Curse

qwe4rty · Nov 10, 2005, 04:19 PM // 16:19

Quote:

Originally Posted by AncientPC

10% armor penetration with a 10% chance is the same thing as 1% armor penetration 100% chance.

Sure if you only take 10 trials you may have 0-5 armor penetration attacks, but with a large enough sample size 10/10 gives you 1% armor penetration.

Thus 1% armor penetration can be accurately used to describe 10/10 sunder mod's effects. We all know that 1 point in strength is 1% armor penetration 100% chance and how little extra damage that yields. In conclusion 10/10 sunder is useless.

10% chance for 200% adrenaline is the same thing as 20% extra adrenaline 100% chance and thus obviously more useful than sunder.

qwerty: Your argument is sunder is good for the chance to spike damage.

What is the chance that two hits will consecutively do armor penetration? 10% x 10% = 1%

What is the chance that three hits will conseuctively do armor penetration? 10% x 10% x 10% = 0.1%

So you're aiming for that chance<=1% to spike a minimal amount of extra damage? You're going to need a better argument for the sunder mod.

I wasn't necessarily saying it will sunder twice in a row, but twice or more per 10 hits.
Sunder = S
Non-Sunder = N

It could be:
S S N N N etc
S N S N N etc
N S N N S etc
S N N N S etc
S N N S N etc

Let's say for example I'm just chargin up all my adrenaline skills on character X. Enemy monks heal X, protect X, etc.
Then, all of a sudden, I turn around and spike Monk A. With sundering, I have a chance to do all that damage I COULD get from Vampiric, but I don't want to have to hit said Monk 5 times, 6 times, etc to do it. I want to do my spike, and move on to another target. Continually change targets

javik · Nov 10, 2005, 04:23 PM // 16:23

so i have been doing the math and the sundering seems disapointing, especially for a ranger, even if it strikes 10% sundering if you doing about avg 40 dmg per arrow, and best case you expect it to happen 1 in 10,(althogh i understand the whole probablility thing) you would only be getting a couple extra dmg per 10 hits which is really sucky. I know the vamp is good especially for short, but if your not going vamp or zealous, is sundering the next best thing. What im really wondering is if a fire or whatever string on certain mob types will give you slightly more dmg? Has anybody ever tested the fire strings in PvP against non fire armor and so on? Or are non vamp/zealous types stuck with sundering that sucks at 10/10 so you buy the 10/9 which sucks even worse?

qwe4rty · Nov 10, 2005, 04:31 PM // 16:31

Quote:

Originally Posted by AncientPC

If you don't understand the math don't call it convoluted.

However sunder works by having a 10% chance to do 10% extra damage. That is essentially the same as 1% extra damage (10% x 10% = 1%).

Actually, sunder works by having a 10% chance to ignore 10 armor.

Quote:

Originally Posted by twicky_kid

throw out all the garbage you have seen trying to calculate dmg or how often it will occur. all of the calculations and probiblity is not the point. the point is that sundering can only trigger on SKILLS. it is not 10% chance on every attack you do.

Hmmm, well, the mod doesn't actually say that its only on skills, but if theres been research/proof done for this, please direct me.
Even if this is true, I think its a bug.

MuKen · Nov 10, 2005, 04:50 PM // 16:50

Quote:

Originally Posted by Sekkira

I was agreeing with you, I just pointed out some flaws. I'm being pedantic.

No, he said right there that "You are doing more damage in 1 in every 10 hits", which signifies that sundering triggers every 10 hits, which it doesn't. I was merely pointing that out.

Ah, ok, just a misunderstanding I guess.

Quote:

I'm sure it's quite possibly it would be used a great deal more if this was the case. Personally, I prefer not to put faith in probability, no matter what the odds.

However, as I mentioned, this is the wrong way of looking at things. You should not be trying to make your DPS as consistent as possible, quite the opposite, you should be making it as variable as possible. If you assume an equal average dps, the more variable it is the harder it is for a monk to deal with it.

