Help needed testing 3.2.x Campaign games!

The projects speaking tube.
Add your two cents if you want to.
Bethrezen
Regular
Regular
Posts: 585
Joined: 25 Sep 2009, 02:05

Re: Help needed testing 3.2.x Campaign games!

Post by Bethrezen » 26 Jun 2018, 22:48

No, it's not. You have included the armour of the medium structure and this is not included in the calculation of the one-third-damage.
doesn't make any difference if you include the armour in the calculation or not you still get more or less the same answer

12 x 0.65 x 0.333333333 = 2.599999997

12/3=4
4x150=600
600x0.65= 390
390 - 7 = 383 per min
383/150=2.5 per shot

12/3=4
4x150=600
600x0.65= 390
390 /150=2.6 per shot

although why you wouldn't include the armour i don't know the one third rule should include both the modifier and the targets armour, just as it is when you are calculating damage normally.
And as you can also see on the formula site of the base guide the one-third damage will always be rounded down, even if it's 2.6 it's not rounded up to 3.
perhaps not but since rounding rules dictate that when 0.5 or higher you round up I would still round up and set it 1 point higher anyway since its a.) such a small difference and b.) will ultimately help the weapon perform a little better in the long run, because in the case of the machine-gun you are only looking at a difference of like 150 damage per min anyway

150x2 = 300
150x3 = 450

which in the grand scheme of things is nothing. About the only time i wouldn't round up is when the difference is going to be significant and there by make the weapon to strong like so for example add like 1000 points of damage more in that case you are better to round down, ultimately whither to round up or down is a bit of a judgment call best made on a weapon by weapon bases depending on what the damage difference would be, and weather it would make the weapon to strong or not.

alfred007
Trained
Trained
Posts: 457
Joined: 31 Jul 2016, 06:25
Location: Stuttgart, Germany

Re: Help needed testing 3.2.x Campaign games!

Post by alfred007 » 26 Jun 2018, 23:19

If you really want to adjust the modifiers you have to do this for every weapon class and must play every time the whole game because if you change the modifier of machine guns for medium structures it has also an influence at the assault gun in beta and gamma campaign. After that, you have to repeat this for hard structures. And after you have finished the anti-personnel weapons you have to repeat the whole testing for cannons, then for anti-tank weapons, then for artillery and last but not least for flamers. If you want to do it thoroughly I estimate you need two years just for that. And if you set the modifiers to a value that the calculated and the one-third damage are the same, you will have no benefit for gameplay after all this work.

It's a little bit funny because now you are doing the math and I argue that the truth is in the game. The balance until alpha 05 is good as it is and I see no reason why we should restart from the beginning. I agree that the one-third-rule is a point we have to look at but we shouldn't make us more work than necessary.

Bethrezen
Regular
Regular
Posts: 585
Joined: 25 Sep 2009, 02:05

Re: Help needed testing 3.2.x Campaign games!

Post by Bethrezen » 26 Jun 2018, 23:30

It's a little bit funny because now you are doing the math and I argue that the truth is in the game.
oh don't think i don't see the irony in that :P

the thing of it is though is that you are using the one third rule as a crutch and that's not what its meant for and doing so is bad practice which is very likely going to end up coming back to haunt us later, and if we are going to tackle this then now is the time to do it before we get to far in to the game, ultimately the fact that changing the modified now has an impact later, is in the grand scheme of things irrelevant since about 98% of the weapons in the game need rebalancing anyway due to being either underpowered or overpowered.

the other thing about this is that currently if you want to make small adjustments to weapons say machine-guns for example currently those changes will have no effect since machine-guns are currently dealing negative damage, so you have to adjust them till they are not dealing negative damage first, and ultimately I'm not actually trying to change the balance we have on the first 5 levels I'm simply advocating for setting a proper base line.

don't get me wrong i get where you are coming from but the thing of it is, is if we are going to be going through the game 1 weapon at a time and 1 level at a time it's not really that much as a stretch when we have the balance we want to just do a quick check for that weapon and make sure that its modifies are set right

its a little more work now due to the fact that we haven't done that for the first 5 levels but once the back log is cleared its not so bad.

alfred007
Trained
Trained
Posts: 457
Joined: 31 Jul 2016, 06:25
Location: Stuttgart, Germany

Re: Help needed testing 3.2.x Campaign games!

