10 replies to this topic

[Cato Zilks]

So, the podcast put out today said some things about some agility twerks coming up in the near future for game balance. That got me thinking that it might be fun to have the mechs handle a bit differently from each other. Give an Atlas great Torso twisting but poor leg turning, while the Direwolf gets good turning but poor torso twisting. Both mechs become more able to adjust movement than they can now, but in very different ways.

Or for other examples, have notorious urban brawlers like urbies and hunchies get great deccel and torso twisting but bad accell and turning. Here is an example spreadsheet for different weight classes. https://docs.google....#gid=2071502881
I am not proposing those numbers be used, they are only meant to serve as an example for what kind of variation we might want to see.

What do you guys think? Would this be a fun way of giving mechs more unique feels?

[Khobai]

the direwolf needs acceleration not turning

a direwolf cant even accelerate out of an artillery radius in time

youre guaranteed to get hit which is just plain dumb

[Cato Zilks]

Thats fine, I don't care whether the a specific mech gets a specific stat, the question is would having different rates of agility stats positively add a flavor to the mech? Right now "agile mechs" tend to get buffed across all agility stats and over performers get nerfs across all agility stats. This means that the mechs all pilot relatively similarly, just to a slower or fast degree.

Almost every mech has the exact same ratio of torso twist speed to turn rate: namely torso twist is from 1.785-1.805 times faster than turning speed. The King Crab and Assassin are the only outliers with higher ratios and the Black Lanner is the only chassis with a lower ratio (its turning rate was buffed) (also, three individual variant are lower as well). You can check all the numbers here: https://docs.google....#gid=1393526485

Edited by Cato Zilks, 12 August 2018 - 01:44 PM.

[Cato Zilks]

Additionally, I had asked Chris Lowry (the PGI balance guy) if the team would ever consider having the mechs feel different like this. Here is his reply:

"It may be worth testing a few methods in the PTS that Paul mentioned, but legibility of the Abilities in the UI as well as consistency with overall performance behavior is always a concern. For example, I've often considered experimenting with the mobility performance curves On 'Mechs to get somewhat different behavior depending on the different 'Mechs. But this presents a few challenges. For example, lets say I reduced the Atlas' low speed maneuverability from where it was now, but gave it back as better mobility at higher speeds. The net effect would result in more mobility at higher speeds, but as the mobility values are often averaged in the UI for many 'Mechs, the numeric number provided would not paint an accurate picture at a glance. You would have to rely more on the mobility performance curve charts to form a more accurate idea in how the 'Mech performs. This also leads to consistency concerns with how values scale in the skill tree. If a 'Mech has better turn radius, but lower Accel / Decel, the progressive nature in the way that the skills are arranged means that progression would be slow going for any Accel / Decel skill investments, but Turn Radius nodes become that much more potent.

Overall none of this is bad in theory, but when we did Engine De-sync we deliberately stuck to the bounds of how things where already functioning in-game to keep things fairly consistent. Both in how players expected 'Mechs to operate going into it as well as to keep skill tree progression consistent between all of the attributes. Since we are still looking into changes to take into PTS on this front I wouldn't mind getting opinions on if any of this sounds like it's worth testing on PTS. We've had enough time away from the older system that I wouldn't mind experimenting with ways to make the handling on each 'Mech feel more unique. But this will mean more reliance on viewing the performance curves to get a more accurate picture of how a 'Mech handles, and it will mean skill tree investments may be weighted differently depending on how far a 'Mech deviates from the baseline template. So by all means sound off if you have an opinion on that front."

Obviously, he lays out the natural consequences for this style change and adds a lot more nuance to the information and how to look at turning radii. So the question really is, are you all interested in have the mechs agility stats more nuanced for the chassis instead of being more cut and paste by weight? Do you want your mechs to pilot differently from each other?

[Cato Zilks]

Any interest? I have saying they are open to feedback on this point and I cannot even get responses?

[TheMundaneYesYes]

Cato Zilks, on 08 August 2018 - 01:40 AM, said:

Thats fine, I don't care whether the a specific mech gets a specific stat, the question is would having different rates of agility stats positively add a flavor to the mech? Right now "agile mechs" tend to get buffed across all agility stats and over performers get nerfs across all agility stats. This means that the mechs all pilot relatively similarly, just to a slower or fast degree.

