The scouting problem

Planetary Annihilation has a scouting problem. Or maybe a stealth problem. Perhaps a fog of war problem. Regardless of what kind of problem it is, it's still a problem nonetheless.

I'm talking about the element of randomness on spawns, and their effect on scouting (and the subsequent effect on the game).

--------

Let me elaborate. When you start a game (say a 1v1), the first thing you do is scout the enemy. Usually this is in the form of building scout planes (why we don't use scout ships or vehicles is a different issue) and randomly sending the planes off in hopes that they'll find the enemy soon.

This is fine for small planets -- it takes only a few seconds to set up a route, and a bit of attention to the plane to see when it finds the enemy.

But what about bigger planets? Naysayers will argue that this is a non-issue 1v1s are best done on small planets (as if big battles require more than 2 players), but this is a game about customizable experiences? What if I want a gigantic, 1 on 1 battle with a fated rival, rather than a little match on a modest sized planet?

On a bigger planet, the enemy could appear anywhere. It could take anywhere from a few seconds to a few minutes (which is a long time in an RTS) to find the enemy, and this number goes up the bigger the planet is! If the enemy finds me in 10 seconds and it takes me 3 minutes to find the enemy, not only am I "wasting" time and resources building more planes and setting their paths, I'm also forfeiting an information advantage which I have no control over. By sheerest coincidence, total randomness, I am now significantly disadvantaged in the battle. This random chance could even be the deciding factor in the war!

So, to sum it up, the element of randomness can be deadly in a match like this. This is a problem. A big problem. If you're into the e-sports side of things, this is a HUGE problem -- it's like playing chess, except one side sees everything after the first turn, but the other side has to wait 5 turns before seeing the opponent's moves. Meanwhile the opponent appears, having built the anti-air to counter the bombers it saw when it scouted you, and some pelters being built on the outskirts of your base. Oh, and they're sending tanks towards your base's backside.

-------

Now, the solution. There is no real clear solution I can see, but there are many ways to go about it. The easiest fix would be to see your enemy's initial starting position. At the minimum, this should be a toggleable setting in the match, and in my opinion, ideally it would be toggled on by default. Of course, you'd only know the spot their commander moved into -- thus you can send scouts right towards them to then see *what* they're building, and continue on the information war from there. In this sense, both players have equal opportunity for scouting, stealth, and sneaky tactics.

What do you think? Is this a good solution? Is there a problem in the first place? Is there a better way to fix this?

perhaps there would be an impact crater formed where each of your commanders starts?
I don't really like this idea that much though, just scout the most viable looking starting spots! lol

I'm not sure if I agree that it's a problem. The randomized nature of the spawns and the need to find your opponent is very integral into the strategy of PA.

Example. I play on mostly 650 to 850 planets for 1v1s. Most of the time my opponent's are on the other side of the planet and I start my build accordingly. It's all fine and good, I have plenty of time. However, I don't know for sure that they're on the other side of the planet. I played a match recently where my opponent spawned very close to me and we expanded towards each other. My fighters (I don't use Fireflies to scout) didn't fly over my opponent until the exact moment my fabricators moved to the same metal spots my opponent was starting to build on. I then reacted and started moving units around an building defensive structures and whatnot. (it was actually a really good match and can be watched here.)

I was rewarded for scouting and my strategy changes rapidly.

If there was some sort of indicator showing where my opponent spawn, then there's not as much of a benefit of scouting. This is exponentially more prominent on large FFAs.

The unknown of where your opponent spawns is a huge strategic element that I really enjoy.

It sounds like you need to adjust how you scout. Whenever I scout, I set up waypoints with my fighters over the entirety of the planet and then go back to what I'm doing. So the longer it takes to find my opponent doesn't really matter at all since I set up the waypoints and forget. Just look around the planet later and 9/10 times I've scouted my opponent with very little effect on the time need to scout. Besides, there isn't too much to do at the start, so it doesn't really effect you too much.

