i have been playing awesomes since i started this game 7+ months ago. in that time the only other assault i have bought is the highlander. i love playing awesome and i consider myself a fairly good awesome pilot, still learning new tricks along the way but even though i have not played other assaults, i can say with high certainty that the awesome are weak in its class.
although the awesome has very good torso twist rate and angle it is very low in armor and fire power compared to the other assaults in this class.
the bone i have to pick is, being 20tons lighter then a atlas, why does the awesome (excluding PB & 9M) only have a max engine rating of 300? (66.8kph). this is only 2.6kph faster then the max engine of an atlas. 2.6 kph is not enough speed to out maneuver stalkers, atlases, highlanders. having much less defense and offence, i would think that the awesome would have some sort of speed advantage to make up for it. 2.6kph is really not enough.
if match making was balanced by weight i might not have such a big problem but currently lighter chassis like the awesome are group with heavier assaults. since there is not much of a difference in speed, the awesome is out classed in every possible way.
so why are the standard awesome so slow?? it is definitely not op to run larger engines or xl. even with all the crazy torso twisting im usually taken out fairly fast in my pb with xl.
i hope you will take a look into it and maybe bump all the awesome engines up to 320+
Edited by King Arthur IV, 14 July 2013 - 06:38 AM.
The awesome is too slow for it's weight in TT so it's too slow in mwo as well. Ideally it would mount a STD320 engine and move 71 kph with speed tweak. Currently my advice is to take a 9M and swap in a STD325 and make it the mech the 8Q should have been in TT.
A fast AS7 has about the same payload as an AWS, plus slightly more armor (and I guess less torso twisting speed + arm movement).
With XL engines, it looks completely different.
But even if AWS were allowed a larger engine, only XL engines would be useful (compare to CTF payload). You can mount a larger engine on AWS-9Ms and -PBs, but there's an unfortunate "trough" at the XL305 engine: engines bigger than XL305 are much less efficient (speed vs payload) than engines below XL300. Additionally, the efficiency decreases faster than linearly, making really big engines incredibly inefficient.
the bone i have to pick is, being 20tons lighter then a atlas, why does the awesome (excluding PB & 9M) only have a max engine rating of 300? (66.8kph).
Max engine rating doesnt matter. What matters is relative speed.
An Awesome with a 300 engine goes 67kph, where an Atlas with a 300 engine only goes 53kph. The Awesome is 26% faster than an Atlas with the same sized engine.
AntiCitizenJuan, on 13 July 2013 - 02:59 PM, said:
Bigger Engine =/= The Answer
The AWS needs a new model, entirely.
With a model change, the Stalker would still be a better pick because it can fill the same role, has a narrower frame, and has an additional 5 tons to play with. If they gave hardpoints actual sizes, they could turn the AWS into a reliable energy platform without touching the number of available hardpoints or the mech's model (it really needs to be done for every weight class).
Base formula:
1 hardpoint = 2 critical slots
Crit space stacks, so multiple hardpoints can result in enough room for PPCs, but you wont be able to match the number of PPCs with available hardpoints
One 2 critical slot hardpoint can carry anything up to a LL; two 2 critical hardpoints (4 critical slots) can carry two LLs or smaller, or a PPC and a single Medlas or anything of comparable size
AWS quirk:
1 hardpoint = 3 critical slots
One 3 ctitical slot hardpoint can carry anything up to a PPC
Again, it is still a single hardpoint, so it can only hold one weapon, regardless of the size
What max PPC loadouts look like with existing hardpoint layouts:
Atlas: RS- 2PPCs
Highlander: 732 and 733p- 2PPCs ; 733c, 733, and HM- 1PPC
It's the only way the awesome is going to be a viable platform (8Q takes the spot of the 6PPC stalker and it's 5 tons lighter!) because none of the proposed changes (heat, hitboxes, model, convergence) do anything to address the role balance within weight classes.
EDIT: I've proposed certain hardpoint sizes a few times and I'm going to put my current version in spoiler tags. None of the stupid heat penalties they are implemented would be necessary if they did this because there would no room for the outrageous builds we see on a regular basis.
What I think the starting base sizes for hardpoints should be:
Energy- 2 critical slots per hardpoint
Ballistic- 4 critical slots per hardpoint
Missile- 3 critical slots per hardpoint
That means mechs get that much usable critical space for each hardpoint they have and the space stacks to allow for heavier weapons where a mech has more hardpoints. These numbers would be used across the board and mechs with specific roles or stock loadouts that do not function properly with these numbers (there are a handful out there) would be given quirks to increase or decrease the base sizes of their hardpoints.
