12 replies to this topic

[ClaymoreReIIik]

Hello everyone, please excuse the wall of text. Complex topic, needs lots of explaining so its not misunderstood.

I think a lot of us agree, that this last attempt at balancing was not on target.

I have been a die hard battletech fan for ages and I have played all mechwarrior titles since mw3 online in the competitive scene, so thats the background on my perspective.

First of all I would like to say that I like about MWO that it has a lot of weapon systems that are playable in the right setting and that there is no "golden one gun to use for world domination".
You have a lot of great ideas implemented for balancing and a lot of things work in the intended way, they are just arranged in a suboptimal way.

The way this post is built up is the following:
First I am going to point out what I see as the defining "Cornerstones" for IS and Clan Mech-Warfare.
Second I am going to utilize 2 "comparable" Mechbuilds that currently don't have many quirks to point out where the above mentioned Cornerstones work and where they currently fail.
Third I am going to explain a couple of "minor" changes that would help balance the differences in gameplay and try to explain why they help.


To the first part:
As a battletech fan I have certain "traits" in the back of my head whenever I think about the Inner Sphere and the clans.

The inner sphere has this "die hard", "never give up", "grit and grind" mentality that despite all the difficulties and challenges they keep being innovative and downright dirty to stay in the fight. Their attempts to close the technology gap are rushed and "fragile".

The clans have this clean, abstracted view on warfare that idealizes 1on1 combat and honor on the battlefield. They have far advanced technology that gives them bigger range, better speed and better heat efficiency.

Lets analyze two 55 ton mechs (one from each side) and try to find out if we can represent that.

To the second part:
I have chosen the wolverine hero QUARANTINE as my IS example because it is similar enough in playstyle to the clan mech and currently does not have significant quirks that could distract from technology differences:
http://mwo.smurfy-ne...02ee37d7bcc895c

The main weapon systems are 3 IS Large Lasers with a base burn duration of 1 second for 9 damage each creating 7 heat each at optimal range of 450 meters. The config is rounded out with 2 SRM4 (with barely any ammo) to improve DPS close up. The mech comes with an XL engine and 2 Jump Jets. The mech has 16 double Heat sinks and 320 points of armor.

This mechs works at ranges 30 to 450 meters in more of a "medium range fire support/sniper"-role.

The clan mech I have chosen is a Stormcrow that is fitted for a "medium range fire support/sniper"-role in a very simplistic configuration. Omnipods are chosen to avoid negative quirks:
http://mwo.smurfy-ne...c744797e3eb8055

Main weapon systems are 2 Clan Large Pulse Lasers with a base burn duration of 1.12 seconds for 13 damage a piece creating 10 heat each at optimal range of 600 meters. The config is rounded out with 2 Clan ER Medium Lasers with a base burn duration of 1.15 seconds for 7 damage a piece at 6 heat each at optimal range of 405 meters. The mech comes with a clan XL engine and 0 Jump Jets. The mech has 19 double Heat sinks and 364 points of armor.

If you compare them at each mechs optimal range, the IS Mech has more Alpha (less pinpoint) and higher sustained DPS and better maneuverability (jumpjets) while the clan mech has better speed, armor, heat efficiency, durability (XL engines) and range.

The clan mech is additionally a lot easier to play, since both weapon systems work in the same way (close in burn duration and firing behaviour, maps encourage encounters between 300 and 800 meters).

So how do these mechs represent my proposed cornerstones?

The Wolverine is representing the "playing dirty" with the Jumpjets it can bring to the table and the SRMs as a "token of love" to the ones charging in on the sniper/fire support. It represents the fragile attempt to cover the technology gap by utilizing an XL that will kill the mech when a side torso goes. It represents the "make lemonade out of lemons" by adding a third large laser to keep up with the better range firepower of the clans.

The stormcrow is representing the clan side by being faster then the wolverine, having more range then the wolverine, having more durability (through more armor and the clan xl death condition). The heat efficiency is better as well.

As it is right now the stormcrow is clearly the better mech, because it is better in all really relevant metrics. A little underrepresented in this comparison is the heat cap. The stormcrow can actually build up heat longer, because it has more heat sinks. Its actually able to outlast the wolverine in alpha-trades.

