269 replies to this topic

[Farmer]

That's what happens when you take an abstraction (RoF in this case) and make it an absolute. All the absolutes that were dependent upon that being an abstraction go to hell. In TT you fired once every round, but that covered multiple shots. It was a complete abstraction. Heat was NOT an abstraction. It was a 1:1 thing. X tons of sinks cooled X amount of heat. X amount of DHS cooled 2X amount of heat.

[MustrumRidcully]

Targetloc, on 08 November 2012 - 03:12 PM, said:

I'd argue it's not an equal increase (which is part of why ER weapons are huge under-performers in MWO).

Average damage shooting a running Hunchback at 210 meters in TT with 4 gunnery:

Medium laser: 0.83
Large Laser: 3.33
ER Large Laser: 5.77


50% more heat makes sense on ER weapons in the board game because there are range brackets where they hit 50% more often than the standard version, which can lead to you doing 30-50% more damage over a few rounds.


Average damage shooting a running Hunchback at 210 meters in MWO if you're 80% accurate:

Medium laser: 4
Large Laser: 6.4
ER Large Laser: 6.4


Or let's try to get Large Laser the same as a base of comparison which means you have 41.66% accuracy.

Medium laser: 2 (vs 0.83 in TT)
Large Laser: 3.33
ER Large: 3.33 (vs 5.77 in TT)


The balance implications are pretty significant.

edit- pizza pizza

I like this post. I really do.

[Draco Argentum]

Protection, on 08 November 2012 - 02:02 PM, said:

I understand your arguments, and I agree with them to an extent, but at this point, can't we just throw away all the tabletop numbers and start tweaking weapons and their attributes to make them more viable, rather than worrying about the archaic rules of a 1980's board game.

Make it so that PPCs do extra spash damage as well (10 to the locaion hit, and 5 to each adjacent location), ER range is unlimited, and Large Pulse deal 50% more damage. Slow Gauss rate of fire and boost Autocannons one last time, and suddenly weapons are starting to fall in line.

Your PPC suggestion would likely be more work than rejigging the numbers of every single weapon. They are totally able to balance the weapons as is with zero coding or graphical work. They should do so instead of pinning their hopes on PPC HUD flicker as they are currently doing.

[Indoorsman]

Just for the lulz I decided to try a 4PPC 1Gauss AS7-RS. It kills stuff real easy... in between overheats. Played it now for a few hours :-o

[MCXL]

the real

Chapter Four: Players, Progression, Power, (And Balancing Around Them)

[Part 1]

Once again, this Chapter will be broken into several sections.

Section One: An Introduction to the Chapter

Before we get started I would like to thank just about everyone who has posted here. Your questions and discussion have elevated this thread to a high level, and this thread in the week or so it has been posted has surpassed some stickies for views. That is pretty darn neato! So, let's keep it up. As long as people remain interested in the topic, I'll keep writing a ******* book on design.

In our previous chapters we have talked about dealing with players as an abstract. How to help them enjoy the game, how to get them involved et cetera. Now we need to talk about the effect of the player on a numerical level.

Yup this is the MCXL balance post.

Now this post will inherently be less reliant on factual figures, primarily because I don't have access to all the cool developer metrics that PGI has. I am quite certain that PGI can see how much damage each chassis takes to each part on average per match, what weapons are doing the most damage, heat maps for specific guns and all sorts of other really useful things when trying to see the numbers at work behind the scenes. That said, as usual the Chapter will attempt to focus on things that aren't about how I would like the game to play per my preferences but rather how my internal designer thinks about the systems in place for game balance.

Also this post won't focus on heat at all, so It might be time for an updated thread title. I'll think about that while I write this, but if you have suggestions let me know. I'm thinking:

ANALYZING MWO'S GAME DESIGN, AND WHY DHS NEVER WAS THE PROBLEM OR THE SOLUTION

So now, to talk about balance.

Section Two: The Fifth Element

(Everyone is a Hero!)

Before we start talking about skill, lets get something out of the way. Just about every player on MWO is going to be statistically much more accurate than you pilots in TT. Targetloc covered it well earlier in the thread.

