Jump to content

Arctic Cheata Hitboxes Analyzed


83 replies to this topic

#81 Marodeur

    Member

  • PipPipPip
  • Bad Company
  • 79 posts
  • LocationBraunschweig, Germany

Posted 04 November 2015 - 05:52 AM

View PostDrxAbstract, on 03 November 2015 - 12:21 PM, said:

I wouldnt say the problem with solely with the ACH, though there are certainly some discrepancies. Rather, the problem is damage calculation as a whole in the game coding.

To give an example: I've emptied 6 alternating salvos of 4 CSplas and 4 CMplas into the front and back of a Blackjack at 150ish closing to within 40 meters that had cherry red CT Internals front and back. The only effort he made to 'spread' the damage was twist his torso roughly 20-30 degrees left and right off center rapidly, resulting in him taking only about 30 damage out of 220, which was registered to both arms (which never, at any point, visually blocked my weapons' line of sight to his CT) and both side torsos while his CT took no damage whatsoever.

The same exact thing happened with a Summoner in a previous game, a Dire Wolf's side torso, Timber's CT, Wolverine's CT, Battlemaster's CT and a BK's side torso. The problem becomes very noticeable when rapid movement of any kind is involved: torso twisting, distance changes, height differences (from falling or using Jump Jets), etc. This game engine's ability to accurately measure, record and convey damage deteriorates significantly as the severity of movement increases--They share a near direct proportional relationship.

Which leads to the realization that instant pinpoint convergence is and always has been a myth in this game because:

Two targets closing distance with one another from say 500 meters to 100 in roughly 6 seconds, alphaing as they go, all things considered, you could have your reticle dead on their CT but your weapons are converging a full second or more behind the actual position of your target so they're going to impact the side torsos despite aiming for the CT. To actually hit the CT in this scenario, you'd have to put the crosshairs on their left torso and fire the weapons on the left side of your mech then rinse and repeat right side to right side. This happens all day every day. If Instant convergence were real in this game then angle, distance and speeds wouldnt matter. They do.

Now that we've established instant convergence is false, consider how the actual geometry of Mechs might affect damage registration in both dealing and taking it:

Example:

Mechs with wide torsos or have far off-center arms are worse at dealing accurate damage the closer they are to their target, coupled with rapid changes in distance.

As for comparing the ACH to the FS9, the comparison favors the ACH.
The ACH has:

1. Equal/better damage
2. Longer range
3. Produce less/same heat per alpha

Because Clan weapons vs. IS weapons; Same damage payload for 3 less tons. Means more JJs and ECM.

That's in a direct build comparison. You can modify the FS9 to match JJ counts or have more heat sinks, but you're going to lose damage and engine size to do it.


Also the damage transfer form destroyed side torsos often doesn't work properly. There are big issues with hit registration in the game, but it is really hard to analyze them, because they are not really reproducable. Sometimes it works good, and sometimes not and it is really hard to tell without any futher information about the surrounding conditions where lies the origin. I think it is a mixture of several reasons.
But with so few real programmers that PGI seems to have, they don't focus on that problem I guess, because it works more or less. And otherwise we never will gett some new content. And I mean real content, not mechs or balance here or there a bit or add something to the UI etc..

#82 Marodeur

    Member

  • PipPipPip
  • Bad Company
  • 79 posts
  • LocationBraunschweig, Germany

Posted 04 November 2015 - 06:04 AM

View PostMcgral18, on 03 November 2015 - 07:39 AM, said:

https://m.youtube.co...h?v=BLZgQfx9kAE

Look at that...when you shoot them they die.


Another test with people shooting each other with Flamers is still needed to test hitreg on a calculation stressed server, but it's clear when you actually hit the cheetah, they take damage with a non stressed server.
People just tend to miss or get partial burns.


Nice video,
and the hit reg is also very good all the time. But as I said, sometimes (may be because of the stressed server, or because of different pings of the people) the damage dosn't register properly. And I think that I have enough experience in this game to judge if I hit my opponent good or not, and my aiming skill is also in a good shape.
But anyway, I apreciate the people that making such an effort and try to objectively analyze such things!

#83 Weeny Machine

    Member

  • PipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 4,014 posts
  • LocationAiming for the flat top (B. Murray)

Posted 05 November 2015 - 03:23 AM

View PostMarodeur, on 04 November 2015 - 06:04 AM, said:


Nice video,
and the hit reg is also very good all the time. But as I said, sometimes (may be because of the stressed server, or because of different pings of the people) the damage dosn't register properly. And I think that I have enough experience in this game to judge if I hit my opponent good or not, and my aiming skill is also in a good shape.
But anyway, I apreciate the people that making such an effort and try to objectively analyze such things!


Exactly that's the problem. I also do not think that it is "only" the hitreg but a combination of many factors: Hitreg-Hitboxes-Speed-Ping

When you use SRMs with their slow flight speed this seems to play a role as well

Edited by Bush Hopper, 05 November 2015 - 03:25 AM.


#84 SgtMagor

    Member

  • PipPipPipPipPipPipPipPipPip
  • Survivor
  • Survivor
  • 3,542 posts

Posted 05 November 2015 - 05:21 AM

Artic Cheetah: Atlas in a half shell. assaults should only take damage like the AC :)





3 user(s) are reading this topic

0 members, 3 guests, 0 anonymous users