So what we are really missing from my explained ideal world is the IS ability to outlast the clans and the clans focus on 1on1 warfare. I just don't feel those right now.

This is what I would like to change on the IS side of things to make the IS mechs "better":
Outlasting the opponent in an armor and structure way only works in 1on1 settings because as soon as you focus more then 1 mech onto a target, the additional structure and armor quirks won't make any difference.

So lets focus the outlasting on the part where you can keep up DPS longer. With some minor exceptions the IS ability to provide DPS is directly linked to the heat efficiency of the Mechs. Most of the mechs that were Meta after "the Quirkening" became better because they brought heat generation quirks (among other things). Most of the OP ones got "pulled back in line" by reducing their heat generation quirks.

To provide more balance (IMHO), 2 changes are needed. First: fix heat cap to 70 for all IS mechs. Thats the current equivalent of 16 double heat sinks and feels about right. Second: let all IS double heat sinks, regardless of location dissipate 2 heat per cycle (yes the none engine ones). This enables the IS to bring smaller standard engines without sacrificing "offensive staying power" and generally to keep fighting longer once in range.

This will (at large) slow down the IS and make them really sturdy with standard engines, or give them the edge in brawls and reward them from getting into range. If you want to go fast and fragile you will not have to dedicate that much space to heat sinks and maybe close the speed gap to the clan mechs at the expense of durability. This opens up more options.

This actually scales quite well in bigger battles, too, since the clan side will have a heat induced DPS dip at some point while the IS will run into that a lot later, giving them a sweet spot and upper hand for a short while. They will have losses before that though.

The IS tools to mitigate the clans range advantage (ER PPC, ER large Laser, AC2) are all difficult to use right now because they run too hot and present a huge disadvantage once things get serious. The above mentioned changes would help with that and give clan ER weapons a reason.

The thing missing from the clan side for me is their focus on the 1on1 combat. This is just not represented. The clan mechs can field more heat sinks (less critical slots needed, lighter weaponry) so in the current system they both have higher heat cap and heat dissipation. The lore wants clan mechs to have better heat efficiency so I would not want to touch their dissipation.

I think it makes sense to lower their heat cap though. Fix that to 50. That is plenty of heat cap to take out a single mech, but you will be riding the red line all the time when fighting the second target. I would keep none-engine heatsinks at 1.4 for the clans, tbh. Most clan mechs I played are strong enough as they are.

What does this do to our 2 example mechs?

1.) The Wolverine can run ER Larges to have comparable firepower at long range.
2.) The Wolverine can exploit its higher DPS and heat cap in a short range engagement.
3.) The Stormcrow can dictate the range with its higher speeds.
4.) The Stormcrow has much higher Damage output at medium ranges, where the Wolverine can not use its SRMs.

Bottom line: Both sides have areas of combat where they are better then their counterpart. Thats a certain form of balance.


A few words on Weapon Quirks, since the above part totally ignores those:
From a personal perspective I liked the currently productive quirk system in the regard, that it provided each IS mech with a unique feeling and role on the battlefield. I think that all quirks above 15% were just too much, though. Velocity Quirks aside here. They make no sense anyway (to me). This should stay in with the following changes (if only to keep people happy who spent MC on a mech for a certain role):

1.) Max quirk is 15%. Huginn, Grid Iron, Dragon 1N all get 15% cool down and Heat Generation but a little added gimick in form of higher run speeds to offset the loss in offensive.Will keep em still fun and make em a lot less overwhelming.
2.) Change the ratio between weapon specific and weapon type quirks. 1:1 seems to restrictive in a way that it forces you to use a certain weapon system instead of focussing you on a slot type. My friends and I feel that a ratio of 4:1 makes a lot of sense. In case of the Dragon-1N you would have a 12% ballistic cooldown and another 3% cooldown if you fit AC5s. Enough incentive to use the AC5, but not enough to make it the "ONLY" viable build.
3.) A maximum of 2 quirks per weapon system. You get heat generation and range, but no cooldown. You get cooldown and range but no heat generation. I think you understand what I mean.


These changes are not hard to implement. They use existing tools you have in place and simplify a lot of things because you will have less quirks total, you will make IS construction rules simpler by not distinguishing between 2 types of IS double heat sinks. The fixed heat cap will be a nice distinction point between IS and Clan that will have a direct impact on gameplay without directly breaking lore (too much).

