Im wondering how most people do the calculations needed for balancing. Personally I use MS Excell to calculate how much damage all towers do in the TDs I make, and how many hitpoints the creeps should have to balance the difficulty of the map itself.
But I cant ever recall to have seen other map makers discussing calculation of balance, so I wonder how other people do this?|||It's more of an art than a science, there's no way you'd be able to account for factors such as buffs and area of effect damage in Excel. Excel is useful for calculating information for you to base your decision on, but you should never just take it as gospel.|||I dont do any crazy math, for I hate math lol. I just make the skill, do some thinking and adjusting untill I think its ok, then test with friends. All ballence issues relly pop up then. And after a few tests I do the changes I found or others mentoned|||Balance... what is that???
Oh, i just generally try it out one player. My current project is vs the computer so i can do it one player...|||trial and error, slloow and tedious testing over and over and over nadover and over andover adover and fefore jfdeire andover anveor avefo
see what happens after a while.
using testers and then debuging as well as the final balance
if there is an easier way id love to use that instead.
but at the endo fof the day you still going to need testers|||Make everything guessing what would be balanced. Play and tweak till everthing seems right. Like Taylor said, its an art, not a series of formulas.
|||i get sum1 smart to play with a hero and if he wups our asses it means he needs to be nerfed :Ptaylors my long lost good side
hehe|||It surprises me actually, that Im the only one (so far in this thread) that do any calculations. In a map with heroes and abilities I agree it would be hard to do any really useful calculations, but atleast when making a TD with lots of towers there can easily be poor balance, and doing calculations on towers damage is very useful as it requires alot less of testing.With a tower with splash you cant calculate the full damage, and you cant calculate accurate with a units attack range, but calculations of damage, speed and cost/damage can save you alot of time that you otherwise would need for testing|||just recently I was trying to balance a footman wars type game... first of all I started off working on an equation, saved it all in a text file, so I will post that here...
The main problem with it is things like movement speed/attack range which can't be put into an equation, so I was working with things to try to put them into perspective, if I can, err, balance the balancing equstion (O.o) then it should be sweet.
I worked at it by working out how long it would take for them to kill themselves (ie. if you have two units which are exactly the same, make them attack each other, they will last for X seconds until they both die, the theory is that if you can then use that value to balance the units... but now I'm just thinking that it would not work as if you have a unit which is exactly twice as strong in all ways, it would take the same length of time to kill itself... but still, I'm sure there is a way this could be changed around that...)
after that, I changed it so that it just ran a simulation (an isle, spawns the units at either end for different players, after 60 seconds it tells them to attack-move at each other)
-----------------------------------------------
-----------------------------------------------
posting that text file now, but beware, it was kinda like a brain storm, so you may have difficulty understanding it... XD
-----------------------------------------------
-----------------------------------------------
DPS = DAMAGE/RATE OF FIRE
DAMAGE TAKEN = .006*ARMOUR
VALUE=DPS*DAMAGE TAKEN*HEALTH*(1+(.1*RANGE/100))
DAMAGE=10
ATTACK COOL DOWN = 2 = .5 ATTACKS PER SECOND (1/ACD)
DPS = 10/2 = 5
ARMOUR = 2
DAMAGE TAKEN = .006*2 = 12%
HEALTH = 100
RANGE = MELEE (LET MELEE = 1)
VALUE = 5*.12*100*1
VALUE = 60 (12)
IF RANGE = 600
THEN VALUE = 36000...
------------WITH NEW FORMULA
VALUE = 5*.12*100*(1+(.1*1/100))
VALUE = 60*1+.001
VALUE = 60.06
IF RANGE = 600
THEN VALUE = 60*(1+(.1*600/100))
VALUE = 60*(1+.6)
VALUE = 96
WOOOHOOO!
IF DPS = 10
VALUE = 120....
IF DAMAGE TAKEN = .24 (4 ARMOUR)
VALUE = 120...
IF HEALTH = 200
VALUE = 120...
NOW THEN... I NEED TO ADD IN CONSTANTS TO MAKE IT ALL WORK OUT OK. BUT.... THESE NEED TO BE FOUND EITHER THROUGH TRIAL AND ERROR OR BY RUNNING EXPERIMENTS... EXPERIMENTS WILL NO DOUBT BE THE EASIEST WAY TO GO! (HAVE A FIELD WITH 2 UNITS WHICH ARE TOLD TO ATTACK EACH OTHER, THEY SHOULD BE 100% THE SAME, WITH NO RANDOMNESS IN THEIR ATTACK DAMAGE ETC.
ON SECCOND THOUGHT, UP ABOVE, A UNIT WHICH DOES 2*DPS GETS A RATING 2 TIMES HIGHER, 2* ARMOUR GETS 2* HIGHER, 2* HEALTH GETS 2* HIGHER... WILL HAVE TO TEST WITH ONE WITH 2* DPS VS 2*ARMOURE VS 2*HEALTH ETC.
