next »

In most (all?) RTS games maps just have a fixed time of day and weather conditions. I thought that we could be original, and introduce a day/night cycle in the game, with changing weather.

Some scenarios:

- A surprise raid on a base at night, when visibility is lessened
- Hundreds of troops storming on a castle and rain starting to fall in large amounts
- A traveling battalion hindered by heavy snowfall
- Exhausted troops due to extreme heat (?)

Certainly weather effects seem right, but I'm always uncertain about randomness.  What if it rains during your decisive offensive to win the game?  Randomly, the other player may have been given an opportunity to change the course of the game.  I don't feel such opportunities should come at the graces of a random event.

Day/Night cycles and effects from weather make perfect sense, however.

Well it would be random, yes, but both players would be affected so I don't think it makes the game unfair. In any case we could keep the effect small (but noticeable)

I believe that in Compagny of Heroes there is something like that, no?

Anyway indeed it's a good idea.

Well it would be random, yes, but both players would be affected so I don't think it makes the game unfair. In any case we could keep the effect small (but noticeable)

The problem is that situationally certain weather conditions could favour one player's current position.  Obviously someone who needs to counter-attack quickly is going to be heavily impeded by a flash snowstorm, giving the other player an opportunity to regroup he might not have otherwise had.  It may be the same effect to both players, but it may grant one player an opportunity or deprive him of one depending on his situation.

At very least we'd need a weather forecast so players could see what was coming ahead of time.  I'd also think that weather should be map-based and keep to only a couple of effects.

IMO a day/night cycle should be implemented. It makes sense and there should be a small effect as to gameplay...  at night, visibility should be reduced.

However, I'm not sure I'd like to have a lot of effects because of weather. While it's certainly realistic, I'm not sure how fun a game is going to be when my attack is suddenly ruined due to a blizzard. You have to consider the line that divides what is realistic and what is fun to play.

Still, I like the idea of weather... the idea that a blizzard can suddenly occur and change the map is very, very interesting and truly a new idea. Let's say a blizzard comes in the middle of gameplay. Visibility should be reduced, obviously. What would be very cool is to have the snow literally cover up areas; changing the terrain and covering up certain landmarks. There should be an effect that makes troops walk slower in the snow. I absolutely LOVE the idea that the map could be changed by a random snow fall during the game, especially if the effect could be random and not always the same - storms may occur randomly, and one snow storm may contain more snow than another... and thus depending on the amount of snow that falls, the map could be changed more/less. I don't know what it would be like to code this, but I think it might require a fair bit of CPU power... I don't see why it can't be done, however.

Personally, the things that really excite me in games (more than graphics) is physics and what can be done to the environment. I love the idea of a storm wrecking havok on the environment and changing the way the game is played... with the snow fall, perhaps several passes could be blocked off, forcing me to head into a chokepoint set up by the enemy, who ends up using the weather to his advantage. What would also be amazing is realistic physics... if I fire a huge, flaming rock into the middle of a town, having the impacted buildings crumble apart in the areas they were hit and catch fire. If I fire rocks at a wall, I want parts of the wall to come off. Of course, this is all beyond what you guys are (realistically) going to do, and also would require a good deal of CPU power. However, that's what I want to see out of future games - realistic physics, fully destructible enviornments. If I shoot a rocket at a wall, I expect it to crumble into pieces.

The problem is that situationally certain weather conditions could favour one player's current position.  Obviously someone who needs to counter-attack quickly is going to be heavily impeded by a flash snowstorm, giving the other player an opportunity to regroup he might not have otherwise had.  It may be the same effect to both players, but it may grant one player an opportunity or deprive him of one depending on his situation.

well that's a property of weather, it just happens and you have to deal with it

At very least we'd need a weather forecast so players could see what was coming ahead of time.

that's possible, though it wouldn't really fit the game

I'd also think that weather should be map-based and keep to only a couple of effects.

can you define map-based please?

Day/night cycles seem pretty unanimous.  Unlike weather, it's totally predictable, which should work wonders.

I do agree that we should have weather, but it needs to be done in such a way that it won't seriously interfere with someone's long-term plan or strategy, and it cannot come on so quickly as to completely throw someone's short term plans to ruin.

can you define map-based please?

In other words, the map would have the weather systems that it will experience as predefined things.  That is, a map is either highly likely to experience a blizzard, or it never will.  This level of predictability, combined with a forecast, is important in order to allow players to accommodate their strategy for what is coming.

In other words, the map would have the weather systems that it will experience as predefined things.  That is, a map is either highly likely to experience a blizzard, or it never will.

That seems logical, yes

Day/night cycles seem pretty unanimous.  Unlike weather, it's totally predictable, which should work wonders.

I do agree that we should have weather, but it needs to be done in such a way that it won't seriously interfere with someone's long-term plan or strategy, and it cannot come on so quickly as to completely throw someone's short term plans to ruin.

can you define map-based please?

