Defensive mechanics suggested improvements

I think there’s a pretty wide consensus that Block is a slightly broken mechanic, since it’s completely sidelined except with the use of one specific unique.

Endurance is a bit more marginal, since it has some uses, but generally it’s eclipsed by health and also imo a little to similar to health in its use cases.

Resistances are not in a terrible state, but imo there is some missed potential, so I will make some suggestions for those too.


Infamously, you need a Bastion of Honour for Block to be worth bothering with as a mechanic. We can see 3 obvious reasons for this:

  • Monopolises a very valuable slot. In many cases you can expect to lose half your DPS just as a price of entry.
  • Even once you’re in, it’s fairly expensive to scale, needing investment in 2 stats which are scattered across many affixes and passives (if your class supports Block at all).
  • Even with that massive investment, it’s generally still quite an unreliable defensive layer.

Accordingly, BoH dominates since it throws the 2nd and 3rd problems out the window. It seems a bit unhealthy for a defensive layer to have such massive problems + then the one way to use it properly is a magic uber-unique that trivially solves them.


  • If we treat the reliability of BoH as the aspirational “perfect” source of block, we need a mechanic that allows other sources of block to approach that perfection in a reasonable and constrained way.
  • Or to put it another way: 50% chance to block an Embermage hit does not feel half as good as 100% chance. It doesn’t really feel good at all. The problem is that when huge burst and oneshots are involved, the gap between “reliable” and even slightly “unreliable” is a huge cliff. What we want is to introduce an idea of “partially reliable”, or “reliable if you apply proper care and tactics”.
  • It’s expected that this should be a large buff for non-BoH shields. Simply put, they are trash. They need it.
  • Suggestion:
    • Guaranteed Block will block a hit that you failed to block normally. This is an intrinsic mechanic for all shields.
    • Guaranteed Block has a cooldown (base 5 seconds)
    • Guaranteed Block can have charges (base 1)
    • Guaranteed Block cooldown efficiency and extra charges can be introduced to affixes, shield implicits and class passives. (I’m imagining the cooldown eff. values could be pretty high, so the cd can get quite low with proper investment)
  • Rationale:
    • Adds a smooth graduation of generic reliability to Block without invalidating the existing design and scaling around Block Chance; Block Chance is still very good since it lets you keep up charges while taking hits from a group.
    • Lets other shields do Bastion of Honour’s “thing” (ie. be non worthless) without actually stealing its “thing” or making it an uninteresting item
    • Promotes a little thinking/tactical playstyle (“I can tank Embermages, but only if I’m not running through a pack of trash at the same time”)
    • Can support some more diverse identity for shield bases and affixes
      • Has some room for further extension with spicy affixes. “eg. 5% less Damage over Time taken per available Guaranteed Block charge”
    • Adds some more interesting itemisation + indirectly helps to shore up other unreliable defensive layers like Dodge


Right now, Endurance plays 2nd fiddle to health. With capped Endurance, 1 point of Endurance Threshold has equivalent value to 1.5 points of max HP. Whereas a point of added health is worth (1 hp * inc health) and (0.2 endurance * inc. health) which sums to 1.3 hp * inc health with 60% Endurance. Meaning you only need 16% or so increased health before Endurance Threshold falls behind health as a source of EHP in one time instance. In practice you can easily hit far higher values of

Endurance does have a distinct advantage over health - while under the threshold, your health recovery and leech also benefit from that 60% DR (or 250% EHP multiplier). However the ability to reach extreme levels of recovery & prevalence of oneshot mechanics make this benefit fairly underwhelming in practice. In practice, you want to have enough sustain to keep your health topped off, so a multiplier to sustain at low health isn’t really impressive.


