

#1
Posted 04 April 2015 - 05:11 AM
I believe one of the biggest issues is the Alpha Strike. We are in the days of 60-90 point alpha attacks that reduce TTK for all but the tankiest mechs to seconds once engaged.
I know for many a short TTK is desirable (give me the wins) and part of lore, as is the Alpha Strike is part of lore.
We've made several concessions to game balance vs lore (some much to my chagrin) so how do we have our cake and eat it too? I do not want to take away the Alpha Strike because it's part of lore, but in lore, it was considered a risky gambit.
I propose a dynamic shift to the way Ghost Heat works. Currently GH works by applying a heat boost when an arbitrary "too many" weapons of the same type are fired at the same time. Don't want GH? Don't take all of the same weapons. For many, it's an easy thing to bypass or mitigate.
In lore (yes, back to lore again), the most dangerous weapons are the AC20 and the LRM20. Why? Because they do 20 points of damage. In MWO, that's a pinpoint 20 (AC) or a spread 20 (LRM). Because these are the most powerful weapons with regards to damage output, I will use them as the example for the next part.
GH should be changed from applying to weapons stacking to damage stacking. Let me explain. Taking the LRM/AC20 as my example with a potential damage output of 20 points, any additional weapons fired within .5 seconds (?) that take that or any potential damage above 20 will be subject to GH. Fire one AC20, you are good to go. Fire two, you get GH, NOT because you fired two of the same weapons, but because your potential damage output exceeded 20 points. Fire 3 Medium Pulse Lasers, you are good. Fire 4 and you get GH. Fire an AC10 + 2MPLS, you get GH.
Now, the amount of GH and that damage threshold (I used 20 points) are up for grabs/debate. It should be exponential increases of GH though.
Thanks.
*I am also in favor of reducing the heat capacity to 30, but that's another story....
#2
Posted 04 April 2015 - 05:19 AM
#3
Posted 04 April 2015 - 05:22 AM
....an echoing voice.....
"....convergence...ence...ence...nce...."
#4
Posted 04 April 2015 - 05:27 AM
Obviously, a light should be doing much less damage than an Assault, so maybe an X amount of damage per ton?
3 for 5? At 25 tons your alpha can be 15 (5 increments of 5, 5x3). At 100, your alpha can be 60 (20 increments of 5, 20x3).
That's still a pretty whopping alpha for an assault, but at the same time....that's kind of their thing. At least in the sense that they're the scariest things on the battlefield. (Maybe 2 or 2.5 would be better, but then lights and mediums start getting some pretty low thresholds).
It's a neat idea. Would allow the Ol' King Crab to actually do it's Clan Busting job when built traditionally.
#5
Posted 04 April 2015 - 05:27 AM
Follow the lore
1: Drop the heat cap (it's currently around 2 to 3 times higher than in lore)
2: Remove ghost heat entirelly
3: Add high penalties for running hot (ghost blips on radar, ammo cooking off and exploding, myomer and actuators slowing down/locking up/etc)
Bam. Entire problem solved like that. It's lore friendly, removes the need for invisible heat, and stops people from alpha-striking because doing so makes your life a nightmare in older games.
Good luck getting PGI to change their minds though. They're so stuck in their ways about ghost heat and the heat cap that they aren't going anywhere but backwards soon.
#6
Posted 04 April 2015 - 05:32 AM
As it stands, MWO is in danger of suffering from the same problem that causes hard-core flight sims to be so unpopular: the complaint usually goes something along the lines of "I flew for half an hour only to be blown away by the first missile! Didn't see it coming, didn't even see where the enemy plane was!!" The analogy here being high-alpha poke. This is something that needs consideration as the game is, and very particularly if PGI are considering dropping MWO in to the chaos that is Steam (reviews). TTK needs to be addressed.
Ghost heat accumulative with damage output could well be part of the solution. I also agree with Stefka that convergence is also a big part of the problem. The third issue is smaller mechs being able to wield enough firepower for an assault mech. Hell there are medium builds that have a bigger alpha, and in some cases larger DPS than most Atlas builds. That's plain wrong by both lore, common sense and game design metrics, but is probably a topic for another day...