So, some scouting tips. Scout in a spiral pattern out of your base. If your opponent is real near you, you need to know asap. Then send your scout to all of the large metal clumps as that's likely where your opponent has spawned.

Setting up waypoints over the entirety of a planet takes just a few seconds.

Also, you don't need to use Fireflies to scout. You can use fighters. So the "I need to build more scouts because I was unlucky and didn't find my opponent quick" isn't really too much of a factor. You need more fighters. Always. I have air factories pumping out fighters for the entire match. So that doesn't effect my gameplay at all.

I don't think there is a problem really, you certainly are able to scout your immediate area fairly easily. There's no excuse for letting your opponent build pelters in range of your base unnoticed.
Scouting is a skill, and knowing where commanders might be is about reading the map, again a skill.

Checking large metal clusters is the best bet. As a bonus, if you find that said large metal cluster is unoccupied, you know where to send your fabbers.

Yeah, universal known metal spots are the best choice for scouting because if they aren't populated by someone else, then do it yourself.

And if you populate all of them, then you have already won, and can afford to build as many satellites as it takes.

Yeah, I agree with brianpurkiss. Scout your vicinity first by spiraling outwards. If the enemy is far away he can't really put up much pressure except by doing air units. Fireflies are actually decent at taking down lone enemy units as well so you can kill lone raiders or fabbers with them.

This is nothing new to the competitive RTS scene. In Starcraft, when there are more than 2 starting positions, scouting the enemy build can depend purely on luck as if you scout in the wrong direction the enemy will have walled off his base or made units to chase off the scouting worker. Knowing the enemy build is a pretty big thing in Starcraft so the game can really be decided by knowing or not knowing the enemy build.
If the enemy is managing to build Pelters on the outskirts of your base on a big map you simply have too bad scouting.
Currently, you can use fighters to patrol the outskirts of your base and in no-mans land. Basically, the only way that the enemy can kill your fighters without losses is to send ground AA or build missile turrets which should give you a good warning of an incoming force or enemy expansion.

Actually, I don't think it is a problem on big maps but rather on small maps. If you don't scout, your fabbers might run into the enemy commander and if your commander is out place you might lose a big chunk of your base before you can put up a fight. The first or second factory for me is always an airfactory because of this. If I do airfactory first, I typically keep making Fireflies until the enemy starts making fighters. If the enemy is scouting with land units I can typically find those and kill them to keep my enemy in the blind and protect my expanding fabbers. It is still very strategical though. Do you try to get a better economic growth and risk facing enemy raiders or a commander push or do you sacrifice some economy to get reliable intel?

As for decreasing the attention requirement for scouting:
A minimap or multiple viewports will allow you to see your scouting progress while you are focusing elsewhere.
The notification system could notify you the first time you see:

• An enemy air unit
• An enemy ground unit
• An enemy structure
• An enemy commander

This would help the player even without a minimap or multiple windows implemented.

So far I'm not buying any of these arguments against it.

You all misunderstand me. I have no problems scouting as a player. I don't need advice. I already know all the tips and tricks to efficiently scouting an enemy. In this thread, I'm making the point that the current system of scouting ITSELF has fundamental gameplay flaws, which become exponentially more possible on larger planets.

What about really big planets? Brainpurkiss gave an example with a size 650ish planet for a 1v1. The randomness of scouting is still there, except it's so low because scouting is relatively fast that it's almost negligible in most games.

However, what if I'm playing on a size 2500 planet? 1v1? It's gonna be an epic match, with giant armies, enormous bases, and unique strategy. PA is about going big, and it should accommodate for large 1 on 1 matches.

In this giant planet, I'm not going to go scout out every cluster of metal, there are probably hundreds if not thousands of these. Making a path will take a while, because a scout can only have so much vision. I'll have to click many times, which will take up a LOT of time. If I have multiple scouts on multiple routes, it will take even longer. And they may take a while to go through their paths. Meanwhile I have a build to manage, new bots to set to different tasks, and this build will take more and more time the longer I go on without finding the enemy.

