Balance 3.1. Is possibly to fix before 3.2?

[vexed]

Just how are you 'fixing' it though ?
Is this all stat tweaks, code tweaks, or what ?
How did you come up with said fixes ?
And if this is code changes, then there are more questions about what is going on...

Raycast #3748 did up some patches, and yet, look how that ended up.
Heck, nobody commented on #3989 either.

If unit A attacks unit B, should the same outcome happen every time (assuming same weapons/bodies/buffs..and so on) ?
How do different tech levels play into all this, is it "balanced" on T2 & T3 as well ?
How can you get valid information if nobody is keeping track of said information--do we just want to go by 'looks right', or do we want to have everything mathematically correct ? Keep in mind that this isn't a lab, so, as in real life, things can miss for a variety of reasons. I really dislike the use of magic variables being used here for balance purposes.
Do we fix the code first, then fix the balance after all the code has been fixed or what ?

Can you explain why in the balance polls we had that most people think that balance is 'good enough' ?
Does that mean, we shouldn't touch anything ? No.
What does it mean ?
Could it mean, that people want a fun game, and don't really care that much, as long as things seem logical in progression ? Even if it is pointed out to them by "pros" that X is wrong, it don't really seem to matter that much, since, as I said, they just want a fun game.

Anyway, once all these questions can be answered, and once we know what kind of 'fix' this is (stats or code or both) then, we can go from there.

[NoQ]

Could it mean, that people want a fun game, and don't really care that much, as long as things seem logical in progression ? Even if it is pointed out to them by "pros" that X is wrong, it don't really seem to matter that much, since, as I said, they just want a fun game.

Yes it's true. Majority doesn't care about balance. Majority doesn't know what balance is. Majority doesn't know the tech tree. Majority has never heard of weapon damage modifiers. Heck, majority doesn't even play multiplayer.

But as long as they don't care, why don't we fix it, for fun of the few who care?

We may even expect those who care to be of more value to the project, expect more contributors from them, not sure, of course. At least they won't throw away their skill that took years to develop, as long as game balance is verified by their skill and matched against it.

NoQ said SC was balanced by playtesting. So this way is right way

Heh, not exactly. I, personally, would like to point out that Battle for Wesnoth also has a very tight balance (with 6 factions in multiplayer, and at least 5 of them are currently tightly balanced against each other). And i think they did some sort of heavy calculations to achieve that, not sure where did i see the prooflink, but they're automatically coming up with unit prices based on their weapons and special abilities.

________________________________

P.S. And maybe release 3.1.1 with this rebalance? There have been many important bugfixes already, and little changes in two months. Maybe we just throw in a stat rebalance and release? I'm ready to write a changelog ^___^

[Per]

I'd be happy to throw in some balance fixes into a 3.1.1 release, assuming they are well tested and not too controversial (since this is after all a stable branch). We need a new release soon anyway to stop the template cheating.

[crab_]

Vexed, thanks for answering.

Just how are you 'fixing' it though ?
Is this all stat tweaks, code tweaks, or what ?

Stat tweaks.
Code tweaks lead to more unpredictable results.
I would be happy to tweak many things in code by my own vision, but i cant.
Code changes is Changes mostly in game mechanics.
Since we have results of playtesting of current game mechanics So we can change things which are 'imbalanced' surely.
Game can be good balanced even with 'bad' inner mechanics.

How did you come up with said fixes ?
And if this is code changes, then there are more questions about what is going on...

I did come up with patch for stat data.

Raycast #3748 did up some patches, and yet, look how that ended up.

Changing game mechanics is not balance tweak. This change touches balance, but we dont know how exactly.
Changes of weapon acuracy is inner game mechanics. Game can be made balanced even with bad accuracy mechanics.

Heck, nobody commented on #3989 either.

ok. I have not seen it yet.
Accuracy is 'instrument' of balancing. I would be happy to fix accuracy, but this requires more playtesting.
We played year with current bad accuracy mechanics. We have results of playtesting with bad accuracy. We dont have any playtesting with another accuracy codes.

If unit A attacks unit B, should the same outcome happen every time (assuming same weapons/bodies/buffs..and so on) ?

