Cavale' Post='3768731 said:
I tried to read this and understand it. As I read, the words seemed to slowly melt and drip away in utter madness as this horrible, horrible word salad of a suggestion melted my frontal lobes.
When you can't find a good reason to reject something, and you won't simply say you just don't like it ... do the above!
Quote
This is garbage. Pure and simple, this is akin to applying Random dice rolls to a shooter, from what little bits of comprehensible words I can divine meaning from.
----
Pretty much, you are attempting to apply a combat system from a turn based RPG to a Multiplayer live action system.
----
This is a bad idea. You are aware that human mechanics, just from reaction time and being unable to control other players own reactions create highly dynamic odds for proper shot placement as it is?
----
You are adding the human to human margin of error to an RNG to create a substantially lower odd of hitting. This is poor design.
You to admit that you couldn't comprehend it, but you still somehow call it "garbage." ... exactly the same thing as saying "this math is wrong" when you've admitted that "I don't comprehend this math."
----
How can you say this when you've already admitted that you didn't comprehend the OP (which wouldn't be bad, if not for your irrationally hateful attitude). "Turn based RPG" - shockingly, I believe our computers can add small numbers and choose between 2-12 and 1-6 in real time - something pocket calculators could do more than 13 years ago:
http://www.pryderock..._battletech.zip
----
It's possible to have 100% of all weapons hit, and the things needed to get this result are known and predictable - and practically doable. Hardware choices, gameplay choices, and human skill all can drive the weapons-connection percentage up to 100% - you just don't get to repetitively "pick your part" as long as you can hold the crosshairs over it and click at the right time. You would be piloting an armored combat unit instead of a large quake 3 avatar.
----
If this were enough of a factor than the armor and damage numbers would never have had to have been changed; and people would not STILL be requesting another doubling of them. Are you aware - even if you don't care - that there is, beyond arms/all else, NO 'mech weapons handling capability modeled in MWO?
IZeratulI said:
Randomized damage spreads decrease the relevance of skill and increase the relevance of luck.
----
FPS gamers as a demographic normally want to win or lose based on their own merits. They don't want to win or lose a game because a dice roll, outside of their control, determined success or failure.
One could just as validly say that perfect weapons convergence decreases the relevance of thought and increases the relevance of lucky clicking.
But that wouldn't be the whole story, would it?
It is possible to add reward into the game for thoughtfulness and training in what choices to make, in combat and out of combat, without removing skill as one of the absolutely necessary driving factors. Shockingly, some people want a Mechwarrior game where the 'mechs perform like ... mechs! Where knowing your 'Mech is of at least equal importance as skill controlling a crosshair on a HUD and timing a shot.
----
Skill in placing a crosshair and choosing when and what to shoot is NOT the only "merit" that exists to be rewarded. "Win or lose ... because a ... outside of their control." Here you are wrong - what spread happens... happens directly because of player choices and aiming skill - and the factors that affect the spread are known, knowable, and predictable. For example, you, as a player, would KNOW that things didn't hit because you were trying to hit a target at long range for your weapon, while your 'Mech was overheating. Not only that, You'd know exactly what to do to fix that situation - either get a longer ranged weapon, wait till the target was closer, and not try and make the shot while you were overheating.
Asyres said:
Systems which are random are inherently unpredictable, except in the macro sense - and even then, you can only predict the frequency of an outcome, not what outcome will occur in a specific instance.
----
I've read your entire OP twice now (thanks, by the way, for assuming it that I hadn't, or that I'd failed to understand it - who's being condescending, again?), and I still think it's a terrible idea.
Either a thing
is unpredictable in any sense, or it
isn't predictable in any sense. If by "random" you mean "unpredictable in any way" ... this is why these discussions are so hard. Nobody explains what they mean. Or ask what others mean.
Thank you for admitting that yes, it is predictable. Meaning it can be known. In this case, known and thus controlled for.
----
I didn't have to assume that you failed to understand it -
you exhibited that you had: "Suggesting that what you've presented is not inherently unpredictable" if a thing is inherently unpredictable - than it can't be predicted in any way - macro, micro, you name it. Yes, I presumed you hadn't read the whole post; because of your missing the point; even if you disagreed with it.
Mystere said:
My order of preference for convergence is as follows:
Manually-adjusted (default set at Mech Lab, adjusted in-match via mouse wheel or keyboard)
fixed (set at Mech Lab)
none
From what I know, automatic convergence cannot be fixed due to issues with the CryEngine itself.
So, what do you think of the 'Mech itself doing the convergence?
Xarian said:
OP, your post describes the system as how it should end up on average after all game mechanics are taken into consideration. That is, if you aim for the head, you should hit the head more often than you hit the feet with various levels of probability (I'm not going into the fine details of your percentages; those are secondary to the overall idea).
However, this in no way addresses the road to actually get there. This is a summary of your suggestion, in the most positive light possible:...