EX: Stalkers have 2 energy hardpoints in each arm (4 crits total in each arm) - that would be enough room for two large lasers or two of anything smaller, or a PPC and medium laser (or something of comparable size). This change takes away the stalker's (and pretty much every other assault's) ability to run the heavy PPC builds that is has come to be known for. The Awesome could be given a 3 crit slot per hardpoint quirk to improve its energy boating ability, making it the only assault chassis capable of carrying anywhere from 2 to 6 PPCs at a time depending on the variant.
EX: The CTF-3d would lose its ability to boat PPCs because it only has single energy hardpoints in each component (forcing them to jump snipe with LLs). The CTF-1x would be able to load a PPC in its right arm, and the jump sniping ability that the 3d lost would be passed down to the lighter Quickdraw that would be able to pack a PPC in its right torso.
EX: Most of the dragon variants have 2 energy hardpoints in their left arm, making them one of the few heavies able to carry PPCs. That change would increase their usefulness.
There are quite a lot of positive (IMO) balance shifts that would occur if these numbers were used. There are also some mechs that don't quite work with this system, but the level of cheese in this game would drop like a rock if this was implemented.
Potential problem mechs that I can think of off the top of my head:
CPLT-K2: Needs another energy hardpoint in the arms to carry PPCs (2 crits per hardpoint) and to give it an edge over the Firebrand who has a very samey, but all around better layout IMO.
Awesomes need a 3 critical slot per hardpoint quirk to effectively dominate the energy boating role (like they are supposed to)
HGN 733p needs an additional energy hardpoint in the left arm to allow double PPCs. It's an energy variant, so it should be given some special attention to better define its role
HBK-4H and Atlases would need increases ballistics hardpoint sizes or more hardpoints so they can carry their stock weapons etc.
I dunno, I think a flat heatcap based solely on mech chassis/variant would be a better boost for the awesome. Give it the highest heatcap in game. It would atleast see use as a PPC boat over the stalker if it cold fire larger salvos before shutting down due to heat.
Give most mechs a heatcap in the 20's-30's, give the awesome a heatcap in the 40's-50's.
Awesome would put out more sustainable firepower, while still being a barn and eating alpha's to the chest.
Stalker's would be more survivable but unable to provide sustained firepower.
Or conversely, Have heatsinks on the awesome be more efficient.
Stalker would spit out massive damage but suffer from severe heat.
Awesome would spit out more sustainable damage.
after reading all these post and some in the forums i am starting to think maybe there is no quick fix to the awesome. the combination of large hix boxes, small fire power, little armor, small engines etc. just dont make it a very good brawling mech.
as a lrm support it is currently my favorite (8R, 4x lrm15) but it lacks close range self defense and it will shortly be indirectly nerfed due to new heat scaling. as a ppc boat or laser boat, its weapons are mounted very wide apart and low, this makes for alot of over exposure before shooting or too may obstetrical in the way.
although the awesome have better quirks eg. torso twist. everything else pulls it so far down that those quirks are negligible.
in the mean time im still going to play the mech because i like playing with the under dog. i hope the devs do not misunderstand why awesome are being played, they are not played because they are balance, they are played because they are awesomes. they are not weak to the point you cant play them but they are undoubtedly weak in its class.
only way the awesome to be relevant would be if hard-point restriction were implemented as it is one of the few mechs that legitimately boat Pee Pee Cees...
With a model change, the Stalker would still be a better pick because it can fill the same role, has a narrower frame, and has an additional 5 tons to play with.
I disagree. The speed difference vs the Stalker shows even with the current bad Awesome model, and chassis other than the 9M. The model is quite obviously the main reason for the "awesome" performance. Just look at what the Victor can survive compared to the Awesome.
Despite the other issues that the Awesome might have, I do agree that the Max Engine Rating should be 320 (or 325 for the extra Heatsink) to allow the Awesome to at least move at 64 kph (and 70+ kph with max efficiencies). Right now the Awesome is at an awkward max speed that doesn't really help it too much.
A fast AS7 has about the same payload as an AWS, plus slightly more armor (and I guess less torso twisting speed + arm movement).
With XL engines, it looks completely different.
But even if AWS were allowed a larger engine, only XL engines would be useful (compare to CTF payload). You can mount a larger engine on AWS-9Ms and -PBs, but there's an unfortunate "trough" at the XL305 engine: engines bigger than XL305 are much less efficient (speed vs payload) than engines below XL300. Additionally, the efficiency decreases faster than linearly, making really big engines incredibly inefficient.
While true, there is nothing more satisfying than chasing down stock light mechs with an 80 ton machine.