Staredit Network > Forums > General StarCraft > Topic: Need testers to help balance my new map
Need testers to help balance my new map
Jul 4 2013, 4:45 am
By: Zacharee  

Jul 4 2013, 4:45 am Zacharee Post #1



I just finished the beta version of my new (first) map :) I'm looking for testers AND COLLABORATORS! Right now the terrain is only a rough mock-up of what I want it to be but I lack the artistic talent and patience to make it look great. My name on bnet is Zacharee (just like it is on this forum) I play mainly on USEast but my name is the same on USWest. If you are interested in helping out send me a message on here and/or add me to your friends list on bnet and send me a message. I'm on almost every night so bnet will probably be the easiest place to get in touch with me.

The map:
(haven't named it yet)

dimensions are 96x128
inspired by Kings & Knights
it is a top vs bottom meleeish (top team is terran, bottom team is protoss) type with a few twists.
unlike kings & knights where you kill critters for money, there are 2 zerg ai in the center area (which is 96x48) that duke it out with each other and you can kill their units for bounty. no bounty for killing the zergs buildings though i wanted to discourage people from wiping out the zerg to early. Also one zerg is allied to each team, however neither team is allied to either zerg so you can farm off either one.
all units and buildings are about 3 times more expensive to necessitate killing zerg as your main source of income. More precisely i multiplied the mineral/gas cost by 3, if there was no gas price originally i made the gas price equal to the original mineral price (for example workers cost 150/50)
Last twist, you can bring 6 of any unit (that has a hero version) to the middle for a hero, similar K&K. This is mainly the part that i need help balancing, as they are right now I think the heroes are a little on the weak side but i'm hesitant to make any changes until i get a few more opinions.

Again, I'm looking for testers who can help me make balancing decisions, since this is my first map and i don't know a thing about balancing. What i would REALLY like though is people who want to get in on the project and bring some fresh ideas to the table, namely new terrain because what i have now is so boring it gives me a headache.

PS I'm new here so idk if this is in the right topic



None.

Jul 4 2013, 3:51 pm Roy Post #2

An artist's depiction of an Extended Unit Death

Quote from Zacharee
no bounty for killing the zergs buildings though i wanted to discourage people from wiping out the zerg to early. Also one zerg is allied to each team, however neither team is allied to either zerg so you can farm off either one.
Couldn't you just make the Zerg building invincible for a certain duration of time? Would the players ever want to completely eliminate either Zerg AI?




Jul 4 2013, 7:11 pm MetalGear Post #3



I've been learning a lot about balancing over the past few months. I'm actually thinking about making a tutorial on it. But I'll give you a quick run down. Balancing, unlike most mappers think, is NOT A GUESSING GAME. From what I've learned, the 'trial and error' method of balancing is totally impractical and you'll never get the game balanced properly. Trial and error is okay ONLY for last minute changes, after you've done the correct balancing procedures.

So, the right way to balance is purely mathematical. You'll need a document to write a lot of figures down and a calculator (just use the one on your computer). Here are some factors you will need to take into account:

[attack speed]
[attack range]
[movement speed]
[unit size]
[damage type]
[unit abilities]

The best way then to measure the units is to draw two tables. The first gives the actual stats on unit traits. The second table assigns points based on those stats. Wherever the unit is lacking (for example, Zealot has 1 attack range), you assign points. Units with more points (in other words, better units) will receive less HP, attack, armor etc when you go to balance the units. It might look something like this...

[Table 1]


WWWWWWRange WWMSWWWDamage type
ZealotWWW1WWWWW6WWWwNormal
DragoonWW6WWWWW5.25WWiiExplosive

[Table 2]


WWWWWWRange WWMSWWWDamage typeWWWTotal points
ZealotWWW.17WWWW0WWWw0WWWWWWWWW1.17
DragoonWW0WWWWW.14WWw.25WWWWWWWW1.39

1.39 - 1.17 = 1.22

So as you can see from the tables above, you assign points to the unit wherever it is lacking. Then transfer those points to a percentage of HP, armor, or attack gain. In this case, the Dragoon's stats (HP, armor, damage) will need to be 22% better than the Zealot, because in theory (from this illustration) the Zealot is 22% better than the Dragoon. So suggesting the default HP is 200, the Zealot will have 200 hp and the Dragoon will have 244 hp.

Hope that helps.



None.

Jul 4 2013, 9:53 pm Zacharee Post #4



Quote from Roy
Quote from Zacharee
no bounty for killing the zergs buildings though i wanted to discourage people from wiping out the zerg to early. Also one zerg is allied to each team, however neither team is allied to either zerg so you can farm off either one.
Couldn't you just make the Zerg building invincible for a certain duration of time? Would the players ever want to completely eliminate either Zerg AI?

