Pilot skill or probabilistic hit locations?
#201
Posted 09 March 2012 - 10:52 AM
If you fire a volley of missiles at a target, chances are you're letting your targetting computer do the work and the missiles will hit within a certain cone of fire.
However if you're using something like lasers then you probably want to take manual control and aim your shot - and this is where player skill comes into play.
#202
Posted 09 March 2012 - 04:42 PM
#203
Posted 10 March 2012 - 02:27 PM
My take: The TT rules are an approximation of some fictional reality. And that reality is not that weapons don't shoot straight. Because they do. We can build helicopter drones today that rapid-fire a mounted sniper rifle with pin-point accuracy. To my mind, the TT rules taken together emulate the fact that the mech pilot might not hit what he was aiming for. That emulation is not perfect, because the game has to stay reasonably simple to play. So instead of declaring the location you are aiming for, factoring in the enemy's orientation, the targeted area's size, and then also movement speeds and distances, the rules simplify it to "X% chance to hit at all, the location is just random".
So my point is, in a computer game, we have a more accurate emulation of the fictional reality. We can directly model the factors that may lead to the pilot missing his aim. Orientation, size of the target area, weapon type, movement speeds, Mechs rocking up and down. Also note that few weapons (if any) score an instant hit with full force: The lasers, as clearly seen in the gameplay footage, fire for ~2s and usually end up "smearing" all over the target. Ballistic weapons have a flight time to their target. Also visible in the footage, I think, is that missiles don't all hit the same spot.
So in my view, this whole argument only applies to the following kinds of weapons:
1. PPCs. Are they insta-hit? Do they do full damage immediately, or do they have a short duration, akin to the lasers?
2. ACs. Right now these seem to fire a single projectile (=> full damage on impact), even though I always thought they were rapid-fire, well, autocannons. So having them fire several rounds (e.g. 20 for an AC20? o.O) in rapid succession would make it suitably hard to insta-head an Atlas outside point-blank range.
3. Gauss cannons. These are obviously single-slug, full damage on impact. But that's their point, isn't it? Counter-balanced by several factors: Very heavy - I presume it's not easy to build a useful Gauss boat. Very slow rate of fire. Projectile flight time.
4. Machine guns. Not sure these are an issue, given low DPS and range?
And to all of the above, we add the inaccuracies caused by convergence delay. Note that convergence affects single-weapon shots as well. It's not about "crossing beams" but about crossing the line from your weapon with the line through eye and reticle.
Some questions remain:
1. Are there wind effects on projectiles?
2. Do ballistic weapons auto-adjust for gravity with a delay similar to convergence?
Edited by pesco, 10 March 2012 - 02:29 PM.
#204
Posted 11 March 2012 - 11:21 AM
Unless the simulation is *perfect* then it is missing lots of information. In a 'real' computer environment we can't simulate the effects of wind, random terrain, ammunition quality, barrel wear, 'sticky' actuators and all of that fun stuff.
Click-commandos who say 'the shots should go where I say they should go' is hooked on this idea that 'perfection' is attainable or even desireable.
I remember this same conversation coming up in MPBT:3025 - do we have the shots go where we aim, or should we use a probability donut to allocate shots? MPBT used 'shoot-hit' where the shots hit the location they were aimed at...no matter what... and it was *awful*.
Why - it severely rewards 'sniping' with no environmental factors making us miss, it is just too damned easy to zero in precisely on the target location (just shoot his head off) from max range. Um, a laser that shoots 'fast as light' and line straight? you would insta-head shot everything.
#205
Posted 11 March 2012 - 11:39 AM
2. Are they direct? Indirect?
3. What kind of weapons is it? Shotgun? Bullets? Rocket? Missile?
If you have ever shot anything, you will recognize how tough it is to actually hit something... from 700m away... never mind target an area... 1 degree of deviation 700m away = not on target...
#206
Posted 11 March 2012 - 04:02 PM
docmorningstar, on 11 March 2012 - 11:21 AM, said:
I remember this same conversation coming up in MPBT:3025 - do we have the shots go where we aim, or should we use a probability donut to allocate shots? MPBT used 'shoot-hit' where the shots hit the location they were aimed at...no matter what... and it was *awful*.
