|
Xendran -> Different but interesting way to improve RNG feel (2/8/2016 20:37:36)
|
This will sound odd at first, but it makes complete sense and keeps the game pretty much exactly how it is now in terms of overall straight skill balance, but improves potential variety and synergies as well as improving the feel of RNG.
It's a far shot that this will be implemented, but I figured I'd throw it here to get you guys thinking about this as well.
The current problem with RNG is that it eats entire turns, and has huge swings. It puts you so far ahead or behind in one go that it really hurts things.
In epicduel, it is the equivalent of starting a match of hearthstone at 10 mana, every card costing 10 mana, and no coin or extra card for going 2nd.
We can use a similar idea to base a system for epic duel to improve its combat.
This will also increase the overall amount of skill usage, and possibly increase potential build diversity because of it. There is more time to pull off synergies of multiple different skills with this setup, but in the end all of the effectiveness of everything remains the same as it is currently.
What we do is we divide turns into 3 actions.
Each thing you do, whether it be a skill or turn, costs actions.
The result of all damage in the game simply has this done to it: (1/3 * actionCost)
This means each action is worth 33% of a turn.
Skill cooldowns have this done to their cooldown: actionCost*3
However, cooldowns are now in actions instead of turns.
This means that smaller skills that are only 1 action can be used and synergized with each other much more frequently than currently, but to the same overall power effect.
Skills with 2 and 3 action costs have proportionately increased cooldowns to increase the investment back to levels we currently have.
Weapon attacks and most skills can cost 1 action. Remember that cooldowns in turns are also instead replaced with cooldowns in actions. This means on a skill with no cooldown and 1 action cost, you could spend your whole turn doing that skill just like you can now. Because each time you do it is multiplied by 0.33 repeating, you end up with a full turn the exact same as now, but the RNG is calculated 3 times. The average effect of the RNG is the same, but it is much less likely that you get put an entire turn behind in one go, and especially less likely that you get put two turns behind in two turns.
Skills that are meant to be more RNG can take more actions, so if we want Berserker to be a 2-action skill because it's meant to be a harder hitting one off skill with more risk, that makes it so you risk two actions of your turn to RNG at once instead of 1.
One thing to note about this, is we can use cooldowns to change the investments. Because cooldowns are cut in 3 to match actions
Ultimates would naturally cost 3 actions.
Most actions are at much less risk to losing an entire turn to RNG, but some actions still have a larger investment, perhaps to make up for other advantages such as lower cost to power ratio.
It's definitely a concept to think about.
A couple of skills may need rebalancing with this, and perhaps stat related skills shouldnt scale linearly due to the nature of stats, but that could be figured out later.
Another thing to note, is you could limit first turn advantage with this.
You start out with 1 action available, then 2, then 3. The player who goes 2nd can start out with 1 extra action.
This causes it to take longer for the full advantage of the first turn to take effect. First they are 0.33 of a turn ahead, then 0.66, then 1.0
We can also choose whether we want to impact first turn advantage.
It is possible to actually entirely eliminate first turn advantage, but because it is a factor of support it may not be wise to completely do this.
The way we do this is by giving the 2nd player extra actions. If we have player 2 start at 2 actions instead of 3, player 1 may only ever be 0.66 of a turn ahead, instead of a full turn.
If we instead give the player a total of 3 extra turns they can use at any time (up to one per turn, three times), we can completely eliminate it. Two to negate it to 0.33 instead of 0.66.
|
|
|
|
