Balance?

What is the best way to go about balancing the stats of enemies and the player characters to make sure it's challenging but not impossible? I have always struggled with this aspect of creation?

---

Happy RPG Makin

Thinking of this, a article for The Hamster Burrow on this subject might be a good idea!

---

Happy RPG Makin

Find out how much damage your heroes can do as a unit, from there, you can figure out how many turns you want to go by until you can slay an average enemy. This is also true for boss battles. Use this same way of thinking for how to set up enemies. How much damage can your heroes take before they are killed? Set your enemies up accordingly.

Another factor is, how much health is the player/enemy going to get restored in the fight? That can also take into account how much health to give enemies or how much an enemy should be able to dish out. You don't want to make a boss fight with a boss that restores 1/2 or more of their life if it takes your heroes forever to do that much, that's not fair to the player, as not many people want to spend that much time fighting a fight just for it to all be undone. But if you can power up heroes to make that kind of move mean nothing then you're on to something. I try to design fights and battles around the limitations of my characters and then you use all of this to not try and make good fights, but how to use your characters to the best of their designed capabilities.

These aren't questions for you to answer to us, these are questions for you to answer to yourself, as this is your game. Hopefully this helps.

---

Check out Red Triangle Games!

To be honest, I have always wondered what sort of techniques people use to balance their games. It always seems like the most magical part of game development, especially for RPGs. If anyone wants to write an article about it, I think that would be great, especially any sorts of math that you do.

I think considering the amount of turns that a battle should take is a great starting point. Pokemon seems to rely on this, with evenly matched 1v1 battles between monsters taking between 3-5 rounds in my experience.

I will go about describing the way I can figure things when I'm not so tired, I'll let you know now that it will take a lot of time and figuring to make it more and more refined and well balanced.

---

dOn'T MiNd mE! i'M jUsT CoNtAgIoUs!!!
Play Orbs CCG: http://orbsccg.com/r/4r6x

It's a fairly complicated issue, but I use a spreadsheet to keep track of it. The important thing to note is that most of your hero's stats can be determined arbitrarily; the rest of the values such as enemy HP, enemy ATK, enemy DEF, and so on, can all proceed from this, but you have to begin with at least some of your variables defined. It is better to do this with hero's stats since there are generally fewer heroes than enemies, and are easier to make broad adjustments to. The important thing here is that the hero's stats should make sense IN RELATION to one another even as they are arbitrary; you are sort of thinking conceptually at this point still, instead of quantitatively.

To try and keep the matter as simple as possible (maybe I'll write a longer article on it later?), assume you have four heroes, and four enemies to stat. The four heroes are Strong Hero, Fast Hero, Magic Hero, and Balanced Hero. You might arbitrarily assign ATK values to each hero thusly:



The actual numbers are unimportant, so long as they make sense in relation to one another: Strong is the strongest, Fast is about half as powerful as Strong, Magic has a weak physical which is about 20% as good as Strong's, and Balanced is a bit better than Fast, but still not as powerful as Strong. This spread kind of holds true for both L0 and L99 stats. Makes sense, right?

From there, you would use the default OHR stat-gain formula to create a column in your spreadsheet showing you what ATK power each hero will have at each level. This will be important in the next step.

OK, with the hero ATK powers defined, we can stat the first enemy. Assume that you have four main dungeons in this game, and one enemy in each dungeon. The enemy in the first dungeon is RED SLIME, and you decide that you're going to want to balance him for a level 1 party.

You look at the level 1 ATK values for your party, and they happen to be:



You can then average those together to know what kind of damage your party is likely to output on a basic turn when everyone uses a default attack. The average value in this case is 50 + 25 + 10 + 35 = 120 / 4 = 30 damage per turn average.

Now, ask yourself: how many turns per round do you want the enemy to survive for? Let's say that a single RED SLIME should take one round to recover. So, its HP should be about 30.

But wait, what about its DEF? The default DEF formula in the OHR is that 50% of DEF is subtracted from average attack to determine actual damage dealt. So, in this case, we want a defensive value that, after it soaks up part of hero damage, still allows 30 damage per turn to get through. It's okay to be slightly arbitrary here as well. Let's say that RED SLIMES have 4 DEF. We then calculate:

x - .5(4) = 30
x - 2 = 30
x = 32

So, change RED SLIME'S HP to 32 and his DEF to 4.

Now, define your hero's HP in the same arbitrary fashion, but where they make sense in relation to one another:



Again, use the default progression formula to give yourself a column that shows what the HP values will be at any given level. Since we're still dealing with a Level 1 monster here, take the HP values for level 1 from each hero, and average them. You end up with a value of 235 / 4 = about 60.

Do the same with DEF values for all the heroes. Let's assume the average DEF value for L1 heroes is 5.

