54 replies to this topic

[Anjian]

Samial, on 05 March 2021 - 08:24 PM, said:

Never even heard of it. Plus i have no consoles sorry.

Not WoT slow, but not COD fast either. Plenty of melee. Not Armored Core stuff.

[Anjian]

Vlad Ward, on 04 March 2021 - 03:16 PM, said:

Mechs are slow. Really slow. We have huge, beautiful maps in which maybe only 5-10% of the playable area sees combat.

Mechs are slow. Really slow. Dodging is not a thing unless you're a 20 tonner going 135+ (and even that is largely because the models are so small).

What if Mechs moved faster? What if there was a good reason to use half or quarter throttle? What if players could actually choose non-central positions and respond to the positioning of the opfor? What if objectives were actually playable for anyone not in a Locust?

It almost sounds fun.

This game is not exactly slow in terms of the speed of mechs from other games. if you limit matches to mechs between 20 to 45 tons, no make it up to 55 tons, you will have a fairly fast game. It will be faster than some mech games. This game as it is, is already much faster and mobile than your typical PvP tank game.

There are mech games out there, and these are even on mobile, where the big mech rumbles around at 32kph. Compared to that, the 48kph speed in many assaults in MWO is a Ferrari.

This game has a big of a schizo personality because it cannot decide if it wants to go after the WoT player market, where game play is slow but strategic --- you think your decisions thoroughly, then live and die with it --- or the Western shooter market, where everyone wings it with skill and speed.

Edited by Anjian, 05 March 2021 - 10:00 PM.

[PocketYoda]

Anjian, on 05 March 2021 - 09:20 PM, said:

Not WoT slow, but not COD fast either. Plenty of melee. Not Armored Core stuff.

What a weird game..

[Fae Puka]

The desync hit assaults in general, far harder than lights; its not just about speed but agility. A 100t of mech should be carrying a shed load of heavy duty servos as part of the basic chassis design. Return to previous incarnations of engine syncing, then the assaults would feel more responsive while keeping speed differentials the same.

[The pessimistic optimist]

Lighter fast moving mechs being more maneuverable with jumpjets would be cool as long as it balanced smartly.

[Dozer6]

Samial, on 06 March 2021 - 03:28 AM, said:

What a weird game..

It is actually a very good game.
Well balanced and endless customization.

I may go play it now...

[Void Angel]

Samial, on 05 March 2021 - 08:24 PM, said:

Its not always potatoes either, its ping related so with 700 ping time (350-350) AU-EU-AU that's almost a full second to hit stuff with a weapon.. Judging every weapon especially PPCs a full second to the target across a map is harder than it looks.. and not to be hit almost a second hit time isn't as easy as players like to believe.

Not to mention the faster things go the more lag shielded they become at high speeds..

Er, host-state rewind means that you get the hits you make on your screen. Network instability related to high ping may interfere with hit registration, but you're not actually trying to lead targets over a second's worth of real time - trust me, I played when you still did.

[Gagis]

Ping is also already the full return trip delay, no need to count it twice.

That guy sure is amazingly inventive with coming up with explanations for their poor performance.

[Wildstreak]

Stuff can be dodged with higher weights including proper use of terrain and positioning.

Map areas not used either need to have certain modes placed there, has come up before years ago, or they learn to not make maps so big. This can also be solved by player strategies, this has been done a few times before, just not many suggest it and sometimes players do not want to do it.

Increasing speed is not needed.

[Dozer6]

Seconded.

[Void Angel]

Gagis, on 07 March 2021 - 12:52 AM, said:

Ping is also already the full return trip delay, no need to count it twice.

That guy sure is amazingly inventive with coming up with explanations for their poor performance.

He thinks he has to add both players' pings together to get total lag. That would only be true if the server had ESP and could see the future, but still waited for his ping to come in before sending the information to my system. That's... not how that works.

[Void Angel]

Samial, on 06 March 2021 - 03:28 AM, said:

What a weird game..

Not that different from MWO. Jump jets, long-range snipers, and close-range brawler archetypes... The only thing really outside of MWO norms is the stun mechanic - and MWO originally had knockdown.

Well, the 'mechs move a bit faster and with more agility as well.

[PocketYoda]

Void Angel, on 06 March 2021 - 09:15 PM, said:

Er, host-state rewind means that you get the hits you make on your screen. Network instability related to high ping may interfere with hit registration, but you're not actually trying to lead targets over a second's worth of real time - trust me, I played when you still did.

Ping (latency is the technically more correct term) means the time it takes for a small data set to be transmitted from your device to a server on the Internet and back to your device again.

Even if its just 350 its a long time in a game. To be fair i see 700ms occasionally to the EU servers.

Void Angel, on 06 March 2021 - 09:15 PM, said:

Er, host-state rewind means that you get the hits you make on your screen. Network instability related to high ping may interfere with hit registration, but you're not actually trying to lead targets over a second's worth of real time - trust me, I played when you still did.

