Jump to content

Re-Examine Jagermech Side Torso Hit Boxes


96 replies to this topic

#61 Noosemane

    Member

  • PipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 461 posts
  • LocationFlorida

Posted 22 October 2013 - 04:48 AM

You can trade in your AC40 for smaller hitboxes.

#62 Bad Andy

    Member

  • PipPipPipPipPipPip
  • 270 posts

Posted 22 October 2013 - 04:49 AM

maybe, but there are many other mechs that have much worse hitboxes like the Awesome and Quickdraw

#63 Dimento Graven

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Guillotine
  • Guillotine
  • 6,208 posts

Posted 22 October 2013 - 07:28 AM

View PostKrivvan, on 22 October 2013 - 02:33 AM, said:

Those hit boxes are working as intended: http://img15.hosting...cataphract2.png

Doesn't stop the Cataphract from being one of (if not the) best Heavy.
I'm sorry, but I don't see how a user/non-PGI employee post "proves" that those hit boxes are 'working as intended'.

From another thread, I did some visual comparisons against the afore mentioned link of known hit box locations against what the 'mech lab presents to us as 'component locations', here's what I found:

Here's what I've found:
Commando - Only difference is the head/torso representation, it's probably over 90% correct.
Spider - Same
Raven - Same, only the head representation is actually a lot closer than the previous 2, meaning it's nearly 100% correct
Jenner - Same, though the head/ct representation is WAY off on this one.
Locust - As far as I can tell, it's 100% correct

Cicada - Only difference is head/torso representation, somewhere in the 90's in correctness though.
Hunchback - Same as above, though the head representation is only slightly off in 'mech lab, it's ~98% correct.
Centurian - Same as above, though it's probably only 90% correct
Blackjack - Same as above, ~95% correct in head/torso region
Trebuchet - Same as above, ~90% correct in head/torso region
Kintaro - Don't own one, can't validate myself.
Shadohawk - Haven't seen the hit box locations published on those yet.

Cataphract - We've already established SIGNIFICANT differences, video was part of the OP
Dragon - Very close, slight difference in head/torso representation, at least 90% correct
Catapult - Same as above
Jagermech - Suffers the same issue as the Cataphract in the arm/side torso region. The arms are represented as having a larger area, and the side torsos are represented much smaller. This needs to be corrected. Interestingly enough the head/torso area is 100% correct as far as I can tell.
Quickdraw - As far as I can tell it's 100% accurate.
Orion - Don't own one, can't check.
Thunderbolt - Haven't seen the hit box locations published on those yet.

Atlas - Almost 100% correct. When clicking the head it appears to be trying to highlight the very center "nose area" of the head in 'mech lab though, which is interesting.
Stalker - Appears to be 100% correct as far as I can tell.
Highlander - Appears to suffer from the same issue as the Cataphract and Jagermech, the arms and side torsos are not represented accurately, the head is 'nearly' accurate, but includes the two side 'windows' when highlighted, it's in the 90 percentile range for accuracy though.
Awesome - The arms and side torsos are really represented wrong in 'mech lab, and the head is close to as wrong as the Cataphract.
Victor - Mostly correct, only the head is represented as taking up slightly more area than the actual hit box location.
BattleMaster - Haven't seen the hit box locations published on those yet.

Basically my point is, either the hit box locations are wrong, or the 'mech lab presentation of those component locations is wrong.

One of the two needs to be corrected.

Besides, denying the possibility that hit boxes are wrong denies the recent history most of us have had to endure with Founders Atlas head hit box issues, the supposed Kintaro hit box issues, just to name the most recent.

Saying you don't want it fixed because it can mount two AC20's or two gauss is spurious reasoning. If the hit boxes are wrong, they need to be fixed. Considering the differences in hit box locations from what 'mech lab shows us is the arm/side torso locations, the difference needs to be accounted for and corrected.

View PostGhogiel, on 22 October 2013 - 03:40 AM, said:

The fallacy of CBT
I'm sorry but your 'one liners' still aren't communicating effectively.

What do you mean?

Why aren't original FASA design specs sufficient justification to have 'mechs of different weight classes occasionally be out of the 'expected' norm?

#64 Ghogiel

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • CS 2021 Gold Champ
  • CS 2021 Gold Champ
  • 6,852 posts

Posted 22 October 2013 - 07:46 AM

View PostDimento Graven, on 22 October 2013 - 07:28 AM, said:

What do you mean?

