Balance Option 2 - Cherry pick in Autobalance (public)

[Iluvalar]

Hello All !
Before I start, I'd like to express how tedious it is to make a balance suggestion like that. I need to discuss with everybody, make an epic load of tests, make sure everybody is happy with that...

1) The first point to that balance is a whole bodies review.

I think most of us know that we use much more the python and the mantis. Part of it come from the research time of course. You should as a rule of thumb chose the latest bodies, but they should not progress faster than the weapon ~dps~ progression. That's what we have here, the heavy bodies make a progression in sawtooth. Even if they have the exact same research time, the python end up with an mysterious overall 10% more strenght than the scorpion. Why bugs that we obtain after are generally weaker ( 2.94>2.96 but still) than cobras ?

On the other hand, there is another bonus in the game that I will call the synergy. That's the bonus you are supposed to get by design. The good weapon with the good body at the good occasion. That bonus is kind of low, most of the time I estimate it at 15-20% . I think we all experiment that in game, but why would I care about a 15% bonus on a scorpions in some occasions when I'm sure to get 10% whatever what happen with pythons ?

Same with propulsions, we get that for a 30 second research (dont tell me you would never research engineering otherwise). Why should the remaining of the game after 2 minutes should be played on htrack ?

As a result, the bodies and propulsion balance will definitively make the design a part of the game again. You might discover that your favorite weapon work not bad on a cobra, you might think about surprising someone by using wheels in the middle of the game.

2) I added a new machine gun for the transporter in the stats. It have the same stats than the normal heavy machinegun (it doesnt get the big boost other vtols get for reloading between attack). I think it's just a trivial fix...

3) Here is the real cherry pick :
Various reasons
Incendiary Mortar (Mortar-Incenediary)* 22%
Incendiary Howitzer (Howitzer-Incenediary)* 23%
Bombard (Mortar2Mk1)* 140% (small change, but go with the incen mortar one, to compensate the lost)
Inferno (Flame2)* 32%
Plasmite Flamer (PlasmiteFlamer)* 27%
VTOL Plasmite Bomb Bay (Bomb5-VTOL-Plasmite)* 55%
VTOL Thermite Bomb Bay (Bomb4-VTOL-HvyINC)* 48%
VTOL Cluster Bomb Bay (Bomb1-VTOL-LtHE)* 436%
Nexus Link Turret (SpyTurret01)* 381% (useless otherwise)
EMP Cannon (EMP-Cannon)* 2470% (useless otherwise. This one will at least be useful as good meat shield)

Emplacements for mortars
Incendiary Mortar Pit (Emplacement-MortarPit-Incenediary)* 148% (big compensation suggested by the balance...)
Incendiary Howitzer Emplacement (Emplacement-Howitzer-Incenediary)* 164% (big compensation suggested by the balance...)

I'm open to suggestions, but those were identified as needing some serious adjustements...

[Reg312]

First, Agreed about bodies and propulsion, my conclusions are same.

Some questions
c1. EMP cannon can be usable, if you add splash effect to it. why not? first time i saw working EMP cannon in your NRS mod and it is intesting weapon.
c2. If you nerfing some (or most of) bodies does it means you making easier turtle tactics?
c3. Nerfing inferno so much can mean it will be useless against scorpion. does it?
i suggest another thing, increase attack rate while decreasing damage to make thermal armor more valuable!
again, can you say why not?
c4. You changed thermite bomb burn damage from 200 to 169, it is not 48%
and burn damage it main part of damage for thermite bombs (about 90% of all damage)

[Iluvalar]

c1 : All my changes follow rules designed to keep most aspects reconizables. I'm open to other kind of change, but i'm not making such decisions.
c2 : Theoricaly yes, turtlers will have the life a bit easier. In practice, since the stats actually pretend all structure share the same body, they decay faster than the droids upgrades. it will just make the structures resist just a couple more seconds. As far as I know, the incen. mortar was a known check point for the turtler in high oil.
c3 : If the opponent make every move : tracks, thermal armor, yellow bodies, the flammer player will need to come with a plan B. I wont say "useless" but suboptimal for sure. However, remember that many players will try to aim the leopard or the panther now. having some flamers at that point might be interesting.
C3-2 : I dont think it's useful, the incendiary damage is many small time in a row, it is already vulnerable to thermal enough.
C4 : Look at the price.

