to amplify the idea, but I stop by now to save you from reading too large wall of text :P I just hope, that it might be useful for developers or reflects future game mechanics in some approximation.



P.S. Sorry for my poor English.
P.P.S. Original map taken almost randomly from Google Earth, all rights belong to respective owner.

Edited by IxxxI, 03 February 2012 - 08:06 AM.


#2 Axxon

    Member

  • PipPip
  • 35 posts

Posted 02 February 2012 - 07:02 AM

yea.. -changing- is a key to build a "not to be boring" game

anything that packs some additional action is a fine bonus.

Maybe the "overlapping" maps can mean that another lance is fighting another battle/mission somewhere near, the same planet, and if they chose their operation time to when we are still in the operation area (OURS) they might show up...? This could stir things up, because, if they are friendlies... they can be helped out. If enemies, the houses can settle scores on an older enemy by an "its a small world" surprise!

One more... you gonna need dozens of people to constanty invent new missions as YOUR battletech universe evolves. new battles, new events. This is what makes a good game. STANDSTILL GAMES ARE DEAD. dead-dead - dead!!

#3 IxxxI

    Member

  • PipPipPipPipPipPip
  • 360 posts

Posted 03 February 2012 - 08:09 AM

I wish someone from the staff said few words on the topic. Am I asking too much :D?

#4 Fiachdubh

    Member

  • PipPipPipPipPipPipPip
  • 971 posts
  • LocationSkulking out along the Periphery somewhere.

Posted 03 February 2012 - 08:17 AM

Like it. Would also be useful if Aerospace ever became playable. Similar to Battlefield 2 while the ground units are confined to a certain area of the map those in aircraft can have a larger area available to manoeuvre in.

#5 TheRulesLawyer

    Member

  • PipPipPipPipPipPipPipPip
  • 1,415 posts
  • LocationChicagoland

Posted 03 February 2012 - 08:27 AM

I like the idea of changing maps, but the big question is how are you going to keep them balanced? Random terrain means that you'll have a significant percentage of battles that favor one side right from the start. It takes a human's touch and a lot of tweaking to map a balanced and fun map.

Edited by TheRulesLawyer, 03 February 2012 - 08:28 AM.


#6 IxxxI

    Member

  • PipPipPipPipPipPip
  • 360 posts

Posted 03 February 2012 - 08:38 AM

View PostTheRulesLawyer, on 03 February 2012 - 08:27 AM, said:

I like the idea of changing maps, but the big question is how are you going to keep them balanced? Random terrain means that you'll have a significant percentage of battles that favor one side right from the start. It takes a human's touch and a lot of tweaking to map a balanced and fun map.


That's right, but it's not a question of map balancing, but a question of maps testing. It'll take much more time to test 'large area map' for possible collisions/dead-ends and so on. It's a doable task from the math point of view, I mean there exists class of algorithms applicable for such tasks. The real question is does Devs have will, time and resources for that or not. It may be, that they have code freeze already and just debug the game.

Edited by IxxxI, 03 February 2012 - 08:39 AM.


#7 Dlardrageth

    Member

  • PipPipPipPipPipPipPipPip
  • 1,198 posts
  • LocationF.R.G.

Posted 03 February 2012 - 09:02 AM

You could just use a set of fixed "tiles" from which to construct every map randomly. Would take a bit of toying around initially, but once done, even a map consisting of only 12 tiles would give a whole lot of different map designs. Multiply that by different terrain types (desert, jungle, urban for example) and you got a whole lot of variety. Yes, stilla slight element of repetition there with tiles, but once you got teh parameters set for those, easier to introduce new ones later on gradually.

The TT and MegaMek veterans will remember how you used to (re-)position different map sheets to get a different battlefield. Now do the same with 3D tiles and you get the idea of what I'm talking about. Pretty straightforward model that is easy to expand upon.

#8 TheRulesLawyer

    Member

  • PipPipPipPipPipPipPipPip
  • 1,415 posts
  • LocationChicagoland

Posted 03 February 2012 - 09:16 AM

View PostIxxxI, on 03 February 2012 - 08:38 AM, said:


That's right, but it's not a question of map balancing, but a question of maps testing. It'll take much more time to test 'large area map' for possible collisions/dead-ends and so on. It's a doable task from the math point of view, I mean there exists class of algorithms applicable for such tasks. The real question is does Devs have will, time and resources for that or not. It may be, that they have code freeze already and just debug the game.


No, its not a matter of just checking for pathing. Creating a balanced map is different than just creating one that works. Even on a fixed map moving a spawn point 100m could have a large impact. Not that its an impossible problem, but you're talking about a fairly sophisticated special purpose AI.

#9 Fluffinator

    Member

  • PipPipPipPipPip
  • The Grizzly
  • The Grizzly
  • 132 posts
  • LocationKY

Posted 16 February 2012 - 01:06 AM

View PostTheRulesLawyer, on 03 February 2012 - 08:27 AM, said:

I like the idea of changing maps, but the big question is how are you going to keep them balanced?

Random is the purest form of balanced. Its balanced over time. Trying to perfectly balance a single fight with set spawn points that don't ever reverse will never be truely balanced unless its an idential mirror.