  • Replace the free Endurance Threshold from max health & replace it with flat Endurance Threshold scaling with your level. This element skews the math and makes Endurance Threshold a deceptively pointless stat relative to health even before we started. Taking it away frees us up to buff Endurance as a better stat/mechanic in its own right.
  • Implement more mechanics that interact directly with Endurance Threshold
    • Directly buffing the efficiency of Endurance would just be flipping Endurance Threshold from “inefficient flat health” to “more efficient flat health”
  • Potential formula change:
    • Damage taken while below your Endurance threshold is divided by your Endurance + 1.
    • This gives Endurance a toned down linear EHP scaling, allowing the cap to be removed. This makes various passives/skills & particularly exalted affixes that give Endurance a bit more exciting.
    • Sources of Endurance could be buffed somewhat, since you need to hit 150% Endurance to have equivalent effect to the current cap.
  • Example of a new supporting mechanic:
    • Iron Guard stacks last for 4 seconds and grant 2% more Endurance. You cannot gain Iron Guard stacks while below your Endurance threshold. Max 10 stacks.
    • (under the “old” Endurance formula, this would apply after the cap, allowing up to 72% endurance)
    • This could be a Sentinel-specific mechanic tied into passives/skill trees comparable to the various Shroud abilities for the Rogue. Sources of stacks should be melee focused, especially melee skills which have to stand still to attack.
    • The idea is that it more strongly characterises Endurance in the role of oneshot protection, since it’s a fairly significant DR but the stacks quickly decay while you are below endurance threshold and actually benefiting from it.

Resistances are not a particularly broken mechanic, but I think the itemisation game could still become a bit more interesting if they were improved somewhat.

Specifically, I see 2 areas for improvement

  • Exalted resistance affixes are worthless, outside of some niche cases where a completely different mechanic adds value. Many skills/passives which provide resistances are also very neglected. More generally, there’s an issue where Health is the god defensive suffix and others can be neglected. We need to spice things up, so all the defenses have a role and it doesn’t just converge to 1 trivial solution.
  • There are various instances of “Reduced damage taken of X element”. This has an awkward overlap with resists and doesn’t fit easily onto character sheets. Yet this secondary defensive layer is much stronger than resistances, a primary layer with similar terminology.
  • Bonus: “X Damage taken as Y” is quite a cool-sounding concept but it’s not worth bothering with in practice. A perfect resistances rework would also address this issue.


  • Instead of subtracting resistance from damage, divide the damage by (resistance+1). (yea, it’s the same as my suggestion for Endurance. The logic is also similar)
  • Remove the resistance cap. It’s no longer necessary, since the benefit of high resists is linear (and in practice high resists get less efficient relative to other defensive stats) rather than “exponential” (not the proper maths term but close enough)
  • When resists are negative, instead multiply the damage by (abs(resistance)+1)
  • “X% Less Y damage taken” can be replaced by either flat resists, or a multiplier to resists (or some combination). This way, the ability appears on the character sheet. And also participates in the resist system instead of making it look dumb by being twice as useful.
  • Intrinsic penetration on enemies could be removed, since this formula accomplishes a roughly similar scaling for the first 75% resistances in a way that’s more transparent to the player
  • Rationale
    • All sources of resistances are now worth looking at (but not overbearing) as a defensive layer. Blessings, passives etc. can be something that synergises with resists on items rather than simply replacing them. Particularly, t6/7 resist affixes can now be kind of cool instead of mostly trash.
    • Since it’s possible to seriously invest in specific elemental defenses, “X damage taken as Y” become affixes much more competitive and interesting. It’s not like it will be OP, those prefix slots are worth a lot of DPS.
    • “Reduced X damage taken” can be replaced with something that more neatly integrates with the existing defensive layers rather than invalidating them.
    • Altho this does probably fairly significantly raise the theoretical ceiling on durability, I don’t think it should be that game-breaking in practice, since any resources redirected to pushing resists will come away from damage or other sources of defenses. It’s also possible to tweak resists affix values once they’re seen in action.

There isn’t, Paladins can get 67% block chance from an Ironglass shield with a t5 block chance prefix, 2 rings with block chance suffixes & gloves with block chance suffix. Plus an additional 12%-15% block chance (for a total of 79% - 82%) if you want to engage with the “fun & engaging gameplay” that is keeping up 4x Sigils of Hope, plus an additional 15% for 4s from Shield Bash (can get up to 100% up time if you want to take the no-cooldown–but-25-mana-cost node) which would put you within a gnats chuff of 100% block chance. Not saying it would be fun & it would be much more “engaging” (ie, hard work) that just equipping a Bastion.

Plus, just because “everybody” believes a thing, doesn’t make it true.

