[Discussion] How do you decide what levels to make encounters/trainers?

Edit: wow, my phone did a number on that. I've rewritten that paragraph as well as writing my own "balance" notes.

When you're developing a game, one of your first instincts is to tell the potential beta testers that you're not looking for anyone to test the game, because they give off that vibe that they just want to play the game early. A good beta tester will test your game with a goal, such as balance testing, which is your best bet for what you're trying to do, or just attempting to 'break' the game(either by finding ways to make it impossibly hard or disappointingly easy or by finding game crashing bugs). While there are a lot of people who volunteer to beta test here who do just want to play your game, I recommend finding a dedicated playtester who demonstrates the ability to document what they do and if at all possible has some design insight of their own for gameplay.

In the short term, though, as you place Pokemon(Trainers', Gyms', and Wild), write up three hypothetical players of your game - one who's an expert on all things Pokemon, one who plays casually but maybe hasn't played for a while or just isn't familiar with your fakemon(most of your players will fit into this category) and one player who has either never played a Pokemon game before or is too young to understand stuff like stat boosts. Generally speaking(but again, this is in the broadest strokes), I like the young player to have an estimated 30-40% chance of winning a battle, an average player having 50-60% chance to win, and an expert having 80-90%. Consider what parties these three players will have at the point you're trying to write encounters for, and build your encounters with those in mind(for instance, if you're in the early game with mostly normal, grass, water, and bug types, odds are that Steel types or Ghost types are going to be fairly difficult to defeat).

Gameplay "balance" (I use quotation marks because really you aren't balancing it so much as shaping its difficulty) is an infinitely deep rabbit hole, though, and I do recommend finding a dedicated playtester or two that can document their process for you, because while the method I described will put you off to a decent start, it's going to always require finer tweaks. And finally, this is just a simplification thing for me, I like to keep all levels as low as possible whenever I can - after all, Pokemon only go up to level 100(or whatever final level you decide on), so once you have that power creep escalate to its maximum, hopefully you're well into your postgame.

What I'd like to do myself is play through the game when I'm making it.
When I want to make a new battle/encounter I just look at what lvl my pokemons are and balance around it.

Juliandroid98 said:

What I'd like to do myself is play through the game when I'm making it.
When I want to make a new battle/encounter I just look at what lvl my pokemons are and balance around it.

Click to expand...

The biggest problem with that is that you're balancing for yourself rather than the general public - your playstyle may be different than Joe Q. Averageguy. What if other players grind more, or catch different Pokemon than you?

Also, you're trying to hit a moving target when it comes to what your party looks like, especially if you change something earlier than your save file has reached. And if you're editing late game stuff, it'd be tedious to try and get through the entire game to gauge what your average team would look like.

I typically come up with some rough estimations based on the route, the routes before it and the level/exp curve. I then test it to see how playable it is and get feedback from others on the level balancing (others being a few designated testers) and go from there.

TBM_Christopher said:

The biggest problem with that is that you're balancing for yourself rather than the general public - your playstyle may be different than Joe Q. Averageguy. What if other players grind more, or catch different Pokemon than you?

Also, you're trying to hit a moving target when it comes to what your party looks like, especially if you change something earlier than your save file has reached. And if you're editing late game stuff, it'd be tedious to try and get through the entire game to gauge what your average team would look like.

Click to expand...

True that, but beta testers are still out there to test if your balance is ok or not.
I just use that method to kinda balance on the fly.

Juliandroid98 said:

True that, but beta testers are still out there to test if your balance is ok or not.
I just use that method to kinda balance on the fly.

Click to expand...

As long as you're getting more than just your own perspective while doing playtests, I think you're on the right track. Sorry if that wasn't clear in what I was saying; I meant that a single perspective isn't as helpful as a range. The mathematical methodology can cover that range, though, like any abstraction, it's somewhat clinical and isn't a magical catch-all.

It's interesting that this discussion came up again; I'm actually working on a design challenge in Chromatic where I'm adding a floor value to wild Pokemon's levels, so you'll be able to find pokemon whose levels are comparable to your own no matter where you go(the design goal there is to make it much easier to try and use new pokemon on your team). At present, wild Pokemon's levels are at minimum 4 * number of badges, which is a rough abstraction from looking at the average levels of gym leaders' pokemon. However, I'm going to have to conduct some playtests to make sure this system doesn't reduce the game's difficulty by making more exp available.

Keep in mind that balancing wild pokemon and trainers' are two very different tasks as well; trainers tend to have multiple pokemon, and offer more experience and also money when beaten. Because of those factors, they tend to have slightly higher leveled pokemon than the wood pokemon around them (with some exceptions)

I just make the trainers a few levels lower than the last boss, except maybe in the late game. For trainer number of pokémon, I generally use the rule:

1 Lv40 = 2 Lv38 = 3 Lv37 = 4 Lv36 = 6 Lv35

Wild pokémon are generally 36-32 using this rule. I also lower/rise level for evolve and unevolved forms, so a level 40 Venusaur is equals a level 38 Ivysaur. The main challenge it's to define the right levels for the bosses. I generally define if I can win battling versus all trainers before, and removing some levels (from my team) to make it easier for the early game and mid game. For late game I even increase the level caps! I almost only use healing items in late game. I also use the previous level gain from previous bosses as template (plus a level or other).

Overall, in my opinion, balancing a pokémon game is VERY easier than a standar RPGs, since most of system (even if you use fakemons and fakemoves) was already defined, but still isn't a easy task.

POKEMON,Minimum Level,Maximum Level(Optional)
Go the PBS and look for encounters and change a pokemon in one area.

Noob Buster said:

POKEMON,Minimum Level,Maximum Level(Optional)
Go the PBS and look for encounters and change a pokemon in one area.

Click to expand...

This wasn't a technical question of how you set an encounter level in the game engine but rather how you decide what levels to make an encounter when designing the game.

Usually, intuition's the key. Since gym leaders are usually spaced up several levels depending, I compare the best Pokemon of the last gym leader to the weakest of the next and balance accordingly. The first gym leader should be around level 9 to level 14, depending on how hard you want the game to be, and the second should be around five levels over.
Say, make the levels move up slowly depending on what route/cave/etc. you go in before you reach the gym. You don't see a trainer on your first route with a level 12 Scyther or something.