Ideally you wouldn't want to eliminate the zerg until the endgame when you're reasonably sure you have enough firepower to destroy your opponents. That being said, the zerg are considered your enemy so you have to kill them to win, but you can do that either before or after you destroy the other team.
What I intend to be a major part of strategy in this game is controlling the zergs rate of growth by killing off drones, overlords, various prerequisite buildings (spawning pool etc) So I don't think making the buildings invincible for any amount of time is a viable options unless its just their starting hatcheries. Each zerg starts with 4 of them, 4 overlords, and 12 drones so they build up pretty fast at first. When I was testing alone the first thing I would do is to kill off a few overlords and then the spawning pools when the finish. Because, being alone, it was far too easy for the zerg who wasn't allied to me to overwhelm me with zerglings very, very early. Although with a full game of 4 players this may not be a problem at least not until they get air.

I'm not sure if i mentioned this in the original post, but each human player starts with 8 workers, a base, and either a hero bc (1st tier, theres actualy 2 bc heroes) or a hero carrier, depending on their race. So one of the major issues i'm having in my public tests is one player always wants to go straight to the zerg and kill everything, so i may be forced to implement invincibility in some way. I'm hoping that if i manage to get any kind of community that likes and plays my game that this wont be an issue though since they'll understand that the early game isn't about brute force and destroying everything.

What i would like is to show you the map and get a more definitive opinion from you :) I'm hesitant to implement invincibility because, who knows what the best zerg control strategy might be? I don't. I could very well end up that killing everything except one hatch might be best. Only the beta tests will tell.



@MetalGear You're post went (mostly) over my head so I want to tell you what my goal is for balance and have you give some specific advice to address that. The idea is that i want the balance to be exactly the same as melee, meaning for the basic units (the ones you build) i won't be changing any stats other than the previously mentioned increase in prices, which are all uniform across both races so i don't THINK they will effect balance. The issue is the heroes, i want them to balance roughly the same as the basic units just 6 times better. correct me if i'm wrong but it seemed to me like simply multiplying all stats by 6 didn't necessarily mean they would be 6 times better. In fact it seemed like they were >6x better. Basically they way i did it (which i already know will change) was i multiplied their health/damage by factors of six, in such a way that the factors multiplied to six. so basically i multiplied one by 3 and the other by 2 depending on what felt right for each unit, but i've already had one public tester say that the heroes weren't worth the cost. also for units that didn't have an attack (as well as the carrier since he has the same old ints as the basic carrier) i just multiplied health by 6.



None.

Jul 4 2013, 11:32 pm Zacharee Post #5



Just finished my first definitive beta test (meaning people actually cooperated! ^.^)

there were a few bugs that i did notice though:

-when my teammate killed zerg units that i go via mindcontrol, the other team recieved money. Not a major bug since you should never have a reason to attack your ally. Thought it does present potential for griefing

-early in the game we killed a zerg hatchery and the computer that owned it froze up and stopped mining. Later though he was still producing combat units, presumably with money he got from killing the other zerg's units. I have no idea how to fix this, so any help would be appreciated.

-at various points in the game i would get bounty from units the other team killed, this is probably because my bounty system uses kill score and it doesn't account for killing units like larvae/eggs and such which do give a small amount of score. This will be easiest to fix once i start using scmdraft. (I'm using campaign editor only to outline the basic mechanics of the map, i do plan on using scmdraft for more advanced stuff)



None.

Jul 5 2013, 1:03 am MetalGear Post #6



Okay let's keep it simple. Just go with me on this. You need to find out how much damage a certain unit deals over time. This will give you a clear perspective of the unit ratios. For instance, let's use 30 seconds as a default comparison time. Suggest we compare two units, Zealot and Dark Templar. Now let's assume you want those units to deal roughly the same amount of damage, let's say 1000 damage in 30 seconds.


WWWWWWWWNumber of attacks in 30 seconds
ZealotWWWWiiii31
Dark TemplarWii24

So here's the math:


Zealot
1000 31 = 32.3 damage

Dark Templar
1000 24 = 41.7 damage

(1000 damage in 30 seconds divide by number of attacks equals damage.)

Now let's assume you want hero units to deal triple damage. You can use the same formula, but change 1000 to 3000 (which is 3000 damage in 30 seconds).



None.

Jul 5 2013, 1:12 am MetalGear Post #7



One more thing, if you multiply ALL stats by 6, that is mathematically flawed. Multiplying 2 stats by 6 times is actually making the unit 12 times better. Theoretically, assuming you are upgrading 2 factors, hp and damage, and you want the unit to be 6 times better than normal units, then you multiply hp by 3 and damage by 3.



None.

