1126 replies to this topic

[AntiCitizenJuan]

Fixing Heatscale > CoF

[Mystere]

Telmasa, on 09 April 2015 - 03:53 PM, said:

On a side note I'd be willing to try 'fixed convergence' for torso weaponry (i.e. shows up as multiple crosshairs on the HUD) & arms without lower actuators - IF and only IF we are given an way to change our HUD's transparency.

Now that is what I call progress.

Telmasa, on 09 April 2015 - 03:53 PM, said:

The amount of clutter on the HUD already bugs me, if you add more reticules to that & none of it is opaque so I can 'see' through them, that'd just make another problem.

It can be as simple as a 2-reticule system: one for Torso/Head-mounted weapons and another for arm-mounted ones.

It can also be a 3-reticule system, with each arm getting it's own. I just don't know how many will be able to control that, assuming they even have suitable equipment to do so.

[Dino Might]

Dimento Graven, on 09 April 2015 - 03:50 PM, said:

Ok, let me pin you down:

HOW DO YOU IMPLEMENT A CoF AFFECT TO APPROXIMATE YOUR "•CEP + Mechanical Slop ≈ CoF" EFFECT, TO WHERE THE EFFECT IS NOT A RANDOM EVENT?!?!

One of out of HOW MANY shots needs to suddenly veer off in a random direction for "reasons" will it take to make you happy? 1 out of 5? Or maybe just the ones headed towards your 'mech?

All your passive aggressive "oh I'm just justifying blah blah blah" doesn't cover the "miniscule fact" you can't implement ANY OF THIS in a computer game without an RNG.

Because as far as the computer is concerned a beam of energy will go in a straight line until interrupted.

As far as this engine is concerned, all 'mech gimbals are perfect, all ammo rounds are perfect, and there's not enough gravity or windage at the distances involved to appreciably affect physical round travel.

So how do you do it WITHOUT an RNG?

Answer please.

You are obviously the more talented MWO player. I bow to your greatness. Now let me for the umpteenth time explain to you why you have no clue what you're talking about.

No kidding we can't model CoF without a RNG - WHY WOULD WE? I'd have to be a blithering idiot to ignore such a powerful tool at my disposal.

The Normal RNG is a MODEL for reality. It very closely resembles reality. Unfortunately we don't have a 5*10^100000000 flop processor yet and a perfect deterministic understanding of every particle interaction in the universe, so it's as good as we can do. And it's actually very very good. I have given you numerous examples of how we use statistical distributions to model reality from nuclear reactions to manufacturing work to weapons testing and even to your example with lasers bouncing off moon mirrors, which you so happily ignored the response to because it again showed you have one overwhelmingly simple misunderstanding that is at the core of your entire diatribe.

Your big problem is that you worry that it will normalize gameplay and make bad players more equal to more skilled players. I have given you examples of why that thought is dead wrong. Again, go talk to anyone who has ever shot at the Camp Perry matches and try to explain to them your thoughts on this. They will be thoroughly entertained.

The perfect precision system is what normalizes player skill to make the bad players closer to better players. If you consider benchrest shooting - the difference between the top shooter and the bottom shooter in your average match is measured in a handful of inches. If you look at Across the Course matches, where you don't have as stable (precise) of a firing platform, the difference between the top shooter and bottom shooter is in inches/feet at shorter ranges than the benchrest shooters were firing. Those guys in the Across the Course matches are still cleaning targets like there's no tomorrow (10s with multiple Xs), while the bad shooters are only hitting the target black with half their shots or so.

I don't know why I'm still trying. You refuse to accept reality when it's staring you in the face. I keep hoping one day you'll back down and admit, you just don't want CoF because you don't like it. It's not an issue that its less realistic, it's just your preference that we don't have it. And once you admit that's the only reason, then you haven't much of a leg to stand on because you recognize that you're demanding everyone acquiesce to your preference without any consideration for other's opinions.

But you can consider me successfully trolled. I cannot stand someone who lacks sufficient knowledge in mathematics trying to discredit its usefulness and application to so many of the things that we do. I am so sick of ignorant people dominating conversations because they yell the loudest. I hope people can see through the petty childish ranting from both of us and make their judgments based on the facts, which are plain as day to see.

