A simple repair to BV2: Double Heat Sinks and 1/2 BV threshold

[Lagrange]

We dont care what each unit is armed with, the point is to illustrate that an unbalanced defense versus firepower (warrior) BENEFITS in the geometric mean compared to the balanced unit (dragonfly).  The units with unbalanced stats should not be receiving even more benefits by switching to a geometric mean.

If I understand right, the argument is that geometric mean valuation is bad compared arithmetic mean valuation because the ratio of imbalanced design swarm units to balanced design swarm units grows and swarms are bad because they are more powerful than represented by BV.

I think there are two factors being combined here: swarms and balance. 

The fact that imbalanced designs (whether swarmy or not) have less value under geometric mean valuations is good in my mind, because it accurately reflects the incapability of an imbalanced design.  For example, based on the BV, you might think the Glass Cannon (below) with a BV of 2653 (and an arithmetic mean of 1327) is likely to win against an inner sphere assault mech.  My expectation is that it performs very poorly in practice against an Atlas (for example).   The geometric mean of 628 better reflects this.

Code: [Select]

Glass Cannon

Mass: 100 tons
Chassis: Composite Biped
Power Plant: 300 XL
Cruising Speed: 32.4 kph
Maximum Speed: 54 kph
Jump Jets: None
     Jump Capacity: 0 meters
Armor: Standard
Armament:
     4 Gauss Rifle
     3 Large Pulse Laser
Manufacturer: Unknown
     Primary Factory: Unknown
Communication System: Unknown
Targeting & Tracking System: Unknown
Introduction Year: 3145
Tech Rating/Availability: F/X-X-X-E
Cost: 23,596,000 C-bills

Type: Glass Cannon
Technology Base: Mixed (Experimental)
Tonnage: 100
Battle Value: 2,653

Equipment                                          Mass
Internal Structure            Composite               5
Engine                        300 XL                9.5
Walking MP: 3
Running MP: 5
Jumping MP: 0
Double Heat Sink              15 [30]                 5
Gyro                                                  3
Cockpit                                               3
Armor Factor                  8                     0.5

                          Internal   Armor 
                          Structure  Value 
     Head                    3         1   
     Center Torso            31        1   
     Center Torso (rear)               0   
     R/L Torso               21        1   
     R/L Torso (rear)                  0   
     R/L Arm                 17        1   
     R/L Leg                 21        1   


Right Arm Actuators: Shoulder, Upper Arm, Lower Arm
Left Arm Actuators: Shoulder, Upper Arm, Lower Arm, Hand

Weapons
and Ammo                Location  Critical   Heat    Tonnage
Double Heat Sink           LL        2        -       1.0 
Double Heat Sink           CT        2        -       1.0 
Large Pulse Laser          RT        2        10      6.0 
Gauss Rifle Ammo (16)      RT        2        -       2.0 
Gauss Rifle                RT        6        1       12.0 
Gauss Rifle Ammo (16)      LA        2        -       2.0 
Gauss Rifle                LA        6        1       12.0 
2 Large Pulse Laser        LT        4        10      12.0 
Gauss Rifle                LT        6        1       12.0 
Double Heat Sink           RL        2        -       1.0 
Gauss Rifle Ammo (8)       HD        1        -       1.0 
Gauss Rifle Ammo (24)      RA        3        -       3.0 
Gauss Rifle                RA        6        1       12.0 

Now consider the swarm issue.  The guiding hypothesis for BV is that it is additive.  If that is inherently untrue, then that should be accounted for in the combination operator.  This was the case with the original BV2, but it was errataed away.  Fixing the combination operator seems like the right thing to do since it directly addresses the swarm issue and because it seems like mangling the per-unit valuation to avoid swarms is problematic.  For an example of the latter, a heavy ASF can be a swarm unit when fighting warships, but if you increase its per-unit valuation to prevent swarms there you end up in an awkward situation in aerial combat where the heavy ASF might be swarmed by conventional fighters.  Should conventional fighters have a much higher BV so they can't swarm ASF which have a higher BV, so they don't swarm Warships?  It seems rather awkward.