Thanks for reading this far. You probably have questions or comments because I tend to be unclear in my first posts about a topic.

Feel free to ask/comment. I will reply as friendly and explanatory as I am capable of.

Flame away/Discuss!!

Regards -
Clay

Edited by ClaymoreReIIik, 14 September 2015 - 11:11 PM.

[Dracol]

ClaymoreReIIik, on 14 September 2015 - 02:53 PM, said:

--WALL O TEXT---

Within the wall of text above a nugget of an idea caught my eye. One I had not personally seen before.

Make Heat Cap for IS and Clan Different
Had not thought about a change like this. Would make a significant impact on balance between the two sides. Personally, I'd rather see Clan's heat cap drop to continue the trend of reducing time to kill. Increasing the heat cap for IS would reintroduce shorter TTK that the removed quirks were fostering.

[ClaymoreReIIik]

Dracol, on 14 September 2015 - 03:07 PM, said:

Within the wall of text above a nugget of an idea caught my eye. One I had not personally seen before.

Make Heat Cap for IS and Clan Different
Had not thought about a change like this. Would make a significant impact on balance between the two sides. Personally, I'd rather see Clan's heat cap drop to continue the trend of reducing time to kill. Increasing the heat cap for IS would reintroduce shorter TTK that the removed quirks were fostering.

The alphastrike of the mechs and the "maximum possible DPS" remain uninfluenced by the heat cap.

Sustained DPS goes way down with lower heat cap. I have no issue with Clans having higher possible Alpha (and with that maximum DPS) since their weapons are more "Damage over Time" then pinpoint. (ACs, LRMs, longer burn durations.) Giving the IS an edge in sustained DPS gives them tactical options though.

Basically the clans have to do damage at range and use their speed to stay in "beneficial ranges" or the IS will tear them to shreds by running them hot...

The "hardpoint inflation" on clan mechs will be mitigated as well. Who would run 12 small pulse if half of em already put you at 75% heat and the other arm will shut you down immediately....
(and the balance tool for the nova would be to give the "nova prime set" a quirk that gives it 50 extra heat cap....the paul giveth and the paul taketh away....)

Solves a lot of birds with one neat little stone.

Edit: fixed some logic hole....its late here.

Edited by ClaymoreReIIik, 14 September 2015 - 03:21 PM.

[ClaymoreReIIik]

ClaymoreReIIik, on 14 September 2015 - 03:14 PM, said:

The alphastrike of the mechs and the "maximum possible DPS" remain uninfluenced by the heat cap.

Sustained DPS goes way down with lower heat cap. I have no issue with Clans having higher possible Alpha (and with that maximum DPS) since their weapons are more "Damage over Time" then pinpoint. (ACs, LRMs, longer burn durations.) Giving the IS an edge in sustained DPS gives them tactical options though.

Basically the clans have to do damage at range and use their speed to stay in "beneficial ranges" or the IS will tear them to shreds by running them hot...

The "hardpoint inflation" on clan mechs will be mitigated as well. Who would run 12 small pulse if half of em already put you at 75% heat and the other arm will shut you down immediately....
(and the balance tool for the nova would be to give the "nova prime set" a quirk that gives it 50 extra heat cap....the paul giveth and the paul taketh away....)

Solves a lot of birds with one neat little stone.

Edit: fixed some logic hole....its late here.

Too long to read or simply no objections?

[DivineEvil]

I personally don't think that fixing anything up is a good course of action in any given circumstances. I'm not gonna waste time expanding perpetual reasoning and try to be short. I've already argued about all the following on this forum section multiple times.

IS mechs for me has a style of blocky, enduring "soviet russia tanks". They have technology, that have been declined for ages and a structure, that have been re-welded ten times over. They're bigger, spacious and made for easy repairs and "stick it somewhere and weld-up" modifications.

Clan mechs are slim, elusive and sophisticated "US jet fighters". They have superior technology, that was refined to do its job as best and nothing on top, while structure is millimeter-fit, edge-to-edge precise modular construction. They're compact, tight, with nothing excessive.