Now, you can ask yourself the same questions, but in reverse. How many times should a RED SLIME be able to hit a hero before killing him or her? Let's say the heroes should be able to withstand about 6 blows from a RED SLIME before falling. So RED SLIME's ATK should be (remembering to factor in DEF):

x = (60/6)
x = 10
x = 10 + .5(5)
x = 12.5
x = 13

So, if RED SLIME has an ATK of 13, about 2.5 of it on average will be blocked, and he'll be hitting for 10-11 damage each turn. The average party HP at L1 is 60, so that's 6 hits.

RED SLIME's stats are now shaping up:

HP: 32
ATK: 13
DEF: 4

You would set up MAG and MDEF in the same way described above. SPD should be a fairly consistent value across the board, with most enemies being about as fast as your balanced party member.

As for ACC and DODGE, again you just set up your hero's ACC in proportion to one another arbitrarily, decide how often you want the enemy to be hit, then use the OHR's to-hit formulas to figure out what a given enemy's DODGE needs to be again, against the average hero ACC, to achieve that hit %. I won't go into the specifics here since the process is similar to above and since the OHR has so many different to-hit formulas possible. I will state, however, that for the most part, I choose to use straight percentages, so that the calculation for hitting is just ACC % vs DODGE %. It's much easier to deal with that way.

Now, there's the question of EXP. Suppose that you want your party to naturally reach about LV3 by the time they get to the end of the first dungeon with the RED SLIMES. Figure out about how many battles you want this to take, and it's a simple formula once again:

The amount required to go from level 0 to level 3 is 226 in the default OHR experience formula. Multiply this by the number of heroes, to get 226 * 4 = 904. So you want the total experience given by all the battles you fight with the RED SLIME to equal about 904. Let's say we want to get there in 20 total battles.

904 total exp / 20 total battles = 45.2 experience per encounter.

Assume there are an average of three slimes in each encounter:

45.2 / 3 = about 15.

RED SLIMES should give 15 EXP.

Gold can be calculated in the same way; figure out about how much gold you want the heroes to have by the time they're done, subtract any major gold chests from it, divide by the total number of encounters you want them to have to fight, divide again by the number of monsters in the average encounter, and there's your GOLD per slime.

Note that this same process easily extends to statting the other enemies now. Let's say that once you're level 3, you're going to be going on to the 2nd dungeon and fighting ZOMBIES.

We use the same formulas as above. Average hero attack at L3 == 50, ZOMBIE should take about 1.5 turns to kill, ZOMBIE's arbitrary DEF is 8.
50 * 1.5 = 75 (1.5 turns of damage)

x - .5(8) - 75
x - 4 = 75
x = 79

ZOMBIE has 79 HP.

Average hero HP at L3 = 80, ZOMBIE should be able to kill an average hero in 4 blows, hero's arbitrary average DEF is 10.

x = 80/4
x = 20
x = 20 + .5(10)
x = 25

ZOMBIE should be have an ATK for 25 to be able to hit for 20 damage per turn after DEF and kill the average hero in 4 turns.

You want to be level 7 by the time you leave zombie dungeon? It should take 30 battles? There are 3 zombies on average in a formation?

Total EXP to go from 3 to 7 = 1410 * 4 = 5640

5640 total exp / 30 total battles = 188 / 3 average # of zombies = 62. ZOMBIE should give 62 EXP.

-----

This may seem like a complex way to go about doing things, but this is why you need to learn to use a spreadsheet. With the spreadsheet, you can set up columns for the hero's stats from L0 to L99. Then, make columns for the enemy's stats and add columns for things like "how many turns to kill", and "how many hits will kill an average hero", and "what level will the party be at when they fight these". For any given monster, you will only need to fill in those additional columns, and everything else can be handled by automation, by simply plugging in the formulas in question. This will allow you to stat an entire game's worth of enemies in relatively little time, and was the method used in both Spellshard and Dungeonmen.

A few things to note:

- The "average hero attack" value isn't going to take into account special attacks and magic. For the most part, this is okay, as these techniques are limited in use, and when you use them, they *SHOULD* feel as if they are outside the bounds of the standard calculations.
- That being said, bear in mind that heroes will invariably use their special kills a great deal against bosses. As such, boss HP should be calculated with this in mind. Add some to the average hero ATK value to compensate for the use of magic and techniques, and then calculate the boss HP from that.
- Be careful when dealing with average hero HP that if you have any extremely low or high outliers, they will likely be able to absorb far greater / fewer blows than the actual average. For instance, if the party's average HP at a level is 100 and the monster does 25 damage because it should kill the average hero in 4 blows, this generally works well. However, if your hero's actual HP stats are like 25/100/100/175, bear in mind that this means that the hero with 25 HP will be killed in one blow, whereas the one with 175 HP will sustain 7 before killing instead of 4. Just something to keep in mind.

