Points of Imbalance.

@FtXCommando said in Points of Imbalance.:

Your tests are the same shit I said.

Heh, you're right. For some reason I read you as refuting the notion that the ML kills its mass in bricks.

BP on t4 should never be touched. An extremely slight stat nerf is the only thing that should even remotely be on the cards.

That sounds reasonable. I'd shoot for 13-14 bricks to kill a ML when micro'd well, and a similar 80-85% mass efficiency for bricks versus all assault experimentals.

There should be an advantage for investing in all that production.

There should be an advantage to not store up 20000 mass until you have your unit.

@harzer99 said in Points of Imbalance.:

There should be an advantage to not store up 20000 mass until you have your unit.

Its about a minute and 16 seconds worth of mass income for a single player if you have a mass income of 260.

Usually i make my first t3 unit when i have a mass income of 260

All units start out off range
All units get into range ASAP
Using move order (No attack move)
No units leave firing range
No vet at start of fight
No unit try to retreat
No stealth or other dirty cybran tricks

ML: 20000 mass, 27500 BT
10 bricks: 12800 mass, 48000 BT
15 bricks: 19200 mass, 72000 BT

ML kills 10 bricks. 12000 hp remains
15 bricks kill ML. 8 bricks remains

30 bricks kill 2 ML. 17 remain
2ML kill 20 bricks. 15000+ hp remain

Just for lulz:
100 medusas (3600 mass) kill ML. 36 remain (5 killed by on death explosion)
Yet microed ML can avoid almost all damage

https://replay.faforever.com/13090487

Did some testing with different number of bricks
@moses_the_red said in Points of Imbalance.:

That sounds reasonable. I'd shoot for 13-14 bricks to kill a ML when micro'd well, and a similar 80-85% mass efficiency for bricks versus all assault experimentals.

Very close to current situation in direct fight
12 bricks die to ML
14 bricks kill ML

PS Attack move ML can die even to 10 bricks cos of attack order sorting and slow turret

edit: added build time to mass comparison
edit2: did some testing with different number of bricks
edit2: mentioned attack move

I miss days of 6-10 Percies killing DirectFire Experimental (NB this wasn’t good balance). But to build on Harzer:

First, I am shitty garbage noob player whose trying to balance with playtest help from far surperior players than I for SCTA(Balance). And in that process I do four things:

1. I use my vague gut looking at Unit DB assigning Stats and relative value. I do sandbox simulations then adjust
2. I use formula Harzer descrihed above to adjust again then do some adjustments. Based on Sandboxing and a couple test games
3. I get folks who aren’t shitty 1k globals to play games and watch both how they play and how the armies fight. Then adjust again.
4. I compare tje formula from 2 to determine relationships again see if anything in raw dps/mass ratio borked things after adjustments. Then I generally don’t do adjustments but keep these in mind as I watch games or sometimes pm folks I know better than me, thoughts or why certain FAF related balance changes were done so I can understand.
5. Well playtest.

As I can vouch for raw bp comparison for balance is actually terrible and results in very strange relationships not accounted for gameplay. One of the issues I ran into using Harzer’s formula’s at end not middle of the process as a fine tuning measure. It either result in units being notably weaker than they should had been. Or vastly stronger. Due to BuildPower relationships in addition to how game naturally flows.

A formula establishes a baseline it does not establish a balance

@advena said in Points of Imbalance.:

All units start out off range
All units get into range ASAP
Using move order (No attack move)
No units leave firing range
No vet at start of fight
No unit try to retreat
No stealth or other dirty cybran tricks

ML: 20000 mass, 27500 BT
10 bricks: 12800 mass, 48000 BT
15 bricks: 19200 mass, 72000 BT

ML kills 10 bricks. 12000 hp remains
15 bricks kill ML. 8 bricks remains

30 bricks kill 2 ML. 17 remain
2ML kill 20 bricks. 15000+ hp remain

Just for lulz:
100 medusas (3600 mass) kill ML. 36 remain (5 killed by on death explosion)
Yet microed ML can avoid almost all damage

https://replay.faforever.com/13090487

Did some testing with different number of bricks
@moses_the_red said in Points of Imbalance.:

That sounds reasonable. I'd shoot for 13-14 bricks to kill a ML when micro'd well, and a similar 80-85% mass efficiency for bricks versus all assault experimentals.

