Making the game competitive. Current and future problems.

I think something else that has to be considered if eSports were to happen is game lengths. Current RTS games (basically SC2) average somewhere around 15-20ish minutes per game, where as a game of PA that's epic enough to go orbital is basically guaranteed to go longer than that. More importantly, when epic games of SC2 do happen, it's a pretty rare sight, and a game of over an hour becomes exciting purely because it's probably breaking some sort of record (remember SC2 clock runs fast as well). A game of PA is not only more likely to become long, but once it goes long, it's going to to go really long, ie. once commanders are in orbit they're not that easy to find an kill.

I'm not sure if the game settings in the lobby do anything, but it might be worth playing with them and seeing if the SC2 trick of just speeding up the clock helps us (that's not uberent's job), but I think gamelength in general is worth some thought.

The good news is games like LoL regularly take at least 30 mins, but that's still on the low end of PA.

30 minutes is actually the average game length in PA, based in data from PA Stats:
http://www.nanodesu.info/pastats/stats
PA Stats cuts away game that are shorter than 5 minutes, too. So the average is not influenced by weird 3 minute games.

Looking at the average game time of 1v1 players shows even lower values:
http://www.nanodesu.info/pastats/player?player=412
http://www.nanodesu.info/pastats/player?player=410
http://www.nanodesu.info/pastats/player?player=420

Yes ofc most 1v1 players do not use orbital. But I am pretty sure that will only change once orbital is more like tier 1,5 which kinda will make it quicker to use and result in shorter games that have orbitals.

What's proper clan support? Like a group system in the game?

how do you join a clan???
please reply

I think that's sortof where my (admitted) postulating about game length comes from. I don't really think a PA where major part of the game are not accessed (orbitals and moon smashing, although the latter should be rarer) is a game that's been fully balanced (therefore I expected games to lengthen eventually). However, if orbital is going to be brought more into the earlier game, then 30 mins definitely sounds fine for game length.

I think we all agree that orbital has to be accessible before a normal competitive game is over and without having crazy long games.

EDIT:

PA has no build in clan system (yet). So I'd recommend you have a look at the clan forums and directly contact a clan you might like.

Not sure if i agree. I like that the game has two fases and that the orbital not always need to be a part of the game or that the orbital game is a late game/face two. But the orbital need some tweaking so we dont end up in those loong boring huntings.

EDIT: ok, since you say accesible maybe I agree anyway

For this game to take off and be an E-Sport then tools for observers and casters are a must. If it isnt friendly for an audience to watch then people just wont watch it.

If I may offer a voice of dissent; I think PA's biggest, unique draw is that it favors long games.

In an age of bite-size, fast-and-furious everything, it's a great boon to encounter a game that does what the RTS of old did; long-term planning, gargantuan battles, and generally being too big to handle easily. Speeding up or shrinking down the game to emulate the current popular e-sports would be a waste of potential.

Why not look at how to use the long format to PA's advantage? As far as sports comparisons go, I would favor a football approach; you don't have to stay for every single play, but it's not boring if you choose to, and the big plays are worth sticking around for. Long formats are also better suited to higher stakes; why not implement Galactic War or something like it as a tournament format, wherein bracket competitors start with smaller matches that get bigger as they go farther up. Small planet battles are perfect bite-size pieces to get viewers hooked, and incrementally upping the ante until you reach a massive interplanetary mega-battle. It'd be totally unlike most esports, where the finals are likely to be just as short as the normal games, and thereby relatively disappointing.

Not to be rude but this thread shouldn't have been made. You are asking if this game that is currently UNDER DEVELOPMENT is ready for competitive play in its BETA STATE. Which of course it is not.

One thing that hasnt been mentioned in terms of competitive play are mods.

Not the cosmetic ones but those who add functionality to the game are my primary concern as they very well can turn the game into pay 2 win.
Ive read multiple times that mods will be able to simulate user input and you most likely will be able to export rendered frames to them. Now its just a matter of time until someone comes up with mods that build a base for you, automacigally expand mexes, react on raids/attacks, keep (anti)nukes up and adjust the buildpower invested there to your economy status,....
Basically mods will do a huge lot of things that the player that uses the stock game will have to do by hand. So the moduser gains an extreme advantage as he can do more things in parallel - or get some done for him - and has to bother with much less "micro".