Edited by Dino Might, 09 April 2015 - 04:46 PM.

[Ralgas]

Dino Might, on 09 April 2015 - 04:34 PM, said:

You are obviously the more talented MWO player. I bow to your greatness. Now let me for the umpteenth time explain to you why you have no clue what you're talking about.

No kidding we can't model CoF without a RNG - WHY WOULD WE? I'd have to be a blithering idiot to ignore such a powerful tool at my disposal.

The Normal RNG is a MODEL for reality. It very closely resembles reality. Unfortunately we don't have a 5*10^100000000 flop processor yet and a perfect deterministic understanding of every particle interaction in the universe, so it's as good as we can do. And it's actually very very good. I have given you numerous examples of how we use statistical distributions to model reality from nuclear reactions to manufacturing work to weapons testing and even to your example with lasers bouncing off moon mirrors, which you so happily ignored the response to because it again showed you have one overwhelmingly simple misunderstanding that is at the core of your entire diatribe.

Your big problem is that you worry that it will normalize gameplay and make bad players more equal to more skilled players. I have given you examples of why that thought is dead wrong. Again, go talk to anyone who has ever shot at the Camp Perry matches and try to explain to them your thoughts on this. They will be thoroughly entertained.

The perfect precision system is what normalizes player skill to make the bad players closer to better players. If you consider benchrest shooting - the difference between the top shooter and the bottom shooter in your average match is measured in a handful of inches. If you look at Across the Course matches, where you don't have as stable (precise) of a firing platform, the difference between the top shooter and bottom shooter is in inches/feet at shorter ranges than the benchrest shooters were firing. Those guys in the Across the Course matches are still cleaning targets like there's no tomorrow (10s with multiple Xs), while the bad shooters are only hitting the target black with half their shots or so.

That might work fine for the real world with like sized targets (see most shooters). It ignores the balance issues and set of kill requirements for mechs and the size differentials of those kill areas involved. No concept put forward yet that involves any form of cof addresses the fact that it heavily favors lights, many even going as far as to give them essentially fulltime lag shield. It also doesn't stop everyone simply switching to lrms/srms to avoid it totally.

Systems designed to "uncouple" some hardpoints but not all only shift the meta mechs, nothing more

Edited by Ralgas, 09 April 2015 - 05:05 PM.

[Mystere]

Yeonne Greene, on 09 April 2015 - 11:27 AM, said:

War Thunder is just doing what real pilots did...and it's awesome.

Fixed convergence distance means you have to make some decisions with your super-quirk lasers. Do you set those ERLL to converge at max range so you can get maximum utility out of their range, or do you set it for some point in the middle to keep it somewhat viable as a defensive munition?

Fixed convergence could even provide impetus for reducing the cone of fire on the MGs, and who wouldn't want that?

That depends on weapon location and 'Mech size. On torso-mounted weapons, I'd set it to around 105-110% of optimal range. On arm-mounted weapons, I'd set it to infinity (i.e. parallel).

[Dino Might]

I'd love a fixed convergence system as well. I think that'd be another decent improvement. I don't think any of our ideas on their own totally fix any of the problems.

If we did a combination though...

Reduced heat capacity
10 sec running average of heat is continuously calculated and current running average directly affects mech
Fixed convergence for torso weapons or CoF system

All together, you'd be able to still hit what you're aiming at (if you're skilled), but you'll have to do a lot more than just point and click. The days of the 360 no scope would be over.

[Mystere]

Boris The Spider, on 09 April 2015 - 02:07 PM, said:

I would totally remove all convergence on torso mounted weapon and have them fire straight forward, and drastically slow convergence on unarticulated arms. If you want your weapons to track and compensate near instantly for distance, you put them on articluated arms.

Torso-mounted weapons do not need to suffer the fate of parallel firing. FIxed convergence (selectable by player in MechLab and in-battle) should at least be available.

Edited by Mystere, 09 April 2015 - 05:39 PM.

