The numerical balancing for a mmorpg (or diablo like arpg) is very complex

I think such topic is rare in this forum or gamasutra.

I am still learning it, and get some understandings. The calculation is complex, because tons of data have been linked togher.

The core of balancing is stable. So which thing should be stable? Fighting cycle time not genius should be stable.

For instance, there is a base charactor, and has HP, MP, AD (physical attack), AP (magic attck), MD (physical defence), PD(magic defence), Hit, Miss, Cri. The player has 60 levels and 4 careers, and different crarrer has different skill ( so different cooldown time)

fighting cycle time = equivalent HP / equivalent DPS
Equivalent HP = charactor HP / ( (1 - damage avoidance rate) * ( 1 - evasion rate) )
You can find many formulars on how to calcualte equivalent DPS.

By far the process is not easy. And equipments and skills can change equivalent HP and equivalent DPS greatly. In MMORPG, equipments always play the bigger role on attributes contribution.

There are 11 types of equipments, such as helmet, necklace, ring, armor.......
Each equipement has 60 levels (or more or less), and each level has five or more colors ( or more, color stands for norma, rare, lengdary...), and each color has 11 qualities ( biigger quality means more enhancement).

Some types contribute to HP, and some contribute to AD...so on so on.

Spreadsheets have been used for these data heavy calculation. You need to make sure the player's fiighting cycle time maintains stable in different level with different equipments and skills.

You have to make more than 50 worksheets to cover above calculations. Finnaly you use VBA to get a simulation.
You choose two careers from 4, and set level, then make simulation. The fighting cycle time should be stable.



I think it is really hard. I feel difficult on how to organize it. And there a lot of details in the process I don't cover, such as how to set the inital data, and how to define the data of euipqments and skills. Maybe somebody can explain more clearly than me.

 

So we're back to this? and an mmo no less..

I really thought you changed direction with your last thread, oh well.. where the popcorn?

 

Here we are. I like how this will go.

@hongwaixuexi Where are you stuck? 50 spreadsheets it's a piece of cake.

I think it will be more fun if you add some Graphic (for example the Flag for this faction) in the Faction column.

 

@hongwaixuexi I appreciate your threads. They actually get people talking. (I say this tongue in cheek but also seriously)

I agree that balancing levels, XP, and skills is a very complicated process and it seems that most games do it over time with an intuitive approach (buff this, nerf that). I'm looking forward to what others think about this.

As for leveling up with XP, you want to have ONE spreadsheet page, with ONE formula, and keep things as simple as possible. The KISS principle.

 

In MMOs they keep fiddling even after the game ships. No matter how much developers try to plan ahead, I guess it's difficult to predict how people will play.

 

Create graphs. Visual aids are far more effective for the human brain than a column of numbers.

 

Dang, I really hoped I was going to agree with the OP on this one. I agree with the title only. Yes it is very complex, but not for the reasons stated.

The problem is a balanced game is a boring game. Balance weakens the impact of player choice. If everything is balanced then there are no unexpected strategies, no real surprises. You'll either have everything able to handle everything (meaning your choices don't matter at all), or a "rock, paper, scissors" approach. With the latter, the player isn't really making choices though anyways. They see rock, they bring more paper, they see scissors, they bring more rock. Boring boring boring. Players will enjoy the game for a while, then once they have quickly mastered rock, paper, scissors, they walk away.

What you really want is to purposely make the game out of balance. Let the players figure out which tactics, which items, which anything, are overpowered. Let that sit for a while until the players have it basically figured out, then you release new content which reshuffles the balance. At that point the players are still playing the same game, but they are somewhat starting over at figuring out how to min/max the game, figuring out what is overpowered.

For example, maybe you have some ship which is objectively better than all others in many common situations. Players figure that out, lots of players focus on that ship, and just when they start to get bored you add a new ship which is a direct counter to the overpowered ship. This starts a whole new cycle of experimentation. Is the old overpowered ship still good at anything? Is any past content in conjunction with the old top dog a good counter to the new ship? Months of debate on what the best strategies are to counter this new threat. When the players finally figure it out, and the game starts to get stale again..... you reshuffle the deck once more, rinse/repeat.

The game is never balanced, the players are always learning the game, and the game doesn't get boring.

 

HP calculation as an example.

First you need two inputs. The startingHP and growthHP per level.
Starting HP 360
Growth per Level 18

A base model is calculated.