Why - it severely rewards 'sniping' with no environmental factors making us miss, it is just too damned easy to zero in precisely on the target location (just shoot his head off) from max range. Um, a laser that shoots 'fast as light' and line straight? you would insta-head shot everything.
Sounds like someone might be bitter because he got killed by snipers all the time.
Convergence is a nice in between, and it adds an extra element to gameplay that's nonetheless controllable to a good pilot. If you're intending to take a shot at an enemy from cover, I can see players lining up a random piece of terrain to set convergence at the right point before exposing to take their shot. I kind of like that.
The one good argument for cone of fire I've seen is with respect to lights. Every 'Mech game thus far has been crippled by lag, but lag also gives forcefields to small, fast targets at range. Abusing that forcefield was an essential part of being a scout.
I'm a little worried that in a game with good netcode and accurate weapons, scouts are stripped of their forcefield and might be too easy to cripple. That's a legitimate objection, as it could remove a major tactical and strategic element. "I don't like snipers", in my mind, is a lot less legitimate.
Edited by Belisarius†, 11 March 2012 - 04:08 PM.
#207
Posted 11 March 2012 - 04:48 PM
Then we factor in the human element of the maintainers along with good ol' Murphy giving them a helping hand...
Steady aim might work for those highly funded, deep pocket House Guard units but for the rest of us some old time guess-timation with some windage thrown in for good measure is just going to have to suffice.
Edited by Morashtak, 11 March 2012 - 05:36 PM.
#208
Posted 11 March 2012 - 05:11 PM
pesco, on 10 March 2012 - 02:27 PM, said:
Um ... yes... and?
It's the 'Mech that actully points the weapons. The pilot doesn't and can't actually point the weapons; he indicates to the 'Mech where it should point it's weapons. Wether or not the 'Mech can actually hit what's being pointed at is a factor of the environment and the firing 'Mech's condition, what the target is doing, and how well the firing 'Mechs pilot can handle the "higher thinking" for his machine.
Quote
And how ... even if this was actually true... is this a valid thing to relate to 'battlemechs, a fictional armored combat unit in a game set in a fictional universe into which we all go to "escape reality?"
Besides which, battlemechs can, with a single weapon, shoot things on the horizon, quite well. It's trying to get multiple non-guided weapons to all hit the exact same point that is the problem; and that is a problem we have not managed to overcome either.
Accuracy: ability to hit a point with a single weapon. Precision: ability to hit a point with multiple weapons.
Quote
The rules emulate for pilot skill and the ability of the 'mech to handle the weapons mounted to it AND the "bench" performance of said weapons; and they also do not make hit location "random" - the hit location rules indicate how precise a 'mech is at handling it's weapons under varying circumstances. They establish a useful set of data on how well a 'mech can handle it's weapons and how well or badly the green to average to elite pilot can handle a 'mech in combat.
Quote
... and the combat section of this fictional reality which is only defined in any way that is useful by the data that the TT rules express. No, I do not mean the form of the TT rules.
Trying to balance a MW video game off of purely arbitrary decisions of which parts of whatever is in vogue for physics at the moment is a recipe for unintended consequences, needless complexity in the back end of the video game, and no real returns in game play that would not be just as well handled on the back end by implementing the data from the TTr into the video game engine.
#209
Posted 11 March 2012 - 05:53 PM
Your own EDIT:("tsk tsk we dont allow this") poor reflexes, nervous twitches, lag, and computational power of your computer
iiiiiiiiiiiiiiiinnnnnnnnnnnnnnnnnn
combination with the movement of any two mechs is all the probabilistic your EDIT:("tsk tsk we dont allow this either") will ever need ingame.
promise.,
if you wanna sit and shoot at each other face to face while crouching, you are more than welcome to attempt to put each shot exactly where uuuuu want it to go.
ahemn
Knock.
and no, i read not a single post in this thread for months.. Just look at poll.
~S~ all.
Edited by Metro, 12 March 2012 - 01:36 AM.
lets try and be constructive =) Thank you.
#210
Posted 11 March 2012 - 06:26 PM
mekabuser, on 11 March 2012 - 05:53 PM, said:
What good reasons do you have for finding it disappointing?