[Mystere]

Ralgas, on 09 April 2015 - 04:58 PM, said:

That might work fine for the real world with like sized targets (see most shooters). It ignores the balance issues and set of kill requirements for mechs and the size differentials of those kill areas involved. No concept put forward yet that involves any form of cof addresses the fact that it heavily favors lights, many even going as far as to give them essentially fulltime lag shield. It also doesn't stop everyone simply switching to lrms/srms to avoid it totally.

Systems designed to "uncouple" some hardpoints but not all only shift the meta mechs, nothing more

Here's the rub. No one has really even given any specifics yet on how large the CoF should be. Who said the CoF can't be smaller than half the width of the smallest Mech at 1000 meter distance?

Some people here just see the mention of CoF and suddenly have terrifying visions of the Devil incarnate. In other words, they won't have any of it ... like a petulant child unwilling to listen to reason.

[wolf74]

Convergence is always a hard item for any game to handle.

As I see it personally. I think going to a Manual Controlled Convergence would be the best option over all. Due to the Fact if will allow for the Pin point shots but you have to adjust for the target distance from you.

Basic Targeting in all mech will correct for Range +-5m from Locked Target.
Targeting Computer can correct up to +-15m (on top of the Basic system for a Max total of +-20m) [only Effects Direct Fire weapons listed below]

Targeting Computer Formula
15 * [Targeting Computer weight / (Roundup(Direct Fire Weapons Weight / 5 (4 if I.S.)))] = Correction amount Granted (Max amount is 15)

Direct Fire Weapons:
Auto-Cannons
Ultra-Cannons
Lasers
ER Lasers
NOT: Pulse Lasers, SRMs, LRMs, and LB-#x Cannons

Non-Locked LRMs Do a High Arc and Land at the set Convergence point (No less than 180m For I.S., Clan No Less than 15m)

Pulse Lasers get a Correction up to +-30m (on top of the Basic system for a Max total of +-35m) (only effect the Pulse Lasers) (There range could be dropped back down closer to the CBT levels if this Happens)

Note the Correction above can ONLY happen on Locked Targets, if you are shooting at one target and locked onto another one it will be trying to correct for your LOCKED target. (Yes this does make ECM everyone Enemy now not just LRMs)

Again the above are just my Personal thoughts

[A Large Infant]

Why do we bother, PGI doesn't care they've had years to at least try something

[Ralgas]

Mystere, on 09 April 2015 - 05:34 PM, said:

Here's the rub. No one has really even given any specifics yet on how large the CoF should be. Who said the CoF can't be smaller than half the width of the smallest Mech at 1000 meter distance?

Some people here just see the mention of CoF and suddenly have terrifying visions of the Devil incarnate. In other words, they won't have any of it ... like a petulant child unwilling to listen to reason.

Heres the bigger rub, you can fit a full lights torso and arms in many of the assaults CT's. Heck i cant remember the range exactly but you can try it now in the testing grounds. Take out a mech with mg's, find the range at which it can keep all it's fire inside the ct of a heavy/small assault. Try the same range on lights and see how well you go. iirc (and i may be wrong, it was a long time ago and i think the cone may have been altered since i did this?) a catapults ct can cause misses on a cicada, let alone trying to hit something as small as a locust.

Assault mech ttk issues wouldn't change (or they would get better and imbalance the "but" even further), but on the other hand every light would get as bad as the borked hitbox annoyances we have now, only it'd be the system and not the underlying code.

As for unhooking ct hardpoints from arms, well that just favors any mech with stacked hardpoints. Alters the meta mechs slightly but not in any sort of more balanced way than what we have now. And ofc that's all assuming it doesn't just start a new lrmpocalyse..... ofc several more clanners being able to stack up 7+ weps in one or the other via omni swaps (+ecm in the case of the hbr) just screams a forum meltdown about clan op down that path anyway.

[Hans Von Lohman]

If you can manually set your default convergence in the mechlab, maybe I would be OK with it.

Right now the Arm Convergence Elite skill does nothing since convergence is set to 0.0 seconds, so accelerating it does nothing.

