34 replies to this topic

[Liveish]

MWO Comp would like to congratulate the 2019 MWO Comp Championship Series finalists:

Grand Final - Majors
EmpyreaL (NA)
1st Jaguar Guards [EU]
EON Synergy [EU]
Majestic 12 [EU]
Games to be played on the Sat 14/Dec/19
Grand Final - Minors

228th Black Watch (NA)
Russian Jade Falcon [EU]
Smoke Adders (NA)
Phoenix Legion [EU]
Games to be played on the Sat 7/Dec/19
https://www.twitch.tv/mwoleagues

Edited by live1991, 03 December 2019 - 07:40 PM.

[Mr D One]

No OCE

I guess no one was koala-fied.

[Appogee]

Dar1ngOne, on 02 December 2019 - 03:33 AM, said:

No OCE

I guess no one was koala-fied.

Our ping does not permit the kind of accuracy necessary for to win at high-alpha pinpoint sniping.

[justcallme A S H]

Appogee, on 02 December 2019 - 10:33 PM, said:

Our ping does not permit the kind of accuracy necessary for to win at high-alpha pinpoint sniping.

There are a number of Aussies on the top ranked teams and have been for years.

Ping is power.

Edited by justcallme A S H, 03 December 2019 - 01:08 AM.

[Appogee]

justcallme A S H, on 03 December 2019 - 01:01 AM, said:

There are a number of Aussies on the top ranked teams and have been for years.

Ping is power.

I've played MWO in 3 different countries ... the US, Dubai and Australia. To cut a long story short, my damage consistently went up by about 30% when I played in the US, where I had a ping of only 45.

Here in Australia I have a ping of 250-300 to the FP servers, which are in NA. While HSR is a pretty good system, there's no way it (or any system) can account for what ends up being a half second (300ms for me plus up to 300ms for my target) difference between where my target appears to be, and which components are exposed, vs where they actually are, if my target changes direction or torso twists ... as good pilots of course do.

Anyway it is what it is. It's not worth migrating to the US to improve it

[Liveish]

Appogee, on 03 December 2019 - 05:13 AM, said:

I've played MWO in 3 different countries ... the US, Dubai and Australia. To cut a long story short, my damage consistently went up by about 30% when I played in the US, where I had a ping of only 45.

Here in Australia I have a ping of 250-300 to the FP servers, which are in NA. While HSR is a pretty good system, there's no way it (or any system) can account for what ends up being a half second (300ms for me plus up to 300ms for my target) difference between where my target appears to be, and which components are exposed, vs where they actually are, if my target changes direction or torso twists ... as good pilots of course do.

Anyway it is what it is. It's not worth migrating to the US to improve it

Yep So Aussies at higher levels would be 40% more OP if we had lower ping then hahaha

We use to play RHOD/LMS before HSR, had to lead 2 to 5cm in front of the mech and never had a problem haha

[justcallme A S H]

Appogee, on 03 December 2019 - 05:13 AM, said:

To cut a long story short, my damage consistently went up by about 30% when I played in the US, where I had a ping of only 45.

Imagine all my 3.5k FP matches were 4.5k matches from some mystical 30% increase. I'd hold some kinda record then for most games over 4k!

Ping is power.

[denAirwalkerrr]

OC players at 14% of their usual ping:

[Appogee]

justcallme A S H, on 03 December 2019 - 08:09 PM, said:

Imagine all my 3.5k FP matches were 4.5k matches from some mystical 30% increase. I'd hold some kinda record then for most games over 4k!

Ping is power.

It's not mystical. It's SCIENCE!

denAirwalkerrr, on 03 December 2019 - 08:45 PM, said:

OC players at 14% of their usual ping:

Or, FP servers relocated to Sydney

Edited by Appogee, 03 December 2019 - 09:11 PM.

[Navid A1]

Appogee, on 03 December 2019 - 05:13 AM, said:

I've played MWO in 3 different countries ... the US, Dubai and Australia. To cut a long story short, my damage consistently went up by about 30% when I played in the US, where I had a ping of only 45.

