73 replies to this topic

[MustrumRidcully]

Zaptruder, on 26 February 2013 - 08:54 PM, said:

The 3L's... hit detection is largely related to the wonkily shaped hitboxes (which you can test with 2 players, firing at each other), which spread the damage over the raven's upper torso a bit too evenly.

Knowing that the Raven 3L and the Commando 2D are currently far and away better than any other light at nearly all levels of play; then what's the solution that can target those two variants specifically, without breaking any other existing system?

Max engine rating + mech quirks (torso twist, turning speed, etc). Adjust values in those areas (without compromising the canon loadout), until variants are on an even footing with other mechs in class... for the vast majority of skill levels.

But why would this help?

We've already noticed in this thread that the Jenner has better hard points than the Raven 3L. Raven 3L and Jenner have similar torso twist range and can reach similar top speeds (the Jenner is theoretically faster even).

Yet the Raven 3L is preferred? Why is that?

The answer is: It can use Streaks. The reason it can use Streaks and the Jenner cannot is that the Raven has ECM, the Jenner cannot. This combination works too well. Before ECM, we did have Streak Cats, after ECM, they were gone and replaced by Splatapults.
So I would conclude - Streaks are problematic. It's ability to use Streaks (in an ECM) world is the only thing that seems to give the Raven a notable edge over the Jenner.

Edited by MustrumRidcully, 26 February 2013 - 11:39 PM.

[Fooooo]

MustrumRidcully, on 26 February 2013 - 11:38 PM, said:

But why would this help?

We've already noticed in this thread that the Jenner has better hard points than the Raven 3L. Raven 3L and Jenner have similar torso twist range and can reach similar top speeds (the Jenner is theoretically faster even).

Yet the Raven 3L is preferred? Why is that?

The answer is: It can use Streaks. The reason it can use Streaks and the Jenner cannot is that the Raven has ECM, the Jenner cannot. This combination works too well. Before ECM, we did have Streak Cats, after ECM, they were gone and replaced by Splatapults.
So I would conclude - Streaks are problematic. It's ability to use Streaks (in an ECM) world is the only thing that seems to give the Raven a notable edge over the Jenner.

I think the point that poster was trying to get across was that if the Raven 3L couldn't do tight circles and twist its torso so far, then the other lights would be able to stay behind them almost 100% of the time.

Imagine a 3L doing a circle 3 or 4 times the size it can do now.........that would be a sizable nerf imo.

Meaning the Raven 3L generally won't defeat (will be at a disadvantage in terms of mobility and ability to get a lock) a decent non-ecm comm / jenner / cicada / or other raven variant solo etc. It would also be a little weaker against larger targets as it will be in their sights for longer.


However, your right in that it doesn't really fix the underlying problems which you demonstrate with the streakcat basically............however.............. those things could also be done to the streakcat which would also be a sizable nerf for them also imo.

Not that this should be the way to do it tho............

Edited by Fooooo, 27 February 2013 - 12:17 AM.

[Ewigan]

Actually i do like Zaptruders idea a lot, i really wonder why it never occured to me :S

Just nerf their max speed, for Com-2D and RVN-3L!

They can keep the ECM, the streaks, no probs. They just do not run 150 kp/h anymore.

[Roadbuster]

Volume, on 26 February 2013 - 07:41 PM, said:

Even if ECM/Streaks end up being useless one day, SRM6's are pretty ridiculous, and this is the only Light that can mount two of them.
They have the highest burst damage/DPS in close-quarters. As mentioned before, the 3L has very high max-engine-speed to close distances with, to use its possibly 2(!!!) SRM6s at an effective range.
It's ONLY weakness is that it can't mount jumpjets, but...Then again, when you have three module slots, ECM, and everything else, I don't think that this one drawback compensates for its other advantages.

Please consider that 1 of the Ravens missile hardpoints only has 1 tube (NARC).
If you take out ECM I'd rather pilot a JJ Jenner with 2 SRM4 which can be fired in 1 shot.

[Zaptruder]

MustrumRidcully, on 26 February 2013 - 11:38 PM, said:

But why would this help?

We've already noticed in this thread that the Jenner has better hard points than the Raven 3L. Raven 3L and Jenner have similar torso twist range and can reach similar top speeds (the Jenner is theoretically faster even).

Yet the Raven 3L is preferred? Why is that?