In version 3.1: in most cases result of fight A vs B is predictable with high precision.
In warzone rarely happen fight of only two units against each other. Mostly we have battle of dozens units.

How do different tech levels play into all this, is it "balanced" on T2 & T3 as well ?

Balance of Tech levels is 'fixable' but...
Main line in MP games is 'no bases' + T1. I want to fix balance for 'no bases T1'.
Though I have some thoughts how T2 and T3 can be improved.

for information: we have 5 tech levels
- 1 - no bases T1
- 2 - med bases T1
- 3 - adv bases T1
- 4 - T2
- 5 - T3
Historically, balance of warzone was tuned good for first option 'no bases T1' and balance is always bad for another options. So MP players prefer 'no bases T1'.

Thinking about tech levels.... I think better set "NULL' tech for med bases T1 and adv bases T1.
T2 and T3 is hardly balanceable for some reasons. But we can try change list of avaible technologies for T2 and T3 and make possible more fair start.

How can you get valid information if nobody is keeping track of said information--do we just want to go by 'looks right', or do we want to have everything mathematically correct ? Keep in mind that this isn't a lab, so, as in real life, things can miss for a variety of reasons. I really dislike the use of magic variables being used here for balance purposes.

For 3.1 I suggest to change only things which were tested in 3.1 and defined as 'imbalanced'.
We dont want to go by 'looks right' (because our vision of 'right' is different)
We dont want to go by 'mathematics'. (But math can be used as advice where to place eye of further playtesting)
We want to go by current results of playtesting of version 3.1.
We want to go by stats tweaks with accurately predictable results.

Do we fix the code first, then fix the balance after all the code has been fixed or what ?

We cannot fix balance after code. Changes in mechanics spoil the results of year of playtesting.
I believe some code changes can be implemented without problems, but better do not touch code right now.

Can you explain why in the balance polls we had that most people think that balance is 'good enough' ?

I think lobby players mostly do not read forum and do not participate in polls.
Most of multiplayer players do no know about our cozy balance subforum.
I think we can make poll like 'Is warzone good enought?' and i'm afraid we will get 'Yes' as result. Then why improve warzone?

What does it mean ?

I think that people should take the test(exam) before they take part in these polls. e.g. defeat nullbot in 15 minutes or defeat NoQ in 30 minutes

Could it mean, that people want a fun game, and don't really care that much, as long as things seem logical in progression ? Even if it is pointed out to them by "pros" that X is wrong, it don't really seem to matter that much, since, as I said, they just want a fun game.

I know many people who complains on game balance though.

Anyway, once all these questions can be answered, and once we know what kind of 'fix' this is (stats or code or both) then, we can go from there.

I hope I have answered all the questions.
Code fixes look too controversal, unfortunately.

I'd be happy to throw in some balance fixes into a 3.1.1 release, assuming they are well tested and not too controversial (since this is after all a stable branch). We need a new release soon anyway to stop the template cheating.

We need some testing strategy. Each balance change can lead to unpredictable results, so changes should be tested carefully.

[NoQ]

defeat nullbot in 15 minutes

10 should be more than enough.

[crab_]

Hey. I'm finished beta version of my patch. Need testers
Actually, testing is most problematic part, I want to prepare a set of tests.
i dont know is my changes controlversal or not

Attached first beta version of my patch. This is untested version and requires checking and tuning after tests.
List of changes below.