[Cannonshop]

If I understand right, the argument is that geometric mean valuation is bad compared arithmetic mean valuation because the ratio of imbalanced design swarm units to balanced design swarm units grows and swarms are bad because they are more powerful than represented by BV.

I think there are two factors being combined here: swarms and balance. 

The fact that imbalanced designs (whether swarmy or not) have less value under geometric mean valuations is good in my mind, because it accurately reflects the incapability of an imbalanced design.  For example, based on the BV, you might think the Glass Cannon (below) with a BV of 2653 (and an arithmetic mean of 1327) is likely to win against an inner sphere assault mech.  My expectation is that it performs very poorly in practice against an Atlas (for example).   The geometric mean of 628 better reflects this.

Code: [Select]

Glass Cannon

Mass: 100 tons
Chassis: Composite Biped
Power Plant: 300 XL
Cruising Speed: 32.4 kph
Maximum Speed: 54 kph
Jump Jets: None
     Jump Capacity: 0 meters
Armor: Standard
Armament:
     4 Gauss Rifle
     3 Large Pulse Laser
Manufacturer: Unknown
     Primary Factory: Unknown
Communication System: Unknown
Targeting & Tracking System: Unknown
Introduction Year: 3145
Tech Rating/Availability: F/X-X-X-E
Cost: 23,596,000 C-bills

Type: Glass Cannon
Technology Base: Mixed (Experimental)
Tonnage: 100
Battle Value: 2,653

Equipment                                          Mass
Internal Structure            Composite               5
Engine                        300 XL                9.5
Walking MP: 3
Running MP: 5
Jumping MP: 0
Double Heat Sink              15 [30]                 5
Gyro                                                  3
Cockpit                                               3
Armor Factor                  8                     0.5

                          Internal   Armor 
                          Structure  Value 
     Head                    3         1   
     Center Torso            31        1   
     Center Torso (rear)               0   
     R/L Torso               21        1   
     R/L Torso (rear)                  0   
     R/L Arm                 17        1   
     R/L Leg                 21        1   


Right Arm Actuators: Shoulder, Upper Arm, Lower Arm
Left Arm Actuators: Shoulder, Upper Arm, Lower Arm, Hand

Weapons
and Ammo                Location  Critical   Heat    Tonnage
Double Heat Sink           LL        2        -       1.0 
Double Heat Sink           CT        2        -       1.0 
Large Pulse Laser          RT        2        10      6.0 
Gauss Rifle Ammo (16)      RT        2        -       2.0 
Gauss Rifle                RT        6        1       12.0 
Gauss Rifle Ammo (16)      LA        2        -       2.0 
Gauss Rifle                LA        6        1       12.0 
2 Large Pulse Laser        LT        4        10      12.0 
Gauss Rifle                LT        6        1       12.0 
Double Heat Sink           RL        2        -       1.0 
Gauss Rifle Ammo (8)       HD        1        -       1.0 
Gauss Rifle Ammo (24)      RA        3        -       3.0 
Gauss Rifle                RA        6        1       12.0 

Now consider the swarm issue.  The guiding hypothesis for BV is that it is additive.  If that is inherently untrue, then that should be accounted for in the combination operator.  This was the case with the original BV2, but it was errataed away.  Fixing the combination operator seems like the right thing to do since it directly addresses the swarm issue and because it seems like mangling the per-unit valuation to avoid swarms is problematic.  For an example of the latter, a heavy ASF can be a swarm unit when fighting warships, but if you increase its per-unit valuation to prevent swarms there you end up in an awkward situation in aerial combat where the heavy ASF might be swarmed by conventional fighters.  Should conventional fighters have a much higher BV so they can't swarm ASF which have a higher BV, so they don't swarm Warships?  It seems rather awkward.