#7
Posted 04 April 2015 - 05:34 AM
http://mwomercs.com/...of-convergence/
http://mwomercs.com/...-heat-mechanic/
I think reducing heat capacity is something far more players would be comfortable with. It also has the benefit of being easy for new players to understand. Number crunching is a popular element of a game like this, but loads of quirks and ghost heat put together make number crunching a lot harder and, thus, less of an enjoyable pasttime for most people who otherwise enjoy tinkering in the mechlab. Err, I mean Smurfy. No one ever tinkers with anything in the Mechlab.
#8
Posted 04 April 2015 - 05:39 AM
I agree that the ability to alpha is the corner stone on which the meta of MWO has built. And no. I'm not overly fond of it. I don't think convergence is the issue, rather it is the heat scale. The solution however is not more ghost heat (ghost heat from the beginning was a terrible idea). Honestly, the solution is just to readjust the heat scale. It's really that simple. Always has been and the refusal to just do it and get it over with is a major reason why the meta has been kind of lurching from one alpha boat to the next. I'd rather the game be like "pew pew zap swoosh" with mechs able to constantly stream fire with proper trigger control and heat management instead of "alpha/semi-alpha, now let me disappear behind this rock and cool off for a few seconds and I'll do it again."
My thoughts on this in general;
Heat sinks would still function the same way. They increase your heat cap. I think that best as really, rebalancing the heat scale this way shouldn't take any significant coding work. it would be as simple as changing the variables on Mech heat and the weapons in game to adjust for it. The work would just need PGI to do a game survey of heat as it stands now, and how to readjust it such that they can remove alphas, or make alphaing a more significant choice than it currently is. The base heat dissipation rate would also need to go up however, as I see Heat acting more like a fast moving limiter that has to be managed from one shot to the next rather than as a buffer for how much you can fire before needing to let it cool down. Rebalance the heat of various weapons (Ballistics in my mind should all virtually produce no heat, missiles would produce so-so heat, and lasers a lot).
I'd honestly like for the devs to work to try and find a sweet spot where it becomes a choice to alpha or stream damage. I.E. the choice of building your mech for burst or for sustain. Enough heat sinks could raise the heat buffer to where a mech could still alpha, albeit at a damage level less extreme than right now, while just a few can give a mech the ability to generally stream damage and then switch to a sort of mini-alpha using the extra room their heat sinks provide.
Basically, I'd rather heat be a resource for us to manage, rather than a fire limiter to be endured.
Edited by Lord0fHats, 04 April 2015 - 05:46 AM.
#9
Posted 04 April 2015 - 05:42 AM
#10
Posted 04 April 2015 - 05:45 AM
It's the Robinhood effect (shoot and arrow into an arrow, does the target take more damage?).
When you alpha strike, why would the one component that is being hit by multiple weapons, not have a some scalling back of damage because, like, holes, and stuff just passing through the area you just hit?
#11
Posted 04 April 2015 - 05:50 AM
Stefka Kerensky, on 04 April 2015 - 05:22 AM, said:
....an echoing voice.....
"....convergence...ence...ence...nce...."
If only PGI had the coding muscles to pack convergeance, hitreg and collisions into an elegant and efficient set of netcode...
Ultimately the solution to TTK, Alphastrikes and GH is a real heat penalty scale.
Example:
0-20% heat - no effect
20-40% heat - 5% speed reduction
40-50% heat - 5% all CD increase
50-60% heat - 10% speed reduction
60-70% heat - 10% all CD increase
80-90% heat - 30% speed reduction and CD increase
90-95% heat - 10% chance to do 0.5 DPS to random internals
95-100% heat - 30% chance to do 1 DPS to random internals
100+% - 2 DPS to all locations internals, random components in each location, automated shutdown
Now this is just an example, not the ultimate balanced version, but you could add this as well as warnings from Betty and markings on the heat scale and now keeping your heat managed is a clear task with real consequences, not just overheat shutdown. 90-95% is not the redzone but not auto shutdown so you have to decide if you want to redline it or run at a safe level.
Edited by Jetfire, 04 April 2015 - 05:51 AM.