Being said that, my proposal is that:
1.) Remove all the specific weapon quirks altogether, keeping everything else from live version. Give sensor quirks to the mechs with smaller total quirk percentages, make them negative for ones, that would stand out strong.
2.) 1st PTS run: double all the internal structure.
3.) 2nd PTS run: get Clan internal structure back down.
4.) Make a public vote to see if players see the internal structure buff on IS as viable way for improving balance.
5.) 3rd PTS run: reduce Clan base heat capacity from 30 to 20.
6.) 4th PTS run: reduce Clan base heat capacity from 20 to 10, and IS from 30 to 20.
7.) 5th PTS run: reduce Clan base heat capacity from 10 to 0, and IS from 20 to 30. Buff all external DHS to 2.0 efficiency, buff all SHS to 1.4 efficiency.
8.) Make two votes to determine which iteration of heat adjustments felt the best.
9.) Evaluate the resulting feedback to develop a complex of changes to internal structure or/and heat mechanics for final release.
10.) Neglect some of the negative/positive quirks if balance has shifted in favor of IS.

These must be performed after the current mech re-balance has been concluded.

Edited by DivineEvil, 15 September 2015 - 11:56 PM.

[ClaymoreReIIik]

DivineEvil, on 15 September 2015 - 11:55 PM, said:

I personally don't think that fixing anything up is a good course of action in any given circumstances. I'm not gonna waste time expanding perpetual reasoning and try to be short. I've already argued about all the following on this forum section multiple times.

IS mechs for me has a style of blocky, enduring "soviet russia tanks". They have technology, that have been declined for ages and a structure, that have been re-welded ten times over. They're bigger, spacious and made for easy repairs and "stick it somewhere and weld-up" modifications.

Clan mechs are slim, elusive and sophisticated "US jet fighters". They have superior technology, that was refined to do its job as best and nothing on top, while structure is millimeter-fit, edge-to-edge precise modular construction. They're compact, tight, with nothing excessive.

Being said that, my proposal is that:
1.) Remove all the specific weapon quirks altogether, keeping everything else from live version. Give sensor quirks to the mechs with smaller total quirk percentages, make them negative for ones, that would stand out strong.
2.) 1st PTS run: double all the internal structure.
3.) 2nd PTS run: get Clan internal structure back down.
4.) Make a public vote to see if players see the internal structure buff on IS as viable way for improving balance.
5.) 3rd PTS run: reduce Clan base heat capacity from 30 to 20.
6.) 4th PTS run: reduce Clan base heat capacity from 20 to 10, and IS from 30 to 20.
7.) 5th PTS run: reduce Clan base heat capacity from 10 to 0, and IS from 20 to 30. Buff all external DHS to 2.0 efficiency, buff all SHS to 1.4 efficiency.
8.) Make two votes to determine which iteration of heat adjustments felt the best.
9.) Evaluate the resulting feedback to develop a complex of changes to internal structure or/and heat mechanics for final release.
10.) Neglect some of the negative/positive quirks if balance has shifted in favor of IS.

These must be performed after the current mech re-balance has been concluded.

I kinda learned that if you want to test features you build small containable testing environments.

You would make(!!ONLY!!) one chassis (shadowhawk for example) available and run a one week test without quirks. Everyone who plays 10 matches on PTS gets 150 MC on production.

That is your control group data.

You then make the same one chassis available and try to quirk the variants in a way your tool suggests to even them out. Another one week run, another 150 MC on production if you play 10 games on PTS.

That is your quirk data.

Then you would have a nice diverse set of data to go from. Your players would know whats up, games on PTS would kick and nobody would get freaked out because "their mech got broken". If its only shadowhawks then everyone will know its not the end of the world, because its only a limited test......

Wishfull thinking, I guess.....

Edited by ClaymoreReIIik, 16 September 2015 - 05:50 AM.

[DivineEvil]

ClaymoreReIIik, on 16 September 2015 - 05:50 AM, said:

I kinda learned that if you want to test features you build small containable testing environments.

You would make(!!ONLY!!) one chassis (shadowhawk for example) available and run a one week test without quirks. Everyone who plays 10 matches on PTS gets 150 MC on production.

That is your control group data.

You then make the same one chassis available and try to quirk the variants in a way your tool suggests to even them out. Another one week run, another 150 MC on production if you play 10 games on PTS.

That is your quirk data.

Then you would have a nice diverse set of data to go from. Your players would know whats up, games on PTS would kick and nobody would get freaked out because "their mech got broken". If its only shadowhawks then everyone will know its not the end of the world, because its only a limited test......