Very close to current situation in direct fight
12 bricks die to ML
14 bricks kill ML

PS Attack move ML can die even to 10 bricks cos of attack order sorting and slow turret

edit: added build time to mass comparison
edit2: did some testing with different number of bricks
edit2: mentioned attack move

I think you missed the replay where I went up against Tagada (currently #3 in the world) and was able to consistently kill 15 or so bricks in an open area with a Monkeylord.

My 1v1 rating is about 700... If a 700 can kill 15-16 bricks consistently against the #3 player in the world... well... its not 13-14.

Its somewhere between 16 and 20. I was unable to kill 20, and was consistently able to kill 15-16. It was admittedly pretty close, so perhaps 16 is the limit, perhaps 17 is the limit. Perhaps with practice I could pull off 18, idk. The point is its too high.

For reference, at launch, it was general knowledge that 10-12 bricks would kill a Monkeylord, so you can see just how significant the T3 balance patch was for assault experimentals. We're not talking about a small change.

Here's the patch itself:

https://content.faforever.com/patchnotes/3696.html

How hard was the Brick hit with the nerfhammer? Pretty damn hard.

Health 9000->7500 ~ 17% decrease
Damage -> 17% decrase
Range -> 8.5% decrease
DPS -> 17% decrease

Its worth mentioning that changes like this are multiplicative. The brick isn't 17% less effective than it was. Its 1.17 * 1.17 * 1.085 times less effective. In real terms this is somewhere around a one third drop in unit effectiveness. You probably need 3 new bricks to be as effective as 2 of the old bricks. Interestingly this seems to mesh with notion that it takes 15-17 bricks to kill a ML where before it took 10-12.

That is a massive nerf to the brick.

And I'm not saying it wasn't justified. I like T2/T3 balance as it is more or less. There's more to that change set than just the balance dynamic between T3 land and assault experimentals.

But that kind of nerf with no corresponding nerf to assault experimentals has left T3 units in a terrible state relative to T4.

And its not just the brick (although perhaps the brick got it the worst) the entire slate of T3 units was hit in that patch.

Given that level of capability decrease of course T4 assault experimentals were going to wind up OP.

@moses_the_red
I've seen your replay. You'll newer be able to micro ML like this outside of vacuum.
Even in your replay you loose if ML start retreating from not ideal vector

But there is simple solution I like (but can't estimate full effect it will have on balance):
Buff brick speed by 0.2

@advena said in Points of Imbalance.:

@moses_the_red
I've seen your replay. You'll newer be able to micro ML like this outside of vacuum.
Even in your replay you loose if ML start retreating from not ideal vector

But there is simple solution I like (but can't estimate full effect it will have on balance):
Buff brick speed by 0.2

Its not just the brick, its all T3 units, check the patch notes for that patch.

Everything was hit.

Percival, Otthum, Loyalist, Harbinger and Titan.

The problem isn't that the Brick is underpowered. All T3 land is underpowered relative to T4 assault experimentals.

We nerfed all these T3 units without bothering to hit assault experimentals with corresponding combat effectiveness nerfs.

I've been talking bout Bricks and Monkeylords, because those are the units I'm most familiar with. I can make a case with those specific units rather clearly... but the problem isn't with just those units.

Also, speed is probably not the stat you want to buff or nerf, because whether it actually matters or not is hard to quantify.

If you buff the speed of a unit, but its still slower than a kiting unit, you really haven't changed the balance between the two.

Speed changes mostly matter at inflection points, points where you have enough speed to do something you couldn't do previously. They do strange non-linear things to balance. Range changes have the same issue.

I'd stick to just health and damage changes if I were the balance team. Those are easy to quantify.

@moses_the_red
If brick speed was 2.5 you would have lost ML in 100% battles of your replay.

