19 replies to this topic

[Gorith]

BV (Battle Value) has been talked about over and over. I do believe a proper BV system would make balancing much easier And that it should be a combination of flat and % based mathematics.

The BV system from TT is not designed for a videogame so would not work. That being said here is what I propose (As I am not in beta I don't know exacts so this is just a theoretical example of what I feel would make a good BV system) Also I apologize if this is hard to read. I am not the best at formatting these things

First I will cover the percentage math as this would act as a modifier to all the flat math to follow

Heat:Taking a 20 second sample all weapons firing on cooldown Whatever percent the of heat would be remaining till the mech shuts down apply half this. Example If a mech would shut down within 20 seconds of full on fire apply a 0% modifier to its end BV where a mech that is "heat neutral" would apply a 50% modifier

Clan tech equipped: +x% As we know NOTHING about clans or their intended implementation I will not hazard just throwing an arbitrary number here. The idea behind this as a percentage modifier is to represent the clans tech advantage and therefore possibly allowing IS players to have clan tech while keeping their canon tech advantage while at the same time not making non clan geared mechs useless.

Engine rating: Find a standard engine size per weight (35,40,45 all being different weights) and every size larger or smaller modifies the end value by +/- 5%, Add an additional 10% for XL engines

Flat Values:

Armor: 10 armor per 1 BV. This is used instead of tonnage to account for FF armor being lighter.

DPS: Yes DPS rather than weapon damage. Over a 20 second period of fire 2x DPS as BV. Average damage for weapons with variable DPS (LRMS, etc)

Extra equipment: Hard to put into a BV without any knowledge of how truly effective they are in game.

Originaly I was going to do an example but realized explaining how to get heat index and DPS numbers would just bore people and really don't add anything to the validity of my suggestion.

Again remember these are all more or less guess work numbers and are just meant to explain how I think a combination of both flat and percentage math to determine BV would help to keep everything nice and balanced.

Edited by Gorith, 15 July 2012 - 03:17 AM.

[Fire for Effect]

there is no apparent reason why it should not work.

pilot skills are not part of BV, if they deviate from the board game they better know what they are doing if they think they know and the deviation is low they can still keep BV or its mechforce equivalent.
Since all mechs (2000+) are already balanced weighted and tagged with BV it would be quite dumb to throw that experience and numbers away and try to invent something new... worst that could happen is something based on weight since that will always fail to make balanced sides especially if the clans are coming.

[Graphite]

Getting a GOOD BV equation is EXTREMELY difficult. The BV2 currently used for TT is the best available, but is still far from perfect, and a replacement is in progress. (BTW there is nothing that makes a TT BV system unsuitable for MWO)

Your proposed system has problems:
- surely the heat percentage applied should be based on the percentage of it's weapons it can keep firing?
- there isn't a standard engine size by weight.
- XL engines would be a huge hazard in say a mech with it's arm/side torso packed with ammo, not a bonus
- wouldn't speed be better than engine weight?
- dps is too simplistic. For example, would you score a PPC the same as an AC10?

I don't mean to just come and spoil the party, but like I said, it's extremely difficult. I like thinking about this too.
If you come up with an idea, you need to test it: apply it to all the known designs in MWO and see if it can at least rate them in the correct order.

[CCC Dober]

OP: The BV per Mech in TT is quite funny. I have noticed, while using HMP, that each Mech's weight is multiplied with ~10 to give the base value. I don't quite understand the reason behind it because this value just represents an empty chassis without weapons. Maybe it relates to close combat effectiveness. However, this would be largely meaningless in MWO because close combat is not implemented yet. Theoretically speaking you could just subtract this BV base value from all TT variants and go from there.

P.S. On another note: I don't think BV as a whole is off, but the way individual weapon BV is handled in TT does not account for time as a dimension in a meaningful way. In a realtime environment it's all about timing and weapons that seem puny in TT can be quite a handful if stacked and fired in an endless chain of death and destruction (see LRM5/10, UAC2/5, machine gun-spam etc.). So I figure we could keep everything as is minus the base BV and weapons BV and tweak weapon BV according to the worst case scenarios in realtime, giving more weight to weapons with potential to suppress targets constantly. Just a thought.

