Game balance and leveling.

[dden911]

Hi!

I have been working on my first RGL, and so far have come across some balance questions.

Leveling system

At this point I think that XP cap. for every level should be twice as much as for previous:
Level 2 - 10XP
Level 3 - 20XP
Level 4 - 40 XP
and so on.
However I am  not sure that such system is good. Can you recommend some articles about such balance things in RGL games?

Scaling
Player grows stronger with time. It gains XP, HP and Attack boosts, possibly he peaks some items and weapons that further buff him. The question is: how I should scale monsters he encounters to maintain balance between easy grinding and hardcore gameplay. Generally I want two main things to be in my game:
-   Player cannot “overgrow” the game universe easily one-hit beating everything, game should be interesting even at high levels of player XP.
-   Leveling speed should possibly decrease not so quick as XP needed for level. In other words, I don’t want to have something like:
Level 2 – 3 minutes of game play needed
Level 3 – 6 minutes
Level 4 – 12 minutes and so on.
It would be good to have less decreasing numbers of minutes.

Obviously my second question may require some math and calculations, however I will be very grateful for any links and articles regarding these problems in RGL game development.

[Vanguard]

Leveling system

What I like best are Fire Emblem-style systems where every level costs the same amount of experience (say, 100 points) but you get a lower amount for defeating weaker enemies and a greater amount for defeating stronger enemies.

That kind of system makes exp gain easily comprehensible.  In a normal RPG, 5,000 experience points means something completely different at the start and end of the game, but in Fire Emblem you always know that 10 experience is 1/10 of a level.

It's also good at keeping grinding under control.  A level 10 player needs to kill, say, 20 level 10 enemies to level up and a level 40 player needs to kill 20 level 40 enemies to level up.  So you never have to worry about leveling time increasing too much as your level increases.  Players who take risks and go after higher level enemies are rewarded with meaningfully faster level ups, and you can prevent players from grinding against trivial enemies by reducing their exp payouts to zero.

Scaling

If you don't want players to outgrow too much of the game's world, you should limit the amount of power they gain from leveling, or slow leveling down, or implement a low level cap.  One method I like is to increase the player's versatility rather than their power as they level up - give the player new abilities (or improve existing abilities), but little to no increase to their health and attack power and such.  That way the attacks of low level enemies are still dangerous at higher levels, but the player has more tools to avoid damage, control the battlefield, and adapt to complex situations.

I'm not a fan of enemy scaling systems.  They can do more harm than good, they're easy to mess up, and they're seldom necessary.  One method that might work is to combine the low power growth method I mentioned earlier with a scaling system that increases the power of higher level enemies (say, a 20% increase to their damage output for each level they have above yours).  That way the player only gains a small advantage over lower level enemies as they level up, but higher level enemies quickly become more manageable.

[Anvilfolk]

Thanks for putting these thoughts out Vanguard. I was struggling to write down how I felt about this, but you just described my favourite system as well!

[Vanguard]

Sure thing, man.  I love these kinds of discussions.

I wish this board was a bit more active, because it's got some sharp people with good ideas.

[dden911]

Thanks for your answers!

I will try to figure out how can this be done. Another question: how should I generate monsters. For example player level is X, how much monsters of level X-1, X and X+1 (just this 3 for simplicity) should be presented on current level? 

[Vanguard]

It's all up to you.  You can generate enemies entirely based on the area (my preference), or entirely based on the player's level, or some combination of the two.  Or do something completely different.  There's more than one right answer here.

[Sandblaster]

I think leveling rate is going to depend on the length of your game, as well as what a level actually gives you.  If you start with only 1 ability and gain a new one every level, you might want to have the player level quickly at first to earn a good repertoire of abilities, before tapering off so that he can't get ALL the abilities in one game.  If each level is a fairly large leap in power, maybe the player should level slowly but steadily.

[Vanguard]

If you start with only 1 ability and gain a new one every level, you might want to have the player level quickly at first to earn a good repertoire of abilities

I know you aren't necessarily advocating it, but I hate this kind of design.  Better to remove RPG elements entirely than to start the player with terrible tools and only allow them to be properly functional after they've done some leveling.

[zasvid]

If you start with only 1 ability and gain a new one every level, you might want to have the player level quickly at first to earn a good repertoire of abilities

I know you aren't necessarily advocating it, but I hate this kind of design.  Better to remove RPG elements entirely than to start the player with terrible tools and only allow them to be properly functional after they've done some leveling.

Seconding this. Though, of course, if someone's beholden to giving an ability a level, you don't have to go all or nothing. Cut out the early part (e.g. start at level 10 out of 50 instead of level 1 out of 50, though of course the game might display levels renumerated so it begins at 1st) and it will be more palatable.

