Tweet
Don't mind me, just thinking out loud.
As I've mentioned in the past, one of my favorite RPG combat systems is the Grandia II system, where each actor in the battle is placed on a "timeline" broken into sections: WAIT -> DECIDE -> CHARGE -> ACT. In WAIT, the actor waits to accumulate enough energy to take an action (rate of accumulation depends on actor's initiative). At DECIDE, they select an action and its target(s). In CHARGE, they wait to accumulate energy again (rate depends on the action, the actor's stats, etc.). At ACT, the action occurs.
Additionally, some actions can move actors about on the timeline. The most obvious of this is that attacks with the "Cancel" attribute will move the target back into the WAIT phase when they land, potentially wasting the target's turn (depending on how well the Cancel attack was timed).
One tricky thing about Grandia II is that it almost entirely abstracts away movement. While positioning on the battlefield is important (most spells hit an area of effect), typically the player does not spend much of any time thinking about where they are on the battlefield. Movement is bundled into the "attack" command (so when you attack, first you automatically run up to the target, then you attack, then you automatically retreat to a random location). Likewise, targets are always actors on the battlefield, not positions on it. Such "smart targeting" simplifies the player's life -- they don't have to try to time a movement command to land them next to an enemy so that they can then, some time later, use an attack command. But it means that area-of-attack abilities tend to have massive areas so that they have a reasonable chance of hitting multiple opponents. And of course defensive positioning is almost entirely nonexistent.
One possible fix for this would be to bundle movement into every action, but make it have an associated cost assessed after the action is completed (by pushing the actor further back along the WAIT line). So you reach ACT, decide to Attack, go through CHARGE, and then you can do some amount of movement combined with the Attack at some point. Then you're pushed back into WAIT, with the distance being (cost of Attack command) + (number of steps taken) * (cost of Move command). Of course, you can assess different costs for moving in different circumstances (terrain, or through enemy-controlled territory, or with different [de]buffs in effect, etc.), and there'd be a cap on just how badly you can send yourself into debt.
Another thing I'd like to see with this kind of system is the ability to spend actions to set up downstream effects. I saw something like this in the Atelier series of JRPGs (well, I watched a few combat videos; I haven't played any of the games myself). In those games actors can take actions that cause several events to happen downstream. For example, you cast a spell, and it occurs once immediately, and then four more times at set intervals in the future. Or you lay a bomb which explodes after a set amount of time.
The overall goal in a roguelike-style game would be to encourage the player to try to predict what will happen in the future. The better they are at this, the more advantageously they can position themselves to avoid taking damage while effectively dispatching their opponents. For example:
1) Enemy starts charging a slow but powerful attack targeted at the player's position.
2) Player hits DECIDE phase; wants to cast a time-delayed fireball, but needs to be out of the enemy attack's area of effect when it goes off.
3) Player does some mental estimates, decides he probably has enough time. Readies spell, targets it where he thinks enemy orc group will be when it goes off.
4) Player's spell finishes CHARGE phase and is emplaced; player gets to move.
5) Enemy charged attack goes off.
Ideally the timings for this will be loose in the early game. Actors would be well-separated into distinct "turns", there'd be relatively few effects that are themselves time-delayed, and the only effects that move actors along the timeline would be Cancel attacks. This gives the player time to get used to the system. The timings should get progressively tighter as the game goes on, so that by endgame, miscalculating things by a single step should be the difference between being hit and not (which in turn could be the difference between success and failure). Likewise, the most powerful effects should all have time-delays on them, so that the player is rewarded for accurately judging what the battlefield will look like in the future.
As I've mentioned in the past, one of my favorite RPG combat systems is the Grandia II system, where each actor in the battle is placed on a "timeline" broken into sections: WAIT -> DECIDE -> CHARGE -> ACT. In WAIT, the actor waits to accumulate enough energy to take an action (rate of accumulation depends on actor's initiative). At DECIDE, they select an action and its target(s). In CHARGE, they wait to accumulate energy again (rate depends on the action, the actor's stats, etc.). At ACT, the action occurs.
Additionally, some actions can move actors about on the timeline. The most obvious of this is that attacks with the "Cancel" attribute will move the target back into the WAIT phase when they land, potentially wasting the target's turn (depending on how well the Cancel attack was timed).
One tricky thing about Grandia II is that it almost entirely abstracts away movement. While positioning on the battlefield is important (most spells hit an area of effect), typically the player does not spend much of any time thinking about where they are on the battlefield. Movement is bundled into the "attack" command (so when you attack, first you automatically run up to the target, then you attack, then you automatically retreat to a random location). Likewise, targets are always actors on the battlefield, not positions on it. Such "smart targeting" simplifies the player's life -- they don't have to try to time a movement command to land them next to an enemy so that they can then, some time later, use an attack command. But it means that area-of-attack abilities tend to have massive areas so that they have a reasonable chance of hitting multiple opponents. And of course defensive positioning is almost entirely nonexistent.
One possible fix for this would be to bundle movement into every action, but make it have an associated cost assessed after the action is completed (by pushing the actor further back along the WAIT line). So you reach ACT, decide to Attack, go through CHARGE, and then you can do some amount of movement combined with the Attack at some point. Then you're pushed back into WAIT, with the distance being (cost of Attack command) + (number of steps taken) * (cost of Move command). Of course, you can assess different costs for moving in different circumstances (terrain, or through enemy-controlled territory, or with different [de]buffs in effect, etc.), and there'd be a cap on just how badly you can send yourself into debt.
Another thing I'd like to see with this kind of system is the ability to spend actions to set up downstream effects. I saw something like this in the Atelier series of JRPGs (well, I watched a few combat videos; I haven't played any of the games myself). In those games actors can take actions that cause several events to happen downstream. For example, you cast a spell, and it occurs once immediately, and then four more times at set intervals in the future. Or you lay a bomb which explodes after a set amount of time.
The overall goal in a roguelike-style game would be to encourage the player to try to predict what will happen in the future. The better they are at this, the more advantageously they can position themselves to avoid taking damage while effectively dispatching their opponents. For example:
1) Enemy starts charging a slow but powerful attack targeted at the player's position.
2) Player hits DECIDE phase; wants to cast a time-delayed fireball, but needs to be out of the enemy attack's area of effect when it goes off.
3) Player does some mental estimates, decides he probably has enough time. Readies spell, targets it where he thinks enemy orc group will be when it goes off.
4) Player's spell finishes CHARGE phase and is emplaced; player gets to move.
5) Enemy charged attack goes off.
Ideally the timings for this will be loose in the early game. Actors would be well-separated into distinct "turns", there'd be relatively few effects that are themselves time-delayed, and the only effects that move actors along the timeline would be Cancel attacks. This gives the player time to get used to the system. The timings should get progressively tighter as the game goes on, so that by endgame, miscalculating things by a single step should be the difference between being hit and not (which in turn could be the difference between success and failure). Likewise, the most powerful effects should all have time-delays on them, so that the player is rewarded for accurately judging what the battlefield will look like in the future.
Comment