You don’t, as I’ve mentioned. It’s just that a Bastion makes it trivial to get 100% block chance compared to the significant investment a non-Bastion block build would require. Do I think that’s unreasonable? Absolutely. Bladedancers can get up to 85% chance to take zero damage from dodge & that’s much easier to get & keep up than a non-Bastion build.

In addition to your suggestion, they could nerf Bastion’s 100% block chance down to ~75% (ish) so that non-block builds would still get some benefit from it but not as much & block builds (ie, Sentinel) would be able to reasonably easily get that up to 100% block (with mainly just passives).

I’m not sure I agree with this. a max-roll T7 resist affix will cap you. Given most masteries can get some resists from passives (which I always do since it makes gearing easier), 1 t6 affix (per resist) would get you capped on those resists.

Skills, yes, but I like the passives that give resists & usually take them where possible, so, sure, buff them if you want. :wink:

This would be a masisve change, given that mobs get up to 75% resist penetration per level. So while your character screen might say you have 75% resist, the game treats it as 0%.

Does it? What happens if you’re undercapped? Or even capped! If you’ve got 75% resist, you’d take 100 / (1.75) = 57% -1 = 42.5% damage, why should I only take 42% damage if I have 75% resist? You’d need to have 300% resist to take 25% of the damage (what you’d expect given the current “75% resist” cap). What would you cap resists at? Would you cap them at all?


I’m getting confused reading your comments on the block thing lol. I don’t think nerfing Bastion of Honour really fixes anything. Imo it’s not so much that nobody uses other shields because BoH is better, rather that BoH (or okay, some theoretical perfect rolled t7 exalt Ironglass that’s even rarer than BoH) is the only shield that’s good enough to make shields worth the opportunity cost. If anything, BoH being so strong is good because it sets a standard for other shields to be raised (closer) to, to the point that they might actually be viable.

The important thing is to understand that the intrinsic pen on enemies is fundamental to the current resist formula. So removing the intrinsic pen is a key part of what I’m suggesting.
Current system (lvl 100 area), with intrinsic pen on enemies

  • Player resist 0 → effective resist -0.75 after pen → 1.75x damage taken
  • Player resist 75 → effective resist 0 → 1x damage taken

Alternate formula, without intrinsic pen on enemies

  • Player resist 0 → 1x damage taken
  • Player resist 75 → 1/1.75x (~0.57) damage taken

If you increase enemy damage by 75% (okay, that’s a pretty important thing I forgot to mention), the alternate formula then outputs the same as the current one(with pen) at 0 and 75 resist. The scaling curve between those values is a bit diffeerent but not in a bad way imo.

Then if all you’re doing is getting back to the same point we’re is at the moment, why bother changing anything? Plus since the enemy area level penetration has been removed you’d get back to the reason why the devs added it in the first place, so that being 1% under cap means that you only take 1% more damage (as against the 4% you’d take in PoE if you had 74% resist compared to 75%).

I can list the amount of block a Pally would get from passives/gear/skills if that would help.

1 Like

I realise I made a big mistake in my post. The point of the alternative formula is to ensure that resistances scale in a sensible (linear) way, so that a cap is not necessary. Ie. the whole suggestion was “change the formula and remove the resistance cap”, but I forgot the important half…

This isn’t a problem - if you do the math, the proposed formula has a low difference between eg. 74 and 75 resist without the need for the penetration mechanic.

Oor you could invest those resources in HP, use a 2h weapon and a titan heart and deal 3 times the damage and probably still be tankier than if you used any shield except BoH. The fact that a certain block chance is attainable has to be interpreted in terms of the opportunity cost which was paid to get there. As it is, block is not worth the opportunity cost except with BoH which has a cost-benefit ratio several times better than any normal shield.

So you want to go back to the protections (or something quite similar to it) system that was in place before the devs moved to resistances. The problem with that is that the stat can’t be a % 'cause if it does it wouldn’t make any sense (if I have 95% resist, why am I not only taking 5% damage rather than the ~48.7% I would using your formula).

Personally I quite liked protections, but lots of other people were confused by it & it made protections/resists pointless for ward builds.

I didn’t think block had a problem (Many people are fine just using a 50% block chance shield). But i really like your solution for it. The other stuff I’m not really sold on.

At the very least your block solution could be it’s own shield but i agree it would be much cooler if you could see those stats in many different places

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.