#12
Posted 04 April 2015 - 05:54 AM
Lord0fHats, on 04 April 2015 - 05:39 AM, said:
I agree that the ability to alpha is the corner stone on which the meta of MWO has built. And no. I'm not overly fond of it. I don't think convergence is the issue, rather it is the heat scale. The solution however is not more ghost heat (ghost heat from the beginning was a terrible idea). Honestly, the solution is just to readjust the heat scale. It's really that simple. Always has been and the refusal to just do it and get it over with is a major reason why the meta has been kind of lurching from one alpha boat to the next. I'd rather the game be like "pew pew zap swoosh" with mechs able to constantly stream fire with proper trigger control and heat management instead of "alpha/semi-alpha, now let me disappear behind this rock and cool off for a few seconds and I'll do it again."
My thoughts on this in general;
Heat sinks would still function the same way. They increase your heat cap. I think that best as really, rebalancing the heat scale this way shouldn't take any significant coding work. it would be as simple as changing the variables on Mech heat and the weapons in game to adjust for it. The work would just need PGI to do a game survey of heat as it stands now, and how to readjust it such that they can remove alphas, or make alphaing a more significant choice than it currently is. The base heat dissipation rate would also need to go up however, as I see Heat acting more like a fast moving limiter that has to be managed from one shot to the next rather than as a buffer for how much you can fire before needing to let it cool down. Rebalance the heat of various weapons (Ballistics in my mind should all virtually produce no heat, missiles would produce so-so heat, and lasers a lot).
I'd honestly like for the devs to work to try and find a sweet spot where it becomes a choice to alpha or stream damage. I.E. the choice of building your mech for burst or for sustain. Enough heat sinks could raise the heat buffer to where a mech could still alpha, albeit at a damage level less extreme than right now, while just a few can give a mech the ability to generally stream damage and then switch to a sort of mini-alpha using the extra room their heat sinks provide.
Basically, I'd rather heat be a resource for us to manage, rather than a fire limiter to be endured.
I'm sold on this idea (LOL at the posted video clip

Were the heat cap to be lowered in that way, in addition to the specified benefits PGI should also hugely increase the benefits for, and earnings from, actual scouting. Under a lowered heat cap the current meta of pack as fast a mech as possible with as many assault-grade weapons as possible would be over. To keep light mechs viable, encourage them to do what their role should have been right from the inception of the game: scouting and reconnaissance, not primary damage dealing. The proposed heat cap lowering would, dare I say it, actually help to introduce... Dun dun duuuun!!! Role Warfare

#13
Posted 04 April 2015 - 06:05 AM
Jetfire, on 04 April 2015 - 05:50 AM, said:
Ultimately the solution to TTK, Alphastrikes and GH is a real heat penalty scale.
Example:
0-20% heat - no effect
20-40% heat - 5% speed reduction
40-50% heat - 5% all CD increase
50-60% heat - 10% speed reduction
60-70% heat - 10% all CD increase
80-90% heat - 30% speed reduction and CD increase
90-95% heat - 10% chance to do 0.5 DPS to random internals
95-100% heat - 30% chance to do 1 DPS to random internals
100+% - 2 DPS to all locations internals, random components in each location, automated shutdown
Now this is just an example, not the ultimate balanced version, but you could add this as well as warnings from Betty and markings on the heat scale and now keeping your heat managed is a clear task with real consequences, not just overheat shutdown. 90-95% is not the redzone but not auto shutdown so you have to decide if you want to redline it or run at a safe level.
Also a delayed convergence. TTK would increase, meta would be less cheesy, we would have a little more realism.
Being asking for this since the beginning (ages ago)
But PGI likes too much instant convergence, 'cause it's easier. And a lot of players too.
#14
Posted 04 April 2015 - 06:10 AM
an entire battlemechs arsenal should never land in 1 pin point location, unless the mech has stopped moving and taken time to gets its balance..... cone of fire even, like the old rainbow six game i had where as you shot more the cone of fire or convergence expanded.
#15
Posted 04 April 2015 - 06:19 AM
I think 30 would be a good threshold value, both because it roughly equals the current thresholds (6mlas, 3LL, 2GR etc) so it wouldn't upset balance so much. And 30 PP alpha is IMO acceptable to keep GH free.
But I actually think the current weapon balance is ok, except for a few outliers being useless (Flamer, AC2 etc). So I'd be very careful about screwing around too much if I was PGI.