The answer is: It can use Streaks. The reason it can use Streaks and the Jenner cannot is that the Raven has ECM, the Jenner cannot. This combination works too well. Before ECM, we did have Streak Cats, after ECM, they were gone and replaced by Splatapults.
So I would conclude - Streaks are problematic. It's ability to use Streaks (in an ECM) world is the only thing that seems to give the Raven a notable edge over the Jenner.

Streaks need to be reexamined; but in the mean time, just speed cap the problem mechs while leaving everything else as is. If other exploitative behaviour is found, then find the common intersecting elements (may be streaks, may be something else), and readjust as necessary.

The problem with high minded goals that take forever to implement is that 1. they're of questionable effectiveness 2. It creates a great deal of imbalance, resulting in negative player behaviour in the mean time, while players wait for things to be fixed.

Demand smaller, faster, easier fixes for smaller faster iterations that can collectively produce better results.

[Erasus Magnus]

I know i`ll make a total ***** of myself asking that but:
what is a state rewind?
sorry i am german and have a hard time translating it into my native.
i cannot picture what it is supposed to do.
can anyone help me out please?

Edited by Erasus Magnus, 27 February 2013 - 12:34 AM.

[MustrumRidcully]

Fooooo, on 27 February 2013 - 12:11 AM, said:

I think the point that poster was trying to get across was that if the Raven 3L couldn't do tight circles and twist its torso so far, then the other lights would be able to stay behind them almost 100% of the time.

Imagine a 3L doing a circle 3 or 4 times the size it can do now.........that would be a sizable nerf imo.

Meaning the Raven 3L generally won't defeat (will be at a disadvantage in terms of mobility and ability to get a lock) a decent non-ecm comm / jenner / cicada / or other raven variant solo etc. It would also be a little weaker against larger targets as it will be in their sights for longer.


However, your right in that it doesn't really fix the underlying problems which you demonstrate with the streakcat basically............however.............. those things could also be done to the streakcat which would also be a sizable nerf for them also imo.

Not that this should be the way to do it tho............

The problem is - this screws everyone not using the "cheese" builds of these mechs over. If you don't use Streaks, you don't have that Anti-Jenner advantages, but you still have the Anti-Streak Disadvantages. This limits build options.

Fix ECM. Maybe actually more importantly: Fix Streaks. Everything else will just reduce the viable design space for mechs, and make the game more boring.

Zaptruder, on 27 February 2013 - 12:31 AM, said:

Streaks need to be reexamined; but in the mean time, just speed cap the problem mechs while leaving everything else as is. If other exploitative behaviour is found, then find the common intersecting elements (may be streaks, may be something else), and readjust as necessary.

The problem with high minded goals that take forever to implement is that 1. they're of questionable effectiveness 2. It creates a great deal of imbalance, resulting in negative player behaviour in the mean time, while players wait for things to be fixed.

Demand smaller, faster, easier fixes for smaller faster iterations that can collectively produce better results.

Fixing Streaks may not require much more than lowering their effective damage output (for example, lowering their cooldown). It will fix Streaks "forever", and if a new mech is introduced that happens to have ECM and Missile slots, the game will not be screwed up, and they won't need to nerf another mech's maneuverability.
If the Devs are using SCRUM, their goal should be to maximize the business value of the stories they implement.

The business value of lowering the Raven 3Ls speed or maneuverability is lower than the business value of rebalancing Streaks, and I dare say, the complexity is very similar. Just lowering the maneuverability of the Raven means that the next ECM/Streak capable mech will need a similar adjustment, adding complexity to introducing new mechs, because there is an additional factor to consider. If they ever decide to fix Streaks, they will have to undo all the changes to other mechs or these mechs will stay gimped. (But undoing this fix may have such a low business value that they won't do it for a long time, leaving people with "crippled" mechs for no reason.)

Erasus Magnus, on 27 February 2013 - 12:33 AM, said:

I know i`ll make a total ***** of myself asking that but:
what is a state rewind?
sorry i am german and have a hard time translating it into my native.
i cannot picture what it is supposed to do.
can anyone help me out please?

Basically it means:
You shoot at a mech. The information that you shoot at him is sent to the server. But this happens with a time delay. By the time the server receives the information, the enemy mech is already moved a little further. So the server calculates back where the enemy mech was when you took the shot (rewinds the state), and uses this information to determine whether you hit or not.
Without state rewind, you basically have to "lead" even with instant-hit weapons, and what you see on your screen is not the same thing the server "sees".
With state rewind, the server knows what you saw when you pressed the trigger button and what you saw will be what the server will also register.