Code (CSharp):

1. Level    Base
2. 1    360
3. 2    378
4. 3    396
5. 4    414
6. 5    432
7. 6    450
8. 7    468
9. 8    486
10. 9    504
11. 10    522
12. 11    540
13. .........
14. 60   1422
15.  

Then you need inputs again. You need to set how much HP contriuted by different color equipments. 60 levels multiply by 5 colors. So you need to input 300 numbers (%). Inputs always need double check, because they are intuitive.


Then reaplce % with real numbers.

We get HP contribued by equipments. Then we can assign this HP to different equipments.

You need inputs how many equipment can affect HP and how much proportion for each equip(total should 100%).
Suppose only shoulder protection and belt contribute to HP.


Then
We get shoulder protection and belt value on HP. The equipment numbers are linked HP system this way.

The equipment also has enhancement quality, and they can increase HP as well. So you need to input how much promotion due to enhancement.

Code (CSharp):

1. Equip Quality    Promotion
2. 0    100%
3. 1    105%
4. 2    110%
5. 3    120%
6. 4    130%
7. 5    150%
8. 6    170%
9. 7    190%
10. 8    220%
11. 9    260%
12. 10    300%

It's obivous shoulder protection has 60 x 5 x 11 = 3300 different values. Even every 10 level, we sill get 6 x 5 x 11 = 330 different values.

Finnaly you can get a HP vs level graph like this.


Then you need to input career multiplier for HP. Wizard has less HP than Solider is a common sense.


Then you get HP for different careers.

Code (CSharp):

1. Level    Base    Solider    Wizard    Pastor    Assasin
2. 1    360    360    306    317    324
3. 2    378    378    321    333    340
4. 3    396    396    337    348    356
5. 4    414    414    352    364    373
6. 5    432    432    367    380    389
7. 6    450    450    383    396    405
8. 7    468    468    398    412    421
9. 8    486    486    413    428    437
10. 9    504    504    428    444    454
11. 10    522    522    444    459    470
12. 11    540    540    459    475    486
13. 12    558    558    474    491    502
14. 13    576    576    490    507    518
15. 14    594    594    505    523    535
16. 15    612    612    520    539    551
17. 16    630    630    536    554    567
18. 17    648    648    551    570    583
19. 18    666    666    566    586    599
20. 19    684    684    581    602    616
21. 20    702    702    597    618    632
22. 21    720    720    612    634    648
23. 22    738    738    627    649    664
24.  

Now HP system is OK, and can be used later.

 

In China, there are some game designers working on numberical balancing, and they have their own title - numerical designer.

I want it simple as well, but unforunately it is still complex.

 

Could you give some examples? Which unbalanced game is interesting?

 

In RPG keep balancing is very hard. If you don't do balancing calculation, then it will be out of balance automatically. So purposely is not necessary.

 

But is it due to the number having a real meaning? Or is it merely that they believe in it and want to prove to themselves that they weren't mistaken in their belief? There is a healthy belief in a concept and then there is an unhealthy one where you seek out ways to justify it when no evidence exists.

https://en.wikipedia.org/wiki/Chinese_numerology

Confirmation bias is an example of a bias where you seek out proof where none exists.

https://en.wikipedia.org/wiki/Confirmation_bias

Weird tangent we got on...

Tyrian was a very well received topdown shooter from the time before Windows. You had a great variety of weapons but they were the most unbalanced mess you could ever imagine. Some were borderline worthless while others were way too powerful for the ease of acquiring them. Yet the game was and remains very popular with people who like that genre.

https://en.wikipedia.org/wiki/Tyrian_(video_game)

 

Numercial designer won't care 13 or 4, and they focus on mathematical model of their game.

 

You'd be surprised. There are days that Asian companies won't ship a product because the number is linked to something else in a negative way. That Chinese numerology link from earlier provides an easy example. The number four is unlucky because the way it is pronounced is almost identical to "death".

 

I am not a phychologist. I can't help you on this.

 

We prefer to 6, 8 than 4, and you can see 4 everywhere in China. That's all.

 

My recommendation is to start learning the basics. You don't need to become a qualified psychologist but having a basic understanding of how humans think is far more valuable than those numbers you love to claim represent emotions.

 

I didn't say numbers reprent emotions in this thread, though I agree varying numbers reprent emotions. But it not the point of this thread, I want to discuss how to balance MMORPG correclty.

 

RPG. Within 10 years examples will be much better.

 

Playtesting. Ideally with a range of different playstyles (casual, roleplay, min max, etc).