[Endorya]

Leveling System:
When I look back at my large list containing all RPG games I've played, I can conclude that Gothic series and Fallout I, II and Tactics series hold the leveling system I enjoyed most, even though I'm not particular fond of having your character's HP increasing like 10 times more over its starting value.

I like the idea of having further levels requiring more XP as you level up, as long everything in the game has a fixed experience reward. I don't like at all having experience rewards based on the character's current level, mostly because it makes parts of the world annoying and uninteresting, i. e. when a certain area becomes only populated by lesser creatures there is no reason for the player to linger about. Not only these areas don't oppose a challenge but the creatures present in the them are not even worth of slaying and also, you never know how much experience you will get from slaying a specific creature. Of course this is heavily dependent on how the game is intended to work. If lesser creatures are replaced by more powerful creatures then I guess, if properly done, it can function.

For my game, I plan to have creatures with fixed experience reward, regardless of player's character current level, making sure each creature you slay doesn't feel like a waste of time or boring as it would happen in Age of Conan Online. But Age of Conan online is a MMO so I understand why it is done in such way. This mechanism was implemented to prevent players from camping at certain areas to farm experience without too much effort; it was implemented to force players to search and to explore the world map for areas in which they can continuously develop their characters. This method can also be seen in other MMO games like in WOW.

So which one would be best? I really can't say that A or B is best for all of you, all I know is that the fixed experience is the one I personally enjoy and that feels more rewarding and fun to play.

Scalling:
I totally agree with Vanguard on this one. I prefer to have the player becoming stronger through equipment and skills than simply increasing its HP, mana, stamina or attack power simply through leveling up. I'm also not fond of enemy scaling systems either. If a goblin gets you 90xp and a certain attack power, than make it it so through out the whole game and don't use levels on them. A goblin is a goblin and not a level 3 goblin or a level 5 goblin or level 16 goblin. Thought all goblins are basically the same, you may have a young goblin, an adult goblin or an old goblin, being the adult one stronger than the older or the younger one. And even adult goblins should have their attributes kinda changed a little bit so not every goblin has exactly 60 HP points. Occasionally, the player should encounter an exceptional goblin with exceptional power and endurance, for a goblin that is. Then you have goblin classes like goblin-warrior shamman-goblin etc. to further increase diversity and challenge.

I truly hate having a system where the world levels up with the player's character. This means that the game gets progressively difficult as the player keeps leveling up. Oblivion has the worst scaling system I have ever played. Players could actually end the game at level 1. The best method I've experienced is the one found in Gothic II, where every area was accessible at any time with all sorts of difficult levels. Some areas would be easy while other areas could be difficult or impossible to play at early states of character development, but at least you always had a way to go around. If a certain area was too difficult the player would explore an easier alternative and come back later to explore the harder path. But this method may represent a problem for games featuring permadeath because if this model is not properly implemented the player can be killed at any time without warning. Fun for some and probably frustrating for many.

In my game, although characters level up, it doesn't mean they get immediately stronger after leveling up, in fact, when they level up, they remain exactly the same; leveling up just means players have earned a certain amount of learning points in which they can spend with trainers or teachers to increase they character's efficiency in a certain skill, profession or ability. This means that in my game, players will not be able to open the character's skill tree to learn a new ability out of nowhere. But this is my view towards it. Anything can function if properly done with personal preferences put aside.

As for the player getting too strong, well, level cap should in fact be considered, though I prefer traditional slow leveling up as long you make sure balancing is kept in place. One solution to make this right is to make leveling up costing about X% more over its previous experience level's requirement, over and over until leveling becomes practically impossible to achieve. Of course that the right value has to take into consideration everything in the game has involving experience rewarding like world map size and conflict level. I prefer this method to level cap because it feels more real. Knowing firsthand your character's level limit kinda feels like knowing when you will die.

[wire_hall_medic]

Honestly, I don't like experience systems much.  I much prefer systems like the old-school Zelda and Metroid games, where permanent improvements are found through exploration.  I do like finding useful gear from combat, so this isn't a perfect analogy, but it's close.  I feel like the reward for combat is a roll on the random loot table; getting XP on top of that is an unnecessary level of reward. 

However, the goal of leveling is NOT to make the player more powerful, it's to give them a feeling of advancement; without feeling like they're accomplishing something, players get bored very quickly.  The difficulty of a game should remain pretty consistent, with perhaps a little less at the beginning.  Having high-level areas accessable from the beginning compounds the feeling of advancement once players can actually handle them.

