Dino Might, on 15 April 2015 - 08:42 AM, said:
This type of curve behavior is what I've been advocating for CoF, but the standard deviation is based on a number of factors including movement and heat at the time of firing.
Heat is no isue you have DHS - there should no influence be within. Same for movement - gyro, mymer etc. will active stabalize your weaponsystems, otherwise the whole Mech concept is a fail by start of introduction as a advanced weaponsystem, if nowadays mil tech have measuresystems to equalize heat and movment deplacment from calibrated point to hit.
Dino Might, on 15 April 2015 - 08:42 AM, said:
I would cut off the tails, so that there is no chance of deviating by more than a certain angle (1 degree? 2 degrees? 5 degress? not sure I'd have to give it some thought). And just push the CDF of the tails back into the distribution.
Sure you can bring the curve by a y-axis correxture under the value and cut of in any point you wanna have so the function doesn't go for the limes. The degree for the max deplament if so needs debated, but it would be more like something in arc minute if not seconds.
Dino Might, on 15 April 2015 - 08:42 AM, said:
If you really want to get complex, you can have total angular standard deviation made up of a vertical and horizontal component based on mech movement, and a random direction component based on heat. Then you get the total magnitude of this vector and that's your standard deviation for your curve, you run the RnG to get a number from that curve that gives you the total angular deviation, and you already have the direction. Shot goes based on the calculation of the resulting true weapon direction.
if you wanna create a unneeded dependence - sure.
Dino Might, on 15 April 2015 - 08:42 AM, said:
Again, this would need to have numbers that would get us to a point where a cool mech not jumping/making snap shots/falling off a cliff and hitting the ground at the instant of firing would put pretty much all weapons to the same component (within a few meters at 1000m or so). But, if you start adding in heat, then movement, and maybe weapon condition, you can see where it would make sense to start having some spread.
I see the condition of extreme stress - if i'm correct while jumping the reticula is displaced - it is introduced in the game right now - however this could be made more accurate, depending on conditions which migh influence accuracy.
Dino Might, on 15 April 2015 - 08:42 AM, said:
Lasers are quite often cited for having no realistic spread effect; however, we get heat from lasers that is vented by (what I assume) liquid cooling jackets; but we only get an average measure of heat across the mech based on calculations of total heat balance. We never consider that firing that ER Large Laser made the mech 10% hotter, but that particular laser is now scorching after firing and will take some time to cool down back to ambient, which will be slightly longer than it takes for the average mech heat to get within measurement accuracy of ambient. So, right when you get back to ambient 5% on the heat scale, that ER Large Laser (housing, lenses, cabling, collimator, excitation bank) may still be sufficiently hotter than ambient to result in some deviation from perfect accuracy.
no or at least, the caused deplacement would be compensated by calibrated data of the BattleMechComputer. Also the weapon cooldown is meant to be that time it needs to get the laser for example back into the state of condition to work without failure like a decreased accuracy. If you wanna bring in inacurracy than you need a override routine, what allow you to fire the lasers before cooldown is over. This actually then could be transformed by such a model into simuating a effect on the hitzone.
Dino Might, on 15 April 2015 - 08:42 AM, said:
There are tons of factors that can be incorporated to get us to reasonable numbers that balance the weapons/mechs, and all of this would happen behind the scene resulting in a completely intuitive experience for the user (seeing the effects of firing at high heat/movement rapidly vs. firing when cool and stationary with more time in between shots). If they wanted to, we could open up the black box and show the calculations.
Sure you can get this deep, however it is much effort to put it in, it can break other thing of the line of code, and the question is do you need it, if you assuem you have a T&T computer in a mech, which do compensate for all the conditional influences, for relive of the pilot, so that he can focus on tactial things and so on. Also it is a question of dataflow between server and client, more data more to break /ping?
If you would have a better damage model with smaller hitboxes, you may could simulate damage on weapons, for example if a laser mold your ac20 barrel, than this could effect the accuracy. Preassumed you don't have sume hit detection software that automaticly recalibrate to the new conditions. Or you are in in tourmaline and your lasers get peneliced for dust, on the optical system with more distance your mech has run over the terrain.
Or you could create a interdependance between runs you made with weapons and the R&R economy system as it was supposed long ago.
However it all needs developmenttime and someone needs to be payed for the manhours to ipmlement.
Almond Brown, on 15 April 2015 - 09:02 AM, said:
That could be our MechLab selectable/locked convergence range value. Only available to the Arm based reticules.
well you could have a button to make a copy movement of one arm to the other for symmetric reflection in the body axis (or in other words a deconvergance of both arms controlled by the user), however this would not be a independent movement of each arm. but at least you could fire left right and with the torso straight.
Edited by Kuritaclan, 15 April 2015 - 09:20 AM.