I've been thinking about MP and its role in combat for a long time, and it kind of frustrates me that I haven't come up with a single rule of thumb that should be used for most games.
Probably the most difficult part of the analysis is that MP might
look like the same mechanic in most games, but actually
constitute an entirely different mechanic between different games. MP is always a "pool" of limited resources that your skills share - I'll get to this momentarily. But beyond that, MP's function varies wildly between games!
- In some games, MP is restored after every battle, so MP simply dictates the rate at which you can use powerful skills during battle (think of it almost like a flexible, global cooldown between skills).
- In some games, MP is limited and doesn't naturally recover, but is easily restored using items or mid-dungeon restore points. My instinct is that this is the kind of game where MP is the most problematic from a design standpoint, because it introduces a psychological "cost" to using your skills which inevitably feels bigger than it actually is. If you blow your MP now, you won't have it for next battle.
- In some games, MP is limited and scarce, and ways to recover it (outside of safe areas like towns) are even more scarce. Here, MP is the player's "trump card" that they can pick and choose when they feel like they need it. It is a chronic (rather than acute) resource that needs to be managed well in a long-term dungeon strategy. In a well designed game of this sort, the optimal strategy should not be to save all your MP for the boss fight.
- In some games, MP is technically limited but the player has so much of it that it might as well not be. Here, the designer should either drop MP entirely and consider other limitations on skills, or should rescale their MP pools/costs to ensure that players need to make decisions about spending MP.
- In some games, MP is restored within battle, for example a bit of MP is restored every turn or a bit of MP is restored each time you land a basic attack on an enemy. Here, MP's role is to dictate the balance and flow of power in combat between your more-powerful skills and less-powerful non-skill actions (basic attacks, guarding, etc.). It forces the player to make themselves vulnerable at certain points, and lets the player feel very powerful at other times. If MP is not also restored after every battle, though, the player may find himself doing un-fun things like skipping turns at the end of a nearly-won battle in order to build up more MP for next battle.
In all of these cases, MP represents a "pool" of a limited resource that is shared between your different tools - you can use a skill as much as you want (as long as there aren't other restrictions like cooldowns), but it means you now have less MP to spend on your other skills. The player thus has a ton of flexibility over their battle strategy, which is great in theory. In the same breath, that battle strategy has a tendency to become stale, allowing the player to pick a single move and spam it over and over, which is not fun or interesting. It also allows the presence of one slightly-overpowered skill to completely crowd out most of the other skills in the game, meaning the player will almost never use those other skills.
Therefore, a system that revolves around the use of MP is probably best for a game where the designer is extremely good at balancing different skills, and has a variety of diverse enemies that actually require different strategies to defeat. A system that revolves around the use of MP could also be good for a game where
expression in combat is a more important
aesthetic of play than challenge or exploration.
---
Cooldowns, unlike MP, are generally limited to individual skills, and therefore the only resource shared between skills is the character's "turn". This has a lot of built-in advantages. It forces the player to use a variety of different skills. It mutes the "crowding out" impact of a single very powerful skill (allowing the designer to
intentionally create overpowered skills with long cooldowns if desired). It also tends to feel good for the player, compared to other limitations: even if you spend the "cost" of a cooldown, it will naturally come back to you without you having to do anything else.
Mechanics like "Over Limits" (temporary periods of enhanced power that occur once you've dealt/taken enough damage, or once something else has built up) also serve a similar design space to Cooldowns if they happen frequently enough during battle.
Cooldown systems encourage medium-term planning and thoughtful use of time/turns. This makes Cooldown systems ideal for games where strategy within battles (as opposed to longer-term strategy across entire dungeons) is paramount. Cooldown systems are also great for designers who are not confident that they can balance their Skills against one another perfectly.
One of the problems with a Cooldown system is that while the player will not spam a single move over and over again, they may fall into the same "pattern" of moves - start with move A (3 turn cooldown), then use move B, move C, and finally move A again; rinse and repeat. Careful design of skill cooldowns can avoid this flaw. For example, if it's natural to use Move B right after Move A, then a good Cooldown amount might be 3 turns for Move A and 5 turns for Move B. If you use Move A on turn 1 and Move B on turn 2, then your next available use of Move A will be turn 4 while your next available use of Move B won't be until turn 7. The player can now choose to either save Move A for turn 6 (and do a lot of other interesting/creative things in the meantime), or can adapt their battle strategy to
not use Move B directly after Move A on the next cycle, in order to use Move A on turn 4.
I'm very torn about the use of MP and Cooldowns together. Some games do this (including one of my own), but it can be slightly uncomfortable to the player to have their options limited by not one but two different systems, and can make battle strategy either unintuitive (if the MP pool is small) or complex without a reason (if the MP pool is very large). On the other hand, the use of MP alongside Cooldowns may be useful to gain most of the design benefits of Cooldowns while forcing the player to engage in tactics other than using Skills every turn - especially if the number of Skills available is large enough that it would be impossible to have everything on Cooldown at once. There could be better ways to encourage the player to do this, however...
---
Limited uses ("Charges") of skills within a battle (or some larger timeframe) would be very difficult to make an entire combat system out of. That's because the power is so front-loaded in these skills and the balance of a battle would shift dramatically depending on the length of said battle; in other words, with Charges, the player becomes more disadvantaged the longer the battle goes. If I were to try and design a battle system around Charges, it would be a puzzle-oriented battle system where the number of turns is small and highly-controlled (either through fixed enemy damage or through a turn limit), so that the player's real decision is which subset of skills they want to use in this limited time. (The use of Charges here prevents them from using the same skill over and over, but in a less restrictive way than Cooldowns.)
More often, skills with Charges are intentionally overpowered and the design wisdom behind giving them a limited number of Charges rather than Cooldowns is to present the player with the decision of
when to make the most of their powerful skill (which can be an interesting dilemma), rather than how to use it most often (which is usually a pretty rote decision of use it as soon as possible, and then use it as soon as it comes off of Cooldown). Often, the number of Charges on a skill is 1 - you can only use it once per battle.
In a system where the player can only equip a limited number of skills, an interesting choice presents itself between powerful skills with a limited number of Charges versus less powerful skills that can be used in perpetuity (assuming enough MP / finished Cooldowns).
Sometimes, you can only use a skill X number of times throughout a dungeon, during a day, or even throughout the entire game. Here, it truly is your "trump card" - you can use it when you really need it to overcome something difficult, but you know that's one less use you have of it later. It's like a more impactful version of rare (limited) consumable items that are collected throughout some games.
The more limited the Charges are, the more powerful you are allowed to make the skill without breaking the game's balance. In the case of Charges per Battle, this also allows the player to use very
fun skills in every random encounter (perhaps even more than once), without the skill making up a significant part of strategy against bosses. In the case of Charges per Dungeon/Game, you can make the skills even more gamebreakingly powerful without risking game balance too much. The player
will feel the impact and rush of using this skill!
The drawback, of course, is that the more limited the Charges are, the more conservative players will be about using the skill. If you want your player to use a skill a lot, Charges (especially any type of Charges that don't restore after every battle) are not the way to go. Charge systems on skills are usually for very specific skills whose design requires it because they are either abusable or intentionally overpowered.
In general, I don't recommend mixing Charges with either MP Costs or Cooldowns on the same skill. If a skill has a Charge system, give it a Cooldown of 0 and an MP cost of 0 (or some negligible amount) to avoid unnecessary complexity. A possible exception could be healing skills - you may need to add Charges on top of the standard MP/Cooldown system for healing skills for balance and skill diversity reasons.