allmhuran, on 08 August 2012 - 10:37 PM, said:
I agree. I think that's pretty much how we draw the line between "required level of realism to avoid complete confusion", and "way too much detail, unnecessarily hindering the design process".
For a game, particularly a game where real people are playing against each other, randomness is a major design no-no. If the finer points of "realism" (like the expansion and contraction of metals as they heat up, the inaccuracy that comes from barrels that aren't just-cleaned, etc) call for randomness to be added, then we should ignore those finer points in preference for better gameplay.
If, on the other hand, a design decision was made to implement a gameplay mechanic like, say, wind effects, then it would be important that the wind effects were applied consistently and to things that "should" be affected by wind, so as not to confuse players or make things unintuitive.
I guess the key here is that anything that realism suggests should be "random" is, by definition, unpredictable. Your intuition about "what should happen" is therefore not going to be of any value anyway. We can therefore safely ignore such aspects.
Realism may dictate that weapons aren't perfectly accurate, but there's no way you can use your intuition to predict the random deflection that would occur due to such realistic models, therefore there's no reason to include it in the game.
Unrestricted, random application of variables would be problematic, but if variables are applied due to intutitive triggers, and these variables are given restrictions on how they are applied and to what magnitude, you can emulate (not simulate) with a believable degree the affects the varibles are emulating.
For instance, you can emulate ballistic gun dispersion of a .50 Browning Machine Gun by using actual dispersion data on the gun. This will provide you restrictions on how the gun will disperse shots. You can then take into account the various conditions that contribute to the dispersion, such vibration induced by the weapon's platform(such as mounted in a wing of a fighter, or atop a Humvee), heat, user proffiency, etc, etc. From that, you can build a reasonable...or at least reasonable believeable facimile of these conditions and how they affect dispersion within the confines of the data.
Since we don't have PPCs a, AC10s, we can take what we know about systems with similar principles and affected by smilar conditions, variables, etc, etc. So we can extrapolate dispersion data from lets say a 20mm, or 30mm or 40mm Bofors to approximate a reasonable dispersion model for in game auto-cannons. Use data from short-barreled 75mm or 100mm anti-tank or tank rounds to approximate largers Auto-cannon dispersions. Of course the actual rounds in mech warrior are bigger, but the conventional rounds of the last centuries tank/armored vehicle munitions would scale up just fine with the those rounds used by mechs...its close enough to be "believeable", pluse we are more inclined feel the attributes of conventional muntions as "correct"...since most have no concept/experience(real-life or in games) of a naval ship size round behaves in a short-range, direct fire capacity.
With respect to lasers, if their alignment is off, they miss their target and loses coherence, thus reduced damage...so a reductn in DoT is a nice emulation. I've actually seen this happen to a laser's beam due to amplification chamber anamolies/defects(it requires a re-tuning of the lense to correct).
Particle weapons, Rail-guns, etc, etc...are essentially ballastic in many respects. Those affects of barrel anamolies essentially apply to these guns too.
The key is building within the game's engine, a model that is based on natural laws and utilizes educated assumptions and input to crank believable and not purely random results.