378 replies to this topic

[Mackman]

cyberFluke, on 15 April 2013 - 07:14 PM, said:

First, following the same example you gave, they would both be moving full speed and it's down to twitch reflex, the rookie panics and hit the CT of the vet and drops him. What's the difference exactly?

Second, I'm not asking for a reduction in the impact your supposedly superior skill has, I'm asking to expand that skill into far more than pure twitch reflex. The current system is too simplistic and arcadey to hold players for a significant length of time. Whether it's a randomised COF (cone of fire) or a similar system I don't care, but the current system *has* to be expanded somehow. That's what this discussion is about, but a lot of people (and I use that word loosely) can't (or won't) see past the word random.

For *anyone* else that wants to just call me a low skill scrub, post *your* stats before doing so, or you're the scrub. I've linked mine *twice* and will happily do so again. I'm aware I'm not an "uber" player, but I'm likely better than your sorry ***.

In the current system, the Vet's faster reflexes and better aim earn him the win every time. In the randomized system, the rookie has the chance to get a lucky kill just by firing first and risking the RNG.

Will it work every time? Of course not. But in a game where you have no respawns and every kill is permanent, a purely random occurrence has the potential to be literally game-changing... and when the teams are closely matched, then the odds of randomness deciding the outcome goes even higher.

As for your "come at me bro" attitude, not once did I make so much as a passing reference to you or your skill. Don't take everything so personal, dude.

[Vapor Trail]

Royalewithcheese, on 15 April 2013 - 06:02 PM, said:

Well, if it's MW2 the vet is probably going to bust out the akimbo G18s, akimbo 1887s, or SPAS for running around corners

Yeah... but the G18s were high damage bullet hoses, and the 1887s had fairly obvious issues in the range department. SPAS was probably the best balanced of that bunch... and yet there was the issue that shotguns were secondaries, rather than primary weapons.

I kinda enjoyed being able to tote an Assault Rifle and a Shotty... but I enjoyed it because it allowed me to multi-range without significant penalty.

Edited by Vapor Trail, 15 April 2013 - 07:43 PM.

[Zeroskills]

Vapor Trail, on 15 April 2013 - 05:32 PM, said:

Like TDGSolid said, that 'Statistically, RNG always favors the underdog.' statement is pure bovine excrement.

False. Here is some math for you:

Lets say the RNG is adjusted to turn 10% of misses into hits and 10% of hits into misses. Player A has 60% accuracy and Player B has 40%. After the RNG gets done with them Player A will be at 58% and Player B will be at 42%. Seems like it favors the lesser skilled player too me. You can use any variation of percentages you want, it will always come out in favor of the less skilled player, the better player has more hits that will be turned into misses and the less skilled player has more misses that will be turned into hits.

[Ansel]

cyberFluke, on 15 April 2013 - 07:14 PM, said:

An interesting idea, how would you set about changing the system?

I would have kept the armor sections and the point totals for those sections, weapons would then only deal damage in a footprint to those sections instead of dealing damage to the whole section.

Each weapon would then have it's own footprint, for example the AC-20 would have the largest footprint of the balistic weapons, and the MG having the smallest footprint, missle and spread type weapons would have very large footprints to go with their lower damage so they generate a lot of overlaping area damage, also making them better for splasing damage into areas that the armor was already breached.

If this were implimented fights would take quite a while longer, so I would probly drop armor point totals to 125-150% of TT values instead of 200%.

[cyberFluke]

Mackman, on 15 April 2013 - 07:34 PM, said:

As for your "come at me bro" attitude, not once did I make so much as a passing reference to you or your skill. Don't take everything so personal, dude.

The low skill reference wasn't aimed at you fella, but the staggering number of posts before yours that offered no input other than to question my skill. Have a look man, it's not even funny.. :/

[Zeroskills]

Ansel, on 15 April 2013 - 07:48 PM, said:

I would have kept the armor sections and the point totals for those sections, weapons would then only deal damage in a footprint to those sections instead of dealing damage to the whole section.

Each weapon would then have it's own footprint, for example the AC-20 would have the largest footprint of the balistic weapons, and the MG having the smallest footprint, missle and spread type weapons would have very large footprints to go with their lower damage so they generate a lot of overlaping area damage, also making them better for splasing damage into areas that the armor was already breached.