I have no idea what you're trying to say here, but I provided an example of a game with little to no balancing that is fun.

 

The mehod I talked here is very useful for games whcih have RPG genre such as LOL, Warcraft, Fallout3 or 4, Borderlands. What will happen if above games aren't balanced?

 

You can't balance MMORPG just by playtesting, though it is helpful, but far below the requirements on how to balance it.

 

Funny how the person asking for advice is telling everyone else that their advice isn't correct... if you think you already know why are you asking in this thread? Playtesting is more than just sending beta keys to everyone. You can start as early as the prototyping stage. Have people play the game, get analytics, and adjust accordingly.

Just keep in mind balanced isn't the same thing as fun. A very well balanced game that isn't enjoyable in the slightest is worse than a barely balanced game that everyone loves to play. Remember you're selling a game. Not a list of numbers.

 

I have an emotion now. I wonder, will this game be fun?

 

Pretty much every MMO which has survived for more than a few years. Take Eve Online for example. They routinely release new content and at the same time make often large adjustments to existing content. This shifts the relative power of pretty much everything in the game. The players theorize on the forums and in their own alliances, test out new fittings and fleet compositions, eventually a new meta forms to max out the power with the latest changes. A few months after the new meta settles down CCP releases a new update and starts the process over again.

The same thing happens in the Magic The Gathering card game, same thing happens in the X Wing miniatures game. The same thing happens in World of Warships. On and on. It is pretty much standard practice with any long running games, including examples above that aren't even video games.

The thing is, a lot of the fun in these games is where you're trying to figure out what are the best strategies, what are the best items, etc. If the game is balanced then you're robbing your game of a lot of the fun, since there are really no right or wrong choices outside of the immediate situation.

Here's another example, PUBG. That game is hugely out of balance on purpose. There are guns that are objectively bad all the time, which are mostly just fun for challenging yourself (crossbow for example), gun power levels are all over the place and spread randomly. Plus there are box drops which include superior weapons to anything that can otherwise be found. This basically ensures that when you encounter another player, one player will have a significantly better load out than the other for situation. And that is all part of what makes the game fun. Pulling out a win in the face of superior firepower is exciting. You balance the game and you rob it of that.

You're really over thinking this stuff. If all your numbers were off by 30%, or you had small jumps mixed with large jumps, it really won't change the game all that much. The players will adapt to whatever these values are as long as they aren't completely broken.

 

This post and the coffe it's yummy. So, let's say we have a Harley Davidson Game. Then you start out
with at PURPOSE unbalanced Harley to challenge others in the Desert Road. When you reach there,
as a Player, with this Motorcycle your grandfather given you as legacy, you test this Motorcycle (sorry my english)
and this Motorcycle sucks. So you keep trying and this Bike sucks. An NPC appears an says to you: Dude,
if you buy my Bike (or Motorcycle, lol) you will go faster. Finally, you will need many $$$ to buy that vehicle.
So, that takes you to find job in the City. BUT! When you enter the City nobody knows you and there's little
chance you get a good Job. Then, you can decide based on Options. Do i still try to get a decent Job
in this City? (an NPC can guide you again and induce you to take actions), for example: Dude, you can find
a job in this place. Or another NPC tells you: If you Join our guild we will give you a Motorcycle for free,
but you know, you need to do some bad jobs for us.

Now, that has some emotion going on: Ethics.

Finally, put the Statistics and Core Mechanics in there: If my Motorcycle goes at 250km-h, some parts
of it will get deteriorated (for example, the engine) and you will need to repair it later. That will cost you some
more $$$.

If you go enter another Region of this hypotethical game, maybe you find some wolves.
Now, you open your Inventory and you found you have a knife and you check your HP (?).

Here we'll have a problem. To a Player this technical writing (hp) can mean nothing or it
will think: oh, this are health points, of course. (my native languaje as you see is not english
so i can confuse when i see just : HP Points). Then, if there's CONFUSION in the UX,
emotion will be just one: zzzzzzzz.

My grain of salt to game design.

SOOOOO, what this has to do with the thread? To do a Diablo or Ultima GDD and write the rulez of the game or whatever it's too complex?

To do a game it's too complex already to add more layers of complexity than you will
really need.

 

It will create fun if the gamer like the feeling of keeping growing, levelling up, collecting weapons and armors, jewels.

 

It's discussion, so don't mind different ideas.

For gamers, it's a game. For designer, it's a list of numbers.