Why aren't original FASA design specs sufficient justification to have 'mechs of different weight classes occasionally be out of the 'expected' norm?

Because using it as a reason for any argument in MWO is a logical fallacy.

#65 R Razor

    Member

  • PipPipPipPipPipPipPipPip
  • The 1 Percent
  • The 1 Percent
  • 1,583 posts
  • LocationPennsylvania ...'Merica!!

Posted 22 October 2013 - 07:53 AM

View PostDimento Graven, on 22 October 2013 - 07:28 AM, said:



Why aren't original FASA design specs sufficient justification to have 'mechs of different weight classes occasionally be out of the 'expected' norm?


Aside from the fact that there is no scale provided in those diagrams that would indicate that all were drawn to the same scale?

How about the fact that FASA didn't design MWO and therefore, as has been pointed out to myself on occasion, it doesn't matter IN THE LEAST what FASA did back then.

#66 MeiSooHaityu

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • Knight Errant
  • Knight Errant
  • 10,912 posts
  • LocationMI

Posted 22 October 2013 - 07:57 AM

The Jager probably takes more side torso damage due to the fact that it's arms are mounted up high. Arms can help act as a shield to side torsos. This is why the OP noticed that Cataphracts tend to take less side torso damage (lower mounted arms) vs the Jager (shoulder high mounted arms).

Basically, it is a trade-off. Less durability for weapons mounted in a more ideal firing position.

If you can live with the Cataphracts weapon height, go with the Cataphract. If you are tired of blowing the tops off of small hills, go with the Jager, but be mindful that your sides are a bit more exposed.

In a nutshell, hitboxes are fine.

Edited by MeiSooHaityu, 22 October 2013 - 07:57 AM.


#67 Dimento Graven

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Guillotine
  • Guillotine
  • 6,208 posts

Posted 22 October 2013 - 08:07 AM

View PostGhogiel, on 22 October 2013 - 07:46 AM, said:

Because using it as a reason for any argument in MWO is a logical fallacy.
Well this is where we agree to disagree then. The MWO designs are based on original FASA designs. That's why so many of the 'mechs that have been brought out are so readily recognizable to BattleTech/MechWarrior fans.

Staying as true as possible to those designs and concepts, for the most part, serves MWO more so than ignoring the past 30 years of history.

Does that mean PGI has to stick "hard and fast" to the exact standards set forth? No, obviously not, even FASA's/WhizKidz design concepts were sometimes more "rules of thumbs" than sacrosanct commandments passed down from on high, so there can be disparity BOTH ways, and I'm fine.

I'm just not one to scream, "The sky is falling" when something doesn't match my preconceived notions exactly.

As a matter of fact, I've played nearly 2000 matches, combined, in the various Jager chassis before I finally decided to post the above, and all told, I've probably got something near, if not over, 10,000 matches played counting all 'mech chassis, pre & post stat reset, so I'm not just talking out my butt here.

There's a problem either with the hit boxes, or with hit detection. If it is the hit detection then all 'mechs suffer it, and any non-light chassis, and any 'mech with a particularly 'thin' hit location will suffer for it more so than others with proportionately 'thicker' locations, so it should be something that's addressed. If it's a few 'mechs that have incorrectly assigned hit boxes, or bad hit box design, then for 'balance' and 'fairness' those 'mechs should be addressed. If it's just that the 'mech lab sets up an incorrect expectation of hit box locations, it should STILL be addressed.

View PostR Razor, on 22 October 2013 - 07:53 AM, said:

Aside from the fact that there is no scale provided in those diagrams that would indicate that all were drawn to the same scale?

How about the fact that FASA didn't design MWO and therefore, as has been pointed out to myself on occasion, it doesn't matter IN THE LEAST what FASA did back then.
No scale? Perhaps you don't see the human standing next to the 'mech to provide some sense of 'scale'?

No, FASA didn't design MWO, but PGI is using the IP mostly originated by FASA to design MWO, so it's not so unreasonable to see the same level of 'variability' in 'mech sizing that's always existed for the past 30 years...

Edited by Dimento Graven, 22 October 2013 - 08:09 AM.


#68 Dimento Graven

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Guillotine
  • Guillotine
  • 6,208 posts

Posted 22 October 2013 - 08:14 AM

View PostMeiSooHaityu, on 22 October 2013 - 07:57 AM, said:

The Jager probably takes more side torso damage due to the fact that it's arms are mounted up high. Arms can help act as a shield to side torsos. This is why the OP noticed that Cataphracts tend to take less side torso damage (lower mounted arms) vs the Jager (shoulder high mounted arms).

