Why hasn't the destrucible environment effect been improved over the years?

[eobet]

It is my understanding that the effects of weapons on the levels in the new Worms games are exactly the same as in the first one, ie they leave perfectly circular, plain holes.

Why hasn't this been improved upon?

I believe that the gameplay could be improved (or at least updated, or added to) by having dirt collapse, slide or be thrown around. The static objects placed around the levels could also collapse, fall or be thrown around.

I know that some classic tank games even have weapons for burying other players under dirt, though that usually makes for a tedious, prolonged game. However, since Worms naturally are good diggers, a new weapon (or feature) to dig yourself out could be implemented and the thrown around landscape could act as means to add damage to an attack. I see new ways of creating traps this way as well (dropping anvils on people, triggering landslides, etc).

Also, why hasn't the crater graphics been improved upon?

To me, the holes in the landscape reveals a fake, cardboard environment, which detracts from the cartoony, yet somewhat volumetric experience I get from a fresh level.

I was very impressed by the following screenshots, which seemed to show a new kind of (be it pseudo) 3D effect in the landscape craters:

http://nofi.wordpress.com/files/2006/12/336871636_1814a598c4_o.jpg
http://farm1.static.flickr.com/155/336871636_1814a598c4_o.jpg

Yet, in all other Worms screenshots I managed to google, I see no evidence of this at all. Not only do I think that the flat, plain holes look outdated, but again, I do think that it makes the otherwise flawless presentation suffer and kills any belief you have that the landscape is real (be it in a virtual universe).

[Distrance]

I believe that the gameplay could be improved (or at least updated, or added to) by having dirt collapse, slide or be thrown around. The static objects placed around the levels could also collapse, fall or be thrown around.

I know that some classic tank games even have weapons for burying other players under dirt, though that usually makes for a tedious, prolonged game.

Like PocketTanks ? :p , in my opinion, I wouldnt want that to happen, Worms is Worms, its not tank game, ok its bit unrealistic when the pieces floats in air but I at least got used to that so I wouldnt want to change that

But changing the hole thingys for example wa would be nice..

[Spadge]

It's been discussed many times. There's a fair amount of tech and design issues surrounding it and these were dropped when we went off and did the 3d stuff. Since we've returned to 2d, we started the thing again without risking totally breaking the game.

In the meantime, some experimentation may well be going on, but something as major as that can have huge implications on how it plays and only time will tell.

[Plasma]

It's been discussed so many times:
1: Not only would the falling terrrain be hard to do, but it would also remove a fun part of the game.
2: Blast holes don't have to be really realistic when it's only a 2D cartoon game.

I was very impressed by the following screenshots, which seemed to show a new kind of (be it pseudo) 3D effect in the landscape craters:

http://nofi.wordpress.com/files/2006/12/336871636_1814a598c4_o.jpg
http://farm1.static.flickr.com/155/336871636_1814a598c4_o.jpg
That hole is part of the terrain at the start of the match, not an effect from the weapons.

[Distrance]

That hole is part of the terrain at the start of the match, not an effect from the weapons.

Really =O, and that pic just made me want to have that worms, I changed my mind now

[Spadge]

Yep, it's all in the graphics - not deforming each and every aspect of the terrain in a kind of 3d way will sorely diminish the actual enjoyment and fulfillment of the social-strategic experience.






[Special offer: Each bottle of wine now adds free sarcasm in every post]

[Plasma]

[Special offer: Each bottle of wine now adds free sarcasm in every post]
Oh dear! In about 5 days, you're going to be worse than Akuryou!

[Spadge]

Oh dear! In about 5 days, you're going to be worse than Akuryou!

Maybe, but I'll be in Las Vegas and there's a lot of stories there that don't make this forum or indeed, my own blog. And not what you might think, either. Certainly no destructible land.

Funnily enough, 1 month from now I'll be in Galway - but don't ask where because you'll hear us all.

[bonz]

In the meantime, some experimentation may well be going on
Hmm... In the meantime? What comes after that time?
The all-new & shiny 2D PC version which is back to the roots? :D
[Special offer: Each bottle of wine now adds free sarcasm in every post]
Oh dear! In about 5 days, you're going to be worse than Akuryou!
Or in the hospital with renal failure. :-/

[quakerworm]

it would be interesting to see some more complex terrain deformation in 2d titles if they ever get back to the pc or high-end consoles. (i have no idea how much processing power one can get out of 360's live arcade) on older machines, realistic landslides would be tricky. some games used simple vertical drop of columns of pixels to simulate it, but that doesn't allow for caves at all. with modern machines, one could create a soil model that would allow for materials of various strength, density, and viscosity. even simple 2d liquid simulations can be ran, allowing for dynamic pools of water to drown worms in. just imagine surprising a dark sider with a flood of water released with a well aimed shot from a little "mountain lake". this is something that would be next to impossible in 3d, but is quite doable in 2d.