 

Yes. This i understand. What about the AI? An MMO will need AI too.

 

Their relationship is the focus. StartingHP can be 100, 200, 300 and other numbers, and growthHP can be 10,20 and other numbers. But the inital data are linked into the whole system. Given lead time, It will affect the number of affect damage.

 

I will be more fun if the AI can read this system and take actions based on this data.

 

If your AI means monster data, then yes.
There are monster spreadsheet and boss spreadsheet, and they are all balanced. 60 levels, monster HP, AD, Hit, Miss....Their figting cycle is steadily growing when levelling up, not fixed at one number.
Monster numbers are also calcualted.

If your AI means FSM or similar things, it belongs to programming work, not designer.

 

Uhuuh! Looks tight!

 

Yes. I was thinking like an AI Chess works. It looks for best movement to defeat the Player.
But i think it's more coder stuff than game design stuff. And i dont have any clue if it's
possible to do that.

 

I don't know either. I am still learning it.

 

There are no 100% balanced MMORPG, and I think the gamer can even find a dominant strategy in 99% balanced game. But that's OK.

80% balanced game or even 99% balanced game doesn't mean unbalanced, though it make the balance unstable.

 

Well technically a 100% balanced game is impossible unless everything is identical, but I assumed you were referring to making it as balanced as possible. What I'm wondering though, is why you even have a goal of making a balanced game? Why is that something you think is a good thing?

In stories for example, people love to root for the underdog - overcoming a superior foe. The lack of balance is one of the main reasons why the story is interesting. In gaming it is little different. Pulling out a win when the cards are stacked against you is much more exciting than if the odds were even. And a loss when you were at a disadvantage doesn't really hurt that much, because you can justify it in your mind as you were supposed to lose. So an out of balance game can bring players higher highs and softer lows compared to a balanced game, as far as the player experience.

Here's a gaming specific example of what I'm talking about with Starcraft. The game is notoriously out of balance. Lets say you play Terran and your opponent plays Zerg, and does the classic Zerg Rush. The Terran player is expected to quickly lose. When that happens the Terran player thinks nothing of it, doesn't really hurt. He/she just jumps into the next game. But when the Terran player successfully fends off a Zerg Rush and turns it around for a win, well that is really exciting. Something he's probably going to tell his friends about, because the out of balance nature of the game says he overcame almost insurmountable odds for a victory. The game is out of balance, yet very popular. Pretty much every popular game is out of balance, which is much of where the fun comes from.

 

Rock, Paper, Scissors... Checkmate!

 

Really?
Rock Paper Scissors Society publishes: rock is chosen 35.4%, paper 35% and scissors only 29.6% of the time.

 

I don't think you fully grasp the idea of balance. It has nothing to do with people choosing a tool or not. It's about the availability.

Exactly. And I think these kind of solutions are more interesting then the plain "balanced" ones. I mean if you have strengths and weaknesses and every tool is better or worse against something. So everything has its best tool to go against it. Instead of everything is balanced and have the same chance on the long run against everything else. Not the primary reason, but one reason is it is more realistic.

 

How to calculate damage?

damage = attack * ( 1 - damage redcution rate);

damage reduction rate = armor value / ( armor value + 100 * level + 400);

100 and 400 are initial data, and can be modified later.

Input base armor value and armor value growth per level.
base armor value =20;
armor value growth =1;

Then get a table for 60 levels

Code (CSharp):

1. Level    Armor value
2. 1    20
3. 2    21
4. 3    22
5. 4    23
6. 5    24
7. 6    25
8. 7    26
9. 8    27
10. 9    28
11. 10    29
12. 11    30
13. 12    31
14. 13    32
15. 14    33
16. 15    34
17. 16    35
18. 17    36
19. 18    37
20. 19    38
21. .....
22. 60 79

Similar, we need to input contribution due to equipment.