If this were implimented fights would take quite a while longer, so I would probly drop armor point totals to 125-150% of TT values instead of 200%.

I really like this idea, but I imagine it would be a pain to code.

[cyberFluke]

Ansel, on 15 April 2013 - 07:48 PM, said:

I would have kept the armor sections and the point totals for those sections, weapons would then only deal damage in a footprint to those sections instead of dealing damage to the whole section.

Each weapon would then have it's own footprint, for example the AC-20 would have the largest footprint of the balistic weapons, and the MG having the smallest footprint, missle and spread type weapons would have very large footprints to go with their lower damage so they generate a lot of overlaping area damage, also making them better for splasing damage into areas that the armor was already breached.

If this were implimented fights would take quite a while longer, so I would probly drop armor point totals to 125-150% of TT values instead of 200%.

That'd be a lot of data to track, transmit and sync to all players in the game... Maybe doable though, if each mech's damage model was stored as an array of 3D arrays, you apply your "footprints" to the model as addition of integers to the relevant elements of the array and a simple "if element value > armour amount then breached" starts applying damage to the internals model. Could even do the internals the same way.

You can then update the paperdoll to a spinning 3D model for example, showing in increasingly darker shades of red the patches on the mech where the damage is....

That would be a much better way of doing things, you're entirely right. Unfortunately it would require a ground up rewrite of how the damage system works... :-(

Edited by cyberFluke, 15 April 2013 - 08:03 PM.

[Ansel]

If it were implimented with the ability to easily change the footprints it would be the best balancing tool the developers have. That alone would make it worth doing

Look at the last missle changes, if the system I detailed were in place all that would have been required would have been a footprint reduction so fewer missles were splashing the same spot, reducing the overlaping damage without even touching the base damage of the missle.

That is what is lacking with a lot of weapons right now, like the LBX-10, where they spray out a lot of low damage shots that mean nothing if they all don't hit the same section.

[Lord Psycho]

sounds like you want to add penalty-to-hit on a tabletop game...wait, is this a tabletop game?

why not just tell them to up the heat values of the laser weapons ?

don't want to be mean but the 4P hunchback...was made for boating

[cyberFluke]

Ansel, on 15 April 2013 - 08:13 PM, said:

If it were implimented with the ability to easily change the footprints it would be the best balancing tool the developers have. That alone would make it worth doing

Look at the last missle changes, if the system I detailed were in place all that would have been required would have been a footprint reduction so fewer missles were splashing the same spot, reducing the overlaping damage without even touching the base damage of the missle.

That is what is lacking with a lot of weapons right now, like the LBX-10, where they spray out a lot of low damage shots that mean nothing if they all don't hit the same section.

Yup, I like this idea more and more. The missile issue is more complicated than just the dmg they do, when investigating the Super LRM bug they found that the pathing code was malfunctioning and all guided missiles are seeking to the CT (If I understand correctly). As a *temp* fix they've dropped their dmg until the code is fixed.

This idea would definitely help add the depth *I* was expecting from a modern MechWarrior game, don't know about anyone else.

Edited by cyberFluke, 15 April 2013 - 08:36 PM.

[Vapor Trail]

Zeroskills, on 15 April 2013 - 07:48 PM, said:

False. Here is some math for you:

Lets say the RNG is adjusted to turn 10% of misses into hits and 10% of hits into misses. Player A has 60% accuracy and Player B has 40%. After the RNG gets done with them Player A will be at 58% and Player B will be at 42%. Seems like it favors the lesser skilled player too me. You can use any variation of percentages you want, it will always come out in favor of the less skilled player, the better player has more hits that will be turned into misses and the less skilled player has more misses that will be turned into hits.

That makes quite a few mathematical assumptions that I'd like to see backed up.

But here's some counterpoints.

First
The RNG wouldn't be "tuned" to make any such thing. The most sane approach I've seen is commonly referred to as "Cone of Fire" where the impact point of any given weapon is within a cone of where the reticle sits at the time of firing. If you're on target with the crosshair, you would be very likely to achieve a hit, and not very likely to miss, unless you were firing many weapons close together, while radically high on heat, while moving at a run, and while having several actuators with damage or destroyed.