Post by alfred007 » 26 Jun 2018, 23:44

Bethrezen wrote:although why you wouldn't include the armour i don't know the one third rule should include both the modifier and the targets armour, just as it is when you are calculating damage normally.
And as you can also see on the formula site of the base guide the one-third damage will always be rounded down, even if it's 2.6 it's not rounded up to 3
perhaps not but since rounding rules dictate that when 0.5 or higher you round up I would still round up and set it 1 point higher anyway since its a.) such a small difference and b.) will ultimately help the weapon perform a little better in the long run, because in the case of the machine-gun you are only looking at a difference of like 150 damage per min anyway

150x2 = 300
150x3 = 450

which in the grand scheme of things is nothing. About the only time i wouldn't round up is when the difference is going to be significant and there by make the weapon to strong like so for example add like 1000 points of damage more in that case you are better to round down, ultimately whither to round up or down is a bit of a judgment call best made on a weapon by weapon bases depending on what the damage difference would be, and weather it would make the weapon to strong or not.
The one third-rule is this way coded in the game. We can discuss it if it should look different but at the moment the damage is calculated as nearo has posted.

About the rounding, you are right that you usually round up from .5 on. But you can code it that way that you always round down. I didn't have a look into the code about this and even if I had I wouldn't be able to see if this is right or not. I have to leave that to Berserk Cyborg.

A difference of 150 is in grand scheme nothing. But in this special situation in alpha 01, you would increase the damage per minute by 50 % (from 300 to 450) and this is a significant change in alpha 01. This will have a remarkable effect on gameplay. It would be like increasing the damage of the machine gun from 10 to 15. Make a test, increase the base damage of the machine gun in weapons.json to 15 and play alpha 01 and alpha 02 and just look how fast you can destroy the buildings.

Bethrezen
Regular
Regular
Posts: 585
Joined: 25 Sep 2009, 02:05

Re: Help needed testing 3.2.x Campaign games!

Post by Bethrezen » 26 Jun 2018, 23:47

A difference of 150 is in grand scheme nothing. But in this special situation in alpha 01, you would increase the damage per minute by 50 % (from 300 to 450) and this is a significant change in alpha 01. This will have a remarkable effect on gameplay. It would be like increasing the damage of the machine gun from 10 to 15. Make a test, increase the base damage of the machine gun in weapons.json to 15 and play alpha 01 and alpha 02 and just look how fast you can destroy the buildings.
well this is why i said its a bit of a judgment call whether to round up or down in order to decide that we just need to try it but of course currently we cant because the machine gun is dealing negative damage so that would need to be addressed first and this sort of a thing is why you don't want to be using the one third rule because then you are unable to test out small changes like this.

alfred007
Trained
Trained
Posts: 457
Joined: 31 Jul 2016, 06:25
Location: Stuttgart, Germany

Re: Help needed testing 3.2.x Campaign games!

Post by alfred007 » 27 Jun 2018, 00:04

Bethrezen wrote:well this is why i said its a bit of a judgment call whether to round up or down in order to decide that we just need to try it but of course currently we cant because the machine gun is dealing negative damage so that would need to be addressed first.
They are only theoretical dealing negative damage. Because of the one-third rule, they still deal a real damage. The one-third-rule prevents that a weapon would deal no or a negative damage. If they would deal a negative damage the HMG wouldn't be able to destroy the NP hardpoints in alpha 05. But they do. Very slow but they do. And that's because the one-third-rule takes over and calculates the damage the HMG deals to the NP hardpoints.

I still see no necessity to adjust the modifiers at this moment. They had no negative influence on gameplay until alpha 05 and if you bring the modifier to a value that the calculated damage is the same as the one-third-damage you will have no difference to the situation we have at the moment. I agree that we have to take care of this for the next levels but I think we can leave that for the situation when we see it has a negative influence on gameplay.