Basically, it is a trade-off. Less durability for weapons mounted in a more ideal firing position.

If you can live with the Cataphracts weapon height, go with the Cataphract. If you are tired of blowing the tops off of small hills, go with the Jager, but be mindful that your sides are a bit more exposed.

In a nutshell, hitboxes are fine.
Only hit boxes for Jagers, Cataphracts, and Awesomes all seem to be VERY different than what is presented to us for the same component locations in 'mech lab. It's odd that these three 'mechs and as far as I've been able to determine, ONLY these three 'mechs, suffer from this disparity between 'actual' component hit boxes, and what's displayed in 'mech leb.

Every other 'mech I've compared arm/side torso locations with in 'mech lab with known hitbox locations provided in the link from earlier in the thread, match up.

This disconnect seems to indicate that there's an issue somewhere.

Maybe it's 'mech lab.

Right now, given the history with other 'mechs, I'm thinking it's the hit boxes.

#69 Ghogiel

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • CS 2021 Gold Champ
  • CS 2021 Gold Champ
  • 6,852 posts

Posted 22 October 2013 - 08:35 AM

View PostDimento Graven, on 22 October 2013 - 08:07 AM, said:

Well this is where we agree to disagree then.



You can't agree to disagree lol. It's a logical fallacy. It's some kind of appeal to antiquity/authority or false equivocation.

#70 Dimento Graven

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Guillotine
  • Guillotine
  • 6,208 posts

Posted 22 October 2013 - 08:38 AM

View PostGhogiel, on 22 October 2013 - 08:35 AM, said:

You can't agree to disagree lol. It's a logical fallacy. It's some kind of appeal to antiquity/authority or false equivocation.
I'm certainly not going to agree with you on your incorrect opinion, and you obviously aren't going to acknowledge the correctness of mine, so therefore we can either continue arguing, or end all arguments by agreeing that we will never agree on it.

Hence, 'agreeing to disagree.'

#71 Ghogiel

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • CS 2021 Gold Champ
  • CS 2021 Gold Champ
  • 6,852 posts

Posted 22 October 2013 - 08:44 AM

View PostDimento Graven, on 22 October 2013 - 08:38 AM, said:

I'm certainly not going to agree with you on your incorrect opinion, and you obviously aren't going to acknowledge the correctness of mine, so therefore we can either continue arguing, or end all arguments by agreeing that we will never agree on it.

Hence, 'agreeing to disagree.'

You don't HAVE an argument because it is based on appeals to CBT.

#72 Dimento Graven

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Guillotine
  • Guillotine
  • 6,208 posts

Posted 22 October 2013 - 08:48 AM

View PostGhogiel, on 22 October 2013 - 08:44 AM, said:

You don't HAVE an argument because it is based on appeals to CBT.
It's not an 'appeal' it is showing the SAME precedent that even PGI is uses for the variability in 'mech size, and therefore when a 'mech is larger or smaller than the non-BattleTech fan might expect, it's not the end of the world, the sky isn't falling, and momma didn't steal your paci....

Hence, I'm not pitching a fit that the Locust doesn't match the FASA scale. It's ok, variability in 'mech sizing has 30 years of precedent, so it's ok...

Edited by Dimento Graven, 22 October 2013 - 08:49 AM.


#73 Ghogiel

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • CS 2021 Gold Champ
  • CS 2021 Gold Champ
  • 6,852 posts

Posted 22 October 2013 - 09:55 AM

View PostDimento Graven, on 22 October 2013 - 08:48 AM, said:

It's not an 'appeal' it is showing the SAME precedent that even PGI is uses for the variability in 'mech size, and therefore when a 'mech is larger or smaller than the non-BattleTech fan might expect, it's not the end of the world, the sky isn't falling, and momma didn't steal your paci....

Hence, I'm not pitching a fit that the Locust doesn't match the FASA scale. It's ok, variability in 'mech sizing has 30 years of precedent, so it's ok...

It is an appeal by definition. You are after all invoking CBT in an attempt to show precedent.