The (sane) RNG approach isn't (repeat, is NOT):
"Oh, wait. You failed your gunnery roll... you can't hit anything with this shot regardless of your aim... you just shoot at the broad side of the planet and miss."

It's more along the lines of this:
"Your weapon being fired is this (base (in)accuracy), your heat status is this (additive, possibly multiplicative threshold values), your acutator damage status is this (additive, possibly multiplicative values), you're moving at this % movement rate (additive, possibly multiplicative values), so therefore your max inherent inaccuracy is "maxVal" arcseconds from the reticle... so your shot's going to be placed at (RNG 1value = {0,1, 2, ..., maxVal - 1, maxVal}) arcseconds from the reticle at (RNG 2value = {0, 1, 2, 3, ..., 358, 359}) degrees direction from a default 0 degrees position."

Second:
Why would someone who's missing 60% of their shots have more of those "missed shots" be close enough to "on target" where the RNG might drag them there? Why would someone who's missing only 40% of their shots have more of them be so far away as to have fewer dragged onto target by the RNG? I would think it would be the opposite of what you state.

I would think that the person who is missing more shots because their point of aim is more often off target would tend to lose more accuracy than they would gain simply because they're further off target to begin with than the person who is not missing more shots.

Basically, the closer you are to "on target" to start with the better the RNG would favor you. In short, I believe your "statistical argument" is flawed outright.

Edited by Vapor Trail, 15 April 2013 - 09:42 PM.

[Mackman]

Vapor Trail, on 15 April 2013 - 09:18 PM, said:

That makes quite a few mathematical assumptions that I'd like to see backed up.

But here's some counterpoints.

First
The RNG wouldn't be "tuned" to make any such thing. The most sane approach I've seen is commonly referred to as "Cone of Fire" where the impact point of any given weapon is within a cone of where the reticle sits at the time of firing. If you're on target with the crosshair, you would be very likely to achieve a hit, and not very likely to miss, unless you were firing many weapons close together, while radically high on heat, while moving at a run, and while having several actuators with damage or destroyed.

The (sane) RNG approach isn't (repeat, is NOT):
"Oh, wait. You failed your gunnery roll... you can't hit anything with this shot regardless of your aim... you just shoot at the broad side of the planet and miss."

It's more along the lines of this:
"Your weapon being fired is this (base (in)accuracy), your heat status is this (additive threshold values), your acutator damage status is this (additive values), you're moving at this % movement rate (additive values), so therefore your max inherent inaccuracy is "maxVal" arcseconds from the reticle... so your shot's going to be placed at (RNG 1value = {0,1, 2, ... maxVal - 1, maxVal}) arcseconds from the reticle at (RNG 2value = {0, 1, 2, 3, ... 358, 359}) degrees direction from a default 0 degrees position."

Second:
Why would someone who's missing 60% of their shots have more of those "missed shots" be close enough to "on target" where the RNG might drag them there? Why would someone who's missing only 40% of their shots have more of them be so far away as to have fewer dragged onto target by the RNG? I would think it would be the opposite of what you state.

I would think that the person who is missing more shots because their point of aim is more often off target would tend to lose more accuracy than they would gain simply because they're further off target to begin with than the person who is not missing more shots.

Basically, the closer you are to "on target" to start with the better the RNG would favor you. In short, I believe your "statistical argument" is flawed outright.

The person who has more skill to begin with has more to lose, for one thing. He has less ability (note that I said less, not none) to directly affect the outcome: He is less likely to hit what he is aiming at.

While the same certainly holds true for the person with less skill, it affects him less than it affects the person with higher skill.

You have also failed to address the situation I laid out earlier. RNG makes it possible for a lesser-skilled pilot to defeat a greater-skilled pilot due to sheer random chance: When the pilots are closer in skill, as in a high-level competitive match, it becomes even more probable for a skirmish to come down to which cone of fire is randomly more accurate: And in a competitive match, the outcome of a 2-3-mech skirmish could easily decide the fate of the entire match.

Heat, number of weapons fired at once, movement speed... the vet is going to be monitoring all of these, and he can still be taken out by a lucky alpha that shouldn't have hit.

[cyberFluke]

Mackman, on 15 April 2013 - 09:42 PM, said:

The person who has more skill to begin with has more to lose, for one thing. He has less ability (note that I said less, not none) to directly affect the outcome: He is less likely to hit what he is aiming at.

