Silkroad Online Forums

What is this?
I saw alot of bladers/Glavirs with that stupid white hawk and when i asked why they said because it increses the attack ratio .
What dose it help?

And
Parry ratio aswell i don't know what dose that mean :X
thanks alot.

Its not entirely clear what parry/attack ratio does atm.

Common hypothesis seems to be that the parry/attack-rating ratio has an effect on the damage distribution. eg: If the defenders parry is much higher than the attackers attack-rating, the damage-distribution would be biased towards the lower end of the attackers damage range. (And the other way around if attack-rating>>parry)

Cant really confirm/reject this. Using the force skill to lower targets parry by 43 does not seem to have much effect for me (same for white hawk buff). It seems the average damage is a bit higher, but i never took the time to do any "real" tests on that.

In my opinion parry/attack-rating has very low influence currently. However, we saw from screenshots that korean sro has "missing" implemented, and i think parry/attack-rating will be quite important once we get "missing" on international sro.

Anyway, I'm just a noob myself, started like a month ago -_- Lets wait for the pro's comments

PS: Any idea when "missing" will work on isro?

what is that missing thing? :X

Attack ratio might be something to do with your attack power or how fast you attack, I don't know. parry has soemthing to do with how much damage you take, the higher the parry the less damage. *I think* Correct me please. ^^

Well, in many games there is a propability for an attack to miss the target. ie: It wont deal any damage at all. Just like blocking works, basically.

I think the propability for an attack to miss will be dependant on the attackers attack-rating, and the defenders parry-ratio. However, thats just an asumption Missing is not implemented in isro yet.

Imo: *If* there is an effect of parry/attack-rating on the damage output, its just a temporary solution until we get missing.

attack ratio and parry explained!!

every attack has a range of damage it can do. a higher attack ratio, or AR, raises the agerave dmg of an attack to closer to the upper end of that scale. ex.: attack does between 150-300 dmg. a higher AR would make each time you use that atttack the dmg will be closer to the 300 end of the spectrum.

parry is the opposite. parry affects how much damage you TAKE. if an attack against you could do 150-300 dmg, then a high aprry on your part would make that dmg closer to the 150 end of the scale.

got that? this has been extensively tested and is pretty much accepted as fact at this point.

Thats what i wrote in my first post. Just saying that its my opinion that the current implementation is only a temporary one.

So its kinda usfull i think.

SG:


Just been searching on the topic a bit. I mostly founds threads like this one: Someone asking what parry/attack-rating does, and some replies explaining the mentioned hypothesis.

But actual tests/discussions i could not find

@SG:
Could you please post a link to the actual discussion, where the tests about this have been posted?

===============================

Just did a short test myself.

Setup:
LVL39 Char, letting 2 Earth Ghosts hit me. I excluded critical-hits.
Parry Buff: Concentration LVL3, +17 Parry

100 Hits With Buff: Average 10.75 Damage, SD 5.97, Min 1, Max 26
100 Hits Without Buff: Average 9.63 Damage, SD 6.3, Min 1, Max 23

T-Test on the samples with:
H0: Mean Damage is same for both.
H1: Mean Damage w. Parry Buff < Mean Unbuffed Damage

That gives us a p-val of 0.9. Uh, maybe there s a fault in my setup. But in those samples there is no significant difference in the mean damage . In fact, the average damage was lower(!) in the no-buff sample, but the difference isnt significant either.

I ll propably do some more testing later (on higher lvl mobs), but from this test i would conclude that parry does NOT lower the average received damage.

plexiq:

to my knowledge, no one has posted anything about these tests on here (and i have something of a photographic memory, so it's likely that i'm right). these have been practical, in-game tests performed by me and others. for example, at lvl57, WITH my white hawk which adds 13 to my AR, i can nuke - nuke - ghost spear a devil yeti nearly every time. WITHOUT it, i can't kill them with that combo EVER, unless the ghost crits (and with a crit 6, this happens infrequently at best). i also have a friend on Athens who has designed a character specifically around AR and parry. his AR buffs and passives add some pretty sick dmg for him. his forum name is Kard if you want to PM him.

^ hmmm light force archer?

light lowers opponents parry,
force lowers opponents parry and attack ratio
bow raises own attack rating

= so high as hell dmg and take low as possible

a sword shield might be funner for it though

@RaiKiRii, Bow/Light/Force, thats my current build

actually, i thought there was a thread somewhere around here saying that tests were done while the 'constant damage' bug was on for a day or so. so if that's true, then it's already proven.

I've killed thousands of Maongs to level up. Then I used a parry scroll (+30% parry ratio) which made me like 250 parry total. During this, I have taken really lower damages and used less HP pots than usual.

@Vintar: The point is, we are all biased to believe that parry does have *some* effect. I also have "the feeling" that i recieve less damage when using concentration buff. But the only way to get unbiased results is do some big scale objective testing.

I was writing down 200 hits, and there was *no* significant difference in average damage.

That could mean:
1) I was doing something wrong? (Ideas?)
or
2) Parry does not have influence on average damage


==========================

Update:
Just did another 100+100 Hits Run with Black Robbers, same setup otherwise.

Again, there is no significant difference in average damage

dude's , girls , just pvp with a lighning user , with your pills off ,
youl see

I wrote down a total of 400 Hits, and there is no significant difference in the average damage, with/without the +17 Parry Buff. Care to enlighten me what my mistake is, when testing?

