Duelyst Forums

Some balance methods

#1

This post will be as objective as possible, and for that porpuse, i am using a standar balance method.
Based on the comments, i have to make some things '‘clear’'
The method i am about to use is NOT omnipotent, it has flaws all over it, but helps to keep in chek the balance at a base level.
To chek if a card is balanced, the sum of health, attack and spell must be apropiate for the mana cost. For example:
-Lightbender: a cost 4 minion. Lest’s chek if he is balanced.
He is a 3/3, so it has a total number of 6, but he has a spell, wich is ‘‘dispell ALL spaces around it’’.

How much does this spell add to the total number? Well, if we take Sunbloom, a 2 cost spell from lyonar, we can chek how much dispell adds.
So, this card can dispell a 2x2 area, 4 spaces. So, dispelling a single space has a value of 0.5 mana, for it to grant a total mana cost of 2 (4 spaces x 0.5= 2), the cost of said spell.
Take into acount that mana cost is not the same that the total number.
For example: total number of a hailstone golem is 10, the add of 4 attack and 6 health. That doesn’t mean he costs 10 mana, he costs 4. Total number is just a number used to value the cards, taking into acount his stats and spell.

Now, we continue with the lightbender, he can potentially dispell 8 spaces, so, the spell adds 4 to the total number. (8x0.5)
Let’s add that 4 to the 6 (from the add of health and attack), and the total number is 10. Wich is a normal number for a 4 cost minion, like the hailstone golem.

That’s the method I will be using to chek other cards.
-Kindred hunter: this card if completely unbalanced.
The stats are: 3/3, so his total nomber is 6. But if we add his spell, ‘‘summon a 3/3 night howler on a random nearby space’’.
So, we add 6 +6, and the total number is 12¡
Of course, this is a bond ability, but that doesn’t change the fact that this is not apropiate for a 3 cost minion, this number is even higher that the one for a 4 cost minion like hailstone. Even close to some 5 cost minions.

-Bloodshard golem: a standard 3 cost minion.
His stats are 4/3, so the total number is 7, this number is ‘low’ for a 3 mana cost minion.

-Silverguard knight: a 3 cost minion.
Stats: 1/5. -total number: 6
Spells: provoke and zeal (gains +2 attack) -number: 2, because provoke has a value of 1 and zeal has 1 too, because of the conditions for it to be usefull, he has to be close to the general.
Also, stat spells has less value that normal stats, because those are dispelleable.
So, the total number of this minion is 8, or arguably 9.
Good, but balanced taking into account that this is a faction minion.

-Primus shieldmaster: a 4 cost minion.
Stats: 3/6 - number: 9
Spell: provoke- number: 1
Total number: 10, the standar for a 4 cost minion.

-Saberspine tiger: a 3 cost mana.
Stats: 3/2- number: 5
Spell: rush.
This spell has a high value, adding 3 to the total number, fot it to be balanced.
Total number: 8, the standar.

  • Punish: abyssian spell that cost 2.
    Now, this should cost 4, the same as entropic decay, why?
    Entropic decay destroys a minion close to the general, ‘‘close to the general’’ is a condition.
    ’‘destroy a minion’’ is the effect.

Punish has a condition and the same effect, why should it cost 2 less?
The condition is, ‘‘damaged minion’’, and we know that’s easy, or equally difficult as ‘‘close to the general’’.
This spell should cost 4.

There are a lot of cards that are high on the balance curve, and some other low.
Please check that.
And don’t be afraid to make changes, if you don’t change things, the player base will get pissed and stop playing.
ONE MORE THING: i don’t manage english as my main languaje, so please forgive my grammar mistakes.
Hablo sobretodo español.
https://www.youtube.com/watch?v=Bxszx60ZwGw (a video donde by a designer talking about this)

#2

Read until ‘Kindred Hunter’ where you forgot that you need to have an Arcanyst in play for it to proc.

If you take that into account and evaluate it as 50% modifier he becomes a 9, idk if it should be 50%, but totally disregarding that made me stop reading.

#3

I know math guys don’t believe this, but not everything in this world can be solved with an equation.

4 Likes
#4

Ok, if you want to reply, please be objective as well, i forgot the bond part for kindred, but that is not a reason for his stats, feelings are not part of balance.

And baharoth, that is a method used by developers, not a magic equation.

#5

I do love me some numerology

“Add double a 2 mana spell to 3 attack and 3 health to get 10 units (of what? who cares) which is a good number for a 4 mana minion.”

#6

what are you talking about? Please be more specific, this post is not perfect, and i said ‘‘my main languaje is NOT english’’, so i can change a lot of grammar issues.