---

SPELLSHARD: THE BLACK CROWN OF HORGOTH now COMPLETE! Grab it today!

A couple quick comments:
Naturally, if you are going to use a spreadsheet similar to what Master Harlock recommends, it is VERY important that you have columns to allow for equipment.

I wrote a series of long and confusing articles on this subject for Hamsterspeak, where I tried to point out something rather different than what Harlock is putting forth. Proper balancing of battles, to me, depends on having properly defined roles for the different characters. The stats are important, but they are easily adjusted through playtesting. What's more important is finding a balance between what sorts of things the different heroes should be good (and bad) at. Any given hero should be good at more than one function, so that he/she is rarely useless. The counterexample is the pure damage-dealer. Dealing damage is (almost) never useless, and so you can have a character be good at nothing but that.

Then once you have roles for your heroes, you make enemies designed to necessitate the use of the various roles. If you have a healer, make sure there exist enemies who do enough singular damage to make the healer useful. If you have a hero with a slowing-down spell, make sure you have enemies fast enough, and survivable enough, to make this spell worthwhile to use. etc etc

At the end of the day though, I think the playtesting is the biggest key. Any spreadsheet will employ many assumptions. Every assumption that is fulfilled will actually just make your battles more 'boring'. Why base random battles on the assumption that the heroes' damage output will be based on them all attacking? Have an enemy that can put a hero to sleep! Have a hero that can boost atk/def in battle! etc etc It would have to be an impressive spreadsheet indeed to somehow account for these occurrences. Playtesting would work just as well, it seems to me.

---

I am Srime

Yeah, definitely. More important to a fun game is, like msw says, having random encounters that require strategic thinking and necessitate the use of all those nice spells and abilities. It should go without saying the use of metrics and spreadsheets is just to give you a basic foundation from which to develop, and to ensure you don't end up with game-killing things like common enemies that pose no threat but take ten turns to kill, &c.

---

SPELLSHARD: THE BLACK CROWN OF HORGOTH now COMPLETE! Grab it today!

For complex systems such as those found in most RPGs, balance is an emergent process and it's nearly impossible to account for everything. This is why you see exploits like the infinite profit crafting in FF9, and it's why you see glitches like Ayla's 9999-damage hit breaking some battle scripts in Chrono Trigger. (But at that point, you're already level 99, so who cares?)

It's easy enough to make the numbers work at a small scale. Harlock described a good method. But as the game goes on, it's tougher to predict the player's power level. Some RPGs, such as FF8 or Oblivion, deal with this by having the enemies scale with you. It's usually a terrible idea. Others, like FF13, gate your growth so that the battles can be designed around a limited set of skills. Most of them, though, just don't care -- if the player wants to grind to level 99 before fighting the first boss, more power to him. (Literally.)

And this method works because it allows players to play at their own pace. If your battles have any sort of nuance to them, then skilled players can go straight through, while less skilled players can eventually scrape by if they beat up enough dudes first.

If you want to have a difficult, interesting RPG, then, two ingredients are required:

- Some amount of tactical depth; and
- Limited rewards.

Tactical depth means giving the player options that make a difference. Some options should be more effective in some situations, others more effective in other situations. No option should act as a silver bullet, or else the player will just spam that and feel cheated when it doesn't work.

Limited rewards means that the player shouldn't feel overleveled or have excess funds if he just goes straight through the game without stopping to kill more dudes. It's easy to do this by just giving insufficient rewards, but then your game is a grindfest. It's important to find a middle ground.

And that is my brief treatise on RPG balancing. Hope that's helpful to someone.

---

Mega Tact v1.1
Super Penguin Chef
Wizard Blocks

Play your game. A lot.

In Skyfall, I've been balancing each section by literally playing it dozens of times. I try running from most fights, fighting around half of them, fighting all of them, grinding, etc. too see how difficult encounters (especially bosses) are and tuning them to suit each kind of player.

In Silhouette there is no randomization and I know exactly how powerful the player will be at any encounter, so I can use a lot more math rather than pure testing.

A different process is needed based on different systems in your game, though. If your game has only level-ups, you might see what level range a player might be by each part in the game, and then tune your enemies based on how much damage you want your players to take at each encounter. If you factor in rewards/healing, and encounter rate, you can get an approximate value for boss fights, as well. (Different design choices here, as well: do you expect a fully healed party or no?)

If you have lots of options, playtesting might be your best bet. Do the obvious stuff, then try the weird stuff that players might do. For example, try every party combination your game allows in the fight.

Of course, not all commercial games even do all this... RPGs can really be balanced by the fact that you can always gain levels or find some "silver bullet" to finish an encounter.