Edited by CCC Dober, 15 July 2012 - 05:18 AM.

[Graphite]

CCC Dober, on 15 July 2012 - 05:08 AM, said:

I have noticed, while using HMP, that each Mech's weight is multiplied with ~10 to give the base value. I don't quite understand the reason behind it because this value just represents an empty chassis without weapons. Maybe it relates to close combat effectiveness. However, this would be largely meaningless in MWO because close combat is not implemented yet. Theoretically speaking you could just subtract this BV base value from all TT variants and go from there.

Yes, I've always assumed the reason for this is charging and melee.

[Nik Van Rhijn]

Charging is in so there would be no reason to take out the base value. From memory XL engines actually reduce BV (although they increase cost) because they reduce mech surviveability. I don't know if BV3 is close to release and if they would be able to use it. Even with its problems BV2 is fairly useable and is a good starting point. I'm sure that there are plenty of people here who could point out (in great detail )
where changes could be made.
It would have to be tweaked to account for the changes that they have made. When they briefly discussed matchmaking they said that they were looking at a composite system. If they wish to take account of pilot skill as well then they will need some form of ELO
Unfortunately the "Minimum Viable Product" may well apply to the matchmaking system as well

[CCC Dober]

Okay, since the 'experts' aren't forthcoming, we might just pinpoint the problems ourselves. Really doesn't take a rocket scientist.
For this reason I have prepared the following Mech variant to highlight some 'interesting' BV calculations.

Spoiler

BATTLE VALUE CALCULATION FOR Cerberus IIC
VALIDATED Design

DEFENSIVE BATTLE RATING CALCULATION:
Total Armor Factor x 2 288 x 2 = 576
Armored Cowl (UK) + 10
Total Internal Structure Points x 1,125 145 x 1,125 = 163,125
Total 'Mech Tonnage 95
Defensive Weapons: Heat:
1 Laser AMS 7 105
1 Laser AMS 7 105
1 Laser AMS 7 105
Total BV of all Defensive Equipment 315
Deduct 25 for Enhanced Imaging vulnerability -25
--------
1134,125

(Max. potential Heat Points - Heat Sink Capacity) x 5 (83 - 40) x 5 = 215
--------------------
215

1134,125
-215
----------
At a full Run of 6, Movement Modifier is + 2 919,125
Multiply by Defensive Movement Factor of 1,2 x 1,2
----------
DEFENSIVE BATTLE RATING = 1102,95
----------------------------------------------------------------------------

OFFENSIVE BATTLE RATING CALCULATION:
Offensive Weapons: Heat:
1 ER PPC 15 412
1 ER PPC 15 412
2 ER Medium Laser 2 x 5 2 x 108 = 216
2 ER Small Laser 2 x 2 2 x 31 = 62
2 ER Medium Laser 2 x 5 2 x 108 = 216
2 ER Small Laser 2 x 2 2 x 31 = 62
Enhanced Imaging BV = .05 x Weapon BV 1380 x .05 = 69
+ 4 Jumping Heat --------
Base Weapon Battle Rating 83 Max Heat Points 1449

Adjust Base Weapon Battle Rating for 'Mech's Heat Efficiency
X = (Total Heat Sinks x Base Weapon Battle Rating) / Max Heat Points
(40 x 1449) / 83 = 698,3133
Y = (Base Weapon Battle Rating - X) / 2 (1449 - 698,3133) / 2 = 375,3434
--------------
Modified Weapon Battle Rating = X + Y = 1073,6566

Calculate Speed Factor
Add Running MP + Jumping MP 6 + 4 = 10

Speed Factor (based on table or calculation) = 1,63
Offensive Battle Rating = Modified Battle Rating x Speed Factor
1073,6566 x 1,63 = 1750,0603

OFFENSIVE BATTLE RATING = 1750,0603
----------------------------------------------------------------------------

Total BV = Defensive BV + Offensive BV: 1102,95 + 1750,0603 = 2853,0103
Add 5% for -1 PS due to Enhanced Imaging + .05 x 2853,0103 = 142,6505
----------
2995,6608