[Spadge]

it would be interesting to see some more complex terrain deformation in 2d titles if they ever get back to the pc or high-end consoles. (i have no idea how much processing power one can get out of 360's live arcade) on older machines, realistic landslides would be tricky. some games used simple vertical drop of columns of pixels to simulate it, but that doesn't allow for caves at all. with modern machines, one could create a soil model that would allow for materials of various strength, density, and viscosity. even simple 2d liquid simulations can be ran, allowing for dynamic pools of water to drown worms in. just imagine surprising a dark sider with a flood of water released with a well aimed shot from a little "mountain lake". this is something that would be next to impossible in 3d, but is quite doable in 2d.

Actually you can have a crack at the 360's full kit on arcade, there's no limit as such other than the combination of costs and complexity given the lower financial model.

We had water way back in 1997, worked ok (if a little low rez) but it just didnt work as a mechanic at all, but I imagine it'd be worth experimenting with again. I'm not sure it would be impossible in 3d, it's just a different solution, but it's a little less complex in 2d.

[quakerworm]

real time fluid simulation in terms of solving pressure/flow equations for each point in space is unfeasible. there are just too many points to go over. simulation over a 1k by 1k by 1k array at single precision floating point would require 16 gigs of ram and a super computer to work with. of course, you can do dynamic surface approximations of liquid behavior. it is easy enough to account for pressure changes in terms of quick volume computations, but flows are much harder to account for. i doubt you'd ever get anything that is close enough for a simulation, but might be close enough for a game. the problem is that you are still stuck with unpredictable surfaces with ever changing polygon and vertex counts. memory management alone would be a nightmare, but i'm sure you are aware of that. you probably tried something similar for destructible terrain for 3d worms before you settled on the poxels.

[Spadge]

real time fluid simulation in terms of solving pressure/flow equations for each point in space is unfeasible. there are just too many points to go over. simulation over a 1k by 1k by 1k array at single precision floating point would require 16 gigs of ram and a super computer to work with. of course, you can do dynamic surface approximations of liquid behavior. it is easy enough to account for pressure changes in terms of quick volume computations, but flows are much harder to account for. i doubt you'd ever get anything that is close enough for a simulation, but might be close enough for a game. the problem is that you are still stuck with unpredictable surfaces with ever changing polygon and vertex counts. memory management alone would be a nightmare, but i'm sure you are aware of that. you probably tried something similar for destructible terrain for 3d worms before you settled on the poxels.

Obviously it all depends on how complex you wish to go - gaming solutions often work better with generalisations, as long as it feels, looks and plays rights. The water solution 10years back was simplified (not simple!) but just didn't work well in game - something we've covered many times, the "swimming worm" animations in the sprite banks are evidence of this :)

[quakerworm]

water in 2d is a lot easier, though. i know it was still a challenge 10 years ago, but most machines today can spare enough power to run full fluid simulation for each pixel on a wa terrain. in 3d it's a whole different animal, but yeah, i'm sure there is a feasible 'feels and plays right' solution.

[AndrewTaylor]

real time fluid simulation in terms of solving pressure/flow equations for each point in space is unfeasible. there are just too many points to go over. simulation over a 1k by 1k by 1k array at single precision floating point would require 16 gigs of ram and a super computer to work with. of course, you can do dynamic surface approximations of liquid behavior. it is easy enough to account for pressure changes in terms of quick volume computations, but flows are much harder to account for. i doubt you'd ever get anything that is close enough for a simulation, but might be close enough for a game. the problem is that you are still stuck with unpredictable surfaces with ever changing polygon and vertex counts. memory management alone would be a nightmare, but i'm sure you are aware of that. you probably tried something similar for destructible terrain for 3d worms before you settled on the poxels.

I assume you've never played Wetrix then? That was on the N64, and had extremely realistic water -- to the point where you could judge the scale of the play area from how it behaved. (The play arena, it turns out, was tiny.)