To better see this, let's take it to extremes. Let's say a few of you are attacking a target and dumping out 150 dps. On the one hand, you have it super consistent, you do lots of little hits that perfectly come out to 150 damage every second, smooth enough to look like degen. No problem, the monks will drop a word of healing down whenever your target drops below 50% and they are operating at super-high efficiency.

At the other extreme, let's say every second your team somehow gets a 10% chance of doing 1500 damage that second. Guess what? The monk is totally screwed, there's no way he can deal with that.

Now obviously both of these examples are extremes that will never occur in practice, but the more your damage looks like the second one, the BETTER off you are. A fixation on doing consistent damage is like looking at the problem with tunnel-vision. Your goal is not to maximize your worst-case damage, your goal is to beat the enemy monk.

Quote:

Originally Posted by qwe4rty

Actually, sunder works by having a 10% chance to ignore 10 armor.

This is incorrect, sunder works by having a 10% chance to ignore 10% of the enemy armor. Against a 60 armor opponent, this is equivalent to removing 6 armor, about 11% damage increase.

Conveniently enough, on a 100 armor opponent, the percentage damage increase for removing 10 armor almost doubles, so since the base damage is halved, the raw number of damage points it has increased is almost the same (a little bit less). Thus, within the range of armors you are likely to run across, small % armor penetrations will tend to put out roughly consistent damage increases regardless of enemy armor.

OrangeArrow · Nov 10, 2005, 04:52 PM // 16:52

well I'm not a fan of sundering myself but I feel someone needs to play Devil's Advocate.

While Sundering is a terrible 1% upgrade in damage it is still an upgrade in Damage in COMPARISON to any other upgrades (except Vampiric).

It does more damage and has more beneficial effects than most of the other upgrades except Vampiric and Zealous.

The fact of the matter is if you wish to keep your damage PHYSICAL there is only 4 Upgrades worth considering: Vampiric, Zealous, Furious, and Sundering.

All the other upgrades that lengthen Condition duration in most cases do nothing. In most cases when you apply a condition 3 things will happen:
1. Target will die before condition naturally expires. (Most Likley in PvE)
2. Condition will get removed. (Most Likley in PvP)
3. You will reapply the condition before it naturally expires.
These 3 Things will happen a large majority of the time so unless your trying to run some kind of mass condition team build or some very specific condition build the condition upgrades are next to useless.

In the hands of a good player the negative regen effects of Vampiric and Zealous Upgrades can be overcome but in the hands of an inexperinced or lazy player the upgrades carry too much of an negative Stigma to be worth using in their eyes.

While Vampiric and Zealous are light years ahead in actual effectiveness your avg player will probabaly chose to use Furious or Sundering because its easier to use and doesnt carry with it any "negative" effects.

The truth of the matter is people are lazy.
Still it is better to use an upgrade with an 1% damage increase is better than no upgrade or an upgrade with no real beneficial effects.

AncientPC · Nov 10, 2005, 05:33 PM // 17:33

Quote:

Originally Posted by qwe4rty

Actually, sunder works by having a 10% chance to ignore 10 armor.

I know, but armor calculations are funky so I just used 10% extra damage to simplify the problem (since 10% of enemy armor depends on each individual enemy).

Quote:

Originally Posted by qwe4rty

I wasn't necessarily saying it will sunder twice in a row, but twice or more per 10 hits.
Sunder = S
Non-Sunder = N

It could be:
S S N N N etc
S N S N N etc
N S N N S etc
S N N N S etc
S N N S N etc

It's true that clumping is possible, but it's still 10% chance per hit. In all the examples you gave above, the chances of 2 sunder hits within 5 attacks is equivalent to a 40% chance sunder.

In reality you have a 5% chance to make a sunder attack within 5 hits. I'm not sure how to calculate the possibility of making two sunder attacks within 5 hits, but it is much much smaller than 5%.