Im 99% sure that parry does not have any effect on the damage distribution at this point. That is, unless someone can point out any mistakes in the test setup.

Parry works - period. It was proven during the constant damage bug.

When I was fighting white spiders recently the effect was beyond obvious. As a L47 pure int, they would hit me for 80-120 everytime without parry. With parry, they would hit me for 1 sometimes.

Like most the buffs, the effect starts to become a lot more obvious the higher level you are (technically, the higher level your skill is).

As far as i know the constant damage bug was everyone dealing max damage?

If the hypothesis with parry/attack-ratio changing damage-distribution is correct, the max damage would not be dependant on parry?

Anyway, would some "parry believer" be willing to do an objective test run on this topic?

A side-note:
This thread might give the impression that i would be glad if parry turns out to be useless atm. As a matter of fact, my one and only char (still only lvl39) is heavily based on attack-rating/parry skills. Realising that these skills are propably useless atm isnt as much fun as it looks like.

But after writing down 400 hit-rounds its more than obvious that +17 parry do *not* change the received damage at my level, against blue and white physikal-damage monsters. If that also holds for higher parry-changes and/or magical attacks, i dont know...but i would appreciate help to find it out.

the damage bug dealt constant damage, not max though. it had something screwed up with the randomizer involved, but the base formula didn't change. so if you attacked something unbuffed, you'd hit for the same every time. if you buff up, it would do the math for that, but then you'd hit for the same -new- damage until the buff wore off. then it would go back to what it was before.

Im not able to reproduce any damage change with parry at the moment. Is anyone else?

I dont know if it worked during the damage bug...now, it isnt working for me. Again, id be very happy if someone could help me w/ this.

Black Robbers, currently orange to me.

skill was lvl 1 concentration, with a bonus of 15 parry. Criticals ignored, as were combo hits.

Without:
N: 12
Min: 163 (3 instances)
Mean: 184.75
Max: 231

With:
N: 12
Min: 163 (6 instances)
Mean: 181.42
Max: 230

Got bored after 12 hits from each, and realized I don't have the pots to keep a very large sample going. Also why I'm not Z/T testing. But, from this minisample, I'm going to take the liberty of claiming the following things:

1. The difference between the two values is so small because A. I hit the monsters lowest possible damage 6 times with test 2 and 3 times with test 1. A higher mob would have seen a bigger difference from the mean, leading to a more significant difference in damage. B. I only have one level in the skill.
2. If these things were changed, as well as the sample size increased, I would see a larger discrepancy between the two tests.

And now i go level some, so anyone who wants further testing, feel free.

Very cool, thanks!

If you post the damage-series as a comma-separated list, i can do the significance testing, if that helps. Thanks again

wel guy's i know for a fact that i deal around 200 more damage with my nukes if i get the lightning debuff going on a target

So lowering parry sure has some effect on the damage your doing

L1: 163, 163, 163, 165 ,169, 177, 181, 188, 198, 204, 215, 231
L2: 163, 163, 163, 163, 163, 163, 174, 178, 195, 209, 213, 230

attack rating works.. i do 3500 avg to a bunwang with my birdie, and without i hit for avg of around 2800.

parry works... its the basis of the full int light sword build... which should be a tiny heads up for all yall who think pure ints should only be nukers

and i dont think theyre gonna make "Miss" a part of isro

Devotia's runs also show no significant difference. Actually, when asuming *no* parry effect, the chance of generating a sample with higher difference in means (noparry avg - parry avg) is over 36%. (Hypothesis is usually considered as significant if this value is below 5%) Although 2x12 Hits might be a bit low for testing

Devotia, Thanks for the help again.

RaiKiRii: You are observing a +700 damage boost? With such a huge difference in average damage, even a quite small sample should be enough for testing. Could you get us like 25-50+ Hits with and without Buff please?

Would be very cool

devotia is also only using the first level of the skill, which only grants 15 parry. if i'm not mistaken, she's a higher level player, so toi her, 15 parry would be rather insignificant. that's like saying the heuksal passive is useless because at level 60 you don't need the 97 HP the first level of it gives you.

trust me, my wife plays a light/blade, and hits for quite a bit more when the lightning is in effect. of course, that's with it maxed to her mastery level. if she had it at level 1 imbue, of course there'd probably be an insignificant difference. not only that, but when we gang up on a giant, it's immediately obvious when the shock is in effect, even if you can't see it because it's blue from my cold. you can tell because the damage rises quite a bit. (i must say that cold/light is a really nice combo, btw)

Mhm, i can see no difference when i use vital spot (target -43 parry) and hawk (+17 attack rating) vs. none of them.

I think -43 parry AND +17 AR should make a recognicable difference?

Actually, I was under the assumption that AR was a proven system, and dodge was questionable. :/ AR, especially with a nuker can be seen almost instantly, while the difference between dodge data points will be relatively small.

As I stated before, I'm aware that my data is flawed, in that the mob I chose to do it on was one incapable of doing enough damage for dodge to really be a factor. 50% of the data points for the skill were at the mob's lowest possible damage. With a higher level mob, all these data points would be even further from the mean. However, since I'm an int user, any higher level mobs will kill me, and that's a price I'm not willing to pay for this :p Black robbers were chosen because I was farming them when I came across the thread, and jumped in game and let a couple pound on me for a bit.

When I get back from class I'll log onto my old STR char and toss on conc lvl 1, and let her get beaten a little. The extra HP will let me actually survive it ;p