Brick is only T3 direct fire unit ML can kite. Percies rape it and others direct fire units are faster

@advena said in Points of Imbalance.:

@moses_the_red
If brick speed was 2.5 you would have lost ML in 100% battles of your replay.

Brick is only T3 direct fire unit ML can kite. Percies rape it and others direct fire units are faster

You're ignoring the fact that the patch affected not just the Brick's balance aganist a ML, but the bricks balance against the GC, Chicken Megalith and Fatboy as well.

It is significantly weaker than it was against all assault experimentals.

If you try to correct this by focusing on Monkeylords and Bricks, you're going to leave a lot of things broken.

Is the Brick speed change also going to reverse the relative balance change between the brick and the Megalith? No right?

And since its a speed buff, you might get unintended consequences. As I said, speed changes aren't going to be linear. Bricks might inadvertently become OP against something else.

If you want predictable balance changes, you stick with health and damage.

Brick does decent against all other T4s. The problem is the vet HP ML gains due to it being the most efficient dps/mass T4.

Experimentals beeing very strong is part of this game

@FtXCommando said in Points of Imbalance.:

Brick does decent against all other T4s. The problem is the vet HP ML gains due to it being the most efficient dps giving T4.

By what metric?

The old balance patch cut its combat effectiveness by something like 1/3. How many bricks to kill a GC?

The point of all this was never to say that brick-monkeylord balance is off. Its to say that when you change a class of units (T3 asasult units) and don't change a competeting class (assault experimentals) you wind up with OP assault experimentals.

None of this will matter if Cybran players just stop building Monkeylords and build Megaliths instead - or switch to Aeon and build GCs.

@herzer99 said in Points of Imbalance.:

Experimentals beeing very strong is part of this game

It is NOW, it was changed.

Perhaps you weren't around before the 2018 patch, but Experimentals - as a class - were generally not mass efficient against T3 assault units.

There were exceptions like the fatboy, but the fatboy has other serious issues (speed, low health) that mitigate the issue.

And to be clear, I'm not advocating that the T3 balance patch be reverted. I'm advocating that it be extended to all the units it should have affected.

You can't just massive change T3 assault units and not change other competing units and declare that all is fine.

By the metric of mass efficiency. ML is the only one where you arguably can lose with equal mass investment. Even this could be argued is a good thing because ML holds the title as the easiest T4 to counter with more out of the box tactics.

Like it's basically the only T4 that can reliably be killed by rambo ACUs, it easily gets murked by any loss in air advantage. Point of it's efficiency in crushing swarms of T3 is that it auto dies to any other form of response.

Not that this issue really exists outside of the brick v ML issue because percy murks ML due to alpha + higher range + shields. Snipers also crush ML pretty bad given vision.

I have indeed just now startet playing this game

@FtXCommando said in Points of Imbalance.:

By the metric of mass efficiency. ML is the only one where you arguably can lose with equal mass investment. Even this could be argued is a good thing because ML holds the title as the easiest T4 to counter with more out of the box tactics.

Like it's basically the only T4 that can reliably be killed by rambo ACUs, it easily gets murked by any loss in air advantage. Point of it's efficiency in crushing swarms of T3 is that it auto dies to any other form of response.

I thought we were all on the same page here. I've been speaking of Bricks and Monkeylords with the expectation that everyone understands that they are just representatives from the classes of units that were thrown out of line. Representatives that are easy to test.

Patch 3696 changed an entire class of units, and did not touch another competing class of units.

Clearly the issue extends beyond the Brick and Monkeylord, or do you think that the balance team just happened to fix T3-T4 balance when it fixed T2-T3 balance by not doing a damn thing - like some kind of miracle everything just worked out perfectly.

Cut all these units effectiveness by 1/3, and lo and behold, that's exactly how off T3/T4 balance was right?

The old game balance was not good. I have no idea why you framed this as though we should be striving to match the garbage late game balance of 2017.

Also the game is balanced for 1v1 games.
Not some weird chokepoint teammaps