Code (CSharp):

1. Level    Armor Value    Equipment
2.         White    Green    Blue    Purple    Orange
3. 1    20    300%    360%    450%    600%    780%
4. 2    21    300%    360%    450%    600%    780%
5. 3    22    300%    360%    450%    600%    780%
6. 4    23    300%    360%    450%    600%    780%
7. 5    24    300%    360%    450%    600%    780%
8. 6    25    300%    360%    450%    600%    780%
9. 7    26    300%    360%    450%    600%    780%
10. 8    27    300%    360%    450%    600%    780%
11. 9    28    300%    360%    450%    600%    780%
12. 10    29    300%    360%    450%    600%    810%
13. 11    30    300%    360%    450%    600%    810%
14. 12    31    300%    360%    450%    600%    810%
15. 13    32    300%    360%    450%    600%    810%
16. 14    33    300%    360%    450%    600%    810%
17. 15    34    300%    360%    450%    600%    810%
18. 16    35    300%    360%    450%    600%    810%
19. 17    36    300%    360%    450%    600%    810%
20. 18    37    300%    360%    450%    600%    810%
21. 19    38    300%    360%    450%    600%    810%
22. 20    39    330%    405%    525%    705%    930%
23. .............................................................
24.  

Replace % with real numbers.

Code (CSharp):

1. Level    White    Green    Blue    Purple    Orange
2. 1    60    72    90    120    156
3. 10    87    104    130    174    234
4. 20    128    157    204    274    362
5. 30    176    220    294    396    514
6. 40    238    300    389    531    708
7. 50    310    414    621    931    1345
8. 60    474    711    1066    1540    2370
9.  

Then assign the numers to different equipments. Helmet, armor and pant contriute to armor value.
So (green color for example)

Code (CSharp):

1. Level    Green    Helmet    Armor    Pant
2. 1    72    33.0%    35.0%    32.0%
3. 10    104    33.0%    35.0%    32.0%
4. 20    157    33.0%    35.0%    32.0%
5. 30    220    33.0%    35.0%    32.0%
6. 40    300    33.0%    35.0%    32.0%
7. 50    414    33.0%    35.0%    32.0%
8. 60    711    33.0%    35.0%    32.0%
9.  

Then get total armor value

Code (CSharp):

1. Level    Charactor Armor value     Equipment Armor value    Total value
2. 1    20    60    80
3. 2    21    60    81
4. 3    22    60    82
5. 4    23    60    83
6. 5    24    60    84
7. 6    25    60    85
8. 7    26    60    86
9. 8    27    60    87
10. 9    28    60    88
11. 10    29    87    116
12. 11    30    87    117
13. ............................
14.  

Now the final important step is to calcuate damage reduction rate because armor value is ready.
damage reduction rate = armor value / ( armor value + 100 * level + 400);

Code (CSharp):

1. Level    Self    Equipment    Total    Redcution Rate
2. 1    20    60    80    14.81%
3. 2    21    60    81    14.10%
4. 3    22    60    82    13.46%
5. 4    23    60    83    12.89%
6. 5    24    60    84    12.38%
7. 6    25    60    85    11.92%
8. 7    26    60    86    11.50%
9. 8    27    60    87    11.12%
10. 9    28    60    88    10.77%
11. 10    29    87    116    13.06%
12. 11    30    87    117    12.68%
13. ..............................
14.  

The reduction rate will be used later.

 

So how do you deal with balancing? Just ignore it or use advanced tools?

 

Imma just watch this thread from afar this time.

 

Welcome to the club
*hands popcorn*

 

Ok, so here's what theoretical balance in a game looks like:
Rogue beats Mage
Mage beats Warrior
Warrior beats Rogue

This is the same as rock paper scissors in that different units have different weaknesses and strengths.
Imbalanced means either than something is out of this loop completely or that something has more advantages than others; conversely, something that is too weak is also imbalanced.

Now how do you determine if something is balanced or not? In simple Game Theory terms, you playtest and see if there are degenerate strategies or Dominant strategies. An example would be that rogues beat mages but can escape/bleed to death the warrior. At that point, the dominant strategy is to kill mages and run from warriors. Warriors cannot beat rogues and mages beat warriors. So this is imbalanced.

But the only way to know if to test it. You can always gather trends, see who kills who most and so on but at the end of the day, you do not know if your tweak will give rise to another dominant strategy or if you have shifted the imba sideways because of a nerf.

So the answer: play test...

 

I'm too weak. I failed.

 

Which direction, left or right? I hope forward.

 

Pace. It's a cycle time, kill monster or npc cycle time. You spend 30 seconds to kill a monster, 30 senconds is the pace. If you spend 5 minutes to kill a monster, while 3 seconds to kill the boss. It is called unbalanced.

You kill many monsters, and they drop very few rare-quality equipments. You kill elites and bosses, and they drop rare-quality equipments, it's called balanced.

 

In design phase, how do you conduct play test?

 

Why talking about Nash equilibrium in balancing design? No one talk about the formulas or the spreadsheets.