#74 Dimento Graven

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Guillotine
  • Guillotine
  • 6,208 posts

Posted 22 October 2013 - 09:56 AM

View PostGhogiel, on 22 October 2013 - 09:55 AM, said:

It is an appeal by definition. You are after all invoking CBT in an attempt to show precedent.
It's not an 'attempt' when it succeeds.

#75 Ghogiel

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • CS 2021 Gold Champ
  • CS 2021 Gold Champ
  • 6,852 posts

Posted 22 October 2013 - 10:03 AM

View PostDimento Graven, on 22 October 2013 - 09:56 AM, said:

It's not an 'attempt' when it succeeds.

It succeeds in being a logical fallacy and as such renders your argument invalid.

#76 Dimento Graven

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Guillotine
  • Guillotine
  • 6,208 posts

Posted 22 October 2013 - 10:12 AM

View PostGhogiel, on 22 October 2013 - 10:03 AM, said:

It succeeds in being a logical fallacy and as such renders your argument invalid.
I don't understand how you can believe that using CBT as a precedent for something in a game, BASED ON CBT, is a 'logical fallacy'.

Either you don't understand the definitions of those words, or you're just being intentionally obtuse.

Damn near every bit of this game is based on CBT.

All the base chassis, and standard variants, and even the hero 'mechs are straight out of CBT TRO's.
All the weapons currently employed are straight out of CBT Rules of Warfare, AND, by the way, most of the base stats of those weapons match, point for point the original CBT numbers.
Movement, for the most part, matches CBT rules.
The ********* time line they're using is based on the CBT Clan Invasion timeline.

Yes, there are differences, but certainly PGI isn't ignoring CBT in their design decisions, and we shouldn't ignore CBT in judging those decisions, either.

So your insistence that using CBT to show precedent in a game based on CBT as somehow 'wrong', is in of itself bringing your 'logic' into serious question.

#77 Ghogiel

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • CS 2021 Gold Champ
  • CS 2021 Gold Champ
  • 6,852 posts

Posted 22 October 2013 - 10:28 AM

View PostDimento Graven, on 22 October 2013 - 10:12 AM, said:

I don't understand how you can believe that using CBT as a precedent for something in a game, BASED ON CBT, is a 'logical fallacy'.

CBT =/= MWO

#78 Dimento Graven

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Guillotine
  • Guillotine
  • 6,208 posts

Posted 22 October 2013 - 10:32 AM

View PostGhogiel, on 22 October 2013 - 10:28 AM, said:

CBT =/= MWO
You still insist on ignoring all the things in this game that are based on CBT.

You are indeed, just maintaining an arrogant ignorance.

I'm going to do my best to ignore you now, you may just graduate to being placed in my forum ignore list...

#79 Ghogiel

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • CS 2021 Gold Champ
  • CS 2021 Gold Champ
  • 6,852 posts

Posted 22 October 2013 - 10:41 AM

View PostDimento Graven, on 22 October 2013 - 10:32 AM, said:

You still insist on ignoring all the things in this game that are based on CBT.

You are indeed, just maintaining an arrogant ignorance.

I'm going to do my best to ignore you now, you may just graduate to being placed in my forum ignore list...

Putting your fingers in your ears and shouting *lalalalala* won't make your argument sound.

American law is based on English law. Trying to show precedent by invoking English case law in a US court doesn't work.

#80 Dimento Graven

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Guillotine
  • Guillotine
  • 6,208 posts

Posted 22 October 2013 - 10:49 AM

View PostGhogiel, on 22 October 2013 - 10:41 AM, said:

Putting your fingers in your ears and shouting *lalalalala* won't make your argument sound.

American law is based on English law. Trying to show precedent by invoking English case law in a US court doesn't work.
And repeating the same stupid thing over and over as you're doing doesn't make your argument any less moronic.

Let me ask you, did you think PGI was stupid when they brought the Locust, Shadowhawk, Thunderbolt and BattleMaster into the game? After all, all 4 'mech chassis and all their variants are straight from CBT.

Where's your posts insisting that PGI create their own fresh 'MWO based' content?

Comparing a video game to 'law' is an inane argument, not really even worth responding to.

I'm not using precedents from some game prior to CBT as a justification for 'mech size variability in MWO, I'm using the same precedents from CBT that PGI has used in their decision making up to this point.

If you can't understand that, then you don't even understand the discussion, and it's not worth arguing with you and you truly can be ignored with zero loss to the discussion.

Edited by Dimento Graven, 22 October 2013 - 10:49 AM.






1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users