In other words, the map would have the weather systems that it will experience as predefined things.  That is, a map is either highly likely to experience a blizzard, or it never will.  This level of predictability, combined with a forecast, is important in order to allow players to accommodate their strategy for what is coming.

The forcecast sounds good, but I'm not sure I agree with making the weather only map-based. Part of the reason to add weather is to make the game somewhat unpredictible, and it's not going to be that when only a few certain maps will have blizzards, and only a few will have rain.

If you take unpredictability too far, however, you leave the door open for gambles.  It sucks to lose to someone who took a risk that paid off, and it sucks to lose because you took a risk that didn't.  Even the safest of gambles can fail, and even the most farfetched can succeed.  I believe in balancing the gameplay for all parties, and certainly the competative aspect requires that the non-human component of the map setting be predictable.  Part of this means making each map have regular weather patterns.  It doesn't exactly make sense for it to rain in the desert, anyways.

Random weather effects would be hated by "pro" players. Instead, weather should be an optional feature for Custom matches. That way "pros" could stay away from nasty game "unbalancing" effects like weather. Yes, it's less realistic if it is not present, but it also is more balanced, as Darvin has explained quite thoroughly.

Instead, if you want "balanced" weather, make it so that either it's predictable (ala a count-down timer when the blizzard will start and when it will end). Alternatively, you could make it so that custom matches allow for players to add a blizzard to the game (a blizzard would be going from the beginning of the match).

Randomness is always shunned by "good" RTS players. That's why nobody who is good at RTS games plays CoH (hehehe).

[Time options:]

Making day/night cycling and weather effects an option would make perfect sense

Time options:
0 Morning (These and below specify the time at game start)
0 Afternoon
0 Evening
0 Night
[] Cycle daytime (Check to make time run, or don't to keep the above time the entire game)

Weather options:
0 Clear
0 Bad (Map-specific)
[] Change weather (Check to make weather change, or don't to keep the above weather the entire game)


In any case, I personally don't really care for the pro gamers, which are such a small part and are never going to play our game anyway. It's not commercial, low-budget and will probably never get into any kind of important tournament

I don't think they're important either. Of course, my dream for Open War is that some day it will have a massive community, to the point where even large community sites like GR start noticing it.

i'm not familiar with GR. if anyone has an account there, feel free to "advertise" openwar, if they have an appropriate place for that

I think we can balance this sort of thing for everyone, not just either pros or casuals.

I do think we can get weather to the point at which it's predictable enough to satisfy pros, yet spontaneous enough to satisfy casuals.  The key is three-fold:
1) all weather effects must have a reasonably high chance of appearing; if a map can have a snowstorm, chances are you will see a snowstorm before it's over, preferably many times.
2) weather forecast so you can see how weather will change in the near future (kind of like seeing the storm clouds on the horizon).
3) faction neutral; no weather systems should inherently favour one faction.

The key here is that if a map is going to have a blizzard, it's not going to be a question of if, but when.  As a result, it's just part of the map, and not a gamble as to whether it will happen or not.  Secondly, because there's a forecast for what is coming players can think ahead and formulate strategies.  Finally, because no weather system favours any given faction, if it randomly "sticks" to one in particular, it won't give a long-term penalty to one player.  If the player is too inflexible to adjust his strategies, that's his problem.


I think if that's how it's done, we can balance both sides quite nicely.

I'm not an embedded member of the GR community, but I do have SOME recognition. How would you like me to go about "advertising"

It'd be much appreciated
You (everyone here ) could start by putting our signature userbar on all forums (where it makes sense)

Forgive my n00bism, but how do I do that?

go to the place where you can change your signature (different on each forum) and enter

Code:

[url=http://www.openwarengine.org][img]http://www.openwarengine.org/images/signatures/OpenWarSig3.png[/img][/url]

Darvin has fine points, and I got little to add to them.

The only thing would be to give the weather forecasts some measure of inprecision. I'm not talking about predictions of rain and a reality of bright sunshine, but rather minor inprecisions with no major effects on gameplay. For example: The forecast is snow in 1 min 10 sec. The reality is snow in between 1 min 0 sec and 1 min 20 sec. Even now weatherforecasts aren't 100% accurate and frequently enough outright wrong, so why would the information you get be perfect?

What I further want to bring to the attention is the amount of lag weather effects can cause. For that reason, not for gameplay reasons, I think the option should exist for players to turn the effects off.

Solinx

I think the forecast could be done as a visualisation of "looking at the skies". So there would be a horizontal bar somewhere on the screen, where e.g. the far left is 'now' and clouds on the right (moving left) mean approaching rain.

About the effects, a simple rain effect is easily achieved by darkening everything, putting a 2d rain overlay over the screen and playing the rain sound.

I think a forecast without a timer would work.  All we need is a qualitative measure of what's going to happen; an unnecessary quantitative measure might be a step towards information overload (something we want to prevent).

Meh, Im too used at things being based on timers, I immediatly thought of that
Qualitative measurement sounds good.

Solinx