VALUE=DPS*DAMAGE TAKEN*HEALTH*(1+(.1*RANGE/100))
LARGER VALUE MEANS A BETTER UNIT, IT SHOULD ALL WORK PROPORTIONALLY, GIVE THEM FASTER/SLOWER RATE OF FIRE, HIGHER/LOWER ARMOUR, HEALTH OR RANGE AND IT WILL MAKE THEIR VALUE TO UP OR DOWN ACCORDINGLY.
FACTORS NOT IN THIS EQUATION--
-DAMAGE TYPE
-MOVEMENT SPEED (IT DOESNT MATTER MUCH DOES IT? UNLESS THERE IS A LARGE DIFFERENCE IN RANGE THEN MOVEMENT SPEED SHOULDN'T MATTER MUCH AT ALL. IF IT DOES HOWEVER, THEN THIS WOULD BE DONE BY MULTIPLYING THE VALUE BY, SAY, .1*MOVEMENT SPEED OR SOMETHING.)
-SPAWN RATES - IF I USE 10SEC AS A BASE, THEN, SAY, 8SEC SHOULD BE VALUE = .8*VALUE.... BUT.... WILL THIS STUFF UP BECAUSE 2 UNITS ARE BETTER THAN 1, OR WHAT? I THINK I WOULD BE BEST TO WORK IT OUT AS IF ALL TEH UNITS ARE JUST ONE (JUST WORK OUT THE DPS AND HEALTH FOR HOWEVER MANY UNITS WILL SPAWN IN 10 SECONDS, THEN DIVIDE BY THE NUMBER OF UNITS AT THE END OR SOEMTHING...)
================================================== ================================================== ================================================== ================================================== =========================
NE-
spawn+
damage+
rate of fire+
armour-
health-
range+
HU-
spawn=
damage=
rof=
armour=
health=
range=
DPS
health/DPS*(1-armour) = how long you will live attacking a unit of your own kind. (1-armour should work out your real dps, ie. if you have .1 dam reduction, then you would be doing .9 of your damage)
*(1+(.1*range/100)) to compensate for any advantage due to attack range... BUT UP AGAINST YOURSELF ATTACK RANGE IS THE SAME ANYWAY... ERHM
VALUE=HEALTH/(DPS*(1-ARMOUR))
WILL HAVE TO TINKER WITH THE RANGE SECTION TO WORK OUT A GOOD VALUE FOR IT.
IF-
HEALTH = 100
DPS = 10/2 = 5
ARMOUR = 2 = 2*.06 = .12
THEN
VALUE=100/(5*(1-.12)) = 100/5*.88[4.4] = 22.7272 SECONDS (IE. YOU WILL LIVE FOR 22.72 SECONDS)
NOW THEN, LETS TRY IT OUT!
W00T! it was very close to 22 seconds! (about 20..) :D:D:D but I think the units were a bit bung due to sides per dice or something...
SO-
NE
IF
DAMAGE = 15
RATE OF FIRE = 1.25 (.8 ATTACKS A SECOND)
DPS = 15/.8 = 12
ARMOUR = 1 = .06
THEN WHAT SHOULD HEALTH BE TO GET A VALUE OF 22.72?
22.72 = HEALTH/(12*(1-.06))
22.72 = HEALTH/11.28
HEALTH = 22.72*11.28
HEALTH = 256.2816... doh! ITS SHOULD BE BLOODY WELL SMALLER THAN 100 DAMN IT! OK, ON CHECKNIG, IT APEARS THAT THIS IS JUST A PROBLEM WITH MY MATHS... LETS TRYI IT AGAIN... HMM, IT ALL SEEMS TO BE IN ORDER, HOWS ABOUT i CHECK IT|||For most maps, balance is hard to define. However, for TDs, there are a relatively small number of factors, so it is possible and often advisable to do at least some basic calculations. Personally, I made spreadsheets for one of my TDs as a starting point (for the other, it took me a while to realize I should use one) to get stuff more or less "balanced." Of course, there are always factors that are hard to quantize, and a spreadsheet is certainly not a replacement for good ol' testing, but I believe that the combination of the two allows relative balance to be achieved most efficiently. Spreadsheet alone presents problems if abilities are involved, or, even if they're not, it can be tricky to balance towers with multi-hit attacks (splash, bounce, line). Testing by itself can be very time-consuming, but its biggest problem is that the initial tests might have some very glaring issues that could have been prevented by analyzing the numbers. Also, there may be some towers that you think are "just fine" and that people don't use them because they don't know what's good for them, and since you rarely see them used, you might not realize that they are horribly underpowered. Of course, once you get those issues fixed, the tests become more meaningful. Just hope that you have people willing to test, regardless of huge imbalances. That isn't to say that with a spreadsheet you can't have huge imbalances; it's just less likely, if you're careful in how you approach it.