Wishfull thinking, I guess.....

Changes that I'm or you're talking about cannot be tested by one mere chassis. They can have varying effects on different weight classes and profiles. Testing on one chassis will achive nothing, as will be evident when result will go live. Wht exactly is the procedure you're suggesting supposed to test?

People ALWAYS expected to realise what testing is. Scale makes no difference.

[Xevius Von Morrigan]

I agree with you, ClymoreRellik.
(The heat cap could depend on chassis class (a bit less for small, a little more for assault) but your analysis is accurate)

[ClaymoreReIIik]

DivineEvil, on 16 September 2015 - 08:55 AM, said:

Changes that I'm or you're talking about cannot be tested by one mere chassis. They can have varying effects on different weight classes and profiles. Testing on one chassis will achive nothing, as will be evident when result will go live. Wht exactly is the procedure you're suggesting supposed to test?

People ALWAYS expected to realise what testing is. Scale makes no difference.

You can test the code and test wether or not Information warfare Quirks influence TTK or mech choices. This last test mainly kicked no matches and created a ton of unhappy customers. People were focussed on checking what happens to their favourite mechs instead of testing the underlying backend systems. Regression testing, is not what this PTS was about...afaik.

They are currently trying to validate Design choices, so creating an environment where these get displayed in small understandable ways makes the analysis of the testing data much easier.

Edit: Let me be a bit more specific:

At this point in time there is something like 200+ different variants of "battlemech" in this game. We have a limited player base and the quirk system needs a revisit. The shadowhawk as a chassis has 7 different variants from ballistic centric through missile centric to energy centric. It can be fast, slow and so on.

If you make all the chassis available in a "step one balance revisit" you have to deal with people having certain tendencies, like having a sympathy for a certain novel character that piloted that mech. Winning a tournament in tabletop with that mech last on the plate etc. pp.

If you give them one chassis to work with that YOU decide you have more control over the variables in the test, because you reduced them.

PGI already realized that their internal testing vastly differs from the production crowd. (I heard that is in no small part through the fact that they mobbed out all the critics from the internal testing team, but thats a different story.) So it makes sense to actually validate a design that passed internal testing by testing it on the PTS. That does not mean PTS sessions need to test 10 different things with 3 person-days of work involved around correlating that data against "test data from live servers".

Unclean statistical work will only lead to bull**** conclusions leading to terrible design choices. Let's not do that please.

Iterative design means you build something small scale that works and then improve on that. Each iteration you have to have a finished working product that passed regression testing.

The absolute only reason to run a "all and everything test" on PTS with all chassis and variants is, if you have a script generating quirks and want to find the outliers where the script parameters created unreasonable results. If you do that, you seriously have to rethink your design approach, because that would mean you think your right (where the fact that you have to do it from scratch, already suggests that you were wrong the last time you thought you were right...). That assumption is (generally speaking) the mother of all failures.

Edited by ClaymoreReIIik, 16 September 2015 - 10:14 PM.

[DivineEvil]

ClaymoreReIIik, on 16 September 2015 - 08:26 PM, said:

You can test the code and test wether or not Information warfare Quirks influence TTK or mech choices. This last test mainly kicked no matches and created a ton of unhappy customers. People were focussed on checking what happens to their favourite mechs instead of testing the underlying backend systems. Regression testing, is not what this PTS was about...afaik.

They are currently trying to validate Design choices, so creating an environment where these get displayed in small understandable ways makes the analysis of the testing data much easier.

I'm not arguing, that first PTS session were a failure - it's an objective fact. Aside from bugs and errors that went live and interference with a running event of significant external importance, it still has not been prepared for the goal it has been meant for. The problem there were is amount and lack of disparity in quirks they're given. It doesn't mattered the least what mechs were favoured if any, because there essentialy were a total lack of individuality left for one chassis or another. An attempt to make variants different without the variation between positive and negative quirks, has erased all chassis distinction currently present on live version. This is what in fact a reason behind the failure, since there's nothing to compare around, and excess of quirks have made practical testing pre-determinally futile endeavour.