Again, I pose the scenario where the random scouting fails: Two equally matched players, Alice and Bob, are fighting on a planet. They are both using an optimal pattern for scouting. Alice finds the enemy in the shortest possible amount of time, while Bob finds Alice in the *longest* possible amount of time. Alice now has an advantage, and she wins the game.

Alice wasn't better than Bob. But she carried an informational advantage for a portion of the game, which influenced her strategy and lead to her victory, since Bob did not have such an advantage until a later time. And Alice only has this advantage because of sheer dumb luck. Tactics and skill didn't win the battle, luck did. This is something directly contrary to the fundamentals of fair competitive play. The system is thus flawed.

Scouting is still important. It will still be a fundamental part of the game. Does the enemy send units to the east or the west? ARe they building tanks or bombers? Do you notice the catapults they are building on the outskirts of your base? Showing where the enemy begins, like every other RTS, INCLUDING Supcom and Total Annihilation, has worked for this reason.

-----

edit: Godde made the point that random spawns is nothing new to RTS's, such as starcraft. Does that mean it's not a flaw? If one person sees the enemy before you do (because of random chance), they get an unfair advantage. Competitive games work on the notion that the game is fair. In almost every other aspect, PA is fair. Random scouting is an element which introduces unfairness. So, even if starcraft does it, it doesn't mean it's fair.

Maybe knowing where the enemy is is a bad solution to the problem. I do agree with Godde that notifications should be made for discovering a new enemy (scouting them). That would cut down on the babysitting you need to do for big scouting operations, and it's not too much to ask for. Perhaps this might be a potential solution to the attention aspect. However the underlying statistical problem is still a major flaw in the game design.

Well isn't a flaw in the games design, but a flaw of not accounting for the games scale.

and the problem is how you create a scalable scouting method that works for large planets with out making the smaller ones game without the fog of war.

Like i have always argued for, is a reduction to the erraticness of scouting. You CANNOT compare the scouting in PA to SC2. SC2 has a nice amount of random, because we are given potentially 3 seperate locations they could be at, but in PA the list is literally limitless.

Besides, what happens if i want to rush an opponent? currently thats almost impossible on all but the smallest of maps. Nobody is asking for a removal of all of the randomness, but to remove it to the point that various strategies and ideas become viable.

The solution in my mind would be to present each player with 5-6 (different potentially based on planet size) potential locations. With one of them being the correct location.

This is a fine solution, but why does there have to be randomness? Is it just the thrill of gambling for the right spot? I don't see any legit reason why there should be randomness, and a very big reason why it shouldn't be random.

This game is inherently random. If we were to remove all randomness from scouting then we need to ask the question of where do we stop?

At wherever the player wants. If I want a planet that is neatly symmetrical with both players starting on opposite sides and fair metal distribution, that should be an option. If a player wants random spawns, random planet, random everything, that should be their choice. The former is well-suited for accurate ranked matches, tournaments, and fair competitions. The latter is for casual players or players who like to introduce randomness and uncertainty in their games. Two different playstyles, but currently only the latter is appealed to. I would like the former to have some recognition.

unfortunately we are not talking about the same thing. I couldnt give two shits about custom games. I am only talking about it for the official gameplay that will be used on ladder. Ladders do not give options.

Tournaments and such always use the gameplay conditions set by the ladder. Any that do not are not worth joining.

Hmm...I dunno. Ideally there would be no randomness on the ladder, planets/systems are symetrical and are equal for all parties involved. This is mostly impossible to pull off due to the design of the game.

So we already have the inherant randomness of planet generation. Hopefully resource distribution is fixed (I believe it's confirmed there will be more work on that). This does have the whole "exploration" about it, but it's still problematic. This would mean that there are freak incidents where a good player gets a bad spawn and loses because of it. That is bad for RTS's, but I think Uber's already sealed their fate on that. In exchange for novel maps in matches, they've sacrificed some degree of fairness.