Bethrezen
Regular
Regular
Posts: 585
Joined: 25 Sep 2009, 02:05

Re: Help needed testing 3.2.x Campaign games!

Post by Bethrezen » 27 Jun 2018, 01:38

ok so just had a go at setting the machine-gun to do 3 damage per shot against there structures which would be

Code: Select all

"ANTI PERSONNEL": {
        "BUNKER": 125,
        "MEDIUM": 85,
}
and actually I can't say I really noticed much if any difference against there normal structures though there was an appreciable difference against there bunkers probably because the difference is bigger since you normally only do 150 damage per min if you are only dealing 1/3 of your base damage but when it was doing 3 damage per hit it would be doing 450 damage per min which is 3 times higher so that's to strong but 2 damage per shot which equates to 300 damage per min against bunkers seemed reasonable of course they are still slower them flamers.

which would be

Code: Select all

ANTI PERSONNEL": {
        "BUNKER": 117,
        "MEDIUM": 85
}
of course the problem with this is that when you get to the twin machine-gun this is way to strong humm....

[edit]
So after putting this into a spreadsheet and playing round with various values I'm not sure this is actually achievable due to the way the game is set up if you increase the modifiers to compensate for weapons that are underperforming, then that makes weapons that are not underperforming to strong against the same target if you lower the armour of the target then you get the same result, if you increase the damage or rate of fire then that makes the weapon to strong vs units since they are not as heavily armoured.

so realistically the only way this would work is if you where able to give each weapon its own unique modifier against units and its own unique modifier against structures, that affects only that weapon.

I guess this must be why the one third rule was implemented to avoid problems caused by using modifiers that affect multiple weapons.

Its a shame pumpkin went that way instead of using unique individual modifiers because unique individual modifiers would give you much more precise control over balancing, although admittedly that would require a bit more work.

plus due to the way the stats system is designed its not even possible to tune some of the modifiers to make units do exactly one third of there base damage.

so taking the machine-gun as an example 1/3 of the machine-guns damage is 150

((12*1.08)-12)*150 = 144
((12*1.09)-12)*150 = 162

as you can it's not actually possible to tune the modifiers to make the weapon do exactly 150 per min since it would require a modifier of 1.08333333333 and so far as i can tell modifiers need to be whole numbers.

so i guess in the end the best thing we can do is what we where doing originally and just set the various values to what we think they should be try them out and then adjust as required.

as always real world experience trumps theoretical math although both are necessary.

User avatar
Berserk Cyborg
Code contributor
Code contributor
Posts: 626
Joined: 26 Sep 2016, 19:56

Re: Help needed testing 3.2.x Campaign games!

Post by Berserk Cyborg » 27 Jun 2018, 17:32

So we can resume onto Alpha 6 then?

@alfred007
Could you try hold with 013daaf4a789c006bf249581aa0c247810955109? Should at the very least stop them from chasing something anymore.

alfred007
Trained
Trained
Posts: 457
Joined: 31 Jul 2016, 06:25
Location: Stuttgart, Germany

Re: Help needed testing 3.2.x Campaign games!

Post by alfred007 » 27 Jun 2018, 19:53

Berserk Cyborg wrote: @alfred007
Could you try hold with 013daaf4a789c006bf249581aa0c247810955109? Should at the very least stop them from chasing something anymore.
As long as the unit gets no new order they are holding their position. But even after a unit moved because of it tried to move another unit out of the way the hold order no longer works. I placed some units in alpha 02 at the entrance of the southern scavenger base and after a repair unit forced them to move they no longer hold their position. Better than in the master before but not even one-tenth as good as the old hold order.

Bethrezen
Regular
Regular
Posts: 585
Joined: 25 Sep 2009, 02:05

Re: Help needed testing 3.2.x Campaign games!