#16
Posted 04 April 2015 - 06:28 AM
cdlord, on 04 April 2015 - 05:11 AM, said:
Uh-huh. Too bad that the laser-vomit Timby has an Alpha in the mid-50s. The dreaded DoomCrow even lower. Even the semi-popular 2xERPPC/2xGauss Dire Wolf is only 50 points with an additional 10 to the flanking areas the PPCs hit. A little torso twisting and several mechs can take 3 or 4 hits from that with no armor breach.
Stop staring at the enemy, don't be the only guy he sees, and for the love of all that is holy, don't be the only guy on your team in front of 3+ enemies, it doesn't end well.
#17
Posted 04 April 2015 - 06:31 AM
Now I am not the most versed in the mechwarrior lore, but am I right in saying this?
Escef, on 04 April 2015 - 06:28 AM, said:

Edited by Virgil Greyson, 04 April 2015 - 06:33 AM.
#18
Posted 04 April 2015 - 06:41 AM
cdlord, on 04 April 2015 - 05:11 AM, said:
I believe one of the biggest issues is the Alpha Strike. We are in the days of 60-90 point alpha attacks that reduce TTK for all but the tankiest mechs to seconds once engaged.
I know for many a short TTK is desirable (give me the wins) and part of lore, as is the Alpha Strike is part of lore.
We've made several concessions to game balance vs lore (some much to my chagrin) so how do we have our cake and eat it too? I do not want to take away the Alpha Strike because it's part of lore, but in lore, it was considered a risky gambit.
I propose a dynamic shift to the way Ghost Heat works. Currently GH works by applying a heat boost when an arbitrary "too many" weapons of the same type are fired at the same time. Don't want GH? Don't take all of the same weapons. For many, it's an easy thing to bypass or mitigate.
In lore (yes, back to lore again), the most dangerous weapons are the AC20 and the LRM20. Why? Because they do 20 points of damage. In MWO, that's a pinpoint 20 (AC) or a spread 20 (LRM). Because these are the most powerful weapons with regards to damage output, I will use them as the example for the next part.
GH should be changed from applying to weapons stacking to damage stacking. Let me explain. Taking the LRM/AC20 as my example with a potential damage output of 20 points, any additional weapons fired within .5 seconds (?) that take that or any potential damage above 20 will be subject to GH. Fire one AC20, you are good to go. Fire two, you get GH, NOT because you fired two of the same weapons, but because your potential damage output exceeded 20 points. Fire 3 Medium Pulse Lasers, you are good. Fire 4 and you get GH. Fire an AC10 + 2MPLS, you get GH.
Now, the amount of GH and that damage threshold (I used 20 points) are up for grabs/debate. It should be exponential increases of GH though.
Thanks.
*I am also in favor of reducing the heat capacity to 30, but that's another story....
concept, not against,.... but feels a bit arbitrary.
Still, worth exploring.
Though I think proper heat effects as you climb the scale would also mitigate a fair bit. After all, in Lore, you get 16% or so up the scale, and you start having minor effects, like reduced top speed, and as you run hotter, you end up with more speed and responsiveness issues, unreliable targeting, potential shutdown and ammo explosions.
I say, keep the heat cap as is, but as you climb the scale, you deal with the effects.
#19
Posted 04 April 2015 - 06:43 AM
Heat is the problem of its own mainly due to TT values of dissipation, and lack of proper heat scale and results in binary state of energy weapons, they are either superior to ballistic weapons or inferior, depends on PGI's mood and their nerfs/buffs.
Mind you, I'm not against your idea OP, I suggested to cap group fire damage several times myself (with no penalty, you just can't fire if you exceed the cap) as it will keep pinpoint fire, won't be frustrating like proposed convergence systems and won't limit build diversity like sized hardpoints but people can't find consensus on what's wrong with the game and how it should be addressed.
Edited by kapusta11, 04 April 2015 - 06:53 AM.
#20
Posted 04 April 2015 - 06:45 AM
Heat is being used as a bandaid on the gaping wound of pinpoint alpha damage. The solution would be a slight spreading of shots when weapons are fired together, which has been suggested many times. We can't have that though, because PGI say it isn't technically possible and would make hit detection even worse than it already is. Plus of course the people who would cry 'but muh skill' and use the false RNG strawman argument if the game was made more challenging by removing deathstar pinpoint.
So we're stuck with a ridiculous ghostheat system until PGI add something better, and they're unlikely to bother unless it makes more money than it costs.
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users