[Suggestion] Random Symmetrical Maps for Ranked Games

On the point of having random maps in a framework of set rules for ranked matches I agree 100%.

On the technical side I can't help much. Imo something like the "Age of Empires 2 Random Map Script" would be cool. With the help of these scripts it allows to go generate random maps with a high level of customizing conditions (terrain features, resources, starting positions etc). Depending on your settings the results are either almost strict symmetrical or totally random worlds. (I'm aware that it's quite possible to be way too much work setting something similar up for PA, but it's one of my favorite examples regarding map generation)

For anybody being interested: http://steamcommunity.com/sharedfiles/filedetails/?id=155256742

Sometimes there is a reason why something is like it is and in this case there are good reasons. Moving "on" would actually just equal to throwing out what we've learned over that past decade.
We've all talked about this before, I think I made my stance clear on this: It's a no go. Playing maps you know actually results in better gameplay. You might as well argue to randomize the balancing for every game.
It's
a) technically not feasible to write an algorithm to generate a good balancing with interesting gameplay and your pseudo code tbh makes me think you are extremely underestimating this issue.
b) pretty much a dice role to play a game whose rules you don't know.

b is pretty much impossible to fix without modding the human brain.
a in theory can be done, but I am doubtful. You are free to actually implement such a map editor, with the new system editor you in fact CAN write your own RNG planet generator as a simple mod.

You can theorize about "what if we can do a", but even that stays theory.

I recommend reading through this discussions here:
https://forums.uberent.com/threads/new-pte-build-up.65586/page-6#post-1027087
https://forums.uberent.com/threads/symmetrical-planets-are-getting-there.65316/page-2
https://forums.uberent.com/threads/randomness-in-early-game.55996/page-3#post-860613

I'd agree to add a single "map" that is a random symmetric map. Just to show people how stupid it will feel compared to other maps.

In the end taking away specialized maps with exactly planned out different gameplay elements just takes away a huge part of competitive game, merging all games into one "let's do the standard build order" type of game, as PA is not competitive enough (to do more you'd have to work on it full time) to learn more than that.

Why is memorization a huge part of the competitive game? That isn't exciting. It's like watching two people quote pi at eachother until one of them misses a number. It can be fun for a little while, but it quickly gets repetitive.

I'm not sure you get it - not knowing what is coming next is a huge part of gameplay in PA. Because you don't know, there is no "best" strategy - it's just how well you can do with what you got. And if you have symmetrical maps, both players got the same thing, and thus covers the *fairness* condition for competitive games.

But we would still keep the innovation and creativity of the average player. I don't know about you, but I love watching other people's games who are better/worse than me and see what their strats are in maps they've never seen before. I used to love watching your games, cola, because your strategy was ridiculously dynamic and risky. It was exciting to watch the game unfold. Nowadays, watching ranked replays is like watching a different version of me making slightly different moves on the same map. It isn't exciting, really.

When you say PA is not competitive enough - I'm not sure I understand what you mean. Could you explain a bit more?

Because "strategy" in the end is nothing more than creating a plan to manage a certain situation. They most effective way to do that is by being able to relate the current situation with past experience, hence memorization.
The reason why quoting PI is boring is not because of the memorization thing but because it is a rather one dimensional thing, playing PA is just much more complex. But the "perfect" player basically has played all possible games already and can reference to them within less than a second to make perfect decisions.
That's btw how the human brain in general works and why people to stuff like repeat the same actions over and over again as "practice".

This is btw the core of why I dislike the argument to make the game more "strategic" as a counter argument to things like repetition and "micro".
Strategy is ALL ABOUT memorization of all possibilities. Thank the setup of your brain for that. Or maybe even the setup of the whole universe.
Just because you don't realize that your brain is working in this way doesn't mean that it is any different and a person who realizes that their brain is working this way and effectively works towards it will very likely become a better player.

I don't know what my opponent is gonna do, that's more than enough. No need for not knowing how to play the game.

That's a balance and a "only 3 small maps with very similar setup" issue.
Ideally you'd have a mappool with i.e.
2 small rush planets, one low metal, one high metal
3 medium sized planets, with one land war planet, one navy planet, one orbital fights system.
2 big planets, again with considerably different setup to "force" i.e. orbital play or extreme expansion style play, etc.

I was thinking of the old GPGnet mappool which had 16 very different maps and searched for the faf mappool, all I could find of the current FAF mappool is this: http://wiki.faforever.com/mediawiki/index.php?title=Ladder_Map_Pool
Seems somebody thought having a lot of ranked maps for variety is good thing there lol.
Anyway I can recognize quite a lot of those maps just from name and I can tell you there are VASTLY different maps in there. Basically playing one map will yield a completely different game than some other map.