Keeping that in mind, I would argue that we should not introduce more randomness into an already unfair system. That could tip the scales in a battle even more. Thus, in response to the question "where do we stop?", I would say "At planet generation". Unless I'm missing other systematic elements of randomness besides random planet generation and the unpredictability of enemy spawns when scouting.

The only issue I can see is there should perhaps be intel units appropriate for large scale games, but I'm not sure how you'd balance those.

I like randomness for the same reason I like procedural maps; it bring about variety and a requirement to think on your feet to quickly assess what are the strategically valuable parts of the map. Not rote builds based on your start location.

Symmetrical is a human construct. Nothing in nature is symmetrical. There are very few competitive sports which are symmetrical.

No game needs to be symmetrical for the benefit of the competitive players. If you can't play with the hand you are dealt, you need to get better. That's what being better means. That you can force a win from a disadvantaged position.

That applies equally to Alice and Bob.

On topic, on scouting, Uber have said that they are going to have area commands. We've seen first pass.

I agree that currently a rush is just an impossible strategy in a 1v1, and I think that is to the detriment of the game.

I think a part of the problem is the huge range of location a person can go ranging from just over pelter distance to the opposite side of the planet. I don't mind the idea of being given so many potential location for the enemy potion.

The biggest thing that makes this game different to star craft is not the randomness of spawns but the access to units that are really good at scouting early like the scout plain and land scout. In SC2 we know the exact time to scout because you know the map ect ect. So in PA we have better scout units but we don't know where to scout even with the knowledge of where best to scout still leaves it very luck based.

What I like the idea of is giving each team a portion of the planet to spawn in so in a 1v1 each player would have 2/5 of a planet to spawn in and 1/5 as a no mans land between them. So each player will have a much smaller area to scout with out giving the location completely away.

The easy short term solution is in the short term is that when a game has 3+ option to spawn every one can see each others spawns. This would be fine as long as you increase the min distance enemy spawns can be.

That seems like an exaggeration. Even I won't say I know all the tricks of scouting currently in PA and from what you have described it seems that your scouting capabilities is not that good.

In the complete opposite I think it could be a problem on small planets but diminishes on large planets.

Here you are talking about the difficulty in executing scouting. This can be remedied by improving your multitasking or by improving the UI in the game. For example there can be a scouting management tool that automatically send out newly built scouts to spiral outwards from your starting location and automatically covering previously unscouted terrain.
Other improvements include minimaps, multiple views and notifications upon enemy unit detection for example.

I'm not sure what an optimal pattern for scouting is. You would have to describe that to me.
If we take spiraling outwards with airscouts, which I think is a pretty good way to scout as you will usually find incoming raiders and expanding fabbers with that, it is almost like an expanding circle. If both players are expending the same amount of resources on scouting they are likely to find each other at roughly the same time.

This "problem", which I really don't think is a problem, will persist throughout the entirety of the game. In an RTS you are always dealing with imperfect information. Later in the game you might find out that the enemy is building a nuke with your scouting force or you might miss it. Scouting requires resources and it is up to the players to decide if they want to know what the opponent is doing or if they think they can predict what the enemy is doing or going to do.
On a large planet there isn't much you can do before the enemy scouts you except an air rush but considering how bad basic bombers are, I consider that a non issue. By the time any type of rush can hit you on a large planet you should have ample of scouts done and be ready to deal with any type of ground rush.

I can compare the scouting in SC2 to PA. I can compare how critical scouting is to the pregame Rock-Paper-Scissors. In SC2 it can completely block off the scouting which means a late scout leaves that player unable to scout the enemy build at all.
In PA as time goes by, it gets easier to scout the enemy as the enemy base is an ever-growing target while outgoing raiders and expanding fabbers are likely to relay the position of the enemy starting location. Also, on large maps there isn't really any starting strategy RPS.

So you are arguing that you should be able to rush the opponent without no scouting?
That sounds like even more pregame RPS to me. If there is a blind choice of going for an early rush that makes the outcome of the game even more volatile and random.