Targetloc, on 08 November 2012 - 03:12 PM, said:

Average damage shooting a running Hunchback at 210 meters in TT with 4 gunnery:

Medium laser: 0.83
Large Laser: 3.33
ER Large Laser: 5.77
...
Average damage shooting a running Hunchback at 210 meters in MWO if you're 80% accurate:

Medium laser: 4
Large Laser: 6.4
ER Large Laser: 6.4

Or let's try to get Large Laser the same as a base of comparison which means you have 41.66% accuracy.

Medium laser: 2 (vs 0.83 in TT)
Large Laser: 3.33
ER Large: 3.33 (vs 5.77 in TT)

The balance implications are pretty significant.

Targetloc points out something that most are aware of but in an interesting manner. I have talked about convergence and accuracy before, but this goes beyond that and points out a critical change in how the games handle dps.

In TT the limiting factor really isn't heat, it's accuracy. There are all sorts of piloting checks, accuracy, movement, knockdown, the list is long, and of course its all RNG. MWO (as most would say correctly) ditches just about every RNG aspect of the TT for a more reliable system, but as we can see from the above numbers, there is a change in how things scale.

Obviously players vary in skill, some are going to be more accurate than others, but accounting for those differences is done though proper matchmaking (later section, we will get there) suffice to say for now;

If we were to convert players back to TT pilots, each player would be a set of loaded dice: some are loaded to a better number than others. Matchmaking makes it so all the dice hitting the table are loaded to the same degree, keeping things balanced.

Back to the math though, again Targetloc:

Targetloc, on 08 November 2012 - 03:35 PM, said:

Medium lasers were always seen as the gold standard of weapons in TT, but without a cone of fire or other accuracy gimmick, they become twice as powerful at the most common engagement range in MWO. And that's before considering their fire rate... as the stats currently stand, 4 medium lasers blow an AC/20 out of the water in terms of efficiency... except for the obscene heat keeping them in check.

Except that now you have a range of potential effectiveness for a medium laser where it might do anywhere between 0.125 and 1.25 DPS depending how many HS you want to assign to it, and the game's high heat capacity means they can front load obscene amounts of damage per ton.

If they gave them a 10 second recycle time, or made them do 2.5 damage for 1.5 heat at 4 seconds then you could say "okay, 3-4 heat sinks and they're maxed, and the burst DPS is relatively sane compared to the sustained DPS. Now let's calculate what people are going to build from there."

I don't 100% agree with his last paragraph, but it's an interesting idea. Targetloc hits the nail on the head though when it comes to the math and the problems it creates. So it's finally time to talk about how to deal with this accuracy problem.

Another thing I want to address is this:

Farmer, on 08 November 2012 - 08:34 PM, said:

That's what happens when you take an abstraction (RoF in this case) and make it an absolute. All the absolutes that were dependent upon that being an abstraction go to hell. In TT you fired once every round, but that covered multiple shots. It was a complete abstraction. Heat was NOT an abstraction. It was a 1:1 thing. X tons of sinks cooled X amount of heat. X amount of DHS cooled 2X amount of heat.

This statement isn't actually true. As weird as it sounds, based on how the original TT rules are implemented, weapons fire once every 10 seconds across the board, (aside from UAC). The reasons we can draw this conclusion are:
1. In one turn, you could only hit or miss, with each weapon, nothing partial
And
2. All damage would be dealt to one area (with exceptions like things that fire spreads, (LRM, LBX, SRM, anything else?)

Based on those two factors, original TT had a 10 second recycle for everything but UAC's (and I guess you could argue the missiles as well, but we don't really know)

Solaris obviously changes this, but that isnt what the basis for this game is, and generally speaking Solaris rules are the "niche" rulebook.

Section Two: Accuracy

(You sunk my Battlemech)

When talking about accuracy there have been lots of threads talking about convergence. For those that don't know the basic idea is that right now, all your guns focus on one pinpoint so if one hits, they all hit (note, this isn't how TT works with a few minor exceptions). A common preposition to help combat this balance problem is to have all torso mounted weapons just shoot straight out.

To visualize it I have drawn an amazing picture.

SICK RIGHT?!?!

As you can see, in the case of the swayback this (in theory makes the amount of firepower that can be applied to one armor slot of a mech substantially lower, especially at close range.