----
1. Player aims at a specific location
2. ???
3. X% chance of hitting cockpit, Y% chance of hitting torso, etc.
Thank you for reading and considering before you replied.
At this point I would be happy as a bobcat locked into the hen-house if the 'Mech's weapons handling were put into the game like it should have been. However, I don't expect that to happen, which is why I said in the OP that it wasn't a suggestion - at least not a suggestion for MWO. I consider that ship to have irretrievably sailed.
----
You're absolutely right. I didn't post the math or the rules they work with. At this point, it's hard enough to get people to understand that human aiming skill and human choices WILL be the factors that control gameplay outcomes. The OP is pretty much strictly an example of gameplay witht he 'Mech's weapons handling performance put into the game, as it should be. The exact specifics are in another post I made a while back in the suggestions forum - a post I need to update with a few corrections.
Quote
You can use a brute-force probability method - figure out which hitboxes a shot hits, then inflict damage randomly with an appropriate formula. This is also the absolute worst way to handle a skill-based game such as MWO or any other first-person game - this isn't an abstract-combat level game, such as tabletop battletech, warhammer, etc. You don't need random numbers to figure out where the weapons hit - they fly out of a gun and hit somewhere on the enemy mech.
"inflict damage randomly" - It's only bad if by random you mean "unpredictable"
and therefore not controllable by player skill and choices - which obviously wouldn't apply here, because the results are predictable - all the percentages are known and all the factors that control them are known - and are used based upon player skill and player choices.
Quote
Inflicting raw probability like this is just going to end up in a situation where you put your gun barrel-to-eye with a non-moving enemy mech, pull the trigger, and hit his left torso. If you adjust the probabilities to take into account other factors, like how fast your weapon and target are moving, distance to target, etc, you end up with the game exactly as it is right now. Those factors are already taken into consideration by the simulation itself, and it's adjusted for player skill.
As I said clearly - the examples listed aren't comprehensive, besides which the damage and armor numbers that will be used are BUILT to be used with said percentages. Even the online real-time video game format HAS to abstract things.
It would be perfectly acceptable to have a math table for for "target immobile, shooting unit standing as still as possible," obviously keeping in mind the damage and armor numbers.
No, you wouldn't end up with the game exactly as it is now - for the simple fact that BattleMechs aren't like other kinds of mecha - they are not capable of getting every weapon to reliably hit that tightly under the crosshairs; and there's nothing wrong with that; and this fact has been confirmed by the horse's mouth.
Quote
So you need to think of pseudo-realistic, in-game mechanics that will help to eventually get you to these hit probability distributions and increase some of the problems with called shots being too easy to hit at medium range. Many examples:
1. Weapon convergence, or lack thereof - torso weapons converge much more slowly, meaning that they will not likely hit where you are aiming
2. Making weapons converge worse when moving quickly or firing quickly (similar to "cone of fire" except making it affect the weapons themselves instead of just making the bullets fly randomly - the weapons would behave realistically based on the conditions that caused them to move/shake)
3. Separate aiming reticles for each weapon with some sort of gimbal (or no gimbal) for various mechs and hardpoints to allow for tweaking convergence
1: Actually, this is already in the lore: Weapons mounted in a 'Mech's torsos don't converge any slower than weapons mounted in it's arms. All weapons pretty much converge at the same speed.
2: "weapons converge worse" - you seem to be missing the point that the 'mech that converges the weapons. This convergence ability is already known in black and white numbers. Some weapons are easier for the 'Mech to converge, and some are harder. Most are baseline.
3: Seperate crosshairs is only valid here for firing arcs that not all weapons can aim into (rear mounted weapons, arm swinging around into a side arc). The crosshairs on the hud don't tell you where any given weapon is pointing at - they tell the 'Mech where to try and get the weapons pointed. Other visual/audible cues would inform the pilot of the quality of alignment.
Quote
You'll notice that what all these methods have in common is the effect of causing physical forces on the mech, rather than just randomly deciding "you hit here because the RNG rolled a 97", or "your projectile files at a 45 degree angle instead of a 43 degree angle because the RNG rolled a 15".
You seem to have missed the point that the percentages represent the 'Mech's physical and computational weapons-alignment capability in any given situation of getting the weapons mounted in it aligned at what the pilot is aiming at.
BattleMechs don't perform like giant quake 3 avatars - they perform like armored combat units out of a specific fictional setting. The units in the BT setting that DO behave like humans toting guns would be... humans toting guns, battle armors and possibly protomechs. If you wanted to try and do the impossible and do full physics (which themselves are a pure abstraction and not predictive) for each and every weapon that fired out of a 'mech, you'd STILL have to resort to those percentages if you wanted a MechWarrior game, instead of a "generic giant stompy bot that shoots other bots" game.
Edited by Pht, 28 September 2014 - 02:46 PM.