MechCommander (the first one, but apparently not the second) had a very cool feature that seems to be very rare in RTS games, yet has the potential to really make things interesting. A little science first: Radar works by sending out radio signals in the hope that some of the EM energy will be reflected off a target and return to the starting point. By measuring the delay between sending a pulse and recieving it''s echo, the system can tell how far away the target is. But radio waves lose their energy as they travel through the atmosphere, so they can only be sent so far before they lose all their strength and become undetectable. Now let''s say that a particular radar system has an effective range of 100km. This means that the signal that is sent out can travel 100km to the target, be partially relected back, and travel the same 100km back to the start, where it will be just strong enough to be detected. So the signal, when it was first sent, had to be strong enough to travel at least 200km (slighty more in fact, since the reflection off the target is not perfect), and hence be detectable at that range. The upshot of this is, of course, that an active radar set can be detected by an enemy''s passive sensors at at least twice it''s own detection range. So there is a wide band of territory where he can see you, but you can''t see him (unless he is also using an active radar set). The designers of MechCommander realised this, and the radar display in that game indicated when you were in that band, and you had choice of using active sensors or just passivly watching the enemy''s radar emissions. The distinction between active and passive sensors is an important one in modern warfare, and yet is largely overlooked by the majority of games (with the exception of submarine sims). The simple choice between using active sensors and remaining passive can have a huge impact on tactics, and I''m really getting sick of playing games that dont give me that choice. In the same vein, many game designers seem to forget that radar uses radio waves, and that radio waves travel in straight lines, just like light does. In other words, it can''t see over hills. When was the last time you saw line of sight rules applied to radar? (OK, yeah I think it has been done, but it''s not common) Intelligence is the key to warfare, and in modern war that means electronic surveillence. When are we going to see games that give us reallistic models of sensor systems? If anyone out there is working on a game that includes some of these concepts, I''d love to hear about it.

Ditto your words of wisdom.

Heavy Gear II also gave you the option of using passive sensors for added stealth on particularly hairy missions. The only drawback is that when you''re overly stealthy, an AI bot will walk up to you and stand next to you with its gun pointed at your face and wait until you move (making you detectable) to shoot, this is not necessarily bad, but it is a misrepresentation that gets on my nerves (because in some missions, some random patrol will walk by your vantage point, see you, and then stand there waiting for you to move before they kill you).

George D. Filiotis
Are you in support of the ban of Dihydrogen Monoxide? You should be!

Yes! Yes! YES!!!

[ GDNet Start Here | GDNet FAQ | MS RTFM | STL | Google ]
Thanks to Kylotan for the idea!

The only correction to your formula would be the application of the inverse square rule. Light and other wave-type emmisions (like RADAR) lose their strength at an exponential rate rather than a linear one. That is, if a signal is strength x at 100'', it is sqrt(x) at 200''.

Anyway, your theory is sound and there should be much better representation of that in many games. However, there are many games that butcher laws of physics and ignore "how things are really done". Even movies are horrible at this... which is what gave us nifty, yet unrealistic, dogfight scenes in "Top Gun".

Dave Mark
Intrinsic Algorithm Development

Plasmadog hit on one of my pet peeves with rts games. The main thing I hate about those games (and mechcommander is a perfect example) is the way your units have supposedly long range weapons but the view offered by the dispay is not scaled to display those types of ranges. New tanks coming out today have 50x optics, folks you can see and identify targets at 15 km with 10x optics. British Cheiftain MBTs were hitting moving targets at over 3000m in the gulf war almost 10 years ago! If i''ve got this 27th century (or whatever) battlemech then I want to at least surpass the abilities of current technology. I think there is a tremendous opening for a well informed ambitious game developer to turn out a super realistic, super fun game.

I agree - but not just for the added realism, but because it adds a strategic element to the game - a decision to be made about when to try and detect enemies. It might be implemented in a fantasy-based game where if a seer tries to find enemies, they might sense that they are being watched etc.

I think scattering and refraction will cause a linear loss of signal strength on top of the inverse square law, but I don''t think anyone would bother using a serious calculation in game.

Realism is a good thing, but like so many other things it can ruin a game. It´s always good to try for realism, but as soon as the gameplay is affected realism HAS to be secondary.

Radar fit with the mechcommander theme, realistic physics did not fit with all the Elite-sucessors.

Most RTS now are all about quick building and quick fun. Radar could give more depth to that, but I´m not so sure that is the developers intention. Mech Commander worked fine, but keep in mind that in MC you couldn´t build units, which made balancing the game a lot easier.

As for the range issue mentioned by the AP, I say let´s forget it. Sure, a Maverick air to ground missile has a range of over 40 km, a modern tank can hit a target a few kilometres away, but it´s not fun.
There are some realistic simulations, but the relative market segment is declining (absolute numbers are about what they used to be). Not many new players like realism to that degree.

I´m all for putting the "strategy" back in RTS, but so far the quick acessibility of a game has been more important.


..like Fox said, "nifty, yet unrealistic".




Rock Hard Games: Games for Trogs.

quote: Original post by Hase
Not many new players like realism to that degree

Because realism so far has always meant a compromise of gameplay.

[ GDNet Start Here | GDNet FAQ | MS RTFM | STL | Google ]
Thanks to Kylotan for the idea!

Ah, yes. Quite right about the inverse square rule, my mistake.

Those who have said that reality can get in the way of fun etc., I quite agree with. However on this particular issue, I''m not talking about extra realism just for the sake of it. A realistic sensor model opens up whole new possibilities in terms of strategy and tactics. As Argus ponted out, there are factors like scattering which do not need to be included, because although they add realism, they don''t affect the gameplay.

I want strat games that are strategy !
Fun vs realism ?
Definatly realism.
I really hate real-time-peon-pushing

the bugle4D engine;
a new dimension in graphics...
[link]http://members.tripod.com/thefivelions/bugle4d/[/link]