#10 Bluey

    Member

  • PipPipPipPipPipPipPip
  • 948 posts
  • LocationAnatolia

Posted 16 February 2012 - 02:47 AM

At the times of Mech Commander Gold I was making custom maps for that game and I can say this it was fun much more fun than having random generated maps.I really mean random is bad..... it kills how atmosphere feels.

If I attack Eden Prime I would like to see same Eden Prime I defended once but under different climate.*Snow* Similar layout and city center.
Now people gonna ask how can developers will built so much planets without randomized maps.Answer is Map editor if developers agree players can create maps give those to devs to consider which ones are suitable for usage and later let those maps used in game.

Thats what make starcraft 2 so popular and same reason how can a 3 faction 30 planet game have so many maps.
I agree on OPs suggestions all the way.

Side note maps shouldnt be 100 percent balanced its damn warfield you never can have perfect battle :)

Edited by Bluey, 16 February 2012 - 02:48 AM.


#11 Zyllos

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,818 posts

Posted 16 February 2012 - 07:19 AM

Nice idea, in concept and theory.

Of course, you would have to also procedurally make sure edges and corners of the boundry selected doesn't cause players to become stuck. Could perform a few calculations and determine common areas of all maps seeing how 2 corners of a generated map makes a chord. With this information, you can make sure this common area is open for spawning within the map.

#12 Exilyth

    Member

  • PipPipPipPipPipPipPipPipPip
  • Bridesmaid
  • 2,100 posts
  • LocationTerra

Posted 16 February 2012 - 08:59 AM

I like OPs idea.

If you made one big static map and spread more spawnpoints than needed evenly over it....
You could use any subsection of that map as the 'actual' map.
Define teams spawn zones within that area, then choose a random spawn point from the teams spawn zone for each player from the team.
Then, the only issue would be balance.

View PostBluey, on 16 February 2012 - 02:47 AM, said:

Answer is Map editor if developers agree players can create maps give those to devs to consider which ones are suitable for usage and later let those maps used in game.


Somehow, this made me think of trick jump maps in other games...
Would be nice if unofficial maps could still be used, but only in training/modes that don't award any XP/c-bills.

#13 TheRulesLawyer

    Member

  • PipPipPipPipPipPipPipPip
  • 1,415 posts
  • LocationChicagoland

Posted 16 February 2012 - 11:44 AM

View PostFluffinator, on 16 February 2012 - 01:06 AM, said:

Random is the purest form of balanced. Its balanced over time. Trying to perfectly balance a single fight with set spawn points that don't ever reverse will never be truely balanced unless its an idential mirror.


You can have balance without having to mirror terrain. I give you that mirroring terrain is by far the simplest way to balance something. Sure completely random is balanced over a very long time. However it also has far far far more outlyers. A normal human balanced map should have a normal distribution curve if you plot some sort of quality of win metric (mechs dead/time to win/etc) vs number. Completely random maps would be completely flat. They would have just about as many shocking wins and losses as close games. Normally games that are close are more fun for most people. The normal distribution curve puts far more games into the "fun" zone even if the average is the same.

#14 Naduk

    Member

  • PipPipPipPipPipPipPipPip
  • Overlord
  • Overlord
  • 1,575 posts
  • LocationAustralia

Posted 17 February 2012 - 07:21 AM

i have had much the same i idea as the OP
i agree 100%

with the game having such a huge emphasis on information warfare
it makes little sense to be fighting over the same chunks of terrain (maps) day in and day out
in a very short amount of time we all learn the most common spots for common actions
so scouts jobs cease to be about finding and tracking the enemy seeing what devious plan they are up to
and instead becomes a simple fact finding mission, are they doing strategy A,B or C
this type of play is so mind numbingly boring
if the game goes down the old school path of static maps it will be alot harder to maintain long term interest

battlefield 3 is currently suffering from this very plight
the game is awesome, combat enthralling and exciting
but the same maps for hours and hours on end
oh , my , god i am so sick to death of wake island

DICE bragged non-stop about the power of their new engine
and its very powerful indeed
but they waste all that potential with stale and static environments

somebody tell me what down side there is to settings like
Random weather
Random time of day
live day night cycle (from random start point)
adaptive map locations (like op's suggestion)

#15 MaddMaxx

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Survivor
  • Survivor
  • 5,911 posts
  • LocationNova Scotia, Canada

Posted 17 February 2012 - 10:23 AM

View PostNaduk, on 17 February 2012 - 07:21 AM, said:


somebody tell me what down side there is to settings like


Random weather - none.

Random time of day - none (day and night fights for sure)

live day night cycle (from random start point) - Match Length (Drop at dusk, match ends before or just at dark. Drop in Dark, still Dark when over)

adaptive map locations (like op's suggestion) - none

And yes, Matches could absolutely be extended by the players on Map, but, will game play warrant it happening enough to place the addition of adaptive day to night cycles as a required additional Development item?

Edited by MaddMaxx, 17 February 2012 - 10:28 AM.