This is already beyond the point where Id want to play it if these mods are free because the game becomes more and more Scripted Annihilation but it wont stop here.
The finest tools will be available for money, and thats nothing more than getting a more or less huge advantage with just spending some bucks. No skill/training/... involved.

To date I havent seen any form of discussion about that and certainly no way of how to solve it.
One cant simply forbid all mods as it would invalidate the games strongest point - mods.
So there has to be a whitelist of whats allowed, but how to enforce it? Will PA feature a security subsytem like VAC, Warden,.... whatever?

I do not think that is actually that big of a problem. It would be a problem in a game like SC2 that is based on the perfect repetition of the same actions over and over again (I don't think bad of that, I like SC2).
PA however does aim to offer a game that is played having the best user interface possible, which can only really be done via a lot of modding.
So yes there will (and there are already) some mods that make it easier to play the game.
The danger of an actual "scripted annihilation" however is VERY rare. Developing a mod that actually "thinks" for the player is extremely difficult. I don't see any kind of AI replacing a human player any time soon.
Only annoying repetitive tasks will be replaced and that is fine.

Of course a player using the stock UI will be at a disadvantage, but mods are free (I don't think we will reach a stage were UI mods cost money) and if a player actually wants to be competitive I think it is fair that he has to think about how to get the best best user interface as well.

I don't like the idea of allowing mods anywhere near competitive play. WoW's proprietary "macro" command system made a wreck of the game, and user mods did even more damage to it. Diluting the essential, unified user experience by splitting the player base between mods and mod-nots is a really, really, really bad idea.

Only thing better than that would be to actually build a UI that can be used competitively and customization remains just for comfort/visuals and not function but free mods that everyone can download arent that big of a deal regarding BASIC UI functions.

Not replacing, supporting.
An "AI" that autoexpands mexes and keeps them up isnt that difficult to do.
Balancing nukeassist to your eco isnt either.

Think of it, you place a nuke launcher and position 20 fabbers around it. Its no big deal for image analysis to identify those 20 fabbers. The program just looks after economy state and commands each fabber to either stop or resume assisting. Theres no moving involved, no pathfinding and its not even Intelligent, just Artificial. Not sure if it even needs image analysis for this one.
Same can be done for factories.
Raiding unguarded mexes with a small botgroup, still no big deal as it just has to recognize their tac icons.
If theres a stray missile tower in there, it can be found by image analysis too. And the program perfectly manouvers the bots around its firing range.

Of course a human player will easily outsmart that but its enought that he HAS to do that whilst his opponent doesnt have to care about it and the loss of 5-10 dox is (most times) no game changing drama if the script messes up. And it can do that on and on without attention. On multiple parts of the planet and even different planets simultaneously. Still not OP?

Autoscouting, easy as cake. Every xx second the script takes one/two/three.. fighters/scout planes and flies over a given territory with slightly different pathes each run. Another thing the scripter just doesnt have to worry about.

No it doesn't. It should have equal balance between both. And because you can manually select map size/complexity, the game is well balanced between a lunch hour game and a 480 hour epic.

For the record, the existing UI is substantially better than SupComs.

First of image analysis is completely off. That never will be in. Ever. Unless somebody does some sort of scientific work on it, which I doubt will happen. It is a much too complex task to do within a mod and if everything goes well it also will not be necessary. The game should provide an API for that.

Mods like that exist in FA, I have never seen anybody really use them to be a "imba" though.

This might be a thing. Even thought I have doubts anybody will implement such a thing quickly and I do think a good player will be able to set up protection for his mex in a way that is too much for the mod.

Doesn't help you much if you are not looking at it. If you do look at it I don't see the problem.