Jul 5 2013, 2:00 am Zacharee Post #8



Quote from MetalGear
One more thing, if you multiply ALL stats by 6, that is mathematically flawed. Multiplying 2 stats by 6 times is actually making the unit 12 times better. Theoretically, assuming you are upgrading 2 factors, hp and damage, and you want the unit to be 6 times better than normal units, then you multiply hp by 3 and damage by 3.

I think were on the same page regarding your first statement, I came to the same conclusion.

So basically what you're saying is that if i did something like hp*4 and dmg*3 the resulting unit would be 7 times better? As in the magnitude of improvement is equal to the sum of the factors you multiplied each stat by? How does this apply to units that have more than one attack? like the bc has a ground attack and an air attack that deal the same damage. When i used my formula for the first tier bc hero it ended up very OP compared to the carrier hero so i only multiplied the ground attack (especially taking into account the yomato cannon)



None.

Jul 5 2013, 9:16 am MetalGear Post #9



No, not exactly. You're nearly there, but you're still a little confused. You multiply the factors by the factors. So if you did hp * 3 and attack * 4, the resulting unit would be 12 times better (4 * 3).

Now, let me lay this out for you so you can understand how to apply it.

Let's say we want a hero unit to be 6 times (600%) better than a normal unit. Assuming a normal unit's [health = 100] and [damage = 10 p/s].

Let's keep this simple for now, and for our first hero, we'll just upgrade the hp. So, if a normal unit = 100 hp, then a hero unit will be 600 hp. Refer to table below.


hp =WWWW100 * 6 = 600
attack =WW10

Easy, right? Now, let's create a second hero. But this time, we'll add 400% to the hero's hp, and the other 200% goes to attack. Here's the sum...


hp =WWWW100 * 4 = 400

6 4 = 1.5

attack =WW10 * 1.5 = 15

Now, this may surprise you. You would have just doubled the attack to make the 200%, right? Nope. That would have made the unit 8 times better (4 * 2 = 8). If you don't believe me, go and create those units in a map. Use a normal Goliath and hero Goliath, and set those stats. They'll die at pretty much the same time.

As for the Battlecruiser question, If its air attacks are the same as the ground attacks, that's fine, no adjustments have to be made. However, let's assume the ground attack = 10 and the air attack = 20, then do this sum...


10 + 20 = 30 2 = 15

So you'll assume the unit deals 15 damage, and balance it from that figure. Now, as for the Yamato Gun, abilities are very tricky to balance. I won't go into the maths for this, its way too complex. You're best just to trial and error this. However, my quick-fix advice is to find out how regularly the BC can regenerates energy to cast YG. Let's assume every 30 seconds the BC recovers the energy for YG. YG deals 260 damage. So...


260 30 = 8.7

YG in this case would be equivalent to 8.7 damage per second. So if the enemy air unit is a Wraith, you'd add 8.7 damage to its attack (assuming it hits once per second).

Post has been edited 1 time(s), last time on Jul 5 2013, 9:43 am by MetalGear.



None.

Jul 5 2013, 10:36 am Zacharee Post #10



ohhhh ok i think i follow you now. but i think i feel like theres a contradiction in your explanation somehwhere.

you said,
Quote
You multiply the factors by the factors. So if you did hp * 3 and attack * 4, the resulting unit would be 12 times better (4 * 3).
so if i multiply hp and damage by 3 then the resulting unit should be 9 times better, so according to this my using the factors 2 and 3 was correct.[/quote]

before that you said,
Quote
One more thing, if you multiply ALL stats by 6, that is mathematically flawed. Multiplying 2 stats by 6 times is actually making the unit 12 times better. Theoretically, assuming you are upgrading 2 factors, hp and damage, and you want the unit to be 6 times better than normal units, then you multiply hp by 3 and damage by 3.
so going by your most recent explanation multiplying both stats by 6 would result in a unit 36 times better.

Whichever is correct i was happier with the results of multiplying both stats by 3. I feel that giving them a unit that is >6x better rewards the player for using heroes which is ideal for how i want the game to be played.

going back to my issue with the bc, if i left the air attack the same and only increased hp and the ground attack what would be the mathematically correct way (in terms of your explanation using percentages) to determine what factors i should multiply the other stats by?

do you by any chance know how long it takes to cast yomato? as well as the bc's attack speed? i'm trying to figure at what point yomato is no longer viable compared to just attacking, only considering dps and the yomato's dps.

