|
On March 05 2015 07:30 Ketara wrote:Well, that's not the way armor is calculated anyway, it's calculated through Riots silly "Effective HP" thing. http://leagueoflegends.wikia.com/wiki/ArmorBut the short version is that yes, no matter how much armor you have or don't have, 0.6 more armor is still just as effective.
But that's plain wrong. I'll just take a simple example to illustrate. Going 0->50 armor means your damage reduction goes 0->33%: those 50 armor take away a third of the damage. Going 50->100 armor means your damage reduction goes 33-50%. Those extra 50 armor takes away a sixth of the damage. 50 armor in each case, are you going to argue that blocking a third or a sixth of the damage is just as effective?
|
Actually, yes.
Lets say I have 1000 HP.
Lets say that attacks are doing 100 damage to me.
If I have 0 armor, it takes 10 hits to kill me.
If I have 50 armor, each hit does 66.6 damage, and it takes 15 hits to kill me, so my armor has made me 50% harder to kill.
If I have 100 armor, each hit does 50 damage, and it takes 20 hits to kill me, so my armor has made me 100% harder to kill.
Every 1 point of armor makes you 1% harder to kill, no matter how much armor you have.
|
On March 05 2015 07:50 Ketara wrote:
Actually, yes.
Lets say I have 1000 HP.
Lets say that attacks are doing 100 damage to me.
If I have 0 armor, it takes 10 hits to kill me.
If I have 50 armor, each hit does 66.6 damage, and it takes 15 hits to kill me, so my armor has made me 50% harder to kill.
If I have 100 armor, each hit does 50 damage, and it takes 20 hits to kill me, so my armor has made me 100% harder to kill.
Every 1 point of armor makes you 1% harder to kill, no matter how much armor you have.
You were dying in 15 hits, now you're dying in 20 hits, that's not a 100% increase.
With your maths, if every 50 armor was removing a third of the damage, all I'd need is to get 150 armor to be invincible.
|
20 is 100% more than 10.
Let me ask you this. If the more armor you had the less valuable armor was, why would any champion ever buy 5 tank items?
|
On March 05 2015 08:04 Ketara wrote:
20 is 100% more than 10.
Yes, except I don't want to how much it's worth to go 0->100 armor but how much it's worth to go 50->100 armor (which obviously doesn't depends on whatever happens if you have no armor).
On March 05 2015 08:04 Ketara wrote:
Let me ask you this. If the more armor you had the less valuable armor was, why would any champion ever buy 5 tank items?
There is still multiplicative scaling with HP and it helps you fill your role.
It may come as a shock to you, but many offensive stats (ad, ap, as, crit chance) have the same issue, each point gives less % damage increase than the previous one. It doesn't mean getting more than one ap item is terrible.
|
On March 05 2015 07:30 Ketara wrote:Well, that's not the way armor is calculated anyway, it's calculated through Riots silly "Effective HP" thing. http://leagueoflegends.wikia.com/wiki/ArmorBut the short version is that yes, no matter how much armor you have or don't have, 0.6 more armor is still just as effective. I think the disconnect here is that Roffles is thinking that 0.6 armor is ass, and 1 damage reduction is not ass. The reality is that they're both ass, but one is ass that scales and the other is ass that does not.
Damage = Base damage / [1+(armor/100)]. So if you have 50 armor you take 1/1.5 if you have 100 you take 1/2 if you have 150 you take 1/2.5
Invert that and you get riots "effective HP" thing which is just another way of talking about armor, albeit one which isn't particularly efficient if you're doing "what should i buy/choose" maximization.
In this case NpG is correct. The effective amount of damage reduced depends on the amount of armor you have.