Quote
iiiiiiiiiiiiiiiinnnnnnnnnnnnnnnnnn
combination with the movement of any two mechs is all the probabilistic your (EDIT) will ever need ingame.
promise.,
This was MW3 in online game play. You might not realize it, but what you're advocating is that the guy with the most money to spend on his computer and internet connection and controller hardware will virtually always win, and the only others that will come close will not be the people who play the game best ... it will be the people that play the netcode the best by lag shooting skill, instead of skill in piloting a battlemech.
#211
Posted 11 March 2012 - 06:35 PM
So do I understand it correctly that some people still lobby for the "perfect aim" shooting, thereby negating all environmental, gravity and other influencing factors? Would make me wonder if they ever so much as aimed or even shot with a gun in real life...
#212
Posted 11 March 2012 - 06:45 PM
#213
Posted 11 March 2012 - 06:55 PM
gravity , of course. ballistic drop....
like i said, even standing still, you are going to be hardpressed to put round on target exactly where u want them. The enemy can still torso twist dissapating the incoming fire.
add movement to the equation and it becomes more than anything the realm of pilot skill. Not cpu power and all that.
forget mw3. Mwll , a modern version of this universe illustrates this perfectly.
Spectate a round or two and take note of what percentage of shots hit.
Its way lower than you would imagine. Most shots miss because most pilots are trying to make hardcore shots.
Lazers with duration further emphasize this point.
ppc, gauss? well there are elite pilots who can do a damn good job with them, but its still not easy to put them exactly where you want against an aware enemy.
The percentage of people who have such elite systems AND good internet is not alot. An elite system is also no indicator of pilot skill, OR that said system is optimized for gaming.
There are many factors conspiring against "pinpoint accuracy"
Id further be surprised if the game is silky smooth in a furball, but then again, a furball its instinct and aggression.
pht. no . about the lag and all that. We play regularly in high ping, aND>.. low ping servers. The way the netcode is in mwll it really doesnt make that much of a difference. And this is on a wonky mod. If the game is running bad enough that your not hitting your target properly more than likely host server, or you need a reboot. Otherwise, its runs pretty damn good.
this is also not mw4, insta lazer death. Which is probably more than anything the complaint against pinpoint. We have duration lazer. Its moot point. Ballistic arcs and lazer duration means that even if everyone had great systems its still not easy to hit your target.
Its almost as if this debate is tied into the NR concept. Add more probability and duration to any battle to maximize the one life you get.
Hmmmm...
Edited by mekabuser, 11 March 2012 - 07:05 PM.
#214
Posted 11 March 2012 - 07:23 PM
Dlardrageth, on 11 March 2012 - 06:35 PM, said:
So do I understand it correctly that some people still lobby for the "perfect aim" shooting, thereby negating all environmental, gravity and other influencing factors? Would make me wonder if they ever so much as aimed or even shot with a gun in real life...
That's not the point. I don't advocate perfect aim that ignores all factors, I advocate controllable aim that takes into account all factors that the game can represent. Cone-of-fire is a cop-out mechanic to hide things that can't be implemented, and therefore can't be mastered.
I would be all for having to account for wind, gravity, movement of my 'Mech etc. In fact I want that, very much. But what I require, and what I think is essential to give the game a high enough skillcap that it's competitively viable, is that a perfect pilot who perfectly compensates for all factors can have perfect aim.
I'm happy for that to be impossibly hard to obtain, as in real life. I'm not happy to throw up my hands and give things to the RNG gods, because as soon as you do that you set a bar that players cannot improve past.
Edited by Belisarius†, 11 March 2012 - 07:27 PM.
#215
Posted 11 March 2012 - 07:41 PM
Granted, weapons systems for Energy, Ballistic and Missiles behave differently, but I would really like to have a choice between feeling like I'm actually "piloting" something in a Mech combat sim, or when I feel like I'm just "playing another flavour of FPS set on a fictional future." The former is more attractive to me personally, especially when you mention the name BATTLETECH/MECHWARRIOR. The latter, I fear "might" cheapen the whole thing when the Devs are claiming that they're taking this how a pilot will actually experience visually.