TOTAL BATTLE VALUE FOR 'MECH = 2.996

DEFENSIVE BATTLE RATING CALCULATION:


Total Armor Factor x 2 -> this factor needs clarification as it favors heavier Mechs
Total Internal Structure Points x 1,125 -> this factor needs clarification as it favors heavier Mechs
(Max. potential Heat Points - Heat Sink Capacity) x 5 -> 'redlining' is not a valid strategy under normal conditions, the penalty is death, not a lowered BV

OFFENSIVE BATTLE RATING CALCULATION:

X = (Total Heat Sinks x Base Weapon Battle Rating) / Max Heat Points -> 'redlining' again

What do you make of these calculations and the highlighted 'questionmarks'?

[Graphite]

CCC Dober, on 16 July 2012 - 12:52 PM, said:

Okay, since the 'experts' aren't forthcoming, we might just pinpoint the problems ourselves. Really doesn't take a rocket scientist.
For this reason I have prepared the following Mech variant to highlight some 'interesting' BV calculations.

What do you make of these calculations and the highlighted 'questionmarks'?

Is this your own system Dober? It doesn't look like the the one in the tech manual...
Armour *2 istead of *2.5
Structure *1.125 instead of *1.5
Heat doesn't affect defensive
X as you've shown it isn't present...

Perhaps this is the old BV1, not the current (and soon to be replaced) BV2?

[Nik Van Rhijn]

Dober, with regard to Armour & IS. Yes it will favour heavier mechs because more armour makes them more surviveable, hence a higher BV. From memory this has been inherent in any form of calculation, tech levels aside, bigger tends to have a higher BV simply because they have more armour(assuming that you max it). Where lighter mechs get their increase from is speed. One thing that is not included in TT BV because it isn't a factor, but could be in a PC game is size/shape. Smaller mechs will be intrinsically harder to hit, especially at range. The non humanoid "flatter" mechs will also be less easy to target. As to how you factor something like that I'm not sure as it's very game specific and would probably need some hard numbers based on in-game scaling.
With regard to heat, it seems, from the videos that we have seen that it spikes more before any cooling cuts in and cooling seems slower. The overall effect being much easier to overheat a mech than the TT figures would imply. All guesswork of course as we don't know how many or what type of heatsinks are fitted. I would imagine that most mechs would tend to have a slight overheat in TT terms unless heavily modified.

[Reoh]

Fire for Effect, on 15 July 2012 - 04:26 AM, said:

there is no apparent reason why it should not work.

pilot skills are not part of BV, if they deviate from the board game they better know what they are doing if they think they know and the deviation is low they can still keep BV or its mechforce equivalent.
Since all mechs (2000+) are already balanced weighted and tagged with BV it would be quite dumb to throw that experience and numbers away and try to invent something new... worst that could happen is something based on weight since that will always fail to make balanced sides especially if the clans are coming.

They're using the canonical values as a starting point but have made some alterations. At the very least there's rate of fire adjustments which may affect the bv of some components within mechs. I think BV should be able to work but it will need to be updated to the MWO changes.


@Nik Van Rhijin

Quote

With regard to heat, it seems, from the videos that we have seen that it spikes more before any cooling cuts in and cooling seems slower. The overall effect being much easier to overheat a mech than the TT figures would imply. All guesswork of course as we don't know how many or what type of heatsinks are fitted. I would imagine that most mechs would tend to have a slight overheat in TT terms unless heavily modified.

Yeah that really changes combat a lot, instead of just adding the surplus you have to deal with the entire alpha-strike's heat before it dissipates. I really like that change, even though it'll hurt the kind of 'mechs I usually pilot (high dmg alpha strike laserboats, that are heat neutral). Should help even the field between types of weapon choice, at least I'm hoping.

[CCC Dober]

Sorry for the wall of text.

Graphite, on 16 July 2012 - 08:03 PM, said:

Is this your own system Dober? It doesn't look like the the one in the tech manual...
Armour *2 istead of *2.5
Structure *1.125 instead of *1.5
Heat doesn't affect defensive
X as you've shown it isn't present...