[Dimento Graven]

Hans Von Lohman, on 09 April 2015 - 07:00 PM, said:

If you can manually set your default convergence in the mechlab, maybe I would be OK with it.

Right now the Arm Convergence Elite skill does nothing since convergence is set to 0.0 seconds, so accelerating it does nothing.

GOOD NEWS!!! It's not set to 0.0.

It's FAST, but it ain't instant.

[Tim East]

Kuritaclan, on 09 April 2015 - 01:42 AM, said:

This argument is a waste of time since kamikaze actions arn't a military strategy with such "highvalue" Technology, beside it was an ambush attack, while the mech was in repair to make it 100% ready for next combat. R&R system was taken out of the game - therefore we now fight like houseforces which get fully operational mechs without malfunction for every battle.

The conclusion that lights up at first with: "Shut up, patched, blown apart, repaired, busted and salvaged over and over and over again." is a fallacy. The techs do all to make the Mech nearly 100% ready. Parts which were damaged get repaired, and destroyed components get replaced, otherwise the technique does not work! Every mil-plattform is based on interchangeable parts to avoid failfunction. This concept is a product of series production even if they are small series. You can argue that a mech in combat have failfuinctions because of damaged parts, but not before. If the mech is not usable, because of any problems, it don't get wasted in a battle. Surrendering before the mech take main damage so it is not repairable was common - fight into a foreseeable death was not a option.

Yeah, the thing about this is that there are a lot of times that you never quite get back to 100% readiness. Once upon a time I knew a command track that suffered, well, let us call it a "maintenance error" and move along. This unfortunate vehicle wound up having no end of additional issues including recurring problems with its hydraulics as a direct result of said incident, and it even caught on fire one night on the way to the field. Sometimes machinery just stops being as awesome and problem-free as when you got it, and you get those little bugs that people inexperienced with that particular machine don't know how to properly handle. You know, like the whole overheating and catching on fire thing.

Thankfully we don't have to put up with that here, despite the fact that battlemech quirkiness is pretty thoroughly established in canon. Aidan Pryde's supposedly jinxed Timber Wolf, for example.
Also this. https://soundcloud.c...attletech-treat

Dimento Graven, on 09 April 2015 - 11:22 AM, said:

The topic of this thread is going for an assumption that there's something wrong with convergence as it functions within the game now.

That assumption is wrong, incorrect, and self-serving to weak players who don't think that stupid play should result in their quick deaths, hence the need to try and direct the conversation to the REALL issue, chain alphas.

It isn't the fact that alpha'd weapons will, most of the time, pin point on the same location when fired, it's the fact that 'mechs can chain alpha repeatedly with very little risk.

Eliminate the ability for no risk alpha firing, and you fundamentally restructure this game on a tactical level.

That isn't actually an assumption, given that for many of us it is based on factual previous experience with delayed convergence and how it made for a more interesting and sim-like game. There should always be a delay between your rangefinder detecting the range to the point your reticule is over, your targeting computer relaying this information to your magical super-gimbals, and then said gimbals traversing a small distance to bring everything to face the direction requisite for your fire to intersect at the point indicated by said rangefinder.

I won't disagree that heatscale and alpha could use a second (or fifth) look, but using that as an excuse to justify nigh-instantaneous alignment of your weapons' trajectories in direct defiance of physics is something of a logical leap.

Dimento Graven, on 09 April 2015 - 11:22 AM, said:

The only people who want cone of fire are the ones regularly dying to pin point.

Ummm, duh? All of the weapons that can reliably kill you are pinpoint. How many people really die to LRMs in this day and age? Heh.

Seriously though, claiming that people want to change something because you personally beat them is a little bit disingenuous. Sure, the people who want to change convergence die mostly to pinpoint damage, but so do you and you aren't arguing for the alteration of convergence.

Dimento Graven, on 09 April 2015 - 12:16 PM, said:

I'm deadly with all weapons systems in this game.

I am pretty darn good playing all weight classes.

Flamer challenge, go! I'll duel you in my Locust against any mech you want to run that uses nothing but flamers. If you prove me wrong and are deadly with flamers, well, then you get the satisfaction of having proven me wrong. Wouldn't that be nice?