Also, ‘‘10 units (of what? who cares)’’. A unit doesn’t have to be specified it there´s no terminology to define it. And in this case, there’s no specification for this method. So i gave just call it ‘‘total number’’, there’s no more i can do.

#7

It’s ok, your English isn’t the problem. I understand you were trying to apply an equation to weird scenarios and the numbers don’t come out right. My criticism was that you said a 3/3 body + 8 tiles dispelled (which you estimated was equal to 2 Sunblooms) would equal a 4 mana package, when in fact by your thinking it should equal 6 mana (2 for each Sunbloom + 2 for the body). However, there are many aspects that you do not consider: Lightbender’s range and versatility is far lower than that of 2 Sunblooms, and 1 card is always better than a combination of 3 cards because they cost less card advantage.

What I think is that you were too interested in making Lightbender’s features add up to 10 (the stat total of Hailstone Golem) and forgot that each card’s viability and balance depends on its relative value at different phases of the game.

#8

Thanks, you are right.
As i said, this is just a method, a form to help balance.
And i can’t take into account all, like card advantage, versatility, or that the card can dispell allies.
But thanks for your interest, and for being polite.

Oh, and that character is from nisekoi, right? xd

#9

the issue I have with this kind of number based balance system is that it is not great, for example, what is stronger, a 2/3 or a 3/2 both with no ability? Everyone with any knowledge of this game will tell you that the 2/3 is better due to the general not being able to take it out in one hit. for an in game example, compare thorn needler a 4 mana 6/4 to hailstone golem a 4 mana 4/6, hailstone is a significant;y better card and sees a lot of play in gauntlet because it has a better stat distribution.

The other issue I have with this system is how do you value abilities? you say provoke is worth 1 because then primus sheildmaster is on par with hailstone, but what about lantern fox, aymara healer, grandmaster zir and other cards with interesting effects? how do you chose what number they match up to? the answer is you can’t without being baised by your own experiences and so this model which you have built to be objective falls down.

Tl;dr this system is bad for working out how strong a minion is due to ignoring stat lines and abilities having numbers which will vary depending on who you talk to

4 Likes
#10

I’m afraid this equation doesn’t provide answers for balancing cards.

With regards to minions, stat distribution plays a powerful role in determining what is played and what isn’t. Take Skyrock Golem for example. To stay objective, this minion is practically never played despite having 5 total stats like other common 2-drops like Healing Mystic and Primus Fist. And if a minion was ever introduced with a 52/1 stat line for 7 mana, do you think people would run it? It has more stats than cards like Spectral Revenant and EMP, but an argument can be made that it is too susceptible to pings to be playable.

Cards are also balanced based on other cards available in the faction. Imagine Songhai with Tectonic Spikes, or Zir’an with Kelaino. After experiencing Rift Mode, we know quite well why certain cards are confined to certain factions.

It’s a nice effort to try to make a universal formula for balance changes, but I’m afraid it’s unrealistic to think a single formula provides all of the answers.

1 Like
#11

An objective way to balance cards… sound good in theory. Let’s see…

Except… Lightbender, on its own, will never dispel more than 7 “enemy” spaces because he needs to be played adjacent to either his general or an allied minion.
So he should be a 9.5, not a 10.

  1. A 6/6 minion is not the same as 2x 3/3 minions.
  2. The “Bond” keyword can’t be neglected. Because without Bond, Kindred Hunter is a 3/3 for 3 mana - which is weaker than Bloodshard Golem (as “number” for a 3 mana minion should be 7).

So a faction minion may go 1 or 2 points beyond the benchmark number and still be balanced ?

Returning to Kindred Hunter and evaluating it’s conditional nature (which is significantly harder to fulfil than “Zeal”)… let’s take it played with and without Bond and use the mean to calculate it’s “number”…

3/3 --> 6
2x 3/3 --> 12

(6+12)*0.5 = 9 - and 9 is a good number for a 3 mana faction minion…

So… is Kindred Hunter unbalanced ?

<<…>>

Unless you present a list with the value of each and every special ability, this “method” of balancing cards doesn’t work. (And even with such a list, a minion’s stat distribution is relevant - e.g. Phalanxar or the aforementioned difference between a 2/3 and a 3/2 minion.)

It can be somewhat useful to roughly gauge a card’s power, but that’s it.

#12

Everything, just rarely one a supercomputer can.

#13

Howdy!