Almost every mech has the exact same ratio of torso twist speed to turn rate: namely torso twist is from 1.785-1.805 times faster than turning speed. The King Crab and Assassin are the only outliers with higher ratios and the Black Lanner is the only chassis with a lower ratio (its turning rate was buffed) (also, three individual variant are lower as well). You can check all the numbers here: https://docs.google....#gid=1393526485

Wow, the KDK-3 was hit brutally with moblity. Did not know the victor has a edge over the HBKIIC as far as agility goes , I rarely use assaults. Still bitter over the nerf to HBKIICs agility . I used to SRM brawl in the B variant now its a ATM boat, yay... Just means instead of brawling I will hit harder to make up for it , something PGI did not intend I imagen.

Edited by Number767, 12 August 2018 - 05:16 PM.

[Cato Zilks]

Right, so some of these mechs were notorious poptarts with tons of pod-space and subsequently had all agility stats pummeled into the ground. Maybe if we don't have all the stats linked, you might still be able to brawl in mechs like the Hunchie-IIC or the Night Gyr. Maybe we could see the Mad-IIC actually be able to accelerate, while not having pre-engine-desync torso twisting that made the Scorch so nightmarishly effective.

Giving each chassis a different mobility profile allows Chris to directly hit certain mobility stats to reign in over-performing mechs while still not making certain play styles obsolete for the chassis. Night Gyr is a great example of this, its dual Gauss poptarting supremacy lead to a mobility nerf... which slightly affected dual Gauss builds but crippled the mech's ability to brawl.

[TheCaptainJZ]

I think mobility is a large enough area that all mechs should be reset to baseline values per tonnage, then changed for specific chassis based on historical mobility (without holding to actual numbers), hardpoints, overall performance/power, etc. Throw that up on the test server with no other changes sometime. I think Engine Desync was great since twisting speed no longer has any bearing on the engine rating, but the problem was it was never tuned afterwards and now Assault pilots (mainly) are upset because they don't twist fast enough to spread damage. Mechs that have fewer hardpoints or are not as able to boat should have better agility. I also would like to see mechs turn faster compared to now when they are very slow or stopped, but not when moving. This is to help track lights in a one-on-one or just for basic moving around buildings or obstacles since the maps are quite cramped in spots. As for more complicated acceleration curves, in general, I'd try to stay away from altering it except in a few cases perhaps to add flavor. But it should be the exception to changing the actual curvature. Change twist speed and overall turning speed which will impact agility as much as necessary.
Edit: Torso twist angle is also an important variable. Less than 90 degrees feels like a nerf.

Edited by TheCaptainJZ, 12 August 2018 - 06:17 PM.

[MechaBattler]

The thing is with varying it up. Is that some mechs have to lose in certain regards. I'm okay with that as a concept. But we'd be relying solely on PGI's discretion for the stats. I mean they're kinda already doing it. The Wolverine had some of the best agility for mediums above 45 tons before the Uziel came along and dethroned it. But high agility hasn't saved the Uziel, with a few people finding it good in it's Solaris division.

[Cato Zilks]

MechaBattler, on 12 August 2018 - 07:11 PM, said:

The thing is with varying it up. Is that some mechs have to lose in certain regards. I'm okay with that as a concept. But we'd be relying solely on PGI's discretion for the stats. I mean they're kinda already doing it. The Wolverine had some of the best agility for mediums above 45 tons before the Uziel came along and dethroned it. But high agility hasn't saved the Uziel, with a few people finding it good in it's Solaris division.

So some mechs will see reductions in stats, but I am not sure that will be a loss. Take the Nova that in TT has a no torso twist quirk because the mech's legs connected to the shoulders. We could massively boost its turning ability while nerfing its torso twist. The net result is not a loss in regards to the ability to quickly turn to face the enemy or to shield one side of the mech, its just that in the Nova you would have to turn your feet to do this quickly as opposed to... i don't know... a Huntsman. The change to the Nova could be a net buff in terms of the mechs ability to turn its torso, just by different means.

Edited by Cato Zilks, 13 August 2018 - 01:31 PM.

[process]

We used to have something like this. Within certain chassis, a variant may have a different engine cap or torso angle. I'm not sure what the rationale would be for this now with the arrival of quirks, but I'm not opposed.

It'll be far more important to get each chassis placed within the entire mech spectrum, e.g. fixing engine desync, before flavoring individual variants further.