When you're watching your experience meter go up, it means you know you're not being rewarded for the vast majority of your fights; just the last one when you level.  I find that if your primary method of improvement is by finding, say, Brogue-style strength potions (or Metroid's energy tanks, or Zelda's heart containers), it keeps the player more engaged because the equivalent of gaining a level could happen around any corner.  If you couple that with a reasonable chance of enemies dropping useful loot, then there's two strong incentives for the player to keep paying attention.

The most engaging type of advancement, however, is new abilities.  Ideally something that can be used both in combat and out.  For instance, a leaping attack that allows you close distance and make an attack, that can also be used to leap pits, thus opening up new areas and tactics.  Even if new abilities are pure combat, there should be something functionally new about them.  Adding knockback is much more interesting than just a damage bonus.

I feel that games should reward good decisions and enforce the natural consequences of poor decisions, so I don't like when a game scales the difficulty to the current level of the player.  Also, I feel that grinding as a game mechanic was useful to make games longer when they came on cartridges with 16 kb of memory; we don't need to do that any more.

[Vanguard]

However, the goal of leveling is NOT to make the player more powerful, it's to give them a feeling of advancement; without feeling like they're accomplishing something, players get bored very quickly.  The difficulty of a game should remain pretty consistent, with perhaps a little less at the beginning.  Having high-level areas accessable from the beginning compounds the feeling of advancement once players can actually handle them.

Leveling also serves to make multiple playthroughs more fun.  In a game of pure skill, once you're good enough to consistently beat a boss enemy, they won't every really be a threat again.  With leveling enemies can stay dangerous until both the player and the player character are ready for them.

I agree with you, though - traditional experience systems are not very interesting.  There are better ways.

[Endorya]

Honestly, I don't like experience systems much.  I much prefer systems like the old-school Zelda and Metroid games, where permanent improvements are found through exploration.  I do like finding useful gear from combat, so this isn't a perfect analogy, but it's close.  I feel like the reward for combat is a roll on the random loot table; getting XP on top of that is an unnecessary level of reward. 

However, the goal of leveling is NOT to make the player more powerful, it's to give them a feeling of advancement; without feeling like they're accomplishing something, players get bored very quickly.  The difficulty of a game should remain pretty consistent, with perhaps a little less at the beginning.  Having high-level areas accessable from the beginning compounds the feeling of advancement once players can actually handle them.

You are describing a type of leveling system that can only function with heavily scripted worlds, having all creatures and corresponding loot carefully measured and set to specific areas and to specific combats, which can indeed function well if the world lacks dynamism. If you move towards a game with large areas of exploration as found in many roguelike games, with most events being dynamically generated on the fly, with endless opportunities for combat and loot, you won't be able to use the system you refer to.

When you're watching your experience meter go up, it means you know you're not being rewarded for the vast majority of your fights; just the last one when you level.

Maybe it is just me but I always feel rewarded whenever I earn XP. I think this is the main objective of the experience meter and the whole point of earning experience points, i.e. to let the player feel rewarded for his actions, whenever he earns some.

However, the goal of leveling is NOT to make the player more powerful, it's to give them a feeling of advancement;

So in other words it would be the same as earning experience. Meaning the character is not getting stronger but having a feeling of advancement. This is how I feel about the experience meter.

I'm yet to be mesmerized by an alternative to traditional leveling up systems. We have Skyrim which although has a leveling system, doesn't offer experience rewards and then you have strictly skill-development-based games like EVE Online without a leveling system.

As much as I enjoy SKyrim, it does not feel as rewarding (at all) as Gothic Series or Fallout Series, having these last two a traditional leveling system and as far as the strictly-skill-based-system is concerned, I think it fails to give you a feel of development. It is really hard to measure the "power-level" of a character when there are no levels. Someone may ask you: "Hey! So how strong is you character now?" You really can't say that your char is at level X, you would need to say something like: "Well, I have 2 points on this, 4 points on that, I also have this ability and that ability, oh and I can also do this and that, so I guess my char is still weaker than yours".

Anyway, I really can't see myself playing a roguelike game without the traditional leveling system, with some minor exceptions. Generally speaking, I believe people prefer the traditional leveling system because it feels very rewarding. It could just be a coincidence but the most profitable game ever developed has the traditional leveling system, I'm referring to WOW. Naturally, this doesn't mean you can't explore other leveling systems or that the traditional leveling system is the best there is. But the truth is that I'm still waiting to be impressed by an alternative...

[Vanguard]

You are describing a type of leveling system that can only function with heavily scripted worlds, having all creatures and corresponding loot carefully measured and set to specific areas and to specific combats, which can indeed function well if the world lacks dynamism. If you move towards a game with large areas of exploration as found in many roguelike games, with most events being dynamically generated on the fly, with endless opportunities for combat and loot, you won't be able to use the system you refer to.