Edited by MustrumRidcully, 27 February 2013 - 12:55 AM.

[Zaptruder]

Erasus Magnus, on 27 February 2013 - 12:33 AM, said:

I know i`ll make a total ***** of myself asking that but:
what is a state rewind?
sorry i am german and have a hard time translating it into my native.
i cannot picture what it is supposed to do.
can anyone help me out please?

Basically:
The client will report the time/lag along with the direction of fire... and the server will take the time/lag stamp and rewind the server state by that much to see if you actually did hit as you thought you did.

Ideally it brings together the best of server side hit detection with client side hit detection; You should hit what you think you hit, but it should be secure so as to disallow client hacks from telling the server "I hit the head of that mech", when in fact it didn't.



So on my screen, with state rewind, when I use my hitscan lasers and start burning the legs of a Raven 3L orange... the server will account for the lag; rewind its state to a point in time equal to my lag... and more accurately gauge that I am in fact brushing my lasers along the legs of the Raven, despite my 290ms lag.

Because right now, when I brush the Raven's legs with my laser; the server will think I'm firing behind the Raven, rewarding me with no damage, even though it's lighting up bright orange on my screen.

[Critical Fumble]

MustrumRidcully, on 27 February 2013 - 12:45 AM, said:

The problem is - this screws everyone not using the "cheese" builds of these mechs over. If you don't use Streaks, you don't have that Anti-Jenner advantages, but you still have the Anti-Streak Disadvantages. This limits build options.

Fix ECM. Maybe actually more importantly: Fix Streaks. Everything else will just reduce the viable design space for mechs, and make the game more boring.

But there are so many trees!!!

[Erasus Magnus]

Zaptruder, on 27 February 2013 - 12:45 AM, said:

Basically:
The client will report the time/lag along with the direction of fire... and the server will take the time/lag stamp and rewind the server state by that much to see if you actually did hit as you thought you did.

Ideally it brings together the best of server side hit detection with client side hit detection; You should hit what you think you hit, but it should be secure so as to disallow client hacks from telling the server "I hit the head of that mech", when in fact it didn't.



So on my screen, with state rewind, when I use my hitscan lasers and start burning the legs of a Raven 3L orange... the server will account for the lag; rewind its state to a point in time equal to my lag... and more accurately gauge that I am in fact brushing my lasers along the legs of the Raven, despite my 290ms lag.

Because right now, when I brush the Raven's legs with my laser; the server will think I'm firing behind the Raven, rewarding me with no damage, even though it's lighting up bright orange on my screen.

thank you very much, now i know whats all the fuzz is about!

sounds promising if it works as advertised. right now i still have to lead my lasers about one size of a raven when one circles me.

[Pihb]

"Before we do anything with the 3L"... <=== Did you guys not read that?

Edited by Pihb, 27 February 2013 - 01:04 AM.

[Xinaoen]

State rewind is a gigantic nerf to SSRMs. Or rather, it's a gigantic buff to every non-guided weapon.

The entire reason that SSRMs are the go-to weapon for Light vs Light combat in the first place is because they're not affected by the lag shield. Take the lag shield away, and Streaks are suddenly much less impressive relative to your other weapon options.

Calling it now: the 3L will have been forgotten within hours of state rewind going live, and we'll all be complaining about whichever ECM boat has the most energy hardpoints.

[Zaptruder]

Xinaoen, on 27 February 2013 - 01:19 AM, said:

State rewind is a gigantic nerf to SSRMs. Or rather, it's a gigantic buff to every non-guided weapon.

The entire reason that SSRMs are the go-to weapon for Light vs Light combat in the first place is because they're not affected by the lag shield. Take the lag shield away, and Streaks are suddenly much less impressive relative to your other weapon options.

Calling it now: the 3L will have been forgotten within hours of state rewind going live, and we'll all be complaining about whichever ECM boat has the most energy hardpoints.

1. While I think state rewind will be a large improvement, I doubt it'll be an instant fix to all the problems.
2. Streaks will still be very effective in their current state; as consistent, will always hit as long as you're breathing; kind of weapons; certainly the optimal weapon for single tube narc launchers.
3. Ballistic and Missle weapons won't be effected much by state rewind; they'll continue to operate as is; where the server confirms that you've fired before the projectile actually shoots out of the barrel. So in those cases; what you see is for the most part, what you get.
4. Lasers will still be lasers; even on slower mechs, when you're both moving around at decent speeds, it can still be tricky keeping the reticle on location 100% of the time (or even a good deal of the time).
5. Other lights will still be inferior to most of the important aspects of 3Ls and 2Ds.