That was, I would assume, done with a height map (and presumably a couple if hidden maps of other things), but it could be modified to work in something like Worms 3D, I'm sure. And it didn't require 16 gigs of anything or a super anything; it required a couple of 32 bit chips and 4MB of RAM.

[quakerworm]

wetrix simulaiton is essentially 2d. it has a terrain height map + water level map. it does not consider the flows in the liquid, but merely checks for overflows. even then, it looks like it isn't more than a 10x10 grid. comparing that to a problem of solving hydrodynamic equations in 3d is absurd, and the later is exactly the kind of problem you'd need to solve for fluids in w3d-like environmnet. poxels aren't heightmaps. in a 3d worms world you can easily have a situation when one pool of water is located in a cave beneath another pool of water. heightmaps for water levels won't work. not only that, but a simple overflow check doesn't work anymore either. the upper pool's support can be penetrated in a number of places, and you would expect water to gradualy flow out of these, not just splash out all at once, with a rate dictated by the number and size of the holes. finally, all that complexity is taken to a much larger scale than the wetrix playfield.

seriouly, andrew, you just sounded like people who demanded w4 mayhem on ds. "but other 3d games run on it!" corners can be cut, of course, but not to that extent.

[AndrewTaylor]

No, but considering the massive increases in computing power and memory since the N64 I don't think the play area size increase would be much of a problem. I doubt if Wetrix' simulation has exponential computational increases with size increase.

And I know poxels aren't height maps, but the top of them can be modelled as one, and that's the only bit the water would ever touch. I know it doesn't consider everything it really ought to, but the waves look pretty realistic to me. I expect it's very realistic (looking) when the water's in lakes (as it usually would be in Worms and always is in Wetrix) and pretty poor if the water is crashing down a hillside, but then anything realistic in that situation will probably be massively (read: prohibitively) complex computationally. Besides, how big is a worm (and by extension, the craters in Worms games)? A few inches? At that scale, water is a much more viscous thing and doesn't tend to come crashing out of things. It oozes, and forms blobs.

For two areas of water, one above the other, what you do is... use two heightmaps. Magic! Problem goes away. Granted a whole raft of new problems arrive, not the least of which would be what happened if you dug a spiral and poured water down it, but how often would that ever happen? I'm not pretending to have thought this through, but I feel fairly sure some compromise system could be reached, and made good enough for jazz.

Honestly, I don't think it is absurd to compare it to solving hydrodynamic equations in 3D. Observe: It would be a hell of a lot faster than solving hydrodynamic equations in 3D, but nothing like as realistic. If you want to pick one as a system to implement, comparing them is essential. I don't know which would be best, but I do know that solving hydrodynamic equations in 3D, on today's computers, just isn't going to happen, because, well, you just told me so. So unless you have a better idea...

The main problems, to my eyes, with any form of water would be that the collision issues (which do exist, even in W4) would become much clearer with pools of water everywhere, and that the water would only flow much when craters had just appeared, which is not really the best time, because that's when shadows are rendering, poxels are being removed, things are flying around... I doubt if the engine as it is would take very well at all to water cascading about at that exact moment.

[Metal Alex]

I can imagine a lava "water"fall in a cave map, coming from the top, acting like a fluid, with it releasing a determined ammount of lava each turn, making lava lakes between turns, or changing the flow with girders or holes... to make different lakes, trhow it over other worms, etc... of course, a slight touch of it acts like fire, too much, the worm gets totally burned (like if water dead, but reduced to ashes). It has a new strategy part... For other schemes, acid (basically the same, but killing), oil (combustion! blowing up liquids...), liquid nitrogen (worms in it get frozen, untill an ally unfreeze him. If only worms from one team are remaining alive and not inside that frozen pool, they win), water (no idea about the effect of it :p)...

I think it can have a LOT of uses for this game, and new ways of strategy appear... insta-buy for me :p

(but I need the weapon variety too :p)

[Squirminator2k]

I recently picked up the Worms Armageddon Strategy Guide from eBay and there's a wonderful little article covering the development of the series from its roots as Artillery right up to the UK PC release of WA. Here'sa little nugget from Spadge concerning early ideas for Worms 2:
We built a huge list of really cool ideas for the game. As these were tried, we found out that a lot of the time, the ideas didn't really gel or they felt out of place, so they were dropped or re-worked. A good example is the interactive object, such as boulders that could be dislodged and moved around the landscape. These just didn't work in the game and were removed.