#16 Naduk

    Member

  • PipPipPipPipPipPipPipPip
  • Overlord
  • Overlord
  • 1,575 posts
  • LocationAustralia

Posted 22 February 2012 - 09:21 PM

View PostMaddMaxx, on 17 February 2012 - 10:23 AM, said:

And yes, Matches could absolutely be extended by the players on Map, but, will game play warrant it happening enough to place the addition of adaptive day to night cycles as a required additional Development item?


firstly yes, its warranted especially if you have already stated your engine can do it (like in the BF3 case)

secondly
why bother with vision modes like low light and night vision if you only ever need them on 1-2 maps from your entire map cycle
in games like BF and hopefully MWO there are advantages and disadvantages to using/equipping items like Night vision
ie, its useless in the day time, but very powerful at night...ect
players that forgo items like night vision in exchange for more power should be worried if the game goes for to long as they wind up in a night fight

the game should take as much advantage of these play elements as possible, not just restrict them to a few maps
where in the end it doesn't change the game at all, because you know your going into that map and equip accordingly as everyone else

surely this is a much more fun and engaging experience to be battling away and realize its now night time and you need to seriously change your tactics and have you really been fighting for that long ? wow

so im not saying they should go to extremes to add such features (altho with the free to play development system, why not at some point)
i am most certainly saying , don't waste the power of your engine
if it can do something , bloody make use of it , don't show off or develop features that you wont ever use

case in point, day night cycles on frostbyte2 , why develop the feature and not use it .... just stupid

#17 MaddMaxx

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Survivor
  • Survivor
  • 5,911 posts
  • LocationNova Scotia, Canada

Posted 23 February 2012 - 10:45 AM

View PostNaduk, on 22 February 2012 - 09:21 PM, said:


firstly yes, its warranted especially if you have already stated your engine can do it (like in the BF3 case)

secondly
why bother with vision modes like low light and night vision if you only ever need them on 1-2 maps from your entire map cycle
in games like BF and hopefully MWO there are advantages and disadvantages to using/equipping items like Night vision
ie, its useless in the day time, but very powerful at night...ect
players that forgo items like night vision in exchange for more power should be worried if the game goes for to long as they wind up in a night fight

the game should take as much advantage of these play elements as possible, not just restrict them to a few maps
where in the end it doesn't change the game at all, because you know your going into that map and equip accordingly as everyone else

surely this is a much more fun and engaging experience to be battling away and realize its now night time and you need to seriously change your tactics and have you really been fighting for that long ? wow

so im not saying they should go to extremes to add such features (altho with the free to play development system, why not at some point)
i am most certainly saying , don't waste the power of your engine
if it can do something , bloody make use of it , don't show off or develop features that you wont ever use

case in point, day night cycles on frostbyte2 , why develop the feature and not use it .... just stupid


I do not disagree, nor did I initially. I was merely pointing out that the Dev have stated that a Match will last (depending on testing) on average, 20 minutes. Barring arriving on a Planet at Dusk, or that the planets Sun is known to literally drop from the sky in 5 minutes after we arrive, how does it become consistent enough to warrant it?

Now, let's give a Mech the requires Module to Enhance its Night vision (before the drop right, no way after), with that Module now installed, instead of some other, do I then get assurances it will be Night at drop time or at some point in during that Match? I have limited Module slots...

How the hell are we supposed to know, unless it is the same all the time on all Day or Night Maps? The Thread topic states Pseudo-Random.

I think it is a great idea. Tell me how folks don't apply Modules that take up space and will be useful on the OFF chance conditions will change if Matches don't last for extended periods? Please.

Did PGI make BF3? (connection?)

Edited by MaddMaxx, 23 February 2012 - 10:47 AM.


#18 Naduk

    Member

  • PipPipPipPipPipPipPipPip
  • Overlord
  • Overlord
  • 1,575 posts
  • LocationAustralia

Posted 24 February 2012 - 07:20 PM

no PGI connection to BF3, only that Developers should learn from the mistakes other Devs make

here would be how i imagine it going down
missions are accepted via contract (confirmed for mercs at least, others im not sure)
contracts would have all sorts of details
-planet
-mission time allowance
-mission time start (day/night ect)
-average ambient temperature
-estimated enemy presence and threat assessments
-field Intel report (if available or purchased)
-satellite scan
-objective list and locations
-extraction location(s) (choose-able by drop commander perhaps?)
-drop weight
-salvage rights
-costs/fees/profit estimates

so with the information presented before you even accept(bid on) the contract
you have plenty of opportunity's to plan the drop
considerations need to be made
say drop time is 6pm, based on intel and threat assessment do you think you can complete the mission before night vision is necessary
or do you not risk it and ask all pilots to equip it anyway
or perhaps drop time is 3am, you would know that night vision is going to be mandatory for the drop

so there are only 24hrs per day and there will be thousands of thousands of drops per day
so it wont be every game, but dusk/dawn missions will be happening all the time

i would say its worth the time for PGI
even with out a active day/night cycle, a randomly selected time of day per mission would be glorious
it could even be pseudo random, by skewing the selection to more day time missions





5 user(s) are reading this topic

0 members, 5 guests, 0 anonymous users