Enter your unit's stats to see which perk gives more EHP.
Assuming the base +8% HP and +12% DEF are roughly equal for your build (use the calculator above to check), the choice comes down to the conditional bonus:
Defence converts to damage reduction on a diminishing-returns curve. The first 100 DEF buys 50% reduction; doubling to 200 only reaches 66.7%. Reaching 90% costs 900 DEF, and 99% costs 9,900. The formula ensures every point of DEF always helps, but each one helps a little less than the last.
A percentage-based DEF perk doesn't give flat returns. At low DEF, 12% of a small number is still small. At high DEF, diminishing returns eat into the gain. The perk peaks around 100 DEF at roughly +2.8% damage reduction, then tapers. This hump shape — unlike a pure derivative — is what makes the perk's value position-dependent.
Substituting the DR formula back into effective HP reveals a perfectly straight line. Every point of DEF gives the same survivability gain as the last — 100 DEF = 2× TTK, 500 DEF = 6×, 900 DEF = 10×. This linearity is the entire purpose of the DEF/(DEF+100) formula: it makes DEF stacking feel consistently rewarding without ever becoming broken.
HP% bonuses are additive with each other, so the TTK multiplier from HP stacking is also a straight line — 50% HP bonus = 1.5× TTK, 100% = 2×. However, because each additional +8% HP is added to an ever-growing sum, the marginal EHP gain per perk diminishes: the first +8% gives 8% more TTK, but at 100% existing HP% it only gives 4%.
With the game's actual perk ratio of 12% DEF to 8% HP, the DEF perk dominates for most mid-to-late game builds. At Hugglebeary's 222 DEF and 68% HP%, the +12% DEF perk gives +8.30% EHP versus +4.76% from +8% HP — DEF wins by 3.54%. The breakeven line shows that above 200 DEF, even 0% existing HP% favours DEF.
At equal 1:1 ratios, HP is far stronger — you need 100 DEF just to break even with 0% existing HP%. At Hugglebeary's stats, the two perks are much closer than in the 12-vs-8 comparison. The devs intentionally gave DEF a 1.5× multiplier (12% vs 8%) to compensate for the +100 constant in the formula always diluting DEF's value.
Simple Defense in HNA passes through three multiplicative layers. First, the attacker's ATK is reduced by the target's DEF via the familiar DEF/(DEF+100) formula. Then each source of damage reduction multiplies independently — a 5% and a 20% DR source give 1−(0.95×0.80) = 24% total, not 25%. k represents other variables, which we will explore in a seperate subsection
| Attacker | Target | DR% | Formula | Expected | Actual |
|---|---|---|---|---|---|
| Muscimmon (119 ATK) | Apollinis (161 DEF) | 0% | 119×100/261 | 46 | 46 |
| Muscimmon (119 ATK) | Apollinis (161 DEF) | 3% | ×0.97 | 44 | 44 |
| Muscimmon (119 ATK) | Apollinis (161 DEF) | 5% | ×0.95 | 43 | 43 |
| Muscimmon (119 ATK) | Apollinis (161 DEF) | 20% | ×0.80 | 36 | 36 |
| Bladehound (204 ATK) | Apollinis (161 DEF) | 0% | 204×100/261 | 78 | 78 |
| Bladehound (204 ATK) | Apollinis (161 DEF) | 20% | ×0.80 | 63 | 63 |
| Bladehound (204 ATK) | Apollinis (161 DEF) | 3%+20% | ×0.97×0.80 | 61 | 61 |
| Bladehound (204 ATK) | Nomnom (55 DEF) | 0% | 204×100/155 | 131 | 131 |
| Bladehound (204 ATK) | Nomnom (55 DEF) | 5% | ×0.95 | 125 | 125 |
| Bladehound (204 ATK) | Nomnom (55 DEF) | 30% | ×0.70 | 92 | 92 |
| Bladehound (204 ATK) | Nomnom (55 DEF) | 5%+30% | ×0.95×0.70 | 87 | 87 |
Damage Reduction sources are 3%/5% Aspect Synegies for Pride(2)/Pride(3) and Apollinis 20% Damage Reduction while skill is active, Nomnom 30% Damage Reduction while skill is active.
Because DR stacks multiplicatively, two 10% sources give 19% total DR (0.9×0.9 = 0.81), not 20%. A single 20% source is strictly better — it gives the full 20% reduction. The gap widens with larger values: two 15% sources give 27.75% total, while one 30% source gives the full 30%. Always prefer one large DR source over two smaller ones that add up to the same value.
HP% gives a 1:1 linear TTK increase — +50% HP means +50% TTK. DEF% is also linear but with a weaker slope that depends on base DEF (at 222 base: slope ≈ 0.69). DR is the outlier — it accelerates. Each additional % of DR gives more TTK than the last, making it the only survivability stat with increasing returns.
To get the same survivability as +8% HP, you need only 7.41% DR — DR is more efficient point-for-point because of its accelerating curve. DEF% requires more than 8% because its slope is always less than 1 (it depends on base DEF).
DR is always 7.41% regardless of existing bonuses — it scales independently. HP% and DEF% both need more the higher your existing bonuses are. This demonstrates why DR is the most efficient survivability stat to stack.
| My unit | Opponent | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Unit | Lv | ATK | DEF | MDEF | DEF% | MDEF% | Unit | Lv | ATK | DEF | MR | DEF% | MDEF% |
| Buffs active | Σ DEF% | Expected | Actual |
|---|---|---|---|
| 12% disposition | 12% | 70.56 | 70 |
| 24% disposition | 24% | 78.12 | 78 |
| 12% disposition + 40% skill | 52% | 95.76 | 95 |
| 24% disposition + 40% skill | 64% | 103.32 | 103 |
Formula: Total DEF = Base DEF × (1 + Σ DEF%). Values truncated to integer in-game.
| Build | Σ HP% | Expected | Actual |
|---|---|---|---|
| Aspect STR(1) | 15% | 2,201 | 2,201 |
| DOM(1) | 23% | 2,354 | 2,345 |
| STR(3) | 24% | 2,373 | 2,374 |
| DOM(3) | 36% | 2,603 | 2,603 |
| STR(5) | 38% | 2,639 | 2,632 |
| DOM(5) | 56% | 2,986 | 2,991 |
| DOM(5) + UGL(2) Mr Avocado Buff | 68% | 3,216 | 3,221 |
Formula: Total HP = Base HP × (1 + Σ HP%). Minor variance (≤5) likely from hidden decimal precision on base stats or buff values.