There are drawbacks to this though, the biggest one being that there would be a learning curve of how to aim each mechs torso to hit things. Center torso weapons would be the easiest to aim, (remaining mostly unchanged on things like the atlas and Centurion) but with a mech like the awesome, at close range it would be very difficult to land both torso shots on smaller mechs.

From a design standpoint this could be seen as a positive or a negative, depending on the feel you are looking for. One definite negative though is an increased learning curve for new players, as it stands now the game is tough to get the hang of. This system wouldn't help that at all.

The other problem with this from a design standpoint, is it somewhat discourages getting up in close, as the paralax between your torso and arms would become greater close in. Another pic incoming:

I'm positive my drawing is not 100% accurate, but the point is that at close range it might be to much of a player turn off to have a system like this. Additionally the system is originally balanced to make getting closer more deadly for everybody involved, and I think that this system might make torso mount powerhouses get weaker (without excessive player skill)

Like it or not MWO is a game that everyone needs to be able to play at a decent level, because players are the lifeblood of a F2P game.

The next possible solution is the Cone of fire. Like in other first person shooters, there would be a range on the screen where your shots could land, EG like COD and many other games:


From a first person perspective:


or

A system like this means that you encourage people to get in close, (which the devs have said they want many times over) and additionally allows us to put another factor from the TT (and many other games) on the table easily: Movement penalties.

Right now there is nothing to discourage players from firing wildly as they charge in, (aside from heat, lol)

This is another one of those things where choice would be essential. You could make it so that when moving a mech at 50% speed you maintain 75% of your max accuracy (like crouch moving in a fps) so that running in like good ol' Leeroy Jenkins isn't always the best strategy for a 4sp, or other brawler mechs. Additionally this system could be weighted based on... Well mech weight, or possibly chassis design (the quirks system PGI mentioned earlier when the YLW came out).

One thing I had to think about though was, obviously you would want different size circles for different weapons, both while standing still and while moving. This could be a way to give a real purpose to the selected weapon in the weapon groups thing at the bottom right. Currently selecting any given weapon (row) is only for assigning the weapon to a group. With this system you could scroll to the weapon hit zone you wanted to monitor, and that would be the circle that appeared. To a degree this might be heavy on the burden of knowledge side, but I think that it adds a game play element that could be really neat.

There is also the option of combining these two systems, and making a sort of hybrid hit zone model:

Now all these things are predicated on tech that might just not be available int he game. I know that a "fps" cone system would likely be simple to implement, but there is also the issue of getting the player and the server to see the same thing when your weapons shoot all over the place. Call it a hunch, but I think that is why right now we are stuck with the scalpel precise weapons that all hit the same point. Of note here, other server based games such as Battlefield 3 seem to manage having a bunch of different 'cone of fire' weapons going at once. "Netcode" in MWO just needs to improve more.

Lots of possibilities.

I think when it comes to accuracy the most glaring flaw with the current system is the lack of any sort of movement penalty. That is sort of an integral element of any MODERN REALISTIC shooter, where you have to chose between accuracy and speed.

Section Three: Different Approaches

(why is there no /dance, MWO PLZ.)

Another possibility is raised by Targetloc's posts. Because you can calculate the average damages at any given range with any given weapon, why not scale damage inside of its "stated range" as well?

The games systems right now have you do the same range out to a set distance, and then the damage falls off over an additional percentage of the range depending on weapon type.

If we are to keep the pinpoint targeting, a possibility would be to reduce the damage a any given stated range along the lines of:

Targetloc, on 08 November 2012 - 03:12 PM, said:

Average damage shooting a running Hunchback at 210 m in TT...
Medium laser: 0.83
Large Laser: 3.33
ER Large Laser: 5.77

Thanks again Targetloc, because your post saves me a lot of work here.

This change actually makes a lot of sense from a realistic standpoint. A laser with enough power to shoot another 200m and still hit armor at the same damage, would logically speaking be more powerful at any shorter range than that. Additionally ballistics lose power from the moment they leave the gun until they stop moving, one way or another.