Here in Australia I have a ping of 250-300 to the FP servers, which are in NA. While HSR is a pretty good system, there's no way it (or any system) can account for what ends up being a half second (300ms for me plus up to 300ms for my target) difference between where my target appears to be, and which components are exposed, vs where they actually are, if my target changes direction or torso twists ... as good pilots of course do.

Anyway it is what it is. It's not worth migrating to the US to improve it

Your statement is not correct.

First, this game is server authoritative, meaning that only ping that matter is your ping to the server. Your experience with regards to aiming and interactions with the world around you is defined by the server which would be around 300ms. You should not count it double.


Second, HSR is fundamentally designed to give higher pings the advantage of no lag gameplay... sometime even to the detriment of low ping player experience.
HSR system in the server will rewind everything in time when you are firing your shot, so that what you see and what you shoot on your screen register as you see it.
Most of the times, low ping players see arms and torsos blown off when on their screens they have already moved behind cover, yet HSR has rewound their movement for the high ping player for his/her shot to be registered.
If you are physically fast enough to snapshot a light mech at a fraction of the second it is exposed, you are going to be rewarded with the damage (although you'll see the reticle flashing red or a kill with a delay).

Low ping players on the other hand should anticipate about 200ms into the future when they are peeking and poking, since all that matters is exposure time... which is going to be transmitted to the high ping player with a delay as it happened as recorded now.



So, I do believe that your statement about 30% higher damage is indeed a myth.
Only thing that degrades your hitreg is unstable connection with pings that jitter a lot. A steady ~300 ping still gives you a good performance.

[Appogee]

Navid A1, on 03 December 2019 - 09:25 PM, said:

So, I do believe that your statement about 30% higher damage is indeed a myth.

How many countries have you played MWO in, and with what variation of ping?

Edited by Appogee, 03 December 2019 - 10:05 PM.

[Navid A1]

Appogee, on 03 December 2019 - 09:43 PM, said:

How many countries have you played MWO in, and with what variation of ping?

Two.

And the first one with internet from 15 years ago by today's standards. (300+ ping to ALL servers)


Specifically tested the HSR system this summer when I went back for a vacation.
Works flawlessly.

And yes, I also have played MWO there before HSR was even a thing... so, I find your statement hard to believe based on my own experience

Edited by Navid A1, 03 December 2019 - 10:06 PM.

[Appogee]

Navid A1, on 03 December 2019 - 10:04 PM, said:

And yes, I also have played MWO there before HSR was even a thing... so, I find your statement hard to believe based on my own experience

And I find your statement hard to believe based on my own lived experience.

I mean, I didn't get 30% better at MWO while I was living in the states. Nor 30% worse when I came back to Australia. Yet my damage scores were consistently significantly better in the US. There has to be reason for this.

Unless, maybe, it's jitter, as you suggest. I haven't noticed any jitter issues when I've been looking at packet loss from time to time though. So that's why I discounted it as a reason.



EDIT: just thinking about HSR some more.

HSR relies on the server to record snapshots of the positions of all Mechs over time. It must do this at some kind of incremental timing, in the same way that the point of impact of hitscan weapons are actually polled incrementally, in 'pings' over the duration of the beam.

I wonder what that interval of recording HSR snapshots is?

The bigger the interval, the less accurately HSR would be able to accurately determine where the friendly and the enemy actually were at the time a shot was fired. This inaccuracy would be greatest when trying to track shots against fast moving Mechs sharply changing direction during the period of latency.

I dunno. There has to be some reason for what I've observed to be true. Lower ping seems to closely correlate to higher damage in my experience.

Edited by Appogee, 03 December 2019 - 11:14 PM.

[Navid A1]

This is how HSR works as a reference

Stable ping = consistent hits



They all can hit you as long as they see you... regardless of where you are now.
Only thing that matters is where you were before.... at their ping delay.

Edited by Navid A1, 03 December 2019 - 10:51 PM.

