Maybe you could approach balance by looking at how other comparable shields reduce damage over time. There are four other permanent shields in the game that only reduce damage and have a fixed cost: shield (generic), diamond/titanium, fog, and dusk. Costs are:
unupped:
1DR: 1
2DR: 4
40%: 2
50%: 6
upped:
2DR: 5
3DR: 6
40%: 1
50%: 4
Average unupped creature damage is just under 4, and average upped creature damage is just over 5 (assuming an even distribution of creatures.) So if we wanted a more complete picture, we could compare these shields based on average damage reduced:
unupped:
1DR: 1
1.6DR: 2
2DR: 4
or 6
upped:
2DR: 5
or 1
2.5DR: 4
3DR: 6
For your shield, you might shoot for an average of about 50% DR for a cost of four
, or for 17% DR (~.6|.8DR), a cost of less than 1
-- which probably means your prototype damage reduction is too low.
Next you would want to think about the time aspect. If you assume that a game lasts some average amount of time after a shield is played -- lets say 5 rounds -- you could balance the damage over time based on this. Compared with other shields with the same overall cost (including
spent), your shield should have a similar average effective damage reduction after five rounds. You could calculate the average by assuming the same incoming damage each turn.
Some possibilities might be:
6
|4
Absorb 80% damage (~4DR). Deals 15% of damage absorbed each turn (~50%/2.5DR after 5 turns).
5
|3
Absorb 60% damage (~3DR). Deals 10% of damage absorbed each turn (~45%/2.2DR after 5 turns).
3
|2
Absorb 40% damage (~2DR). Deals 5% of damage absorbed each turn (~35%/1.8DR after 5 turns).
Of course you could also switch the wording around:
6
|4
Absorb 4 damage. Deals 1 damage each turn for every 7 absorbed.
5
|3
Absorb 3 damage. Deals 1 damage each turn for every 10 absorbed.
3
|2
Absorb 2 damage. Deals 1 damage each turn for every 20 absorbed.
When you change it to numbers rather than percentages, you can see that the
skill makes more sense for a higher absorb shield, that the stacking idea makes more sense for a flat-reduction shield, and that the damage over time should be relatively high for the
skill reduction to make a significant contribution (at least 1dmg/turn).
