Homebrew system help

[Intregus182]

Playground I need some advice/direction!

I have been making my own RPG and I have character creation down, the basics of the game complete and now I’m building monsters to test against player characters to balance everything so I can come up with a formula for how to create monsters.

First let me just say this part of game design never crossed my mind and I am finding this to be challenging.

What is the best way or are some good ways to run numbers to help me figure out the balance of monsters vs PC’s?

My current strategy is cumbersome but its the only idea I have. What I’m currently doing is creating multiple different kinds of player characters at level 1 and then running them through “white room” combat encounters to see how much damage PC’s are doing vs monsters, how long combats go, how difficult is a level 1 monster vs a level 2 monster for a level 1 PC to fight etc.

And then I’m thinking I need to level up each PC and do it all over again etc, etc, through level 20.

There has to be an easier way than this. What am I missing?

Any advice? Should I abandon this and instead copy/paste or base my system off a different systems damage, defense, hp system like Pathfinder 2e? That seems to look like the best option in the long run but IDK

Overall feeling a little discouraged any insight would be greatly appreciated.

[brian 333]

The primary issue with game testing is, you are only one guy, with one outlook. You can create a perfect game system and some other guy will think of something you didn't that breaks your system.

To address this, get as many play testers as you can then have them help come up with ideas to test. Then play test untill you roll the corners off of your dice. (Ever notice how many authors and play testers are involved in creating even a basic game?)

Secondary to this is character balance. Test against not only monsters, but other characters. White Room testing is good, but even better is terrain that favors tactics. There should be situations that render your glass cannon vulnerable to your tank, and vice versa. Then test them against your monsters to determine if your monsters are tracking differently.

Some games take a rock-paper-scisors-lizard-spock approach to character design. Each defeats one and is defeated by another based on strengths and weaknesses. Original D&D took a 'build-a-team' approach which has to this day never completely been put into the rock-paper-scisors mold that some players tried to make it fit. Character balance is only important if players expect it to be.

As an example, 1st ed Thieves were weak compared to other classes. They even had a level progression that pushed their levels up ahead of other classes, while Wizards had the worst exp/level ratio. They were 'balanced' for a team, in that a group of characters of similar exp could keep up with each other against the campaign foes. In character vs. character solo cage matches, they were horribly mismatched. But the point was, each had abilities that complimented the others', and as a team they could accomplish what none could alone or with a team of similar characters.

So, your design philosophy matters as much as the actual mechanics. If you want solo campaigns, then you want Swiss Army Knife characters. If you want cooperative play you want specialized characters. And if you want PvP Arena characters, then you want balanced characters none of which have an advantage over the others.

And all of this applies to monster design as well.

[NichG]

I'd instead start with general numbers, rather than do all pairs and balance in detail. Like, how much damage per round will a basic character with no optimization do at the start of a campaign? At the end? How many rounds do you want fights to last - okay, now you roughly know how much hp (or how many copies) you need of 'featureless bags of hp' to make that happen.

Then you can use those as touchstones to say 'ah okay, this enemy has the ability to reduce its damage taken by half, so it's going to count twice towards that' or 'ah okay, this enemy might play keep-away due to movement modality differences, so I'll take that into account'.

But like, trying to make a formal CR system that is balanced to within +/- 1 level across 20 levels is usually pointless, because degree of optimization, strategic play on the players' parts and the GM's part, etc, etc will introduce more variance than that. Instead if you have a rough ordering of power of monsters and a rough idea of the scale of gaps or particular danger points (like x3 crits at low level in D&D, or something with a save or die, or something that can fly) then you can make notes of those things and generally go up or down the hierarchy if a party is wildly overperforming or underperforming. Especially in a more sandbox style of game, you can offload some of the responsibility for choosing appropriate challenges to the players - just make monsters, let them know the stats, and then part of the gameplay is for them to decide 'can we take this yet?'.