Steady hands, of course, play an important role, but that's practically the same for someone driving a tank, a fighter plane or even a ship. Your pilot's hands aren't doing the real aiming (Unless they decide to do a control system inside the battlemech that basically mimics the pilot's movements --- like that Robot Jox film in the 80s). Your pilot's effectively "driving" the machine. A computer translates your control manipulation to a series of commands that will make parts of your said machine to behave a particular way. All those simple independent actions culminate into accomplishing the act of "aiming". Your pilot can only adjust to the visual cues that the machine will feed back to him/her in the form of the HUD. Many things can happen along the way between your pilot manipulating the controls, the machine following the translated instructions and the actual hardware trusted to behave the same way constantly... elevation changes, weather changes, wear and tear, battle damage. The entire equipment can only be expected to behave whithin those accepted lines. And let's not get started with adjusting the distance where your weapons will converge on the same dot on your targeting reticle, because that's another related (and lengthy) matter to take into account.
steady hands should be rewarded, yes. But only to a certain extent where the probabilities will now affect it. This isn't Unreal Tournament, Battlefield, Call of Duty, etc. This is Battletech/Mechwarrior.
My other concern with the first poll choice is the threat of "aimbots". I'm not party pooper, but I've had my share of playing online/multiplayer games where the aiming is reliant on the user specifically... someone gets lucky/smart and develops a little tool to make things easier for him and there goes the love for the game for the rest of the players. (Irony if and when this happens? Oh look, they're relying on targeting computers, afterall)
these are just my Php 0.02 on this matter. Value will change vary from player to player
#216
Posted 11 March 2012 - 07:55 PM
Belisarius†, on 11 March 2012 - 07:23 PM, said:
That's not the point. I don't advocate perfect aim that ignores all factors, I advocate controllable aim that takes into account all factors that the game can represent. Cone-of-fire is a cop-out mechanic to hide things that can't be implemented, and therefore can't be mastered.
It also can be (has been) a means to limit the sheer avalanche of calculation processes a "realistic" modeling of all factors would require. Yes, of course a RNG is somewhat of a crutch, but even with modern technology these days it is not always possible to simulate exactly everything in a given situation, unless you got multi-million dollar systems and lots of time.
Quote
Maybe I misunderstood you here, but are you saying that you expect that theorethical "perfect pilot" to account and compensate for all environmental factors? Suspension of disbelief might become a bit strained here, even with a relatively simple factor as wind. To have an impact on a missile or ballistic weapon shell you don't need the wind to change from a light breeze to a full-grown typhoon. Let's push it to the limit here then. Even an educated metereologist can't tell you pinpoint-exact all the time how the wind will exactly behave in 5 minutes. And you expect your fictional pilot to pull that off? For this one of many factors?
Moreover... I recall a thread where we discussed damage modeling and how you'd need very complicated multiple vector-based calculations for each shell to even get an approximately realistic mechanic. When factoring in impact angles etc., which admittedly goes way beyond usual BT damage modeling like we know it. I think I'd come to a similar conclusion here like I did there, that all what is theorethically doable is not necessarily a good thing to implement. Because the raw amount of data could still overwhelm/lock down the one big chokepoint that still exists - the connection between server and client. Thus - albeit grudgingly - I admit that some RNG-based simplification might be necessary here and there for the sake of playability.
Quote
I sure hope it doesn't come to that. The sheer ridicule that RNGs make of gameplay, like in WoT sometimes for example, where it often becomes a "RNG to rule them all", is a guaranteed letdown for many players. Thus I'd personally like to keep the influencing factor of the RNG modest to slim, but I don't think you can completely rule it out. Because that would either make the game too "massive" for decent online playability, or simply unrealistic by negating too many influencing factors. At the end of the day even the "perfect pilot" who aims perfectly cannot guarantee the perfect hit at, say, 500m+, because he is not the wind god and cannot simply command all environmental factors to do his bidding.
I still hope for and think it possible to have some sort of RNG to simulate effects not actually calculable in detail for gameplay. As long as it isn't one who makes me miss a tank the size of a small house from 40 meters out and have my 122mm shell fly up into the skies when I aimed for his tracks...
#217
Posted 11 March 2012 - 08:00 PM
Pht, on 11 March 2012 - 06:26 PM, said:
OOI: Is this also a problem in MW:LL?
Pht, on 11 March 2012 - 05:11 PM, said:
I'll admit to an only cursory knowledge of the boardgame rules. However: Hitting an Atlas in the head, rather than at all, should be much easier at short range than at 1000m. How do they reflect this?