I'm sure that the true TT purists are hissing at em and going "Nooooooooo!" but as outlined earlier, there is a problem right now where range isn't working even close to the same as far as balance, and this could be used as a measure to solve it if weapon spread isn't on the table.

Another potential approach would be to give weapons other types of bonuses based on range, such as improved crit chance the closer your are, and likewise degraded crit from far away for example.

The point is, right now players can hammer one part of the mech to effectively, which means that the devs have had to limit player power via other means.

Section Four: Measuring Killing Power

(Battle Value 3.0)

This is where access to better data would really help, but lets look at TT BV and think things out from there.

In TT Battle Value is basically, a numerical measurement of the power of your force. How hard it hits, how fast it moves, how resilient it is, etc etc.

The tabletop game is measuring minute factors, and every little interaction adds to a unit's BV score. Looking at any of the mechs pages you can see how even small variant changes can be quantified by the system.

Of course when talking about a player driven game things become more muddled because a good player can amplify or degrade equipment. The most OP!!!! build in the hands of a really bad player is still going to be ineffective, and even less effective builds in the hands of technically brilliant players will perform better.

So adapting BV to MWO means understanding the interaction between pilot and mech.

This is where a ranking system needs to come into play. By measuring the killing power of the player, and using that as a multiplier of the power of the mech you can draw an overall combat effectiveness rating. Or:

BV/ELO=Total score.

I think that doing matchmaking in a combined way like this could be useful, in a solo context:

• Higher skill players can handicap themselves, making the challenge more prominent, but (if the system works right) also allowing them to remain combat effective when they play right.
  
• Lower skill players can compensate for a lack of power by buying a higher value mech, without gaining a competitive advantage.
  
• Properly assessed items could be numerically upgrades, maintaining the TT tech progression, without destabilizing matchmaking, (because it can measure these things)

Obviously a big disadvantage here is that matchmaking becomes a heck of a lot more complicated for solo algorithms. Not because the function is hard, but measuring the power level of each thing.




Measuring the overall power of a mech isn't that complicated, it certainly isn't a cakewalk, but understanding the overall statistics of the mechs means that you can extrapolate their Battle Value. For instance, a simple equation could read something like:

(Damage*effective range*Heat efficiency*Move speed*Armor+bonuses)/1000

Bonuses would be things like ECM, Jump Jets, etc.

When you can understand the values of the items in game, you can calculate their effectiveness. Of course in order to do that you need data. There is a pretty slick and obvious solution:

Use a rating system on every item in the game.

That is, measure how much damage each weapon is doing, how many mechs are using it and winning, and whatever other metrics are in place, and each patch revise the BV of items in mechs.

Complicated, but also pretty ******* neato if they could pull it off from an engineering perspective.

Then start measuring player skill, and you have your multiplier factor. The average player is a multiplier of 1.0, better player get a higher multiplier, worse players get a lower one, and the game matches the number from there.

This is of course, only an Idea, and the complexities of maintaining a system like this might not be worth the time investment, but it would fulfill the spirit of the table top battle values system.

Could be neat though.

Section Five: The Matchmaker

(Current Problems)

This section will be pretty short but I am going to outline some small issues I have with matchmaking current implementation (and before you make an assumption, no, I don't want 8 man stomps back)

Right now the matchmaker forces each team to have an identical team comp or rather it always matches lights for lights, heavies for heavies etc.

The problem this leaves us with is simple, it makes team stacking very likely.

The simplest way to show this...


Basically, matching groups so rigidly generally will lead to more groups on one side than the other. The bigger the groups, the harder it is to match them with other groups.

For large groups there shouldn't be any sort of "power" matchmaking, (and from what I gather the stage 2 will not enforce mirror matches)

All that said, matchmaking is probably the most feature incomplete part of the game, and so again, making actual judgement about it at this point are premature.


I do have more I want to write as part of this chapter, specifically about progression, and associated power, butsome things got in the way to slow me down tonight, I'll make another post shortly. (Probably tomorrow)

Until then,
-MCXL

Edited by MCXL, 09 November 2012 - 07:08 AM.

[Indoorsman]

You didn't really address if you take away the BV system if TT would still be balanced. Because if BV was what kept TT balanced then scaling TT for MWO won't work.