Sure you can.  If you want your game to have, say, 20 heart containers, just generate 20 appropriate randomized dungeons in randomized locations and put the heart at the end.

If you wanted to, you could even add items like the hookshot and tie progression through the game to them.  All you need to do is put the hookshot in one of the tier 2 dungeons, and make sure that you can access tier 1 and 2 dungeons without it, but require it for some of the later ones.

Mage Guild is a traditional-style roguelike in most ways, but character growth is based on finding power up items the game spawns in the dungeon.  It works really well.  Try it out.

It is really hard to measure the "power-level" of a character when there are no levels. Someone may ask you: "Hey! So how strong is you character now?" You really can't say that your char is at level X, you would need to say something like: "Well, I have 2 points on this, 4 points on that, I also have this ability and that ability, oh and I can also do this and that, so I guess my char is still weaker than yours".

Why do we need quantified power levels?  It's not like a level 10 barbarian is always exactly equal to a level 10 wizard in the first place.  All you really need to know is that your character is strong enough to beat the phantom knight if you play carefully, but you'll still need to run away when the cyberlich finds you.

[Endorya]

y, 20 heart containers, just generate 20 appropriate randomized dungeons in randomized locations and put the heart at the end.

If you wanted to, you could even add items like the hookshot and tie progression through the game to them.  All you need to do is put the hookshot in one of the tier 2 dungeons, and make sure that you can access tier 1 and 2 dungeons without it, but require it for some of the later ones.

Mage Guild is a traditional-style roguelike in most ways, but character growth is based on finding power up items the game spawns in the dungeon.  It works really well.  Try it out.

That's exactly what I was referring to. You need scripting things up to implement wire_hall_medic's method, which begins to break dynamism. I don't want to have a world with exactly 20 heart chests but generating a world where heart chests have X probability of spawning. So in the end, I could have a world with 0 to 20 heart chests (roughly speaking).

But even if I add specifically (scripting) 20 heart chests spread all over the world the results won't be the same as found in Zelda game, because in an open world the player will roam at will without following a certain path, making the chances of finding all heart chests in a huge world, next to zero, while in Zelda your path is incredibly scripted, making your character develop itself in predefined way set by the developers as they know that the player will find the 20 heart chests at a certain milestone. Sure you may have at some point in Zelda game the choice of following path 1 or 2 but all is heavily scripted and taken into consideration just to make sure that in the end the player's party has a certain measurable power to face the last boss with a proper difficulty level. I'm really not fond of RPGs-on-rails nowadays.

If I was to implement characters' power level based on loot in my game, its game-play would be completed ruined as I would need to add the right amount of loot (scripting). Even with the "right amount" of loot, the game's difficult level would remain inconsistent because the player could be lucky finding too much too early or unlucky and not finding nothing at all even after long periods of game-play. The only solution for this would be: scripting, which could destroy the surprising factor during exploration.

Sure I could solve this problem by spawning the right loot only after a certain level or after a certain amount of game-play time, but this also means that I would never be surprised in finding occasionally great loot. Skyrim has this method, I mean loot quality completely bound to the character's level. It works well I might say but skirim is skill based and loot based. However, this method does bring the inconvenience of me predicting the loot type, I know for instance that a dwarven type armor will not appear until my char doesn't reach level 12.

I want my game to be skilled based. Not only it feels more realistic but also avoids the player relying on luck to find that exceptional piece of armor or weapon to ease down the game's difficulty level.

Why do we need quantified power levels?  It's not like a level 10 barbarian is always exactly equal to a level 10 wizard in the first place.  All you really need to know is that your character is strong enough to beat the phantom knight if you play carefully, but you'll still need to run away when the cyberlich finds you.

I agree, it is not like a level 10 barbarian is equal to a level 10 wizard but I also think that everyone knows that races and roles play differently by now. But by knowing you wizard is a lvl 10 you automatically know that is better than a lvl 9 and worst than a lvl 11, excluding the player's skills and gear. No need for comparisons or revising your character's skills / abilities to understand if he can beat a mob with no level information. If a mob is lvl 10 or lower, you know for sure you can beat him.

Having level information automatically sets you at a starting point related to a possible level power. EVE Online uses Skill Points and not levels. As you see even a level free game such as eve EVE Online uses a "power level" meter, in this case it counts the skill points developed which tells everyone how long the character has been training and how powerful it is, if we further look into the character's skill point count in each skill group.

In my opinion their sure be a power level indicator in every RPG game, even if it is just informative.