The problem in layman's terms, with FSM was that it created an immediate answer: optimize your entire BV into a single unit, and if anyone shows up across the table with more than a single unit, they have to be handicapped until BV matches.  Which is great if you're the single unit recreating Przno River, but since it's a simple answer for everyone, what you get is two optimized designs standing at medium range rolling dice until the random numbers drop one, and if the tech bases are dissimilar, it's almost always an automatic win for the Clantech unit, where if it's uniform value, the units end up being identical with 0/0 pilots for a BV that should be sufficient to field a company of units.

why? because FSM actively punishes you for filling the gap with infantry, or taking more than one unit.

WITHOUT the FSM, you can have varied scenarios and good fights, with it, you might as well go watch someone play nintendo for the hour or so instead of playing battletech, or catch up on your soap operas, or go play warhammer 40k, because you've incentivised boring play.  That's why it was dumped-it didn't actually address the problems it was supposed to address, it simply created an immediate exploit you'd have to be a moron not to use-that is, dumping all your BV allowance into a single 'mech and not taking anything to go with it.

[DevianID]

In the example mech provided, the glass cannon has a clan tech engine, giving a geometric mean of 627.  With an IS XL, it is 555.  Since these are averages, to compare to an existing BV scale, you are looking at 1254 v 1110.  The difference in the normal BV sum is 2619 versus 2653, so because the lower defensive BV provided by the inner sphere XL engine despite only having a 34 BV difference the effect is magnified to 144.  This is the whole issue with geometric means: the further you go down either F or D, the cheaper you get.  Your glass cannon's counterpart is a mech with max speed and armor and no weapons, for an assault mech jumping 8 for the same BV as a light mech thanks to massive discounts by not taking weapons--such a unit would only exist because of geometric mean discounts, as normally the value of its defense doesn't get a massive discount for not having weapons.

This means your glass cannon has a close BV to the Vulture IV B, with 2649 BV.  In the geometric mean, it has 1275, so it would be facing 2 or more of your glass cannons.  In a duel, this might be fair--they cripple or destroy each other at the same time, though the vulture has to split fire.  As part of a force though, the Glass Cannon at the edge of range or in hidden positions, despite being a comically bad mech, will still scare the pants off anyone.  Thus, in standard BV, the Glass Cannon is bad and a niche design, and loses 1v1 to the Vulture.  In *root scaling, it gets a massive BV discount making it probably the best ambusher in the game, and can beat the Vulture in a straight fight some of the time.

[Lagrange]

The problem in layman's terms, with FSM was that it created an immediate answer: optimize your entire BV into a single unit

Yeah, what existed does seem problematic.  I'm not sure what the right approach is---it just seems better to look there for a solution rather than monkeying with individual BVs to solve a systemic problem (if it exists).  TigerShark tells me the MegaMek uses front loaded initiative, and that alone is enough to reduce the temptation to swarm.

This is the whole issue with geometric means: the further you go down either F or D, the cheaper you get. 

I'd say this is appropriate, but I also support putting some floor on how low you can go so a near-zero isn't available.

This means your glass cannon has a close BV to the Vulture IV B, with 2649 BV.  In the geometric mean, it has 1275, so it would be facing 2 or more of your glass cannons.  In a duel, this might be fair--they cripple or destroy each other at the same time, though the vulture has to split fire.  As part of a force though, the Glass Cannon at the edge of range or in hidden positions, despite being a comically bad mech, will still scare the pants off anyone.  Thus, in standard BV, the Glass Cannon is bad and a niche design, and loses 1v1 to the Vulture.  In *root scaling, it gets a massive BV discount making it probably the best ambusher in the game, and can beat the Vulture in a straight fight some of the time.

I'm not sure what you are trying to say here, I guess?  2 Glass Cannon vs. 1 Vulture sounds about right.  It certainly could scare the crap out of someone, but that's not what BV is based upon.  Against someone who knows the design's weakness, a force full of Glass Cannons seems likely to lose even with the imbalance discount implied by using the geometric mean.  Certainly, I don't see it as an unfair fight in favor of the Glass Cannon.