For non-TD maps or any map in which many variables are difficult to control or quantize, massive testing and good judgment is the most feasible solution.|||When I first did Taylor�s Hero Siege it wasn't well balanced, it wasn't particularly over or underpowered in any of the phases but there was no real difficulty curve, just a random assortment of difficulty jumps - I opened Excel and got a formulae for health, damage and bounty increases for the different unit types (melee, ranged and rare) and entered that - in spite of my efforts to balance units with spells the first go I found myself getting overwhelmed in the later phases and needed to manually edit a lot of units, then the bounty was way too much and it was the enemy who was getting overwhelmed, etc. In the end it was mostly balanced by hand but the Excel values made it a LOT easier. So I'd never do it by simple tweaking alone as some are suggesting, the simple formulas I had in Excel took 5 seconds and provided a solid foundation for my own balancing decisions.|||I make the thing, play the thing, then change the thing until it's balanced enough.|||The hardest part is balancing out the heroes spells. Especially if you change the amount of levels for each spell. This gets really bad like, in one of my test map I had a summoned bear which did 50-60 chaos damage 1500 hp and 2 hp regen with bash and blink. This thing was ugly.|||Quote:
The hardest part is balancing out the heroes spells. Especially if you change the amount of levels for each spell. This gets really bad like, in one of my test map I had a summoned bear which did 50-60 chaos damage 1500 hp and 2 hp regen with bash and blink. This thing was ugly.
Are you telling us you had trouble mounting your own bear?|||I've used unit balancer on this page a couple of times (http://maps.worldofwar.net/misc/damagecalc)|||i do pagan rituals until the blood of the goat i sacrificed forms words on the floor, works for me
-The evil
is everywhere-|||Quote:i do pagan rituals until the blood of the goat i sacrificed forms words on the floor, works for me
-The evil
is everywhere-must be an NZ thing.
*drys off knife*|||Quote:
must be an NZ thing.
*drys off knife*
where are you from?, and change you avatar back to AAAHHH!! that was coo|||Quote:
where are you from?, and change you avatar back to AAAHHH!! that was coo
I'm from NZ.
as for the avatar - the mouse has had its turn, I have three other new ones to go through first... then no doubt I will be making new ones :P|||I do rough calculations in my head, do lots of multiplayer testing, and make adjustments accordingly. I'm sure using Excel wouldn't hurt, though.|||I don't very think, so I let excel do it for me.|||Well on my map there also needs to be some balance, cause the hero could get very strong and to not let the creeps be to easy I just did some random calculations in my head and then tested the thing, then found out they were still to weak ... and so on, cause it is very hard with hero baced maps. If you want an AOS kind of map you could easilly make the units as strong as eachother (same skills and stuff) and then let the player have the opportunity to upgrade the units (with money or smth) cause then it is certainly balanced in the beginning, and after a while you would get the upperhand (also you could make really strong hero's but the problem with that is that it will get just slaugtering at a certain moment.
So it really depends on the map what kind of calculations and stuff you need to do. I never made a TD (mostly because there are so many and I want to make new stuff, Although I play them soith a Tmetimes) and I agree that with TD you should/can use calculations wisely.|||So far every map I've made/started has all players equal from the start of the map's creation, so I don't really need to balance.|||I use map balancer, I import the units, and then I let them mindlessly collide, muhahahawa
.Don't have time for spilling goat blood, which is probably the next best thing :y-sneaky: ;).|||Quote:
I'm from NZ.
as for the avatar - the mouse has had its turn, I have three other new ones to go through first... then no doubt I will be making new ones :P
Im from nz too :Y Hello:|||Quote:
trial and error, slloow and tedious testing over and over and over nadover and over andover adover and fefore jfdeire andover anveor avefo
see what happens after a while.
using testers and then debuging as well as the final balance
Exactomundo. Working on balancing out my multiplayer RPG campaign between hero stats, skills, and starting equipment, then adjusting monsters to match, and THEN increase monster difficulty compared to which heroes were picked and how many... essentially it'll be a different monster set up every time you play as long as you choose diffrent heroes, it'll adjust to the types of heroes you have, late game it'll almost all be monsters who'll be effective against your heroes, so it'll be a bad idea to have all of one sort of hero, because you'll be owned if you do. o.O
The hardest part to make/balance will be the magic system, initially I had presets like everyone else, but now I've got a completely different method that'll be based off of what day it is, each hero will (hopefully) have a different magic type, though that's not too certain right now... it depends on how many heroes will end up existing in the game. The 'level' of magic will be determined by Int, about one level per 25 Int (until I find a better method, perhaps a skill system for both magic and weapon skill, but the whole project is on hold until I get my computer back from home)
But yeah, balance is a pain, I perfer the trial method... makes it harder to screw up that way when -other- people are playing.
- PyroticTiger, 1337 Cat|||I compare moves to melee and make them like them accordingly. It's all just guessing for me.|||Balance in hero maps is making sure both sides have the same skills in some form or another .... wait?!
Really what I do in any game (since I haven't done a map in war3 before this project), is make sure armor, reduction, defensive survival abilities have equivalent ratios to direct damage spells and summons over a 10 minute period.
Somehow you have to factor in that melee heroes will eventually always kill everyone at the end while leaving them playable in the beginning. My major issue is accounting for summoners as their pets can be focused early on & die easily making it difficult against good players.
I was always shocked by how great any hero can be against casual players & how easy it is for experienced players to smoke 2 or 3 players.
That's what I strive to avoid - no one guy should ever be able to take on 2 players and kill both.
No comments:
Post a Comment