With that mindset, you can even embrace the variance. It's fine for a party to fight something '4 levels weaker than them' but with a movement modality they have no good answer for, or to beat something '8 levels stronger than them' using the environment and their knowledge of the creature's abilities. Then you can even stop thinking in terms of specific level appropriateness entirely, and just think in terms of what sorts of abilities would pose interesting tactical scenarios.

[aimlessPolymath]

What you're describing is a pretty important part of the dev process, but it's usually done a little later on, during playtesting.

Step 0 here is to measure some benchmark player numbers- how much damage can different characters, at different levels, do to a training dummy? What are their attack rolls like? Defenses? Usually, you should be able to do this without actually building a character- for example, in D&D 5e, you can generally estimate someone's key stat for their attack rolls, as well as their proficiency bonus.

Then, you can attempt some simple math to give you a starting point to work from. Start from a few main data decisions:
-What's the 'average' success rate you want for actions in combat, assuming people are executing their gameplan effectively? (This also sets up how you assign monster attack and defense numbers- compare to 'benchmark' player numbers from the same level.)
-What's the average damage players do per successful 'damage' turn, assuming they're executing their gameplan effectively?
-How many turns or rounds should an 'average' foe take to take down? (This varies depending on if you're assuming a 1vparty setup, or equal numbers on each side)
-How much time is spent doing setup, control effects, support? For example, if you expect one character in a 4-person party spends most of their time performing support and control effects, or if a character needs to spend time doing setup before they're fully 'online', then that would reduce the party's ability to close out the encounter.

Then, you can work backwards to decide character statistics.

HP = Turns * Average damage / Success rate / (% of actions spent directly on damage)

Monster damage is largely calculated in a similar fashion to HP, though your decision for 'turns to take down' may differ between players and monsters.

These numbers won't be perfect, but the important thing is that they get you in a position where you can start testing and iterating.

The other big qualifier here is that in play, encounter difficulty is very contextual and party dependent- a group that's all wizards will find very different types of encounters hard compared to a party that's mostly fighters. Parties heavy on healing and control effects will generally have much longer encounters than parties with lots of damage-dealers. Tactics (ex. focusing fire) will also affect this!

Re: playing out combats, it gets a lot easier if you have some playtesters to test with and split the load- you can run a short adventure of 1 or 2 combats. You don't need to test at every single level, but having a few 'checkpoints' to test will let you get an idea of how it plays at different levels- for example, running an adventure at levels 1, 5, 10, and 15, for example. Once you have people make characters, you might find that your benchmark player stats were off, and can revise your numbers before actually running the combat.

Also- a full 20-level system is a pretty large scope to work on, and I'd actually suggest starting by picking a particular level range to focus on for the near future, rather than building a full monster roster for 20 different levels.

[brian 333]

Spoiler

Show

I'd instead start with general numbers, rather than do all pairs and balance in detail. Like, how much damage per round will a basic character with no optimization do at the start of a campaign? At the end? How many rounds do you want fights to last - okay, now you roughly know how much hp (or how many copies) you need of 'featureless bags of hp' to make that happen.

Then you can use those as touchstones to say 'ah okay, this enemy has the ability to reduce its damage taken by half, so it's going to count twice towards that' or 'ah okay, this enemy might play keep-away due to movement modality differences, so I'll take that into account'.

But like, trying to make a formal CR system that is balanced to within +/- 1 level across 20 levels is usually pointless, because degree of optimization, strategic play on the players' parts and the GM's part, etc, etc will introduce more variance than that. Instead if you have a rough ordering of power of monsters and a rough idea of the scale of gaps or particular danger points (like x3 crits at low level in D&D, or something with a save or die, or something that can fly) then you can make notes of those things and generally go up or down the hierarchy if a party is wildly overperforming or underperforming. Especially in a more sandbox style of game, you can offload some of the responsibility for choosing appropriate challenges to the players - just make monsters, let them know the stats, and then part of the gameplay is for them to decide 'can we take this yet?'.