Quote:

Originally Posted by MuKen

To better see this, let's take it to extremes. Let's say a few of you are attacking a target and dumping out 150 dps. On the one hand, you have it super consistent, you do lots of little hits that perfectly come out to 150 damage every second, smooth enough to look like degen. No problem, the monks will drop a word of healing down whenever your target drops below 50% and they are operating at super-high efficiency.

At the other extreme, let's say every second your team somehow gets a 10% chance of doing 1500 damage that second. Guess what? The monk is totally screwed, there's no way he can deal with that.

Now obviously both of these examples are extremes that will never occur in practice, but the more your damage looks like the second one, the BETTER off you are. A fixation on doing consistent damage is like looking at the problem with tunnel-vision. Your goal is not to maximize your worst-case damage, your goal is to beat the enemy monk.

This is getting off topic, but you can argue the opposite as well.

Let's say the enemy is at 200hp and you do 50 damage per attack. You are guranteed to kill him in 4 hits.

If your damage was variable, let's say 10-90 damage (still averages to 50 per attack) you have the possibility to kill the enemy in as little as 3 hits or as much as 20 hits. The variable can work against you as well.

When the damage is constant (this applies more for spells since you can spike), killing an enemy hero with a monk lying around depends more on timing your attacks and the monk's reaction time. When the damage is variable, the attacking player has less control of the situation and it's just a matter of getting lucky and hopefully clumping enough high damage hits together to kill.

In my opinion, having constant damage is preferable to variable damage.

On a side note, doing 1 damage for 100 hits is preferable over 100 damage in 1 hit because there is less overkill (ignoring absorption and both attacks take the same amount of time).

MuKen · Nov 10, 2005, 09:11 PM // 21:11

Quote:

Originally Posted by AncientPC

This is getting off topic, but you can argue the opposite as well.

Let's say the enemy is at 200hp and you do 50 damage per attack. You are guranteed to kill him in 4 hits.

Like I said though, your goal is NOT to make any kind of guarantees on your DPS, it is to beat the enemy monk. The situation you describe is one where the enemy monks have already been disabled, in which case you've pretty much won anyway. In a scenario with monks, you simply won't ever take down that 200hp doing 50 damage per attack. However if you have an increased variance, you increase the likelihood the enemy monk will screw up.

Quote:

If your damage was variable, let's say 10-90 damage (still averages to 50 per attack) you have the possibility to kill the enemy in as little as 3 hits or as much as 20 hits. The variable can work against you as well.

Once again, your model is flawed because we are not trying to get to a limit of 200 in a minimum of hits, we are trying to arrange our hits so as to throw of a competing source of 'negative' damages. To this end, the variable is never a bad thing; there are two ways you compete against an enemy monk.

One is in spikes, making your damage unpredictable and hoping to catch him off guard. Toward this end, variance is a good thing, because it makes it harder for him to properly spend his energy on the spikes. The other is in damage over the long term, and if you know anything about statistics, your variance will rapidly smooth into the expected average over a long enough period, so it doesn't matter.

Quote:

When the damage is constant (this applies more for spells since you can spike), killing an enemy hero with a monk lying around depends more on timing your attacks and the monk's reaction time. When the damage is variable, the attacking player has less control of the situation and it's just a matter of getting lucky and hopefully clumping enough high damage hits together to kill.

Your view is too 'self'-centric, you AREN'T in control of the situation. This game is all about the monk and how he reacts to your spikes, not about you and how you react to his healing. You make periodic spikes, and he tries to react to them, it's not like he makes periodic healings and you try to time your spikes around them. Thus, the more random the power of your spikes are, the harder it his for him to react appropriately.

Quote:

On a side note, doing 1 damage for 100 hits is preferable over 100 damage in 1 hit because there is less overkill (ignoring absorption and both attacks take the same amount of time).