[Kamikaze Viking]

Appogee, on 03 December 2019 - 10:43 PM, said:

HSR relies on the server to record snapshots of the positions of all Mechs over time. It must do this at some kind of incremental timing, in the same way that the point of impact of hitscan weapons are actually polled incrementally, in 'pings' over the duration of the beam.

Instead of guessing how it works, read exactly how it works. Directly from the programmer who wrote it.

https://www.gamasutr...rior_Online.php

[Rustyhammer]

Appogee, on 03 December 2019 - 10:43 PM, said:

I mean, I didn't get 30% better at MWO while I was living in the states. Nor 30% worse when I came back to Australia. Yet my damage scores were consistently significantly better in the US. There has to be reason for this.

Maybe your moved to AU when PGI implemented changes to the total damage calculation?
Arm damage is no longer added to the total damage when you do torso destruction. Looks to me like the 20-30% damage reduction you're talking about.

[Appogee]

Kamikaze Viking, on 03 December 2019 - 11:16 PM, said:

Instead of guessing how it works, read exactly how it works. Directly from the programmer who wrote it.
https://www.gamasutr...rior_Online.php

He confirms there is some kind of incremental tick to the recording of positions, but doesn't say how long it is.

But he also explains why a higher ping player could experience worse hitreg with projectile weapons:

He says: "We also need to make a few more assumptions about the motion of targets. We assume that the motion of the target between its rewound position and current position on the server can be approximated relatively well using a straight line and that the velocity of the target through this period is also constant."

The higher the ping, the faster the target moves, and the more it changes direction during the rewound period, the less accurate this assumption of a straight line and constant velocity will be. The server may in effect rewind the enemy Mech through positions where the enemy Mech never actually was. Similarly, its exposed components could be in a different place to where they appear to the attacking player, its leg animation may be up instead of down, and its torso may be twisted differently.

Unlike the neat straight line animation shown earlier, what's going on looks more like this...





Net of the above is: a lower ping will enable more accurate shots more of the time, because the amount of recall required will be lower, and the margin of error due to HSR's assumptions will be correspondingly lower.

This would particularly be true when shooting at faster and smaller targets, where the distance travelled is longer, leg animation cycle is faster, speed of torso twist is faster. Against larger, slower moving, untwisting targets, HSR's assumptions would generally be much more accurate.

Rustyhammer, on 03 December 2019 - 11:49 PM, said:

Maybe your moved to AU when PGI implemented changes to the total damage calculation?
Arm damage is no longer added to the total damage when you do torso destruction. Looks to me like the 20-30% damage reduction you're talking about.

Reasonable guess, but nah, I had moved back before the damage nerf.

Edited by Appogee, 04 December 2019 - 01:01 AM.

[Navid A1]

HSR has to save multiple snapshots, since it has to facilitate consistent hits for an arbitrary ping value which can vary from player to player.

Between now and 300ms earlier, my guess is that it's much more than one!!!

[Appogee]

Navid A1, on 04 December 2019 - 02:33 AM, said:

Between now and 300ms earlier, my guess is that it's much more than one!!!

I'm sure you are right.

However, Neema explicitly states that HSR doesn't perform iterative calculations, as that would create too much server workload. HSR makes one calculation based on where the target is versus where it was when the shot was fired, and assumes the attacker and target took a straight line at constant speed from wherever they were to wherever they are now.

The higher the pings, the less accurate HSR's assumptions will be... particularly where agile, small, fast moving Mechs are involved.

Edited by Appogee, 04 December 2019 - 03:37 AM.

[Kamikaze Viking]

Appogee, on 04 December 2019 - 03:21 AM, said:

The higher the pings, the less accurate HSR's assumptions will be... particularly where agile, small, fast moving Mechs are involved.

Yup this is what I've always said.
Its generally not noticeable until the target mech is going 140-150+ and its at is worst when a low ping fast light is hassling a high ping assault and crossing his path, particularly when close and zigzagging.