While the methods is not wrong per say, there are too many variables to consider while considering these cards for something like this to work. From interactions with other cards and effects, to un-quantifiable abilities, there are just to many ways a card a slip through this system. It’s like trying to modelling a biological system with newtons laws; technically it could work, but there’s no use trying.

Some general rules:

  • snowball cards are extremely hard to eqaluate. They can be insane like owlbeast sage, or garbage, like grow cards and black locust
  • cards that combine effects are more valuable than the individual effects taken alone
  • the power level does not scale linearly with mana cost.

Examples to counter your theory:
Nimbus:
5 mana 3/8,
When damaged, summon (a 0/4 obelisk that destroys anything that damages it)

For 5 mana (expected 14-15 “points”)
You get a 11 point body and an average of 2 obelisks that would see play at 3 mana, so I’ll assume 8 points each. Added together, you get 27 points. However, nimbus rarely sees play due to its slow and snow-ball nature, on top of being weak to removal.

Makantor Warbeast
6 mana,
4/4 rush + frenzy

Assuming rush is worth 3 points and frenzy is worth 2 (diretide frenzy), Makantor Warbeast is worth a measly 13 points. That’s only 1 more than kindred hunter, so surely it’s trash, right? Well, the keywords play off of each other very well, making this the best 6 drop in the game

These examples are pretty extreme, but I think you get the point; qualtatative point systems tell you buggerall about a card’s actual power. A better way to look at cards is using the good old data analysis and look at performance over a large number of games.

#14

According to your math, Fire Spitter is broken. Quick refresher for everyone who forgot about this card, it’s a 4 mana 3/2 with ranged. It’s stats are a 5, which is half of the recommended value. However, every time Spitter attacks, that’s effectively a free cast of Phoenix Fire. It can cast infinite Phoenix Fires, so it’s “number” is infinite. Spitter is literally the most broken card possible, I guess (Astral Crusade is broken too, as it has an infinite amount of GF casts).

1 Like
#15

Everything can be solver by an equation.

If the equation its optimal enough to be solved by humans methods on a acceptable time its another topic.

#16

Wouldn’t kindred hunter be a 3/6? Since it can do 3 damage and take 6 damage. Sometimes. Depending on how the damage is distributed. Even if you could boil down all the abilities and stats for a more basic card into one number, for a card like kindred hunter you’d need a probability distribution where it can be anywhere between a 3/3 and a 6/30 (like if it ate 2 excelsious swings).

#17

Ok, thanks for all the coments, i see why i am wrong.
But let my say, this is just a method, not an absolute solution. And yes, it cannot be applied to instantly balance ANY minion, there are MANY thing to consider, this just helps the base level of balance.
Also, for everyone saying that kindred hunter is not OP, i know they are just saying that based on a subjective opinion, because they use that card, and have grown attatched to it.
To show that, i will have as an example:
-Orb weaver: a 2/2 for 3 mana.
Spell: summon a copy of this minion on a nearby space.
So, the total number is: 8. And that is high, compared to bloodshard golem.

Are you really telling me that ‘bond’ reduces 3 from the total number?
Take into account, that bond is not made to be used in a single turn, you should stablish another minion before that turn.
And, having two 3/3 in the table, whit only 3 mana is crazy (if you use bond correctly). You can have a very good advantage early in the game.

And, let’s compare it to a 5 cost minion.
-Brightmoss golem:
Stats: 5/9, total number: 14, only 2 more than kindred.

-Dancing blades:
Stats: 4/6- total number: 10
Spell: deal 3 damage to ANY minion in front of this.
So, this could potentially add 3 to the total number, but it has some conditions and negatives, condition: ‘in front of this’, negative: ‘any minion’.
While this are ‘easy’ things to avoid, it definitely changes the way to play it.
So let’s say it adds 2 to the total number.
Total number: 12, or maybe 13.
The exact same as kindred, or 1 plus.

Thanks for the comments, i appreciate it.

1 Like
#18

what you are saying is that 2 3/3’s is the same as a 6/6 right? which is completely wrong, at most it is the same as a 3/6, as the amount of damge per health is the same. there are obviously differences with single target removal and aoe affecting them not the same, but you get the idea. a poorly stated 3 drop with the potential to be an ok 4 drop for 3 given a certain board is reasonable.

1 Like
#19

It is not reasonable, you cannot play two 2/3 with 3 mana, even less two 3/3. And, it is better to have two 3/3 than a 6/6. Because there are many single target removal spells in this game.
And, one 3/3 can be buffed it it survives, wich is easy to archieve.

But i know you just like that card, and you will not concede to the facts.

#20

And what about playing three 2/3 with 5 mana ?