Post by Bethrezen » 28 Jun 2018, 08:09

Question: Why do we need modifiers when you can achieve the same effect by simply manipulating the armour value of the target?

Now I know that might seem like an odd question but bear with me.

At face value the answer to that question seems simple, we need modifiers because we want the same weapon to deal different amounts of damage vs different types of target, however the answer isn’t quiet that simple, because if it was then you could achieve that by manipulating the armour value of the target in question.

So for example say the machine gun does 12 damage per shot and of has a rate of fire of 150 which equates to 1800 per minute if we want that weapon to only deal 2 damage per shot or 300 damage per minute then you would simply set the armour value of the target to 1500

However there is a flaw in doing things that way say you are using a bunker buster that does 125 damage per shot and say you want the bunker buster to deal 100 damage per shot per shot vs a bunker but only 50 damage per shot vs a tower, in this circumstance you can't simply increase the armour of the tower because that would throw off other weapons making them weaker against towers than they should be, so therefore we need a modifier which allows us to manipulate the damage we deal.

So that’s case closed right?

Wrong!

Given all the back and fourth we have had about damage calculations and understanding the math involved and about balancing it got me wondering is there an easier way to go about this, which brings me back to my original question.

Why do we need modifiers when you can achieve the same effect by simply manipulating the armour value of the target?

Well as we have already established things are not quiet that cut and dry and we need modifiers to allow us to manipulate the damage we deal so we can't just get rid of them, HOWEVER .... !!! There is still room here for making things simpler.

Now currently the damage calculations look like this

Code: Select all

((damage x modifier) - armour) x ROF
The problem with this however is that it can cause weird calculations that would result in weapons doing no damage or negative damage, which then requires a fail safe in order to deal with situations like this and that fail safe is the one third rule which looks like this.

Code: Select all

damage x modifier x 1/3 or 0.333333333
However what if we flipped things around and instead of applying the modifier to the weapons damage, what if we applied it to the targets armour?

Now given that flipping the damage calculations round like this would require a bit of work to implement you may well be asking why would you want to do this ?

Well actually there are several reason you might want to do this.

1.) it makes the damage calculations easier to understand
2.) it makes the damage calculations easier to do
3.) it makes the damage calculations more precise
4.) You don’t have to deal with weird calculations that result in no damage or negative damage
5.) you don’t need the one third rule.
6.) Most importantly it makes balancing easier.

So what would this look like?

Well instead of the damage calculations being

Code: Select all

((damage x modifier) - armour) x ROF
or

Code: Select all

damage x modifier x 1/3 or 0.333333333
where the damage calculation would result in damage being less then one third of the weapons base damage.

It would become

Code: Select all

Armour = base armour x modifier

Actually damage dealt = damage x ROF - armour
So already that is much easier to understand

So what would the modifier look like? Something like this perhaps.

Code: Select all

Structure modifier.json
{
    "BUNKER":{
        "bullet":
        "shell":
        "rocket":
        "laser":
        "artillery":
        "bunker buster":
        "flame":
    }
        "HARD":{
        "bullet":
        "shell":
        "rocket":
        "laser":
        "artillery":
        "bunker buster":
        "flame":
    }
        "MEDIUM":{
        "bullet":
        "shell":
        "rocket":
        "laser":
        "artillery":
        "bunker buster":
        "flame":
    }
        "SOFT":{
        "bullet":
        "shell":
        "rocket":
        "laser":
        "artillery":
        "bunker buster":
        "flame":
    }
}
so what about vehicles

Code: Select all

Weapon modifier.json
{
    "legs":{
        "bullet":
        "shell":
        "rocket":
        "laser":
        "artillery":
        "bunker buster":
        "flame":
    }

    "wheels":{
        "bullet":
        "shell":
        "rocket":
        "laser":
        "artillery":
        "bunker buster":
        "flame":
    }

    "halftracks":{
        "bullet":
        "shell":
        "rocket":
        "laser":
        "artillery":
        "bunker buster":
        "flame":
    }

    "tracks":{
        "bullet":
        "shell":
        "rocket":
        "laser":
        "artillery":
        "bunker buster":
        "flame":
    }

