Why 100% reduced enemy stun threshold makes you stunn every attack?

"

ScrotieMcB написал:

Exactly, it's a number used to calculate. In the case of damage/defender_effective_max_life, a zero results in an error; in the specific case of 100% reductions, the game often adjusts the value to 1. I remember seeing Mark say exactly the same thing about stun threshold reduction in a previous thread.

Ah, decent point. Though, if they're using IEEE floating point values for these numbers, so long as the damage isn't also 0, there won't be a problem (you'll just get Infinity and the comparison will always be true). If the damage was 0 as well, and assuming a non-signalling floating point implementation, then the division would produce NaN, and all comparisons with that value produce false (according to IEEE floating point), so presumably you wouldn't get a stun in that case.

However, you might be right about the edge case being treated separately just to avoid abusing the semantics of floating point arithmetic like that.

"

MesostelZe написал:

"

ScrotieMcB написал:

Exactly, it's a number used to calculate. In the case of damage/defender_effective_max_life, a zero results in an error; in the specific case of 100% reductions, the game often adjusts the value to 1. I remember seeing Mark say exactly the same thing about stun threshold reduction in a previous thread.

Ah, decent point. Though, if they're using IEEE floating point values for these numbers, so long as the damage isn't also 0, there won't be a problem (you'll just get Infinity and the comparison will always be true). If the damage was 0 as well, and assuming a non-signalling floating point implementation, then the division would produce NaN, and all comparisons with that value produce false (according to IEEE floating point), so presumably you wouldn't get a stun in that case.

However, you might be right about the edge case being treated separately just to avoid abusing the semantics of floating point arithmetic like that.

As a programmer, I would always treat the edge cases like this separately if I had even a hint of an inkling that it might possibly occur:) Better safe than sorry!

This is one of those binary things.

Or it works absolutelly up to 100%, or its not worth it.

Back when stun threshold just multiplied your damage for the purpose of testing if a stun would occur, being independent of enemy life it sucked big time.


The cool thing to do though, is not use ground slam or maces or two handed weapons.

Try it with your minions, totem or bow.

15% rare belt =
10% facebreaker = unique gloves
15% broadstroke = unique quiver

Total = 40%

3 5% global links near marauder = 15% =

Total = 55% (for just 7 passive points out of the way, or none if one goes for Blood Magic)

stun gem = 39%

Total = 94%

You can then surpass the 94% if you find Stun Threshold on the Bow.

Just two medium rolls on weapon and belt, plus the 2 uniques and the stun gem should be enough.


For maces its easy, but then again, maces are not lightning arrows... uhmmm

"

Interesting написал:

The cool thing to do though, is not use ground slam or maces or two handed weapons.

Try it with your minions, totem or bow.

Although I didn't test minions or Ranged Attack Totem, I have tested spells (well, spell... Freezing Pulse) and, similar to life leech, they don't receive reduced stun threshold bonuses from generic passives or gear (the Stun support still works).

However, with a maxed Stun support, unique quiver, unique gloves, all three 5% marauder-area passives, and a max belt suffix, you can get 99% reduced stun threshold on bows, which should still be enough to stunlock everything ever (if you're doing less that 0.5% max enemy health per hit you might have bigger problems).

I'm guessing the 100% is reducing it to 1 instead of 0, unless you want to divide stuff by zero (which you shouldn't).

"

Interesting написал:

You can then surpass the 94% if you find Stun Threshold on the Bow.

Just two medium rolls on weapon and belt, plus the 2 uniques and the stun gem should be enough.


For maces its easy, but then again, maces are not lightning arrows... uhmm.

I dont think bows and wands can roll reduced stun threshold mod.

Tried to find one, wasn´t able to.

This is just good, cause split arrow, chain + 100% threshold reduction would be way too broken.

I Think you can get too close though..

"

ScrotieMcB написал:

However, with a maxed Stun support, unique quiver, unique gloves, all three 5% marauder-area passives, and a max belt suffix, you can get 99%

That´s 94%. Still close though.

"

MesostelZe написал:

"

ScrotieMcB написал:

"

Mark_GGG написал:

Stun threshold reduction reduces the defender_effective_max_life compared to their actual max life. If that value is reduced by 100%, then by definition it's reduced to 1, and any hit stuns.

That's what you meant, right Mark?

If you reduce something by 100% of its value, it's 0. The defender effective max life here isn't the actual life of the mob, it's just a number used to calculate whether a hit is going to stun.

If this would work for energy shield recharge too...

Stun threshold should have the same kind of diminishing return as energy shield cooldown recovery has, since the latter becomes progressively useless the more you stack and the former gets exponentially more useful up to the sweet spot where every action stuns which is absurd.

Whats confusing people is this statement:

"If the stun chance would be less than or equal to 25%, it's ignored, so you need to deal more than 12.5% of effective maximum life to have a chance to stun."

They are reading that to mean: If you reduce effective_maximum_life to a point where stun_chance is 25%, it caps out at 25%........and effective_maximum_life does not go down anymore.

In other words, effective max life can never drop below 25%. So you would have to do 200*(12.5% of effective_max_life)/effective_max_life to have a "chance" to stun.
(200*12.5%/25%=100)

Than assuming "chance" in the statement is a typo, and means "100% chance".

The problem, of course, is this only makes sense in %, and adding real values makes this make absolutely no sense at all. And now that I see it, I can see how it confuses people. It makes sense that a mob would cap out at 25% of there life......it would allow scaling that directly relates to the damage you do, which is in-line with a lot of other POE mechanics (burn/shock/freeze). But this would make high life mobs un-stunnable.

I admit, when I first read the mechanics thread, I thought the same thing until I switched the percentages with real numbers and realized how silly I am...... that's why I know what the OP is saying.

-----------------------------------------------------------------------------

And to clarify for the OP, that statements means:

"If the stun chance would be less than or equal to 25%, it's ignored, so you need to deal more than 12.5% of effective maximum life to have a chance to stun."

If you hit a mob for such a small amount that you don't do at least 25% stun_chance (or 12.5% of effective_max_life) the stun will not have a chance to land ever. Reducing stun threshold reduces the % of mob life it takes to have a chance at stunning.