Dimento Graven, on 09 April 2015 - 01:06 PM, said:

When do "I" have a problem with convergence, you ask?

1. When it's not working 'fast' enough and my shots past on either side of the intended target, or partially miss a target and it survives longer than it should have.
2. When someone, unbeknownst to me, is aiming at my 'mech and destroys something vital I was planning on using later.

Issue 1 - That's how it should be, sometimes I need to take more time switching between targets.
Issue 2 - I have to try harder at being a more difficult target.

I have literally never had that first thing happen to me. The closest I've come to that is having one set of my lasers hit one leg and the other set hit the other leg, and that only happens at the most extreme close range. What range are you doing that at that you are missing on either side? Did you take your crosshairs off of the mech when you shot, or aim them between the target's legs? Either of those things can cause your rangefinder to misread the distance for convergence. Also, was your target moving? If they're at a very high speed, the projectile flight time can influence whether or not you attain perfect convergence, since the weapons are set to the point you were aiming when you fired, not when the projectiles hit the enemy. Though again, I've literally never seen this happen with lasers. They're hitscan and tend to insta-converge the moment they wash over things.

wolf74, on 09 April 2015 - 06:04 PM, said:

Convergence is always a hard item for any game to handle.

As I see it personally. I think going to a Manual Controlled Convergence would be the best option over all. Due to the Fact if will allow for the Pin point shots but you have to adjust for the target distance from you.

Basic Targeting in all mech will correct for Range +-5m from Locked Target.
Targeting Computer can correct up to +-15m (on top of the Basic system for a Max total of +-20m) [only Effects Direct Fire weapons listed below]

Targeting Computer Formula
15 * [Targeting Computer weight / (Roundup(Direct Fire Weapons Weight / 5 (4 if I.S.)))] = Correction amount Granted (Max amount is 15)

Direct Fire Weapons:
Auto-Cannons
Ultra-Cannons
Lasers
ER Lasers
NOT: Pulse Lasers, SRMs, LRMs, and LB-#x Cannons

Non-Locked LRMs Do a High Arc and Land at the set Convergence point (No less than 180m For I.S., Clan No Less than 15m)

Pulse Lasers get a Correction up to +-30m (on top of the Basic system for a Max total of +-35m) (only effect the Pulse Lasers) (There range could be dropped back down closer to the CBT levels if this Happens)

Note the Correction above can ONLY happen on Locked Targets, if you are shooting at one target and locked onto another one it will be trying to correct for your LOCKED target. (Yes this does make ECM everyone Enemy now not just LRMs)

Again the above are just my Personal thoughts

Interesting thoughts. Might need work, but has the potential to be very cool.

Dimento Graven, on 09 April 2015 - 07:45 PM, said:

GOOD NEWS!!! It's not set to 0.0.

It's FAST, but it ain't instant.

What IS it set to then? How do you know? Is it slow enough that you can measure it with hitscan weapons like lasers, and if not, how did you come to that conclusion? Accessing that game's code? Are you certain that elite convergence does something, because I've never noticed a difference. You know, aside from double basics, and that's completely unrelated.

[ROSS-128]

Apparently some people just want to play a very different type of mech than what is actually in the game.

Aggregating the various "TTK too short" complaints it seems they want to play a slow, undergunned, over-armored, beat-up piece of junk that struggles to hit a skyscraper from a couple hundred meters, doesn't carry enough ammo to kill a copy of itself, has zero degrees of traverse on its weapons, and overheats every time it fires. Not entirely sure what kind of gameplay they expect from that though. Perhaps when rounds routinely end in a draw due to both sides running out of ammo before they run out of armor, TTK will finally be "long enough"? Then again I don't see much complaining about XL engines, so I suppose I can scratch "slow" from that list. Set engines to ramming speed?

I still maintain that given the large size of our targets and the short ranges we are engaging from, the errors that you would want to base your CoF on are so small that they would be negligible. Is a CoF a few pixels wide really worth the effort of modeling? Sure, each of those pixels might represent a couple feet of target zone, but our targets are so large it'll still hit the same component. And if difficulty hitting these three to four story tall robots from just a few hundred meters *is* par for the course, somebody's engineers need to be smacked upside the head.