    "hover":{
        "bullet":
        "shell":
        "rocket":
        "laser":
        "artillery":
        "bunker buster":
        "flame":
    }

    "vtol":{
        "bullet":
        "missiles":
        "shells":
    }
}
So how would this work in practice? Let's take a bunker as an example and let's use a bunker buster and a machine-gun since we know bunkers are strong against bullets but weak against bunker busters

let's also assume the following values

Machine gun
Damage 30
Rate of fire 107

Bunker buster
Damage 125
Rate of fire 10

Bunker
Health 300
Armour 500

Modifier

Code: Select all

"bunker":{
    "bullet":513
    "bunker buster": 1
}
Bunker Buster
Armour
500 x 0.1 = 50

Damage
125x10-50 = 1200 per min
1200/10= 120 per shot

Machine gun
Armour
500x5.13=2565

Damage
30x107-2565 = 645 per min
645/107=6.028037383 (rounded to 6) per shot

so you can see from this simple example doing the damage calculations this way gives you the same result as the current system but is way way simpler to work with.

now I'll grant you that implementing this would require a bit of work not least for the fact that currently the modifiers are hard coded instead of being loaded dynamically and of course doing this would brake the game and you would have to rebalance every weapon for both the player and the computer but ultimately we are already doing this anyway so maybe its something worth thinking about.

Bethrezen
Regular
Regular
Posts: 585
Joined: 25 Sep 2009, 02:05

Re: Help needed testing 3.2.x Campaign games!

Post by Bethrezen » 28 Jun 2018, 08:37

So we can resume onto Alpha 6 then?
I'll need to give alpha 5 another go because i was actually having a few issues now this may well be down to bad tactics on my part so I'm going to give it another shot and see how i get on.

one thing i can say though is that on insane at least you may well need a little more time because with the reinforcements coming from off screen plus units coming from the new scav factory and from the new paradigm factory that actually really slows you down when trying to get into the computers base coz you keep having to stop to deal with these additional units, i've also found that i need additional troops to help me deal with these additional units where before i could do the level with just 18 heavy machine-guns and 18 mortars or 36 heavy machine-guns

now I need minimum 3 squads but ideally 4 squads one squad of 18 light cannon to deal with the new paradigm units coming from of screen one squad of heavy machine gunners to park at the bottom of the ramp leading to the new scav base plus a squad of mortars to knock out there turrets and a squad of light cannon to protect the mortars from the units coming from the new paradigm base.

which of course requires time to transport in

now i can probably get away with not using mortars and instead just use 2 squads of light cannon and 1 of heavy machine gun or i can knock off the scav base and just use 2 squads of light cannon, but what ever way you go about it you are probably going to need a little more time.

oh one other thing i noticed the off screen reinforcements kept coming even after taking out the new paradigm command centre not sure if this is intentional or not but normally once you knock off the enemy command centre off screen reinforcements stop coming.

alfred007
Trained
Trained
Posts: 457
Joined: 31 Jul 2016, 06:25
Location: Stuttgart, Germany

Re: Help needed testing 3.2.x Campaign games!

Post by alfred007 » 28 Jun 2018, 17:02

You don't need so many units. I could destroy every NP ground reinforcement with 10 half-tracked light cannon units. You don't have much space to manoeuvre and if you use 18 units the risk is very high that you get your foremost unit too late back to repair. I destroyed the newly produced scavengers in the south with just four mortars attached to a sensor. Every time a unit reached the rally point I destroyed it. And for the NP base, I used 16 mortars attached to a sensor with a lot of repair units between the NP units and my sensor. In this way, the NP units attacked my repair units and I had enough time to kill one after the other from a safe distance. From the beginning of my attack at the middle scavenger base until the factory of the NP and all troops were destroyed I needed less than 10 minutes and had 16 minutes left to research everything. And except the last repair facility upgrade, I could research everything.
I agree that the ground reinforcements should stop as soon as the Command Center is destroyed but suggest then to move the artifact from the Command Center to the research facility.