[Vapor Trail]

On the cone of fire implementation. There's several way's of doing that. A "single" cone, where all the weapons in a mech vary based on that single cone... but there's at two reticles. So two cones of fire make sense, one for the arms, and one for the torso. Then there's the fact that the left and right arms can be different (due to actuator differences, damage, etc)... which would mean three... and if you extrapolate that for each of the torsos (damage only, though leg damage might affect all cones)... that's five seperate cones for a single mech. Then you could actually implement a cone for every weapon hardpoint.

All of those would mean an incredibly (and increasingly) complex visual representation...

I think I'd like the added complexity. However certain weapons (long range weapons) would need less of this CoF effect to be effective at the ranges they're supposed to be.

Balancing is going to be a stone cold... breeding age female canine.

Edited by Vapor Trail, 09 November 2012 - 04:36 AM.

[Draco Argentum]

"I think when it comes to accuracy the most glaring flaw with the current system is the lack of any sort of movement penalty. That is sort of an integral element of any shooter, where you have to chose between accuracy and speed."

This just isn't true. Modern pseudo realistic shooters (with totally real regenerating health) have those penalties. Old school run and gun classics like Quake and Unreal Tournament don't. If you were good enough to flick rail in Quake 2 you could do it while also running. It creates a very different dynamic because it doesn't reward camping as much as movement based accuracy penalites do.


Your BV suggestion is essentially doing Elo scores for equipment. It will work and the (presumably) massive number of matches would help to settle the scores out in a reaonable time frame. It still fails at the high end of town though. Top players will have to use top ranked equipment to stay top ranked. Whatever is too powerful because you didn't bother balancing will dominate in high level play.

Another issue that I believe will occurr is equipment carrying. Whatever the high level players use will have its rank boosted by being carried to victories in the same way that you can carry people to high rank. Since people can self select their equipment it will happen preferentially to better items. Gauss in the current game would end up with a high rank, and should. What I expect to happen is that its rank would continue to rise beyond its true rank. Then anyone using it would be effectively given a matchmaking handicap that they don't deserve.

BV is a bad system. Simple tonnage with the equipment designed for balance instead of varied power levels wold be more effective.

[Squidhead Jax]

Thontor, on 08 November 2012 - 05:14 PM, said:

the problem with nerfing medium lasers to account for convergence, is you are making just having 1 or 2 of them useless.. they would need to be boated to be effective at all

Aye, which is why any solution to boating needs to be boating-specific.

[MCXL]

Draco Argentum, on 09 November 2012 - 06:07 AM, said:

"I think when it comes to accuracy the most glaring flaw with the current system is the lack of any sort of movement penalty. That is sort of an integral element of any shooter, where you have to chose between accuracy and speed."

I forgot the words modern and realistic. (When I talk about realistic I'm actually thinking about ARMA)

Indoorsman, on 09 November 2012 - 04:07 AM, said:

You didn't really address if you take away the BV system if TT would still be balanced. Because if BV was what kept TT balanced then scaling TT for MWO won't work.

Its also fundamentally impossible for me to answer this. Because of the fact that the real limiting factor in TT when it comes to team comp is what the group allows, its difficult to AS*ES weather the game would function properly without BV. Obviously there wouldn't be a direct way to see the power of units without it, but generally speaking, perfectly matched forces are rare anyway.

EDIT: I CANT SAY A S S E S

WTF?!

Edited by MCXL, 09 November 2012 - 07:13 AM.

[MustrumRidcully]

MCXL, on 09 November 2012 - 07:08 AM, said:

I forgot the words modern and realistic. (When I talk about realistic I'm actually thinking about ARMA)




Its also fundamentally impossible for me to answer this. Because of the fact that the real limiting factor in TT when it comes to team comp is what the group allows, its difficult to AS*ES weather the game would function properly without BV. Obviously there wouldn't be a direct way to see the power of units without it, but generally speaking, perfectly matched forces are rare anyway.

EDIT: I CANT SAY A S S E S

WTF?!

Doesn't the word you want to say not also include 2 s at the end? I haven't assessed yet if you'd be allowed to write that, though.