Code: Select all

1) Flamers nerfed.
 - thermal armor of yellow bodies increased greatly (bug +70%, scorpion+40%, mantis+30%)
 - flamer damage reduced, but rate-of-fire increased proportionally
 - thermal armor upgrades moved earlier
2) Machineguns nerfed
 - impact of rate-of-fire upgrades reduced -20%
 - reduced MG damage to structures and defenses
3) Incendiary Mortar nerfed, rate-of-fire -20%
4) Try to increase effectiveness of early artillery in low-oil games
 - mortars, bombards, pepperots was sligtly improved. (damage +15%, price of mortar turret reduced)
 - first Howitzer improved (damage +20%, price -20% (Ground Shaker and Hellstorm was not touched)
note: mortar pit was nerfed to satisfy high-oil games (construction time +35%, price +10%)
5) Nerf Rockets   (because they should become OP after we nerfed MGs and flamers)
 - rockets damage -10%, price +10%
6) Nerf bunkers (because we nerfed flamers)
 - bunkers price +25%, construction time +50%
note: bunkers mostly used as meat shields in turtle tactics in high-oil game.
7) Try to nerf rush strategies in early game in low-oil matches
 - reduced production time for Mini-Pod and Light Cannon
 - reduced effect of oil drum from 100 to 50
 - truck speed reduced (now trucks 25% slower on off-road in early game)
8) Try to bring scout units into game
 - sensor now avaiable without any research
 - reduced production time of sensor turret
 (tried to make sensors avaiable without HQ but failed)
9) Hover propulsion slightly nerfed, speed -20%
10) Try to increase effectiveness of defensive strategies in low-oil matches
 - price of cannon hardpoints -25%
 - removed requirement to have researched MG Bunker before start research MG Hardpoint
 - improved Mini-Rocket-Array Battery
 - improved Tank Trap
11) Reduced effect of ROF upgrades of AA-turrets
12) Cannons slightly improved
 - light cannon, splash radius increased
 - medium cannon, price slightly decreased
 - HPV-cannon, rate-of-fire increased +14%
 - assault cannon, decreased production time and price.
13) Heavy Rocket Bastion nerfed, range -20%
14) Try to fix Vtol bombs 
 - improved Phosphor Bomb Bay, Cluster Bomb Bay (damage +40%)
 - nerfed Thermite Bomb Bay, Plasmite Bomb Bay (damage -15%)
note: thermal bombs also was nerfed by improving thermal armor of yellow and green bodies.
15) Some overpowered researches changed
 - impact of power generator upgrades was reduced 25% (this touches only 3 first power upgrades)
 - slightly nefed first upgrades of research (Synaptic Link Data Analysis mk1, mk2, mk3)
16) Some late-game weapons and bodies improved to make them more usable
 - Flashlight, Pulse Laser improved, rate-of-fire +15%
 - Needle Gun, rate-of-fire +15%
 - black bodies: decreased research requirements, decreased production time
 - Leopard body: moved earlier in research
 - EMP Cannon: moved earlier in research
17) Transports weapons nerfed. Strategy of transporters-abuse was removed. 

[Iluvalar]

I can defeat NoQ and I defeated Nullbot in 6:30 minutes in vision. and you have my seal of approvement for that change log.

Only 2 points :
1) Are you sure the flamers dont hit the 1/3 minimum damage with your change. If so, the change could make them even stronger "rate-of-fire increased proportionally" bit.

2) Nothing here adress the DPS increase provoked by the accuracy bug. That one still need a fix.

[crab_]

1) Are you sure the flamers dont hit the 1/3 minimum damage with your change. If so, the change could make them even stronger "rate-of-fire increased proportionally" bit.

Not sure 100%, i plan testing and checking exact value of damage on different research stages.
33% of damage it is 33% from base damage with propulsion modifier.

Abstract Example: flamer with 50 damage and 1.0 ROF against mantis with 20 armor: deals 30 damage per second. Flamer with 25 damage and 2.0 ROF against mantis with 20 armor: deals 25/3 = 8.3 * 2 damage per second.
Lowering damage is making lower 33% of this damage.

2) Nothing here adress the DPS increase provoked by the accuracy bug. That one still need a fix.

I think it is hardly possible to recover to 2.3.x balance data (in what version of warzone accuracy bug not existed?)
My change log based mostly on results of playtesting of 3.1 version (and beta11).
I mean: we played with current accuracy bug long enough. Fixing this bug may lead to more unpredictable results and requires more playtesting.
As far i understand weapons with low accuracy became stronger in 3.1. (Machineguns, Mini-Pods)

[NoQ]

We could probably nerf the single machinegun (maybe even decrease ROF a bit).
As for flamer ROF, i doubt it can be too powerful, as damage caused by flamers is not proportional to ROF: they just either make their targets burn or not instead, burning them second time wouldn't double the effect. Only if completely different enemy units get burned the second hit would make a difference, and even in this rare case it would still be at most proportional to ROF.