At best, I see that the viability of non-missile/streak/ECM light mechs will improve.

[OpCentar]

I think you all forgot how the lag shield isn't something that is purposely built into the game, it's a damn bug!

So state rewind will not nerf (or buff) anything. It will simply fix an exploit everyone was using as it was on all the time.


The idea is that you can outmaneuver anything heavier than your light mech so that you can keep clear of the enemy mech firing arc, not run around in circles laughing while enemy fire doesn't register any damage on your light!

That's why we don't need any nerfs or buffs to specific chassis only fixes in the game engine, mainly the mentioned lag shield and collisions.

[Voidsinger]

A few views from me.

The fact Commandos go pop very easily more than makes up for the extra missile hardpoint.

Streaks. They need to lose lock-on after firing. In TT they did, as stated in the initial description. Also need a lockon lag for multiple launchers.

To get rid of the Streak/ECM issue based on PGI's implemetation of ECM: If you carry a weapon that requires a lockon, then due to the limited rangebands available to targeting sensors, you lose the ability to stop enemy mechs locking on to you.

[Xinaoen]

Zaptruder, on 27 February 2013 - 01:30 AM, said:

1. While I think state rewind will be a large improvement, I doubt it'll be an instant fix to all the problems.

(snip)

At best, I see that the viability of non-missile/streak/ECM light mechs will improve.

Streaks will still be perfectly viable post-SR, but they'll no longer be The Only Brawling Weapon that Doesn't Force You to Guess Your Target's Ping in Order to Deal Damage ^™. Will they still deal consistent damage? Sure. Will they be able to lay down as much raw DPS as well-aimed SRMs or lasers? Not likely.

At its heart, the Streak is a weapon that trades damage for the sake of being easy to connect with. That's a fair compromise for players with poor aim, but the min-maxers and munchkins are gonna turn away from the scrubcannon as soon as they can count on getting what they pay for with the point-and-click weaponry. That's my prediction, anyhow.

Edited by Xinaoen, 27 February 2013 - 01:55 AM.

[MustrumRidcully]

Zaptruder, on 27 February 2013 - 01:30 AM, said:

1. While I think state rewind will be a large improvement, I doubt it'll be an instant fix to all the problems.
2. Streaks will still be very effective in their current state; as consistent, will always hit as long as you're breathing; kind of weapons; certainly the optimal weapon for single tube narc launchers.

Other weapons will catch up a bit, but Streak Cats were not just dangerous Lag Shielded Lights - every mech had problems with that. A Streak Cat wasn't fast enough for (notable) Lag SHields. But it sure had the manevuerability and the speed to make aiming non-trivial, and giving it the option to to torso twist to direct attacks at less vital areas, while maintaining lock was not as difficult as consistent aiming and hitting weak spots on the Catapult.

[Khobai]

Quote

You missed the COM-2D in this bit: "The 3L is still the only platform that can mount ECM at that speed and have missile hardpoints (sorry, Cicada). It is the only light that can, which is important considerin"

Nope. Commando-2D can't go 151kph and still have all the weapons it needs to be competitive. Commando-2D can only reasonably use a 200XL engine.

I really hope one of the quirks they give Commandos is a passive 5% speed boost to help mitigate the fact it can't go 151kph.

Edited by Khobai, 27 February 2013 - 02:25 AM.

[Draigan]

Zaptruder, on 26 February 2013 - 08:47 PM, said:

Just nerf their speed. 115 w/ speed tweak max. Same for the Commando 2D.

You want ECM? Sure. You want Streaks? Go ahead!

You're going to have trouble keeping up with lights and lag shielding yourself against heavies and assaults.

Simple tweak that at worst, nerfs 2 mechs heavily, leaving the rest of the systems and mechs untouched.

I don't think it that would placate as many people as you think i run my 3L 1x medium 2x medium pulse and 2 ssrms with a max speed of 112kph and I still get plenty of whinging from people

[CECILOFS]

ChrisOrange, on 26 February 2013 - 07:59 PM, said:

The only lag shields that I don't like are the euro/australian players that think it's a good idea to keep an American client/account.

Give us Oceanic servers and I will gladly play there. Otherwise there's little difference between US and European for me.