Perhaps this is the old BV1, not the current (and soon to be replaced) BV2?

Rest assured, it's a straight copy and paste job from the heavy metal pro sheet and yes, it details the BV1 calculation. The point however stands: there are a lot of factors involved, as I have bolded/highlighted, which appear to favor heavier Mechs, granting them a higher BV by default. For no apparent reason I might add. From what I saw of BV2 on sarna (barring the actual, detailed calculations), it is clear that the gap widened a lot more in this respect. Heavier Mechs have even higher BV2s than before. To keep it short, I don't think that a lot of these factors make sense in a realtime, skill based game. Also, I recall that Light Mechs were favored with lower BVs in the old BV system and simply gave more bang for the buck than heavier Mechs. Now with BV2 they are even farther ahead ... sense this does not make.

Nik Van Rhijn, on 16 July 2012 - 11:15 PM, said:

Dober, with regard to Armour & IS. Yes it will favour heavier mechs because more armour makes them more surviveable, hence a higher BV. From memory this has been inherent in any form of calculation, tech levels aside, bigger tends to have a higher BV simply because they have more armour(assuming that you max it). Where lighter mechs get their increase from is speed. One thing that is not included in TT BV because it isn't a factor, but could be in a PC game is size/shape. Smaller mechs will be intrinsically harder to hit, especially at range. The non humanoid "flatter" mechs will also be less easy to target. As to how you factor something like that I'm not sure as it's very game specific and would probably need some hard numbers based on in-game scaling.
With regard to heat, it seems, from the videos that we have seen that it spikes more before any cooling cuts in and cooling seems slower. The overall effect being much easier to overheat a mech than the TT figures would imply. All guesswork of course as we don't know how many or what type of heatsinks are fitted. I would imagine that most mechs would tend to have a slight overheat in TT terms unless heavily modified.

Agreed on the geometric shape of the Mech. Back in MW4 there were several extreme cases that showed how much of an influence hitboxes have on Mech performance. I'll give a couple of examples as to trigger some memories in this regard.

1. Templar -> HUGE, weakly armored shoulders, possibly bigger than the complete average Light Mech torso (seriously bad news)
2. Zeus -> see Templar
3. Fafnir -> see Templar, except Mektek artificially boosted maximum armor levels on all 95+ ton Mechs, which helped a lot
4. Madcat -> the whole torso was practically CT, until Mektek reworked the hitboxes, which gave the Mech a fighting chance
5. Centurion -> very thin, flat and reasonably fast/agile, which was a huge problem, even without lag

The way I see it, most Mechs scale more or less favorably with increasing weight. Some of them however don't (see examples above). Barring geometry modifications and/or rescaling, there are other, proven methods to harden them, but the very same methods are heavy on maintenance whenever tweaks to the Mechs or competitors are made (proxy buffs/nerfs). If that is not taken into consideration, then we will end up with Mechs that will always be considered 'shelf material' in favor of other Mechs doing the same job with a more favorable geometry. This does threaten the viability of current and future content. That is something I have seen while working with Mektek and there was considerable effort and resistance to tweak parameters in order to 'un-shelf' said Mechs.

P.S. Sorry for taking so long to answer, I've been catching up with this show called 'Breaking Bad'. Watched them all in the last couple days =)

Edited by CCC Dober, 20 July 2012 - 06:46 AM.

[Graphite]

CCC Dober, on 20 July 2012 - 06:26 AM, said:

Sorry for the wall of text.



Rest assured, it's a straight copy and paste job from the heavy metal pro sheet and yes, it details the BV1 calculation. The point however stands: there are a lot of factors involved, as I have bolded/highlighted, which appear to favor heavier Mechs, granting them a higher BV by default. For no apparent reason I might add.

Ah, BV1. BV2 definitely isn't perfect, but it's a lot better than BV1.
Larger mechs do more melee damage, more crit free HS for a given speed, and have more internal structure points.

Quote

P.S. Sorry for taking so long to answer, I've been catching up with this show called 'Breaking Bad'. Watched them all in the last couple days =)

Great show that!