I defeated Nullbot in 6:30 minutes in vision

By the way, technically, there are at least two bugs in 3.1.0 (both fixed in 3.1-git) that break NullBot's early build order. One of them is on the game code side (makes the AI instantly throw a truck on derricks), other (more minor, appears in 1/4 of all games) can be workarounded by getting a newer NullBot package. I don't doubt you, of course (:

(tried to make sensors avaiable without HQ but failed)

This should be done in rules.js via enableTemplate() call (relies on templates.txt to provide an actual template).

[crab_]

We could probably nerf the single machinegun (maybe even decrease ROF a bit).

Do you want make rush strategy more weaker or you want nerf MGs more?
I think it is normal if rush strategy will be slightly stronger than other strategies.

Some thoughts:
- if rusher successfully rushed enemy base then he wins
- if technologist successfully survived first attacks then he wins (half-tracks makes obsolete rushing units very fast)
-----------------------------------------------------------------------------------------------------------------
+ rush strategy should be stronger when it was not expected by victim of rush
+ when you know about rush - you have to be able to resist rush.. but how?
A - pure rush
B - half-rush half-tech
C - pure tech

A stronger than C
C stronger than B
B stronger than A

So when you expecting rush - make B strategy.
When you expecting tech - make A strategy.
When you expecting half-rush - make C strategy.

Lets define A,B,C
A - pure rush - F-F-R-CC-F. Only one research with damage upgrades
B - half-rush - F-F-R-R-CC-F - Two research. 1 - damage. 2 - half-tracks or power module
C - pure tech - ??? R-F-F-R ?

I think price of first upgrades should be increased. To make strategies more clear. In current balance rusher can make some research later and do enough research.
Hmm. it is just thoughts....

As for flamer ROF, i doubt it can be too powerful, as damage caused by flamers is not proportional to ROF: they just either make their targets burn or not instead, burning them second time wouldn't double the effect. Only if completely different enemy units get burned the second hit would make a difference, and even in this rare case it would still be at most proportional to ROF.

Flamers have direct damage, not just burn. And in my patch direct damage was nerfed more than periodical damage.
Testing. Testing. Testing. Testing.

This should be done in rules.js via enableTemplate() call (relies on templates.txt to provide an actual template).

I have tried this. Unsuccessfully.
Have you working example?

[crab_]

@NoQ
What you think if i move Mini-pod and Light cannon just a bit earlier in research? Then tech tactics will work better.
I see following task: tech-strategy should be stronger than 'half-rush' but weaker than 'pure-rush'

See:
1) Rush
- invest into factories. Whole base earlier, derricks earlier. More oil.
- sacrifice 2-3 minutes of technologies (rusher can buy this tech because he have more oil because he built up derricks earlier)
2) Tech
- invest into research labs
- sacrifice energy and progress in expansion (loses derricks in middle of map)

I think Rush strategy gives more advantages. I would vote for paid derricks. But i think it will make some players angry
Technologist have better techologies but he need more time to recover oil resources and expand.
Technologist have 2-3 minutes when he have mighty half-tracks, but after this 2-3 minutes he does not have any valuable advantage.

[crab_]

i have got some comments and criticism on my balance patch.
They say "too much changes". is that is real problem?

Some first time testing of patch was done and this testing also revealed some issues.

List of revealed issues
1) Machineguns are too strong still.
2) Bug body has bad resistance to flamers. Need increase thermal armor of bug.
3) Usage of scorpoin body is not panacea from flamers.Themal armor upgrade appear too late. Need make earlier first upgrade of thermal armor.
4) Some research appear too early. Fixed.

Now i'm working on second version of the patch.

[NoQ]

I have tried this. Unsuccessfully.
Have you working example?

[crab_]

@NoQ It works. Thank you very much
Now i see in my try i've made unavailable sensor template.

[Iluvalar]

You are wrong crab_

Pure rush beat h-track and rushers all together.

Since the accuracy bug the mg got an accuracy buff of around 25% against viper and 100% against structure.

The result is that if you build an extra research lab for something else then mg in the first 6 minutes, you are doomed.

Note that i distinguish "pure" rush, from normal rush in which the player usually still care about a second research lab eventually.