Basically if you introduce random maps like that you just go: "okay to play perfectly a player has to memorize all possible maps and gameplay setups". Since the generator is supposed to generate only super awesome maps (let's assume it can) it will be very likely that with enough work put into it you could work out a set of different strategy and builds for every map type the generator can generate. A lot of work. But if PA would be played like i.e. chess is you would actually see people do this just to win.
PA however only is small niche RTS, the players are hobbiest that at most maybe win a few hundred $ each year by winning tournaments.

I think adaptation shows more skill than map memorization which is more technical application.

I'd say yes, to symetrical maps in ranked. Must be:
-symetrical map
-symetrical metal
-symetrical spawns
-200 size spawns to give large spawn variety and differention. Rather 1 200-300 size spawn than 2 reg. Commander walk distances a real imbalance.
-300-500 size. 1v1 is ideal like that.
-optional: Dual planet spawn, symetrical planets and planets identical. Gas giant. 100% water. Metal planet devoid of metal spots.
-each player gets a map veto. In case they prefer single or double or naval or lava or whatever.

now, a variety map pool would help short term. But "good" procedural generation will keep people playing longer. I like shooters on familiar maps, but I am one of few that can grind on 5 maps with no unlockable metagame, just see RenX playerbase for reference. But **** everyone, I could play walls and field 8 hours straight, have before. I can also play Pavlov on CoD classic in 15 minute free for alls about 4 hrs a day 5 days a week. But not everyone's a creature of habit and the game would eventually pay for it.

I had to fix something for you

300 is very small. 450 is double 300 no? For that matter, 450 becomes too big for t1 radar which still needs nerfed to 550 radius so its less than half planet. 600 is tops.

1000 is your personal preference and in fact is over 5x the surface area of 300? The maps should be within 1 magnitude of the median. With half size or twice size. No map should be 4x or more another.

they had much luck so far in GW custom procedural layouts. Gw layouts are in fact perfect for ranked, if symetrical was forced on all generated planets, metal was between 30 and 70, and it was capable of randomly being 1 planet, 2 planet, and/or include a gas giant or metal planet without metal.

you give their current generation too little credit when its excellent given tweaks like symmetry.

the current biggest ranked map is 500. It is a crazy rush fest for the most part with only "lets take a few mex" expansion type of play.
600 is not tops, 600 is like the lower range of what a medium planet is. 1000 sure is extreme, but definitely more playable than 300. twice the size of the median? Well how about having a 300 a 600 and a 1200
Seriously I don't understand why anybody would argue for 500 sized ranked maps as a 1vs1 standard. That basically is loudly asking for repetitive single base build order rush fests.
All really epic games of the past have been played on 600 and up. Avg probably around 750?

If I had designed the current mappool than there would be one single 600 planet added with the thought "well let's have a small one for people who like rush play" xD

you can't rush play on a 600. A 400 can be smallest. That leaves a 600. And then a 900. Now Colin, I know you know maths, so the playable surface area wise, 1000 is double 667 no? Surface area is VERY touchy at high numbers. Also, low end machines suffer at 800, mine does, I can play but 1200 takes up just about the max. Sadly, I can load 4 350s.

bias aside, I can see a 1000-1200 as viable as 100-200 size planets. Yes, that's combox, may well combox than play an rts game of "Go" with air and teleporters.

Can't balance the game around ... erm what kind of machine suffers at a single 800?

min specs brah. 4g ram. Mine has 8, but 4 ram is faster than other 4, and at an 800 or more than 2 300s, it goes from less than minute to 3 minute load time.

again, you know math. If 800 strains, 1200 is double. 1600 is double, I couldn't even generate. 2400 is double, a good comp sometimes can't generate. Which is why the very high end is only capable of over 2400.

it opens up just little more access to use 400-600-900. 1000 is almost another combox surface area larger, can't you barter at the halfway?

I am not asking for 1600 or 1200

well again, a bargain would be no 300s in exchange for no 1000s. So 400-900 it is.

I am fine with a single 300 for the lulz if we have at least 8 ranked maps.

with a map pool, I am more ok with a 300 than a 1000. If forced to choose both or neither I'd neither.

with a map pool, I'd also like a 400, 600, 900, all land. Then a dual 400 with gas giant, a dual 400 with small metal planet no metal spots, a dual 400 metal planets around a gas giant, a water only single 550, a 900 with small islands in lava, and a 4 planet collision system where each longer living planet gets half size and third metal (200 no metal, 300 14 metal 35min, 450 50 metal 20 min, 550 120 metal 7 min)