Hello, some feedback from me, along the lines of improvement suggestions:

overview of units: icons on the screen, to see what buildings and how many are on this/each planet. the same with units, to see how many of them (it's advantageous to find the commander, quickly and easily)
for the buildings, by clicking on the icons you zoom in on the building and by repeated clicking switch between them.

Make Set distances from the planet, to easily find and select satellites(and other orbit units), without zooming in and out to find them. For example: a bar on the left, showing which zoom distance one is at and click to choose which. Like "Ground", "Orbit", "Planet+Moons", "Solar System". But keep the manual zoom with the mouse-wheel, that's fine!
If i select a planet or moon in the Solar-System and scroll in, it sometimes switches to another planet/moon, which is in the way. So it would be easier with the "Distance Bar" only two clicks needed.

Being able to freely move in the solar system/space without jumping from planet to planet, to see/find moving units/enemies/missiles

A dotted line preview of the route of "Astreas"(?) for example or missiles, plus a time estimate (count down?). When giving a command to "Astreas" to fly to another planet/moon, enable solar system view to choose a planet without manually zooming out, select a planet and zooming in to select a specific spot to land (it's just going to orbit anyway...). So this way you can see the preview line and then a message like "Commander has landed on XYZ".

AI-Settings: Be able to set difficulty levels. Maybe even a special setting for AI Commanders to be focused on different tactics e.g. Orbital-Commander, Nuke-Commander. Annihilation-Commander, Bot-Commander, etc.

For the Warning Messages: If something is destroyed, please add the name of the planet, where it happened. Enemy contact warnings would be good too. Same for nuclear impacts/detonations.
And a Activity Log, to see each message/warning, with a filter to find specific ones.

This game is awesome, even though it's in beta! Nice work, love playing! Hope it's not too much to read at once.

ok... devellop?

"In" what? In PA - yes will never be in.
But what hinders me to pass a rendered frame to an external program and interface that to a mod?
Of course it would be more efficient to extract the data of what unit is doing what and where straight out of the client.

FA could view the whole battlefield in one piece and didnt take place on multiple planets.
Beeing everywhere at the same time was less of a topic than it is here in my humble opinion.

PA is much more about situational awareness and not having to bother microing stuff around. Its time you have for other things your oponent doesnt. Each one is a small part but together they add up for a formidable advantage.

The time this takes to get done is just a matter of interest. Once there is a real competitive scene with leagues its a matter of days or weeks Id expect and at least for the MexRaider its far from a scientific project.

Do you always set up suiteable protection for all mexes? Or just when you know your opponent uses such a mod? Which of course, you dont know beforehand.

Im more into uC programming and some x86 ASM for fun, the biggest "image analysis" Ive ever done was tracking a simple tennisball through a webcam stream so thats not exactly a strong skill of mine. But even back then it wasnt too hard to track stuff like a tacicon through a frame with background that lacks anything similiar to the feature to be tracked and differentiate between multiple icons. Id strongly expect some pretty advanced libraries to be available for that now.
The right people could get this up frightingly fast and usually dont make it public.
Id really love to do that myself but I certainly lack not only time but also the skills required.

Basic stuff like progress on a nuke or the status of anti can be displayed as alerts, no need to look at it.
Nuke at location xxx now at 10%.
Alert: Nuke at location XXX now at 90%!.
AN launcher at location YYY empty.

Important information you get displayed without wasting a single second on obtaining it.
It certainly wont be one thing that makes it "imba" but a load of different ones that just make it somewhat unfair.

Even though a script that micros bots through arti fire could be considered quite close to the former.

If you are going down that road you might as well be concerned that somebody builds the same thing for SC2 or the like. Has not happened so far to my knowledge.

I really do not think this all will be an issue. Simply because I think the things that are possible are good for the game and not bad. Sure too much automation is bad, but the level I would consider "too much" is beyond what I expect any mod to ever implement.
If PA ever reaches a level of competition where those really smart people start to write really crazy mods and don't even make them public I will be happy about it. Because honestly I am afraid PA will never reach that kind of popularity. It is not called starcraft after all.