[Reg312]

с1. so. it sad to hear you cannot just fix some things in stats, cause i know few things which should be fixed. For example check build time of assault gun hardpoints.
I feel you are prisoned in your script, which never balance rate of fire and some other things
But rate of fire is better than damage to balance because it is not breaks damage-armor substraction. For example adding 10% to rate of fire is 10% in all case. are you agree?
c2. i understand. closed.
c3. Again. i ask may be you can increase research points for flamers and balance (increase) damage this way?
i just remembered example of twin assault cannons. It was nice weapon, but when if was nerfed by 10 damage (65 - 55 or so)
it become "balanced" and now people rare use it.
You nerfing inferno much more than TAC was changed.
c4. If one unit can destroy whole army, +100 energy to build is not scary.
but i noticed thermite bombs became weak in master without any stats changes.
need testing. closed.
c5. why not fix cannon fortresses. May be It is balanced on low oil. but on high oil is very unbalanced (range like laser, splash effect like artillery, damage like 4 heavy cannons)

[Iluvalar]

c1 : Yeah but you'd be really surprise... even angry if the medium cannon was suddenly shooting twice as fast. Or your tmg was shooting slower than the light cannon because of "balance". That kind of change must be chosen, not be automatic.
c3 : I detect the TAC as still 30% too strong, see how Oddity reacted on the other thread. Also, the flamers use wheel for speed, kill wheel due to modifier and are more potent vs ligher body due to penetration and area of effect.
c5: The fortress are not so OP, it's by comparison to other structures from same time that they are real good. They were also falling just behing the "worst" things that I was aiming in that change.

[Reg312]

c1. you can set it by hands and recomplie auto-balance.
but you refuse to set something by hands, even useless EMP
c3. so if TAC 30% stronger. what with heavy cannon? load of heavy cannons also strong
if you decrease armor then you improve TAC. in current version TAC is not so strong.
c5.i play high oil games and i can say fortresses overpowered. Players build it in 100% games, if researched.
and it just requres minimum micromanagenet (what increase OP value)
turtle high oil games have two defense checkpoints:
- flame mortar
- cannon fortress

so, as expected all my suggections refused. You're right, and all, no way to suggest something.
i'll prepare list of adjustments and post here.
I have feel most changes and whole "balance" refer to 1v1 or 2v2 low oil games (6-10 oil res per player)

Short conclusion:
1) Auto-balance is good way to detect certain things and adjust "balance", but better make changes by hands, to provide round numbers.
2) Auto-balance is Iluvalar's black box. No one know what inside. For example what value have build time of units?
Should make it open. not in style "56*74*543 = 349034 what mean 48% and we should do %%%%"
3) I agree with changes, except inferno (changes looks too sudden), i dont want another unusable weapon.
4) I have feel what high-oil-flate-crap games become worse after this changes, but is too long to take such games on master version

[Iluvalar]

so, as expected all my suggections refused. You're right, and all, no way to suggest something.

You are not the only one around who suggest things. I can't just accept every suggestion and manually change the stats. You need concensus. I take the feedback of everybody into account dont worry. The Balance option 3 is comming I guess.

2) Auto-balance is Iluvalar's black box. No one know what inside.

It's just a big equation that take about everything relevant into account. I just checked and the equation in around 200 operation long in the core. But you can guess safely it's a mix of damage, HP and cost.

[effigy]

I asked about this in IRC, but don't remember getting an answer:

Why do you lump all changes by percent? I think it's deceptive.

For example: "EMP Cannon (EMP-Cannon)* 2470% (useless otherwise. This one will at least be useful as good meat shield)" I'm left to assume this means EMP Cannon has 2470% more HP. I've no clue if you're talking about the turret, hardpoint, or pit, though.

Another example: "Incendiary Howitzer (Howitzer-Incenediary)* 23%" What about it changed by 23%? Range? Thermal Damage? Kinetic Damage? HP? All stats changed?

[Iluvalar]

it's irrelevant here, but i gave the full changelog as well, I dont think I need to copy it here. But my usual method is to change HP, armor, cost and all type of damage simultanously (if relevant). I do the changes after making a full template transformation. This is assuming there is a body beneath that weapon before applying the changes. I chose those stats because they will change minimaly the main game style.