I don't understand that last sentence.. I play every night.

Void Angel, on 07 March 2021 - 06:31 PM, said:

He thinks he has to add both players' pings together to get total lag. That would only be true if the server had ESP and could see the future, but still waited for his ping to come in before sending the information to my system. That's... not how that works.

I had bad info i was told thats how it works my apologies. Still do see 700-1000+ ping times on here occasionally.

Edited by Samial, 07 March 2021 - 07:46 PM.

[Void Angel]

Samial, on 07 March 2021 - 07:36 PM, said:

I don't understand that last sentence.. I play every night.

I had bad info i was told thats how it works my apologies. Still do see 700-1000+ ping times on here occasionally.

No worries; I'm glad to help - there are cases where your enemy's ping matters too, but you don't have to worry about accounting for that directly most of the time. Adding your latency to your opponents' does tell you how long it takes you to see an action he initiated, but that's not important information when you just want to shoot him in the face.

Now, about the sentence that was confusing to you, "you're not actually trying to lead targets over a second's worth of real time - trust me, I played when you still did," I don't mean to imply that you don't play enough now. I meant that I played way back in the early days of the game when the devs hadn't fixed Host-State Rewind.

If you didn't know (and not everyone does,) host-state rewind is a technique where the server "rewinds" the game by applying damage or effects that you inflict on an enemy 'mech based on your screen, not the actual state of the server when your rounds land. Or, to put it another way, HSR is the system a game uses to make sure that what you see in the game actually happens. Without HSR, you'd have to lead every target a little extra based on your ping at the time. For something that's not moving, there's no difference, and for slow movement (like an Atlas under full steam,) it's only a bit for most pings - but if you had a Raven cooking along at ~150kph, you'd have to shoot well in front of it to hit things with lasers.


It was... "fun."

PS: Extremely high latencies or unstable internet connections (which often come with high latencies) can screw with both hit registration and HSR, but unless you're looking at one of those huge 700-ish pings, I wouldn't worry. Even then, the problem is annoying, but not too pronounced, in my experience.

[YueFei]

The6thMessenger, on 04 March 2021 - 04:18 PM, said:

I think we should try out the size first. Mechs are just too damn big. If they were smaller, the distance between them would feel longer, the lower height above the ground would make mechs feel like they are moving a bit faster.

It's actually possible to calculate precisely the bare minimum scale+agility which is needed to enable a mech to at least juke incoming fire to adjacent hitboxes, in order to spread incoming damage effectively.

If we were to take the Gulag rescale of the Atlas, for example, shrinking it to 86%, then instead of each of the torso hitboxes being 2.4 meters wide, they'd be 2.06 meters wide. This would then mean that to juke a shot aimed at the center of a torso hitbox would require lateral displacement of 1.03 meters in roughly 250ms.

Let's suppose our Atlas pilot knows to "slope step", and leg-turns + twists + accel/decels against incoming fire. We would want to give the Atlas a combination of leg turn rate, twist rate, and acceleration/deceleration, such that it adds up to a total lateral displacement of 1.03+ meters in 250ms.

Example:
• Leg turn rate of 60 deg/sec.
• Torso twist rate of 110 deg/sec.

In 250ms, combining the above, you'd change angle by 42.5 degrees. The hitbox profile would no longer be 1.03 meters, but instead be 0.759 meters. If you were moving at 50kph, the arc in your turn would also contribute to a displacement laterally of 0.452 meters. That leaves just 0.307 meters to displace in 250ms. If you can accelerate at 35.36 kph/sec, then you can do it.

So the values for Atlas would be:
• Accel/decel of 35.36kph/sec.
• Leg turn rate of 60 deg/sec.
• Torso twist rate of 110 deg/sec.

I note that the Gulag's proposed agility changes for the Atlas put it at:
• Accel/decel of 20kph/sec.
• Leg turn rate of 55 deg/sec.
• Torso twist rate of 90 deg/sec.

Which is less that what I'm proposing above, but depending on variance caused by human error / human imperfection, the Gulag's proposed values for the Atlas might very well be enough. Or it might only be good enough at lower skill levels, but still leave the Atlas unable to compete at higher skill levels (where players' aim is better). It might be that an Atlas pilot may have to invest in the Skill Tree to increase accel/decel turn rate up to the levels I proposed, rather than having it start that way as a baseline.

While empirically play-testing would of course ultimately demonstrate whether certain agility values are sufficient, what I'm showing is that this can be approached from a theoretical basis to determine how much agility is needed for each mech, depending on the width of its hitboxes. This would allow initial attempts to tune agility to at least arrive within the correct ballpark figures, with further tuning done based on play-testing.

Edited by YueFei, 10 March 2021 - 12:31 AM.