Edited by Graphite, 20 July 2012 - 06:55 AM.

[CCC Dober]

Graphite, on 20 July 2012 - 06:53 AM, said:

Ah, BV1. BV2 definitely isn't perfect, but it's a lot better than BV1.
Larger mechs do more melee damage, more crit free HS for a given speed, and have more internal structure points.



Great show that!

Alright, let's expand on that. Melee damage is a compelling argument, so this definitely makes sense. Althought the way TT handles this, kind of implies that even the smallest Mech can uppercut an Atlas, which is ... not very realistic, all things considered. Heavier Mechs and those with a longer 'reach' certainly have a decisive advantage, since they don't need to facehug their opponents into submission and have more weight behind their blows. Triple Strength Myomer is also very beneficial in this regard.

More crit free HS appears to be linked to reactor size, meaning the higher the reactor output/volume/space inside. Now you could argue that said HS are crit free, but the reactor weight increases progressively, which means the free crits are bought with rapidly increasing reactor weight, so they are definitely not free in a general sense. Meanwhile a much heavier reactor on a smaller Mech can easily double the top speed, whereas with a heavier Mech it barely makes a difference (in combat at least).

The internal structure points kind of make sense given the increasing height/bulk of Mechs with increasing weight. While making perfect sense in TT, height/bulk is a huge disadvantage in terms of surviveability when you consider the pc as a platform. I would even go so far to penalize heavier/bulkier/larger Mechs BV wise, just because they make for such easy targets.

Edited by CCC Dober, 20 July 2012 - 08:31 AM.

[Nik Van Rhijn]

To be honest I would prefer an imperfect form of BV in the matchmaker at launch, eith continual refinement, to none
With regard to the size effect on the larger and usually slower mechs, it usually seems to be considered to be balanced by their heavier armour/greater weapon capacity. Maybe the ideal mechs will be fairly fast sniper mediums
What I would hope is that we can see how the matchmaker works rather than having to work it out by inference, once the game launches.
We also have no idea as to wether they will have some form of pilot ELO. Even more difficult to "balance" will be the organised teams. You can have pilots with "average" individual stats yet are excellent teamplayers. Personally I'm not sure if you should even try to take those factors into account.

[CCC Dober]

BV certainly could be useful in the beginning, even an imperfect version. There can be no doubt that it would deviate from TT at some point in that certain parameters/factors are added/removed/tweaked. This is due to the fact that certain mechanics are simplified for TT use, whereas they are played out differently in realtime. Maybe it is time to think about an additional set of characteristics that define a Mech's worth in combat.

A Mech's hitbox or bulk certainly is an important factor, as the MW4 experience has shown me. Less is more in this regard.
Another important factor is weapons loadout and placement, which is beneficial when the (heaviest) weapons aren't clustered in one or two locations, but spread across the whole Mech, so as not to lose them in one fell blow. The other benefit is when those same weapons are placed either in line or above the Mechs cockpit. This allows to engage targets almost instantly when breaking cover (think of hills to hide behind), instead of having to wait until the weapons have LOS on target, which in return allows the target to shoot back and hide first. Having a better weapons placement than your opponent allows you to minimize your exposure and shoot first in such a standoff situation. Coupled with superior hitboxes and agility/mobility, this can be downright game breaking, just to give a perspective here.

Then you have other soft stats, detailing torso twist range/speed, acceleration, deceleration, steering response, knock(down) resistance and a couple more heat related parameters that can make the difference in a fight. I know this to be true since I saw the impact on Mech performance while working with Mektek. In other words: BV as we know it from TT does not account for the majority of these factors, which are driving factors when you want to assess Mech performance in a realtime environment.