Quote
we'll add 400% to the hero's hp
just a syntax correction, if you add 400%(of 100) to 100 you get 500. i think you mean to say "we'll increase the hero's hp to %400" ^.^



EDIT - i just made a simple map to test this with using ghosts and hero ghosts, the heroes ghosts attacking 6 basic ghosts 1 at a time. i found that the heroes closest to being truly equavalent of 6 basic ghosts were the ones that used factors 2 and 3. one hero that had hp*3 and dmg*2 successfully defeated 6 ghosts with only 7 hp remaining. the hero that had hp*2 and dmg*3 finished with 13 health remaining. the hero with both stats multiplied by 3 finished with what i believe was enough hp left to take on 3 more ghosts (i didn't actually test this), supporting my hypothesis that he would be 9x better. an interesting side note it appeared that the basic ghosts had +1 range over the heroes, however this may be because my heroes were approaching a stationary target and units in sc are sometimes slow to stop and attack.

Post has been edited 1 time(s), last time on Jul 5 2013, 11:49 am by Zacharee.



None.

Jul 5 2013, 11:15 am Zacharee Post #11



MetalGear is it possible that we can set up a time when i can actually show you the map? I feel like you're the right person to answer a few of the questions i have, but i don't want to ask them without you knowing exactly what the situation is.



None.

Jul 5 2013, 3:00 pm MetalGear Post #12



Sorry, ignore this...

Quote
One more thing, if you multiply ALL stats by 6, that is mathematically flawed. Multiplying 2 stats by 6 times is actually making the unit 12 times better. Theoretically, assuming you are upgrading 2 factors, hp and damage, and you want the unit to be 6 times better than normal units, then you multiply hp by 3 and damage by 3.

Go by my later post. I must have been half awake when I wrote that. You're right to pick up on the contradiction. And yes, multiplying both stats by 6 each will give you a unit that's 36 times better.

Battlecruisers hit 24 times in 30 seconds. So before I answer your question about the BC, do you already know what it's worth? I'm not entirely sure what you're asking. Have you already set the stats (hp/damage/armor) and just need to find out what the unit's worth compared to other units?

I'll show you how I'd calculate the BC.

First we need a unit of comparison. You should have already chosen by now default rates. Earlier I used 100 hp and 10 dps as default rates. So I'll use those rates again for this example...

Assuming the BC is a strong unit, I'll make it 900% of a normal unit.

Let's start by increasing the hp to 300%


100 * 3 = 300

Now the unit is 3 times better than normal units. Now we are left with (9 3) = 3. We can make the unit 300% better for the remaining stats.


Yamato Gun

260 30 = 8.7

8.7 10 = .87

The calculations show that Yamato Gun increases standard damage by 87% (standard damage being 10 per second, and YG dealing 8.7 per second).

So, 3 1.87 = 1.6

Now we have 1.6 (60%) left for the remaining stats. Let's find the damage.


10 * 1.6 = 16

16 damage p/s * 30 seconds = 480

480 24 = 20 damage
(BC's hit 24 times in 30 seconds.)

Here's the results of those calculations...


HP = 300
Damage = 20
Yamato Gun = usable every 30 seconds


And yes, its probably best we skype chat. I'll be free all weekend, but I live in Australia, I'm not sure about the time differences.



None.

Jul 6 2013, 4:02 am Zacharee Post #13



ya... i'm not sure which of my questions you're answering about the bc, at this point i've decided to regard yomato as irrelevant stat wise and treat it as the bc's counter to the carriers superior speed. as for my question regarding the bc dps and yomato dps, i'm simply trying to find the point that basic attacks are >= yomato dps. we'll if you're still on right now i'm about to log into bnet, tomorrow i'll be on about 6pm american central time. my name is Zacharee (just like it is here) on useast.

Post has been edited 1 time(s), last time on Jul 6 2013, 4:15 am by Zacharee.



None.

Jul 10 2013, 12:28 am Zacharee Post #14



sry MetalGear i accidentally left my sc logged in when i went to sleep when you tried to contact me. i just released the map in map showcase but i may move it to development



None.

Options
  Back to forum
Please log in to reply to this topic or to report it.
Members in this topic: None.
[09:12 pm]
IlyaSnopchenko -- Miau
[09:11 pm]
Suicidal Insanity -- "Open wide"
[07:24 pm]
IlyaSnopchenko -- Will do... Though only tomorrow i guess
[07:19 pm]
Pr0nogo -- check images.dat, compare that to other attacking buildings
[07:19 pm]
IlyaSnopchenko -- Guess I'll have to spend more time on this
[07:18 pm]
IlyaSnopchenko -- Yeah, so far it just... Doesn't attack :) even though even the button gets highlighted :)
[07:16 pm]
Pr0nogo -- experiment and see what works, i'm not making a new attacking building for a long time
[06:43 pm]
IlyaSnopchenko -- I just might need one for a map I'm creating now (though could perfectly live with a trigger attack, still undecided)
[06:42 pm]
IlyaSnopchenko -- So what would I need to create an attacking building, apart from editing the iscript, the command card and the building orders (and a couple of flags)?
Please log in to shout.


Members Online: Roy, Nekron, jjf28