If you do argue for the use of localised testing environment for checking a specific system as Sensor quirks, then I'm in an agreement with you. Still, PGI telling us now, that it was about "just testing how sensor quirks were" is, unfortunately, looks like a mere excuse for trying to achieve more, but achieving less as result.

[ClaymoreReIIik]

DivineEvil, on 16 September 2015 - 11:34 PM, said:

I'm not arguing, that first PTS session were a failure - it's an objective fact. Aside from bugs and errors that went live and interference with a running event of significant external importance, it still has not been prepared for the goal it has been meant for. The problem there were is amount and lack of disparity in quirks they're given. It doesn't mattered the least what mechs were favoured if any, because there essentialy were a total lack of individuality left for one chassis or another. An attempt to make variants different without the variation between positive and negative quirks, has erased all chassis distinction currently present on live version. This is what in fact a reason behind the failure, since there's nothing to compare around, and excess of quirks have made practical testing pre-determinally futile endeavour.

If you do argue for the use of localised testing environment for checking a specific system as Sensor quirks, then I'm in an agreement with you. Still, PGI telling us now, that it was about "just testing how sensor quirks were" is, unfortunately, looks like a mere excuse for trying to achieve more, but achieving less as result.

Amen.

We agree on that. I still think heat cap is the only way to balance clan vs IS. Structure and Armor Quirks mean next to nothing unless they make you survive half a round (more) of "4 mechs focus firing on you". At an average alpha of 30 points of damage that would be adding 60 points of structure/armor to the kill locations. That would be an unacceptable shift in 1on1 situations.

[DivineEvil]

ClaymoreReIIik, on 17 September 2015 - 01:36 AM, said:

We agree on that. I still think heat cap is the only way to balance clan vs IS. Structure and Armor Quirks mean next to nothing unless they make you survive half a round (more) of "4 mechs focus firing on you". At an average alpha of 30 points of damage that would be adding 60 points of structure/armor to the kill locations. That would be an unacceptable shift in 1on1 situations.

Well duh, just imagine for fun that right now, on a live server, every IS mech would get double internal structure (equal to mech's maximum armor), and you put quirks on top of that. Do you realize how much impact it is alone going to have? Heat capacity is a good addition, that can be mixed and matched with armor/structure changes to reach a conclusion, that everyone would agree upon. It's just a better method to adjust the overall low TTK and alpha-strike metagame. IS can handle heat changes better than Clans, so it's risky to push the heat changes alone too far to complement all three issues at the same time. Like I said, IS mechs are generally more bulky and bad at avoiding the damage, so double IS structure is a good way to skew the balance in proper direction. Newer players would also have better time getting used to the game with IS toys, before going for slim, elusive and powerful, but more demanding and brittle Clan machines.

[ClaymoreReIIik]

DivineEvil, on 17 September 2015 - 02:11 AM, said:

Well duh, just imagine for fun that right now, on a live server, every IS mech would get double internal structure (equal to mech's maximum armor), and you put quirks on top of that. Do you realize how much impact it is alone going to have? Heat capacity is a good addition, that can be mixed and matched with armor/structure changes to reach a conclusion, that everyone would agree upon. It's just a better method to adjust the overall low TTK and alpha-strike metagame. IS can handle heat changes better than Clans, so it's risky to push the heat changes alone too far to complement all three issues at the same time. Like I said, IS mechs are generally more bulky and bad at avoiding the damage, so double IS structure is a good way to skew the balance in proper direction. Newer players would also have better time getting used to the game with IS toys, before going for slim, elusive and powerful, but more demanding and brittle Clan machines.

Structure Quirks have a huge problem, in the way that they do not protect the equipment. Do you remember the critical hit issues with MGs when they instantly critted every equipment in open locations? The Ember was unstoppable because it did not have to do much damage to make you useless.

Have you ever played MWO with 3025 stock mechs? The german community ran a couple of tournaments with that tech level and stock mechs and MWO is actually extremely fun that way. (That was before Quirks.) With a few outliers there was actually quite good balance there. The main thing was that everyone reached his heat cap about 10 seconds into the fight an after that time the game turns into a real tactical shooter, because you have to seriously start playing smart.

I do not believe the first guy getting the focus fire should survive it or even have a chance to, but I think that the guy after that should see a lot less fire because everyone ran hot.

Edited by ClaymoreReIIik, 17 September 2015 - 02:19 AM.