[Nik Van Rhijn]

@Dober - I agree with everything that you've said on the additional parameters. The problem is if they haven't already started to code for this that it will take a lot of time and effotrt as they will effectively have to start from scratch.
The above factors were all known to any of the better league teams and spreadsheeted for every mech in the game. It's why only a small number of mechs were considered "effective", depending on the role and hardpoints. I have absolutely no doubt that this info is availablr already to all those teams that have members in beta. Whether they have seen fit to share their knowledge with the devs is another matter, I would like to think so.
Making an individual mech more, or less, squishy or tanky to balance it out is very important in ensuring some chassis are not ignored as liabilities due to the artistic design. The cats "ears" are another thing that springs to mind. The devs seem an experienced team and much of this has been mentioned before so hopefully they have been working on it. It depends on how much of a priority it has been given. Personnally I don't think it will be anywhere near finished at launch.

[CCC Dober]

If this is true, then there will be hell to pay when the Match Maker is trying to balance Mechs on both sides.
Here's a quick example: if the Match Maker would try to balance the Catapult (65tons) with a JagerMech (65tons), you can already see where this is going to end. The Cat loses both ears quickly to concentrated longrange Gauss fire and the Jag wins by default. Now you could argue that the Cat is supposed to have less face-time, but it still runs on ammunition and depends on other Mechs in its primary configuration. It's rather helpless on its own and definitely needs protection.

The situation above is just one of many examples where simple tonnage balance fails hard because it doesn't account for the Cat's vulnerability, namely the ears, where most of its firepower is concentrated. This weakness is in fact so severe that even a Light Mech can eat the Cat all on its own with nothing more than a PPC, some average manoeuvering skills and a bit patience. On paper the Cat has the firepower of a Heavy Mech and the armor to go with, but its hitbox and weapons layout renders both advantages into liabilities given the tonnage they are paid with. I can only see it work in its intended role as the K2 variant where the missile boxes are replaced with sleeker guns, which at least don't eat every random shot fired in their general direction. Both variants are worlds apart, although they share the same chassis. A simple BV value would never tell you that.

I do hope that the dev team had sufficient MW4 time to identify said parameters and can work them into a formula. If not, then we're in deep sh!t if the Match Maker goes out like it does in other games, assuming all Mechs within their class and weight category are equal, when in fact they are not.

If this is not given enough priority, the Match Maker will bomb at least as much as it did in WoT (worst case). Alternatively we'll see a lot of Mechs sitting on the shelf doing nothing but looking pretty while being utterly useless at the same time (in a competitive environment most of all). We have seen the latter happen in MW4 and the same factors above will also be the make or break of Mechs in MWO because the rules of the environment are nearly identical. It is true that not all Mechs can't be made equal and that would be boring, but it is possible to account for their differences with an intelligently designed Match Maker. This goes a long way to assure that the players make the difference in a fight, not their Mechs.

Edited by CCC Dober, 21 July 2012 - 11:23 AM.

[Graphite]

CCC Dober, on 20 July 2012 - 11:50 PM, said:

Then you have other soft stats, detailing torso twist range/speed, acceleration, deceleration, steering response, knock(down) resistance and a couple more heat related parameters that can make the difference in a fight. I know this to be true since I saw the impact on Mech performance while working with Mektek. In other words: BV as we know it from TT does not account for the majority of these factors, which are driving factors when you want to assess Mech performance in a realtime environment.

Agree. They have their work cut out for them there.

[Nik Van Rhijn]

A lot of these factors were discussed in detail before Christmas, with a lot of comment from those MW4 league vets. The devs were very active on the forums at that time, so hopefully this work has been at least partially implemented. Certainly BV won't do anything about the Cats vulnerabilities. Its why nobody expected BV (which has trouble working on the TT) would be the total answer for a PC game.
The trouble is for some of us the Matchmaker is perhaps more important than more mechs in game, or a number of other features. Things like maps, mechs etc are just putting time in. The MM needs serious thought and balancing.

[CCC Dober]

Well, if it helps to keep the perspective, individual Mech performance has a direct impact on MM performance. The Cat seems to fulfil several roles with different variants (most of all Prime vs K2), which could be taken into consideration by MM when teams are balanced. The Prime variant obviously needs help, being more of a fire support Mech than anything else, whereas the K2 can easily walk it alone. Other Mechs have similar (and other) 'problems', so this might help to find a working solution. The MM could try to place those Mechs in need of help with suitable/stronger partners in the same team.