Alright, the meow has slept, and it's time for a breakdown on his way of figuring balancing.
=====
Step 1: Figure out how big of numbers you want for your game
=====
When making an RPG, do you want your numbers to be large? As in, your attacks start off doing around 100-200 damage a hit and end up doing 99999 by the end of the game?
Or perhaps you desire something more simple, like Paper Mario where you do 1 damage with a regular attack for about half the game, and have a small assortment of skills and items to do a bit more, like the fire flowers that do 4 fire-based damage to all foes, or the pow blocks that do 2 but flip certain foes.

The big the numbers, the more chaotic and a little less manageable it will be, meaning it'll take a bit more work than normal. But, inversely, as easily manageable as low numbers are to work with, just giving something just 1 more point of attack is a drastic boost in power.
=====
Step 2: Hero Points and gear
=====
Once you have figured what number size you desire, figure out a generally rounded system for how you want points to be distributed.
I'd do something, for instance, of saying "I have 700 points to distribute between Attack and Magic Attack, and I have 600 to put between Defense and Magic Defense", something like that, for heroes that is, as it'll make your heroes a little more balanced themselves, where you don't have one all-round ALMIGHTY SMITERMAN!...unless you want one.

Now, assuming you want equipment in the game, I'd build them around hero stats as well. For instance, your hero gains about 6 points of defense total per level (that's including both physical and magical), you could make it where each tier of equipment is about a half-level in growth, being that your level 1 hero will have 6 defense points, and your first armor you can obtain has 3 defense points, distributed as you please.

Does this make sense? You can do the same with weapons. You can figure other special things for gear by, say, reducing some stats (like drops in speed) for more power, or perhaps using one of those points stated for elemental resistance.

=====
Step 3: Enemies
=====
This may be the tricky part, but do not get overwhelmed! I advise basing your enemy strengths pertaining to your heroes stats and levels. Figure with some math as to what your hero will have for attack, defense, and HP at the time of encounter. Now, keep in mind that normal attack calculation on the OHRRPGCE is "Attack vs 1/2 defense of target", so unless you choose other calculations, imagine the amount of your defense as 1/2 of what it really is. This is how much damage they will block on average.
Now that you've figured these things, you will also want to consider whether or not players will have the availability to use healing items and skills. Perhaps status buffs as well. Now you can build foes up off of this.
Start around making sure how damage you want a foe to inflict, for example:
Heroe has 50 HP, 20 Atk, and 20 Def, and he has some heal potions that restore 10 HP, and no status changing skills.

Squibbles has 15 Atk, so he will do about 5 hp on average, so Heroe can only survive his attacks about 10 times, but since he has heal potions, he can heal up after an encounter.

The Angry Washrag has 25 Atk, so he'll do about 15 damage average. Heroe won't be able to last too long on his lonesome, but he can at least survive a battle.

Terrifying Mr. Munchie has 40 Atk. He does about 30 damage an attack. There's so no way Heroe can survive without some status effects.

For this instance, Squibbles seems to be the ideal choices for, say, an overworld area or lengthy dungeons. The Angry Washrag will probably be good as a midboss or warning to a player to not head into an area too soon, and perhaps to train beforehand. Terrifying Mr. Munchie is far too powerful to handle. You will want to lower his attack, for sure. Or you could save him for a later area.

Now that the hard part is overwith, you'll want to make sure to see that your hero will have a relatively simple time dispatching weaklings and have a somewhat harder time dealing with the more powerful foes.

Heroe still has that 20 Atk, and he only has normal attacks, let's say.

Squibbles has 12 HP and 20 Def. You'll do 10 damage on average, but if you level up you'll be able to dispatch them in one hit! This is good for a weak foe.

The Angry Washrag has 21 HP and 20 Def as well. This is a slightly more challenging foe, but he should still be able to be dispatched with relative ease.

Terrifying Mr. Munchie has 70 HP and 35 def. You'll only be doing about 2 or 3 points of damage on average. Again, this is not something you'd want on a foe with such HP.

But doing little damage isn't always a bad thing!
Let's say there's a Glimmering Squibbles, a rare form of squibbles.
It, too, has 35 Def, BUT it has only 8 HP. It will last only about 3 or 4 turns. These type of creatures in rpgs can typically flee and give good exp if defeated. Defeating such a foe feels really rewarding, especially since it usually is.

=====
In conclusion of this stuff, though, there are a ton of little factors that you as the designer can choose to deal with, from special types of attacks, to elemental properties. As Mystic stated:
This will help greatly in the ways of figuring things out. Especially since mathematically calculating how things should go isn't a 100% fail-proof tactic, it'll at least help you shave off the grand amount of fat surrounding the deliciously balanced layer of beef underneath.

I hope this helps! - Meowskivich

---

dOn'T MiNd mE! i'M jUsT CoNtAgIoUs!!!
Play Orbs CCG: http://orbsccg.com/r/4r6x

Oh this is all such brilliant material. I think I'm going to start digging through the forums to find other helpful threads and make an article with links to them. I would hate for this stuff to get lost in the noise.