While the same certainly holds true for the person with less skill, it affects him less than it affects the person with higher skill.

You have also failed to address the situation I laid out earlier. RNG makes it possible for a lesser-skilled pilot to defeat a greater-skilled pilot due to sheer random chance: When the pilots are closer in skill, as in a high-level competitive match, it becomes even more probable for a skirmish to come down to which cone of fire is randomly more accurate: And in a competitive match, the outcome of a 2-3-mech skirmish could easily decide the fate of the entire match.

Heat, number of weapons fired at once, movement speed... the vet is going to be monitoring all of these, and he can still be taken out by a lucky alpha that shouldn't have hit.

Your argument is based on the assumption that all the weapons will be directed at the same random point and not to their own floating aimpoint. The large reticule is merely an indication of the area in which the weapons may individually strike.

[Brilig]

I'm a little late to this particular party, but figured I would weigh in.

I do think convergence is an issue, but I do not like the idea of RNG. It doesn't add skill, it adds a random chance to hit that I would rather not have to deal with.

I have nothing against boats, or alpha strikes. What I don't like is that its super easy to put 5 LL or 6PPCs into the same exact spot all at once. It makes boating the most viable tactic, and makes variants capable of boating significantly stronger than the others.

It's also not about boats being unbeatable or anything like that. Its about getting some variety back into mechs. I'd like the mechlab to be for more than figuring out how many of 1 weapon type you can cram in.

I've thrown this idea around a few times, I know a few other people have some similar ideas, so I'll throw it out again. Also I know how tanks work now adays, I am sure future tech wizards could make every gun converge on 1 spot with no issues. That being said this is a video game, and the aforementioned tech wizards haven't figured out how to make a half ton machine gun shoot for more than 90 meters.

Brilig, on 29 March 2013 - 02:26 AM, said:

My suggestion to "fix" that would be to change the cross hair system around. You can keep things skill based, and avoid pinpoint damage issues from heavy alpha strikes.

For instance here is what the Hunchback 4Ps cross hairs could look like.

That would keep the weapons from being pinpoint, without the random number generator frustration. Each torso based weapon or weapon grouping could have its own cross hair. Nothing too extreme, but enough to spread the damage around. People can keep alpha boats, but they aren't quite as devastating.

The big drawback is having multiple crosshairs to worry about. I figure you can alleviate some of that by putting them close enough together to make it easy to deal with, but still far enough apart that getting all your hits on one section wont happen.

[cyberFluke]

Brilig, on 15 April 2013 - 10:16 PM, said:

I'm a little late to this particular party, but figured I would weigh in.

I do think convergence is an issue, but I do not like the idea of RNG. It doesn't add skill, it adds a random chance to hit that I would rather not have to deal with.

I have nothing against boats, or alpha strikes. What I don't like is that its super easy to put 5 LL or 6PPCs into the same exact spot all at once. It makes boating the most viable tactic, and makes variants capable of boating significantly stronger than the others.

It's also not about boats being unbeatable or anything like that. Its about getting some variety back into mechs. I'd like the mechlab to be for more than figuring out how many of 1 weapon type you can cram in.

I've thrown this idea around a few times, I know a few other people have some similar ideas, so I'll throw it out again. Also I know how tanks work now adays, I am sure future tech wizards could make every gun converge on 1 spot with no issues. That being said this is a video game, and the aforementioned tech wizards haven't figured out how to make a half ton machine gun shoot for more than 90 meters.



The big drawback is having multiple crosshairs to worry about. I figure you can alleviate some of that by putting them close enough together to make it easy to deal with, but still far enough apart that getting all your hits on one section wont happen.

Another solution I'm more than willing to entertain, would add more skill to getting all the damage in one spot, which is what I'm after. I quite like this actually, means you'd have to alter your aim quickly to place shots from different weapons rapidly into the exact same location.

[Vapor Trail]

While a completely parallel firing pattern would probably be viable, and simpler to implement... I'm thinking that it might swing things too far in the direction of mechs with weapons ports that are physically close together, and possibly mechs with primary weapons locations in the arms. Examples would include the Swayback for sure for the torso, the Stalker (for the arms, each arm would effectively be one weapon location, same deal with the Jenner.