[BATTLEMASTER]

Using '+' instead of multiplication creates quite different dynamics.

I agree, and that was probably a typo on my part, as in it's probably really multiplication and not addition.  It's been a long time since I came up with that code for MegaMek.  I can't take credit for making it an actual option though - that was someone else's doing.  It was all inspired by some other BV2 tweaks TigerShark demonstrated at the time several years ago.

[TigerShark]

I agree, and that was probably a typo on my part, as in it's probably really multiplication and not addition.  It's been a long time since I came up with that code for MegaMek.  I can't take credit for making it an actual option though - that was someone else's doing.  It was all inspired by some other BV2 tweaks TigerShark demonstrated at the time several years ago.

BTW, not that this is the correct thread, but I'd love to see my "Alternate Pilot BV" turned into the BV1 table instead. The suggested I made (reducing piloting from 15% per level to 10% per level) was implemented in canon/errata. So the canon BV1 table (i.e.: only 5% per Piloting level, and no multiplication with increased Gunnery levels) would be a nice replacement.

[DevianID]

I'm not sure what you are trying to say here, I guess?  2 Glass Cannon vs. 1 Vulture sounds about right.

I am saying you took a design, tried to make it as bad as possible (defense wise), and the end result costs as much BV as a Vulture IV B.  The Vulture spanks it, and it should--the Glass Cannon is meant to be bad as possible.
Then in comes *root, and now this mech, designed to be as bad as possible, is suddenly beating the much better balanced Vulture thanks to >2v1 odds.  This is a mech designed to be just the worst, and yet now it is benefiting.  That is not balance, as other bad mechs that are bad but balanced in F/D dont get this insane discount.  The glassiest of cannons is 1110 adjusted *root BV, which puts it at the same level as the balanced F/D Hachetman 5DD.  So the *root makes your 100 tons of 100 damage at pulse range the same as a 45 ton Hachetman--the Poor Hachetman, or pretty much any mech in the same *root BV, cant take 100 damage at range 20, nor can they usually kill the glassiest cannon at that range.

As an aside, Tigershark to address the OP, what did you think about my math about movement heat on the spider and phoenix hawk?

[Lagrange]

I am saying you took a design, tried to make it as bad as possible (defense wise), and the end result costs as much BV as a Vulture IV B.  The Vulture spanks it, and it should--the Glass Cannon is meant to be bad as possible.
Then in comes *root, and now this mech, designed to be as bad as possible, is suddenly beating the much better balanced Vulture thanks to >2v1 odds.  This is a mech designed to be just the worst, and yet now it is benefiting. 

Is it suddenly beating?  What scenario did you have in mind?  In general though, I expect it is always possible to pick something which consistently loses.  Consider an alternate loadout for the Vulture IV: ERPPC + 2 ERLL+targeting computer + double heat sinks.  This has almost the same BV.  Do you think it's consistently losing?

That is not balance, as other bad mechs that are bad but balanced in F/D dont get this insane discount. 

I don't find this compelling, because there are many ways to misdesign things, not merely misbalancing F&D.  Given that, progress in defining a BV is inherently incremental---you can't solve all problems with any simple change.

This does leave a question: what is a compelling argument?  Well, suppose we have two different mechs mutually optimized against each other that have defined-different ratios of F&D.  What happens?

[TigerShark]

As an aside, Tigershark to address the OP, what did you think about my math about movement heat on the spider and phoenix hawk?

My apologies, I just saw that. :) I can certainly see what you're saying. The absence of JJ would make the weapons less expensive, as they're not discounted by movement heat. When JJ are added, it provides a discount.

That would require jumping to have a different effect on the heat calculation for OBV, in order to avoid that pitfall. I'm guessing that's where the arbitrary "6" arose at the front of the canon calculation.

[NavPoint]

Hi, late to the party here. Wondering if the people interested in this thread could take a peek at this specific post:

https://bg.battletech.com/forums/fan-designs-rules/alternate-battlevalue-formula-preliminary-results/msg1757101/#msg1757101