E.G. suppose you go from 50 to 51 to 52 armor and take 100 damage at each point. At 50 you take 66.66 damage at 51 you take 66.22 damage. At 52 you take 65.789
Or, going from 50 to 51 you reduced damage by .4451. A .666% increase in durability over 50 armor
and going from 51 to 52 you reduce damage by .4356 A .662% increase in durability over 51 armor
Either way. 5% more bonus stats is actually decent at higher levels if you're buying those stats. Going from 150 to 155 armor is about 2% more durability and as you continue to add bonus stats this increases to 5% more durability asymptotically . (plus its defenses and so is also an effective regen/healing boost)
But its probably not something that you should consider on a mage or over the hard regen from HP pots like biscuits or the early game defense of the flat damage reduction
|
I'm not going to involve myself too much into this because I really can't be bothered but the tl;dr is that Ketara is right. The difference between 98% reduction and 99% reduction is greater than the difference between 10% reduction and 70% reduction.
|
On March 05 2015 21:17 the p00n wrote:
I'm not going to involve myself too much into this because I really can't be bothered but the tl;dr is that Ketara is right. The difference between 98% reduction and 99% reduction is greater than the difference between 10% reduction and 70% reduction.
What do you mean by greater? I don't get how doubling your survivability is greater than tripling it.
|
On March 05 2015 21:17 the p00n wrote:
I'm not going to involve myself too much into this because I really can't be bothered but the tl;dr is that Ketara is right. The difference between 98% reduction and 99% reduction is greater than the difference between 10% reduction and 70% reduction.
No. Ketara is applying a fixed reference point where applying one does not make sense. The fixed reference point he is using is the champions base health with zero armor. Which we don't care about because we care about the margin. And at the margin we're at +some amount of armor.
The 5% more bonus resistance gives you more survivability like so
Survivability With/Survivability Without = Damage Taken Without/Damage Taken With
Damage Taken With = DMG/(1+base armor + Bonus armor x 1.05)
Damage taken without = DMG/(1+base armor + Bonus armor)
So the ratio is (1+base armor + Bonus armor x 1.05)/(1+base armor + Bonus armor)
As you buy bonus armor it goes up. As you gain levels it goes down (as base armor is more of a component). Asymptotically this moves to 5%, though you typically won't get that much armor/mr ever.
We can however look at two examples. Level 13 Viktor with Scaling Armor runes and Zhonya(I.E. the KurO vs AD assassin build). 69.5 bonus armor 66.5 base armor. 5% more bonus armor is worth about 1.47% more survivability. Pretty clearly not worth it over getting the bonus HP early from pots.
Level 18 Maokai with Frozen Heart, Randuin's Omen, Sunfire Cape, and Thornmail, scaling armor yellows, flat armor quints, 21 defense in the middle of 6 enemy champions. 96.7 Base Armor. 383 Bonus armor = 3.3% more survivability. Might be worth it there.
|
On March 05 2015 07:50 Ketara wrote:
Every 1 point of armor makes you 1% harder to kill, no matter how much armor you have.
Everyone 1 point of armor makes you 1% harder to kill relative to your base amount without any armor, but not relative to your previous armor level.
50->100 is the same as 150->200 relative to being at 0 armor, but 50->100 relative to 50 armor is more than 150->200 relative to 150 armor, which is what we actually care about.
|
No, what you actually care about is a mastery page for Viktor, and somehow this has turned into a giant semantics argument that I refuse to continue to take part in.
|
On March 06 2015 07:21 Ketara wrote:
No, what you actually care about is a mastery page for Viktor, and somehow this has turned into a giant semantics argument that I refuse to continue to take part in.
Sure but you described the effect of the mastery page for Viktor incorrectly.
|
Armor doesn't have diminishing returns and %increase in damage isn't as important as flat increase in damage. There's so much misinformation in this game on how multiplicative scaling works.
|
On March 07 2015 14:57 obesechicken13 wrote:
Armor doesn't have diminishing returns and %increase in damage isn't as important as flat increase in damage. There's so much misinformation in this game on how multiplicative scaling works.
Short answer: No. Armor has diminishing returns
Long Answer: It depends on how you define "returns". If we define it from the static point of "your current HP" then no, armor doesn't have diminishing returns. You gain the same EHP per point of armor. But this has problems. Because then we have an issue that Health increase the returns from armor. Our point really isn't static
If instead we define based on the margin then armor does have diminishing returns. One point of armor does not increase your tankiness as much as the prior point of armor. This had advantageous properties. The biggest is that the armor return no longer depends on the amount of HP i have. If i go from 20 to 30 armor i increase my EHP vs armor by 8.3% (assuming no pen). I can then compare that to buying health directly. If i have an item which increases both because they're unitless increases i can product rule them together.