[KuroNyra]

Dimento Graven, on 09 April 2015 - 03:44 PM, said:

That's because quite a few of you who are disagreeing with me ARE bad players.

Pretentious much?
Bad player in your opinion, yet some of them could kick your ass if you continue to underestimate others who disagree with you.
You have the attitude typical of a 15 years kid. "It's not me who is bad, it's the others. And all the ones who don't agree with me are bad! Na!"

[Boris The Spider]

Mystere, on 09 April 2015 - 05:15 PM, said:

Torso-mounted weapons do not need to suffer the fate of parallel firing. FIxed convergence (selectable by player in MechLab and in-battle) should at least be available.

Mystere, I wouldn’t exactly call it 'suffer' most torso weapons are arranged in pods so have a pretty tight grouping anyway. The two problems I have with a fixed convergence on torso, rather than direct forward are, 1^st off, aesthetically, lasers are going to appear to cross every time you fire them, 2^nd its going make aiming projectiles even more difficult. Say I'm using a HBK-4G with an AC20, with a direct forward fire, I know, wherever I line my centre reticule, my shell is going to hit 1.5 meters to the right of it, every time without fail. With a fixed convergence, I then need to start taking into account the range to know if its right or left, too hard, not intuitive enough. That's before I even think about adding a mixed projectile load-out to it, say AC2/AC5 combo.

[Quxudica]

Ralgas, on 09 April 2015 - 06:49 PM, said:

Heres the bigger rub, you can fit a full lights torso and arms in many of the assaults CT's. Heck i cant remember the range exactly but you can try it now in the testing grounds. Take out a mech with mg's, find the range at which it can keep all it's fire inside the ct of a heavy/small assault. Try the same range on lights and see how well you go. iirc (and i may be wrong, it was a long time ago and i think the cone may have been altered since i did this?) a catapults ct can cause misses on a cicada, let alone trying to hit something as small as a locust.

Assault mech ttk issues wouldn't change (or they would get better and imbalance the "but" even further), but on the other hand every light would get as bad as the borked hitbox annoyances we have now, only it'd be the system and not the underlying code.

As for unhooking ct hardpoints from arms, well that just favors any mech with stacked hardpoints. Alters the meta mechs slightly but not in any sort of more balanced way than what we have now. And ofc that's all assuming it doesn't just start a new lrmpocalyse..... ofc several more clanners being able to stack up 7+ weps in one or the other via omni swaps (+ecm in the case of the hbr) just screams a forum meltdown about clan op down that path anyway.

This is why convergence alone isn't the only change needed, the heat system also needs to be rebuilt. Cap drastically lowered, dispersion per second increased and penalties added for running hot for extended periods of time. Combine these two overhauls and you arrive at a much deeper more considered combat model

[Doman Hugin]

E Rommel, on 09 April 2015 - 08:38 PM, said:

Apparently some people just want to play a very different type of mech than what is actually in the game.

Aggregating the various "TTK too short" complaints it seems they want to play a slow, undergunned, over-armored, beat-up piece of junk that struggles to hit a skyscraper from a couple hundred meters, doesn't carry enough ammo to kill a copy of itself, has zero degrees of traverse on its weapons, and overheats every time it fires. Not entirely sure what kind of gameplay they expect from that though. Perhaps when rounds routinely end in a draw due to both sides running out of ammo before they run out of armor, TTK will finally be "long enough"? Then again I don't see much complaining about XL engines, so I suppose I can scratch "slow" from that list. Set engines to ramming speed?

You've just described a BattleTech TT game.

But this being a computer game, I'd settle for something between that and what we have now.

[xImmortalx]

I'm genuinely curious just how many of these people who think TTK is too short peeked a 6-man firing squad, got ganked and went to the forums immediately to QQ about not being able to make big mistakes and live to tell the story. My TT memories are pretty blurry but I'm sure that having 6 mechs shooting at you in TT would instagib you just as fast if not faster(hello random crits).