[Squidhead Jax]

MCXL, on 09 November 2012 - 07:08 AM, said:

EDIT: I CANT SAY A S S E S

WTF?!

The word you're looking for has 2 S at the end.

Assess.

Bet it's blocked too...
edit: apparently not!

Edited by Squidhead Jax, 09 November 2012 - 07:36 AM.

[MustrumRidcully]

Slowpoke Jax, on 09 November 2012 - 07:36 AM, said:

The word you're looking for has 2 S at the end.

Assess.

Bet it's blocked too...
edit: apparently not!

I have assessed that it's okay to use the correct writing. Fasscinating, isn't it?

[Squidhead Jax]

MustrumRidcully, on 09 November 2012 - 07:37 AM, said:

I have assessed that it's okay to use the correct writing. Fasscinating, isn't it?

Ey, I was only a minute behind you!

[Asatruer]

MCXL, on 09 November 2012 - 03:15 AM, said:

Chapter Four: Players, Progression, Power, (And Balancing Around Them)

Excellent job. Thank you for covering the issues of range and convergence.
I had hoped for at least a little mention of the disparities of the different ballistic flight qualities of different weapons and aimed shooting versus guided weapons, but maybe next time?

Edited by Asatruer, 09 November 2012 - 09:30 AM.

[Indoorsman]

MCXL, on 09 November 2012 - 03:15 AM, said:

Section Two: Accuracy

Did the other mech warrior games have convergence or were there random hit factors?

[Vapor Trail]

MustrumRidcully, on 09 November 2012 - 07:37 AM, said:

I have assessed that it's okay to use the correct writing. Fasscinating, isn't it?

The assassin assessed the assessment of assassinating the assimilating jackasses.

Gotta love the censorbot. So many small words that it eats. *** (n-i-p) for one... and as far as I know no one's used that as a common insult in thirty years.

Edited by Vapor Trail, 09 November 2012 - 09:51 AM.

[Vapor Trail]

Thontor, on 09 November 2012 - 09:56 AM, said:

I assume its because *** is short for ******.

Thanks for making that more clear


"N-i-p" as a slur probably actually originated in the Second World War. It's a shortened form of "Nipponese" meaning a "person of Japanese descent."

Like I said, probably hasn't been used as a common slur for thirty years.
Doesn't make it less of one if the word is used that way...

But censoring it really screws up the saying "*** this in the bud."

Edited by Vapor Trail, 09 November 2012 - 10:07 AM.

[MustrumRidcully]

Thontor, on 09 November 2012 - 09:56 AM, said:

I assume its because *** is short for ******.

******? That's the password I always enter for the forums! How did you know?

Indoorsman, on 09 November 2012 - 09:45 AM, said:

Did the other mech warrior games have convergence or were there random hit factors?

I don't know about all the games, but I am pretty sure Mechwarrior 3 and 4 have used convergence as well. It may be on some level the simplest way to implement them. I never played MW3 in multiplayer, but I would expect that coupled with the general "gunbag" + model approach of the game (e.g. all mechs are basically just defined by tonnage and their model, but you can equip whatever you want) it made boating energy weapons were attractice. Convergence, Hit-Scan, Identical reload times - headcapping or legging mechs would be easy with such a model.

Peresonally, I can probably live with convergence, but it requires a different balance models for mechs. In the table top, just having a high damage value alone meant the weapon had an inherent advantage - even a random hit could deal serious damage to an important hit location. But with convergence, you actually have to balance the damage per hit advantage on the question "how many weapons can I boat together": For example, you may not really be able to boat more than 2 AC20s - so it's "single hit damage" is 40. 8 Medium Lasers can yield the same damage - and they weigh much less than 2 AC20s, and some mechs can carry more. (Though we're not quite there yet. don't know if we have to wait for the Nova to get a mech that can field 12 lasers...)

Edited by MustrumRidcully, 09 November 2012 - 10:13 AM.

[Vapor Trail]

Oh... Gotcha. A pap, teat, part of a breast, boob, knocker, funbag, etc. The business end of the mammary glandular system.


Now here's an interesting one... ***** (g-l-a-n-d) as in 'mammary *****' is censored... WTH?

Edited by Vapor Trail, 09 November 2012 - 10:35 AM.