Whether or not you want to explicitly use the second framework, as soon as you start maximizing with regards to armor and HP you will find that you're working under the effects of that framework anyway, you're just doing more work for no reason
Edit: maybe a concrete example. You think you care about EHP multiplier, which is X = 1+ armor/100. This is linear. It's first derivative is positive (positive returns) and its second derivative is 0(constant returns)
But don't actually give a fuck about that. We care about is DR which is (1-1/x). The first derivative of this is 1/(x^2). Positive, so positive returns. The second derivative is -1/(x^3) negative, so decreasing returns. So it looks like armor has decreasing returns.
|
On March 08 2015 18:00 Goumindong wrote:Show nested quote +On March 07 2015 14:57 obesechicken13 wrote:
Armor doesn't have diminishing returns and %increase in damage isn't as important as flat increase in damage. There's so much misinformation in this game on how multiplicative scaling works. Short answer: No. Armor has diminishing returns Long Answer: It depends on how you define "returns". If we define it from the static point of "your current HP" then no, armor doesn't have diminishing returns. You gain the same EHP per point of armor. But this has problems. Because then we have an issue that Health increase the returns from armor. Our point really isn't static If instead we define based on the margin then armor does have diminishing returns. One point of armor does not increase your tankiness as much as the prior point of armor. This had advantageous properties. The biggest is that the armor return no longer depends on the amount of HP i have. If i go from 20 to 30 armor i increase my EHP vs armor by 8.3% (assuming no pen). I can then compare that to buying health directly. If i have an item which increases both because they're unitless increases i can product rule them together. Whether or not you want to explicitly use the second framework, as soon as you start maximizing with regards to armor and HP you will find that you're working under the effects of that framework anyway, you're just doing more work for no reason Edit: maybe a concrete example. You think you care about EHP multiplier, which is X = 1+ armor/100. This is linear. It's first derivative is positive (positive returns) and its second derivative is 0(constant returns) But don't actually give a fuck about that. We care about is DR which is (1-1/x). The first derivative of this is 1/(x^2). Positive, so positive returns. The second derivative is -1/(x^3) negative, so decreasing returns. So it looks like armor has decreasing returns.
Yeah sometimes you buy health when you have enough armor. That's fine, but the discussion in this page so far has been talking about how a 25% reduction in damage (0->33 armor) is better than a subsequent 15% reduction in damage (100->133 armor) when in reality the flat increase in ehp is the same.
Diminishing returns doesn't care about %increase/decrese it's about flat increases/decreases.
Finally your post is going on a tangent. Everyone knows you buy resists and health together.
|
I... I just explained to you, in both math(i literally did the derivatives to show you!) and in practice how it has diminishing returns. I don't know what else to say.
|
If you have 1000 HP, and are taking 100 physical damage per second.
At 0 armor it takes you 10 seconds to die, at 100 armor it takes you 20 seconds.
Going 0->33 takes you from 10 seconds to 13.3 seconds.
Going 100->133 takes you from 20 seconds to 23.3 seconds.
In either case 33 armor made you survive 3.3 seconds longer.
There are no diminishing returns.
Yes, 3.3 seconds is a smaller percentage of 20 than it is of 10, but that has absolutely nothing to do with the discussion, which is why I called it a stupid semantics argument. It has derailed the thread, is of no use to the champion discussion, and really needs to stop.
|
I thought this was the goddaman Viktor thread.
How about people create a math thread if they want to talk about that instead? Also that there aren't diminishing returns in terms of absolute values don't necessarily mean that relatively speaking you can't start getting less out of your buys.
In other news, after some more tests I'm 0K with augmenting Q after sorcs+morello if I don't immediatly need zhonya's or void staff to be able to do something. If it lines up with the roaming/skirmishing part, or if you get a match-up that lets you shove then move to gank sidelanes then the ability to chase/kite is really nice.
|
|
|
|
|
|
EPT