User avatar
Berserk Cyborg
Code contributor
Code contributor
Posts: 626
Joined: 26 Sep 2016, 19:56

Re: Help needed testing 3.2.x Campaign games!

Post by Berserk Cyborg » 28 Jun 2018, 19:12

Reinforcements depend on HQ and the HQ artifact has been moved to the research lab on Alpha 5. Alpha 6 seems to work, as far as the new AI research tree and the player research. I have not played through that mission yet.
camBalance.wz
(110.26 KiB) Downloaded 31 times

alfred007
Trained
Trained
Posts: 457
Joined: 31 Jul 2016, 06:25
Location: Stuttgart, Germany

Re: Help needed testing 3.2.x Campaign games!

Post by alfred007 » 28 Jun 2018, 19:41

I'm also just waiting that Bethrezen said that he is satisfied with alpha 05 to move forward.

Bethrezen
Regular
Regular
Posts: 585
Joined: 25 Sep 2009, 02:05

Re: Help needed testing 3.2.x Campaign games!

Post by Bethrezen » 30 Jun 2018, 19:20

You don't need so many units. I could destroy every NP ground reinforcement with 10 half-tracked light cannon units.
The difference though is that I'm not using stupid amounts of mobile repair units I only use 2 with each squad.

In addition I'm not destroying the new scav base or the new paradigms command centre till my research is finished and I'm ready to end the level so that my units can continue to gain experience while I wait for my research to finish.
You don't have much space to manoeuvre and if you use 18 units the risk is very high that you get your foremost unit too late back to repair. I destroyed the newly produced scavengers in the south with just four mortars attached to a sensor. Every time a unit reached the rally point I destroyed it. And for the NP base, I used 16 mortars attached to a sensor with a lot of repair units between the NP units and my sensor. In this way, the NP units attacked my repair units and I had enough time to kill one after the other from a safe distance. From the beginning of my attack at the middle scavenger base until the factory of the NP and all troops were destroyed I needed less than 10 minutes and had 16 minutes left to research everything. And except the last repair facility upgrade, I could research everything.
Oddly enough after trying several different ways of doing this I've found the quickest and easiest thing to do is to just crush them with overwhelming force, and simply use 3 squads of light cannon tanks on half tracks for a total of 54 tanks supported by 6 mobile repair units 2 for each squad, that's a little over kill I'll admit 2 squads would probably be enough but what can I say the computer was annoying me :annoyed:

I tried using mortars but there just to slow and I consistently ended up coming up short by about 10 minutes, which left a fair amount of research left to be done.

More over the computer kept overriding my orders and my mortars keep charging in to enemy fire and getting destroyed when I tried setting them on guard to control the scav's because they wont obey the hold order when there attached to a scanner which is bloody irritating even more irritatingly for some reason the new paradigm reinforcements that come in from off screen kept ignore the units attacking there base and kept going after my mortars which is kind of odd, one would assume the new paradigm reinforcements would go after the units attacking there base but noooo the computer has to be awkward so the combination of those 3 things effectively renders mortars completely useless on alpha 5

Flamers aren't really much cop either, there to lightly armoured and don't inflict enough damage fast enough even with 3 damage upgrades so there probably going to need some more adjusting if they're to maintain there usefulness.

I think my only real gripe with alpha 5 at the moment is that you are more or less shoehorned in to just brute forcing your way through the level since you are kind of out matched, flamers are no help because they are to lightly armoured and not powerful enough and mortars are to lightly armoured and can't defend them selves which necessitates bringing in extra troops to protect them from attack which requires extra time you don't have which means they are to slow if you want to have enough time to actually complete all the research, which only really leaves heavy machine guns and light cannons as viable choices, but otherwise it's not to bad, although them medium tanks are a bit of a pain in the back side, as is the computers fetish for my mobile repair units, and the constant crashing wasn't doing much to improve my mood either.
Warzone2100.RPT
(237.63 KiB) Downloaded 24 times

Post Reply