As far as random chance determining outcomes, if it's that close a game, then the vagaries of ping and internet traffic are probably going to have at least as much effect on the outcome as any single shot.

Random chance does have a place. Granted, it's not supposed to be the primary determining factor in a game, but no one is (or should be) arguing that... However, it shouldn't be removed (or its addition dismissed out of hand) simply for the sake of removal of random chance. There's a reason a lot of sporting events still happen in outdoor arenas, where the vagaries and random effects of wind, sun, and weather have sway, and not hermetically sealed, soundproofed, climate controlled down to the last percent of air composition, experiment boxes. It's because that last little bit of randomness adds to the challenge.

Random chance exists in the real world, and no amount of skill can eliminate it completely. Keeping a small amount of random chance simply adds verisimilitude to a video game.

If a game is close enough fought that the result comes down to random chance, then perhaps the guy who threw the hail mary chose the correct course of action, and the guy who hesitated to did not?

[Alienfreak]

cyberFluke, on 15 April 2013 - 07:27 PM, said:

Here's an example:

Take Counterstrike, a game we're all familiar with, and has a lot of similarities with MW:O, a lot of differences too, I'm aware, but still.

Make every weapon pixel perfect accurate every shot from the hip. What would happen?
There would be two weapons to use to be competitive, the AWP (highest single shot dmg) if you were confident of your aim, or whatever has the highest rate of fire to damage ratio if you're not so confident.

That's basically what we have here IMO.

Exactly this.

BUT OMG RANDOM ROLL YOUR DICE AT HOME SCRUB!^!!

[Alienfreak]

Brilig, on 15 April 2013 - 10:16 PM, said:

The big drawback is having multiple crosshairs to worry about. I figure you can alleviate some of that by putting them close enough together to make it easy to deal with, but still far enough apart that getting all your hits on one section wont happen.

Yeah but perfect convergence of all weapons at all ranges has to go.

[Brilig]

Vapor Trail, on 15 April 2013 - 10:52 PM, said:

While a completely parallel firing pattern would probably be viable, and simpler to implement... I'm thinking that it might swing things too far in the direction of mechs with weapons ports that are physically close together, and possibly mechs with primary weapons locations in the arms. Examples would include the Swayback for sure for the torso, the Stalker (for the arms, each arm would effectively be one weapon location, same deal with the Jenner.

I thought about that, and I suppose it would depend on the mech. The crosshairs would have to be customized for each mech to keep any one from being significantly better than others. The Hunchback 4P for example could have a rather large box for its right torso mounted hunch weapons. Where as a Cataphract with its 1 energy point in the left torso, and 1 ballistic 1 energy in right torso could have their torso crosshairs closer together.

Mechs with arms could pose an issue to, but I think if you put the same restrictions on them it would keep them from being too powerfull. Ill use the Atlas RS for an example since it has 2 energy spots in each full articulated arm. Red circle for arms and blue squares for torso mounts.



A single weapon port in the right torso for the ballistic is off set just a little bit to the right. The two missile ports in left torso are just a little off to the left and stacked 1 on top of each other.
The crosshairs don't have to match the weapon positions on the mech exactly. A Hunchback 4P could have 6 small boxes all stacked on top of each other to keep the damage spread out. A Cataphract with its 2 weapon slots in the right torso could have one small box right on top of another.

The arm mounted weapons would have to be able to articulate just like the do now depending on the actuators, but we can still separate the weapons out a bit. Each circle represents an arm mounted weapon 2 for each arm. They will all move as 1 unit like the current arm crosshair does, but still be far enough apart that it does not grant a large advantage over mechs with fewer arm mounted weapons. Or if that's too much the Atlas RS could just have 2 circle for both arms 1 above the other because they could both articulate enough to converge on the same spot. A mech with 1 weapon on each arm like a Hunchback 4G could just have 1 circle.

[smokefield]

you know what i want...shaking when my mech eat an ac slug...the bigger the slug the more shaking...and i want recoil when i fire projectile weapons..

..bigger crosshair ? no - we are in a precision instrument of speed and weaponry...i am sure we can have an accurate targeting system regardless of the situation..be that you are running or jumping or sliding. it is not a fps where you play a human.