From a different angle: Say I'm at point blank range to that Atlas and manage to aim for its head, pulling the trigger in just the right moment. Would you like a die to be rolled in the background and the shot go to the toe instead?!
Back to the big picture: If a Mech is supposed to be able to hit anything at all at max range of its long range weapons, any imprecision in the weapon must be very near negligable at shorter ranges. If it turns out that impossible head-sniping occurs too often, sure, let's have a few millidegrees of randomness in the firing vector. I'm just not sure these will (and should) have any net effect on much else than your max range large laser or Gauss cannon shots. (Not counting the "convergence" effect we already have, and which I support!)
#218
Posted 11 March 2012 - 09:01 PM
Dlardrageth, on 11 March 2012 - 07:55 PM, said:
Then we're on the same page and we differ only in what we think is acceptable variance. I think there's enough variance in the existing mechanics to make a cone-of-fire unnecessary, especially if you added simple, static wind and gravity vectors. I would not be opposed to a small RNG variation if that was the only way to stop someone getting headshots at 1000m 100% of the time, but it isn't.
Firing at long range, even if you perfectly predict travel time, wind, gravity, recoil, whatever, you're at the mercy of your enemy's movements during the time your bullet is in the air. What was a headshot might instead clip his shoulder if he turns. That's the kind of variance I'm happy with; the kind I can explain if not always correct for. Even my perfect friend can't correct for that one, but he could - for example - know that the sound of a missile lock will cause his opponent to twist away pre-emptively, making a follow-up gauss shot more predictable.
I would, at this stage, be opposed to a cone of fire that could cause my shots to miss by more than my opponent's movements already do. And because such a small cone of fire is, by definition, useless as a balancing mechanic - because the other guy could instead just learn to move without crying imba - I'm opposed to CoF in general.
The only exception to this in my mind would be lasers on light 'Mechs at long range, as I've already stated. If lights turn out to be trivially easy to kill even when pulling out all the evasive stops, I would reevaluate cone of fire. But before I did that I would look for other, controllable factors (like convergence, which I think is a great idea) that could help a light survive while avoiding a situation in which my own noob butt is as good at light-hunting as the perfect lancemate I keep talking about.
Edited by Belisarius†, 11 March 2012 - 09:09 PM.
#219
Posted 11 March 2012 - 09:02 PM
pesco, on 11 March 2012 - 08:00 PM, said:
If I were the GM of that particular campaign, I'd say frakk that. I wouldn't be stupid enough to stick to the book at what the dice result would be. I weight in a batch of common sense on it. If the roll's successful, I'd let the head hit clean. If there's a certain degree of failure, I'd let it grace the head or even hit the torse and only graze the head. Only a silly GM would allow that go to the Legs/Feet/Toe if the aimed shot failed when all reason points clearly that the weapon was aimed upwards at close-range. (I'd adjust the hit locations properly for the table/chart when the player will roll for that given the circumstances... I will not let him/her roll for the entire hit chart where it doesn't seem plausible.) <--- this here is a proper use for degree of probability.
-------------------
But that's beside the point of discussion in this thread. XD
I've read everyones' posts so far...and the polls are swaying between option 1 and option 2.
i.e. There's only so much aiming you can do at a point-blank shot, especially when your weapon of choice is hard-mounted to your mech's chest/torso and has very little angle/arc of adjustment. Best you can do is title your torso upwards and hope the distance isn't too close for the rest of your target's anatomy to make the shot difficult/impossible.
That's one of the few situations where any of the three aiming choices above will fall on their faces on. There's stil the physical limitations brought on by the mechs/weapons themselves that are taken into account by the gameworld/gamesystem that neither anyone nor any HUD targeting reticule will be able to compensate for.
These kinds of situations are merely examples of how and why well-aimed shots will never always land where the pilot wants it to. Nothing's perfect, even in a fictionally futuristic and state-of-the-art world where warfare is their life's definition. May it be long-range or short-range... the degree of deviation will vary, of course.
Edited by Kleine Vidan, 11 March 2012 - 09:15 PM.
#220
Posted 12 March 2012 - 01:37 AM
2 user(s) are reading this topic
0 members, 2 guests, 0 anonymous users