[Metal Alex]

uhm... those boulder things give me new ideas... like a mode with indestuctible crates, making them act like those boulders when shot... or a "move this object by shooting at it to x place" training mission...

but, in the game, as part of the place, it wouldn't work so well... Anyways, I'm glad they had such a variety of ideas... and them tested. Few people do that.

[AndrewTaylor]

"move this object by shooting at it to x place" training mission...

You can do that with a landmine. WWP's Smart Mines let you do it with no risk of it exploding along the way.

[Metal Alex]

I meant something like that, but this way, different materials could be chosen for the object, not only a mine... (rubber, stone, etc...)
Just so it looked different, and being able to crush worms beneath :p

[Preasure]

I meant something like that, but this way, different materials could be chosen for the object, not only a mine... (rubber, stone, etc...)
Just so it looked different, and being able to crush worms beneath :p
You'd then have to implement a way for worms to burrow out of a crushed situation, or you'd be griefing players by trapping their worms under things with no way out.

[Spadge]

The thread should read "Why haven't new idea for destructible landscape been tried over the years?" - since changing it is no guarantee of improvement :-)

[quakerworm]

And I know poxels aren't height maps, but the top of them can be modelled as one, and that's the only bit the water would ever touch.
every poxel is a convex object. always. that means no pool of water can ever form on a poxel. furthermore, most poxels are not on the surface. giving each a height map would kill the system, that's why poxels are not rendered by themselves. the only thing that makes sense is to use a generated mesh surface of each lattice. but you can dig a cave within a single lattice, so it will have two surfaces that can support water. and that's just talking about pool formation. now lets talk about dynamics.

I know it doesn't consider everything it really ought to, but the waves look pretty realistic to me.
waves are peanuts. in a deep water approximation, it is easy to show that a 3d flow equation is replaced with a 2d second order equation for surface behavior. solution to one of these is trivial. of course, these equations cannot account for surf, but we don't need them to. for waves, they are fine.

but waves are not the important bit here. you want water to flow right. that's the important part. you can have rather complex flow paths in worms. there can be tunnels, caves, various objects on the hill side. these are conditions when flow momentum and local pressures matter. the problem just became 3d. just imagine a big bowl of water that got penetrated at the bottom. water flows out, picks up some minimum rotation from uneven shape of the lattice surface, coriolis forces increase non-radial component of velocity, and before you know it, you have a vortex. now picture yourself how much computation goes into maintaining a vortex.

now, i'm not saying that the worms game must have sufficient simulation for vortex formation, though, i'm sure somebody will say that. this is just one little example where liquid dynamics suddenly goes very complex. the thing is, yeah, you can come up with a wetrix-like simulation, where water simply spills out over edges, and the pools are maintained as numerous heightmaps with simple wave simulation. but without the rushing flows, crashing waterfalls, and rotating vortexes, you just took all the fun out of it.

[Metal Alex]

You'd then have to implement a way for worms to burrow out of a crushed situation, or you'd be griefing players by trapping their worms under things with no way out.

I though about it... but again, I was talking about some kind of mission, so maybe, it's an instant kill, or any other thing.

Anyways, I'm telling my ideas, and also knowing that they will not get implemented, or at least not because I stated them here... *sigh*

Well, I can allways dream... *stares*

[AndrewTaylor]

I didn't mean pools on one poxel. I meant, for each pool, you find the heightmap that corresponds to whatever poxels are directly beneath it. It's 2AM, so I'm not going to work out how to deal with underwater caves or anything. I'm going to go to sleep now. But I feel sure that that kind of system has far more potential right here and now than a vortex-forming 3D fluid dynamics solution, on account that it would run.

Why on Earth are we having this discussion anyway?
now, i'm not saying that the worms game must have sufficient simulation for vortex formation, though, i'm sure somebody will say that. this is just one little example where liquid dynamics suddenly goes very complex. the thing is, yeah, you can come up with a wetrix-like simulation, where water simply spills out over edges, and the pools are maintained as numerous heightmaps with simple wave simulation. but without the rushing flows, crashing waterfalls, and rotating vortexes, you just took all the fun out of it.
Well, some of the fun. It could still be used as a tactical item. As long as it behaves predictably, looks passably real, and runs without crashing or slowing the game down too much, it can at least be used in a game. Whether it's fun is a whole other consideration. Personally, I think it would just make the game too arbitrary; anyone could start with worms conveniently placed below the huge water tank...