With that mindset, you can even embrace the variance. It's fine for a party to fight something '4 levels weaker than them' but with a movement modality they have no good answer for, or to beat something '8 levels stronger than them' using the environment and their knowledge of the creature's abilities. Then you can even stop thinking in terms of specific level appropriateness entirely, and just think in terms of what sorts of abilities would pose interesting tactical scenarios.

Spoiler

Show

What you're describing is a pretty important part of the dev process, but it's usually done a little later on, during playtesting.

Step 0 here is to measure some benchmark player numbers- how much damage can different characters, at different levels, do to a training dummy? What are their attack rolls like? Defenses? Usually, you should be able to do this without actually building a character- for example, in D&D 5e, you can generally estimate someone's key stat for their attack rolls, as well as their proficiency bonus.

Then, you can attempt some simple math to give you a starting point to work from. Start from a few main data decisions:
-What's the 'average' success rate you want for actions in combat, assuming people are executing their gameplan effectively? (This also sets up how you assign monster attack and defense numbers- compare to 'benchmark' player numbers from the same level.)
-What's the average damage players do per successful 'damage' turn, assuming they're executing their gameplan effectively?
-How many turns or rounds should an 'average' foe take to take down? (This varies depending on if you're assuming a 1vparty setup, or equal numbers on each side)
-How much time is spent doing setup, control effects, support? For example, if you expect one character in a 4-person party spends most of their time performing support and control effects, or if a character needs to spend time doing setup before they're fully 'online', then that would reduce the party's ability to close out the encounter.

Then, you can work backwards to decide character statistics.

HP = Turns * Average damage / Success rate / (% of actions spent directly on damage)

Monster damage is largely calculated in a similar fashion to HP, though your decision for 'turns to take down' may differ between players and monsters.

These numbers won't be perfect, but the important thing is that they get you in a position where you can start testing and iterating.

The other big qualifier here is that in play, encounter difficulty is very contextual and party dependent- a group that's all wizards will find very different types of encounters hard compared to a party that's mostly fighters. Parties heavy on healing and control effects will generally have much longer encounters than parties with lots of damage-dealers. Tactics (ex. focusing fire) will also affect this!

Re: playing out combats, it gets a lot easier if you have some playtesters to test with and split the load- you can run a short adventure of 1 or 2 combats. You don't need to test at every single level, but having a few 'checkpoints' to test will let you get an idea of how it plays at different levels- for example, running an adventure at levels 1, 5, 10, and 15, for example. Once you have people make characters, you might find that your benchmark player stats were off, and can revise your numbers before actually running the combat.

Also- a full 20-level system is a pretty large scope to work on, and I'd actually suggest starting by picking a particular level range to focus on for the near future, rather than building a full monster roster for 20 different levels.

Both of these posts are correct. My post presumed the rough math part was done and theoretical modeling had been considered.

Scaling is an issue. A complaint by D&D builders is that Fighters begin powerful, but do not scale up nearly as well as Wizards, who begin weak then become massively more powerful than fighters. This is a valid complaint if your goal is to compare solo characters to each other. It is less so if your goal is to create team play, in which the wizard requires the fighter to survive and the fighter requires the wizard to finish off powerful foes.

[Intregus182]

I do have the benchmark numbers down. I know exactly what bonuses to hit are possible at every level how much HP PCs can have at any given level same for defenses and damage etc.

My thought was before I get players to play test this (which won't be a problem) I wanted to run these "white room" simulations just to see if my first draft at making monsters is even close to balanced or how I want it to feel.

My game is set up Into 5 tiers (every 4 levels) my goal was to balance one tier at a time not all 20 because that would be a nightmare.

So I'm just trying to get the first 4 levels balanced currently. Unlike DND for example players don't get up every level but an increase in every tier (just to keep HP numbers down so ho bloat doesn't become a thing)

I super appreciate everyone's responses!