Again, you're taking the wrong view, trying to make small maximizations to your damage efficiency over a large number of kills. Doing 1 damage for 100 hits you have NO chance of making that kill, doing 100 damage in 1 hit does give you a chance at that kill. The most efficient thing you can accomplish is to GET that kill in the first place, resulting in a DP, a huge loss of energy for the target and his potential future contributions, and time spent trying to rez him. Besides all that, the loss of your 'overkill' is meaningless.

qwe4rty · Nov 10, 2005, 09:39 PM // 21:39

Quote:

Originally Posted by MuKen

This is incorrect, sunder works by having a 10% chance to ignore 10% of the enemy armor. Against a 60 armor opponent, this is equivalent to removing 6 armor, about 11% damage increase.

Conveniently enough, on a 100 armor opponent, the percentage damage increase for removing 10 armor almost doubles, so since the base damage is halved, the raw number of damage points it has increased is almost the same (a little bit less). Thus, within the range of armors you are likely to run across, small % armor penetrations will tend to put out roughly consistent damage increases regardless of enemy armor.

Ahhh, thats what I meant. I knew that, just at school, typing fast, dunno what happend

qwe4rty · Nov 10, 2005, 09:44 PM // 21:44

BTW, one last point thats bothering me.
I don't have time to do the math now, but I'm pretty sure the argument

10% sunder of 10% armor penetration = 100% sunder 1% armor penetration
is flawed

In this scenarior, lets say you do normal 50 damage a hit. Now lets say you do 100 damage with a skill, AND you sunder. W/ 10% armor penetration, that adds a lot more damage, but with 1%, it doesnt add as much.

While, if your damage is consistent, the 1% could apply, I'm pretty sure when you take into account criticals and attack skills, this idea is flawed.

Savio · Nov 10, 2005, 10:15 PM // 22:15

Quote:

Originally Posted by qwe4rty

BTW, one last point thats bothering me.
I don't have time to do the math now, but I'm pretty sure the argument

10% sunder of 10% armor penetration = 100% sunder 1% armor penetration
is flawed

In this scenarior, lets say you do normal 50 damage a hit. Now lets say you do 100 damage with a skill, AND you sunder. W/ 10% armor penetration, that adds a lot more damage, but with 1%, it doesnt add as much.

While, if your damage is consistent, the 1% could apply, I'm pretty sure when you take into account criticals and attack skills, this idea is flawed.

Sure, there's the random chance that you'll get the 10% AP on a good hit. But they're not consistent. You could end up with 10% AP every single time you hit, or you could never get it at all. On average, though, you're only going to be getting 1% AP out of it, which basically gives you 1.05% more damage. Vampiric doing 3 points of extra damage per hit far outperforms it.

Sir Skullcrasher · Nov 10, 2005, 10:24 PM // 22:24

so a 10/10% Sunder would increase your chance of higher damages by 1%? or 10%? How much damages does 10/10 mod really do?

I seen them being used before and it seem to take longer to kill someone.

Daegul Mistweaver · Nov 11, 2005, 08:30 AM // 08:30

Quote:

Originally Posted by Sekkira

Yes yes, but I still stick firm to saying you can't accurately predict DPS on an item which has a chance to do a varible amount of damage. Which is what this is all about, isn't it?

Sorry, this made me chuckle. Last time I looked we didn't have any weapons that did flat damage; they all had variable damage like 11-22. If this is "what this is all about" then this thread and all like it are pointless; since it would mean we can't make any predictions on any weapons, eh?

Alfresco · Nov 14, 2005, 03:43 PM // 15:43

Quote:

Originally Posted by sir skulkcrasher

so a 10/10% Sunder would increase your chance of higher damages by 1%? or 10%? How much damages does 10/10 mod really do?

I seen them being used before and it seem to take longer to kill someone.

10% means 10 chance to do a 10% arm penetration.. if a guy has 100 armour and your damage is 150.. u will do 50 damage.. and with 10% penetration.. it means it ignores 10% of 100 means 10.. and means u do a total of 60 damage and equalvent that the guy has only 90 armour.. =D