Defense Skills; OR, everything you ever wanted to know about Evade but were too lazy to search for

Doer

Marauder
Joined
Nov 29, 2005
Posts
7,004
Location
Muddlin' Through
Society
Rangers
Avatar Name
David "Doer" Falkayn
Edit 2010: If you are new to this thread, it would be best to start your reading with the Entropedia entry on Evader/Dodger which summarizes this thread and two others to present an overview of defense in EU.

If you would like to help out by adding new datapoints to the plots on Evader vs. mobs, see if your Evader or Dodger level is different from those already plotted with the mob you have in question and, if so, go out and count misses and hits. There are several ways to do it; the important thing is that you repeat it enough times to be sure it's accurate.

_______________________________________


Questions and misconceptions about the defensive skills and effects thereof come up frequently on EF. I decided to do some research and combine some good sources of information as a resource. If anyone with higher or lower Evader pro standing would like to do some tests of their own to share here so i can add the data points to my plot it would be nice, but even as-is the plot is useful. Edit: Thanks to Steffel i think there's enough data, even if not very rigorously controlled, to call this well enough. I added an additional plot with his data included by scaling and it fits a curve pretty well.


The Skills

To evade is to avoid something, for example, an attack from a mob. Not surprisingly Evade is the principal defense skill. This skill increases as you use it; so, as you are attacked by mobs that do melee damage. For firearms the relevant skill is called Dodge -- you will gain that skill as you are attacked by other players with pistols, rifles or explosives, or by drones, warriors, and troopers. Ranged attacks that are not from firearms (e.g. arrows from TskTsks, electrical bolts from bristlehogs or Tezlapods) all fall under the Evade skill.

VU ~9.0: This VU changed the division between the two defensive types. Dodge now applies to ranged damage. It appears that any mob with an attack range over a certain value must be dodged, while mobs with shorter range must be evaded. While this generally means that ranged attacks are dodged and melee attacks are evaded, mobs with a long reach (e.g. falxangius) fall into the Dodge category, even though their attack is melee. Conversely, a launched attack like the acid from the snablesnots has a short enough range to fall under Evade.

Avoidance is a hidden or unlocked skill that applies to both dodging and evading. It is unlocked by reaching a professional standing of 25 in the Dodger or Evader profession.

Quickness is unlocked at level 55 Evader.

Intuition is most likely unlocked at level 80 Evader.

The Professional Standings (what really counts)

These two primary skills have corresponding professional standings (Evader and Dodger), and those professional standings are the real measure of your avatar's defensive ability. The pro. standings don't just depend on evade and dodge, however: they are a composite of several skills. Therefore, it is important to consider more than just Evade and Dodge when improving your defensive abilities. For example, jdegre's thread established the following skill contributions to the Evader professional standing:

- Evade: 25% (Dodge constitutes 31% of Dodger)
- Agility (*20): 8%
- Athletics: 14% (contribution to Dodger is just 12%)
- Combat Reflexes: 11% (contribution to Dodger is just 10%)
- Courage: 7%
- Alertness: 5%
- Serendipity: 5%
- Avoidance: 6%


The other 19 percent comes from the unlocked skills associated with defense like Quickness and hunting skills like Combat Sense. Knowing these percentages can be very helpful when trying to calculate the least expensive way to chip up to a certain level. As an example, starting from 0 skills it would take 10k levels (10,000 x 25% = 2500) of Evade alone to unlock Avoidance (level 25 Evader), but it is typically unlocked at around 3900 Evade when skilling up naturally because of the contribution of the other skills gained in the process.


How they work

Professional standings in EU range from 0 to 100, they determine when unlocked (hidden) skills are obtained, and they are the numbers EU uses to determine your success at the profession. To use an example most players are familiar with, the hit ability (HA) of a weapon (from 0 to 10) is how likely you are to hit a mob with that weapon. It is also equal to the professional standing in the (Hit) for that weapon type, divided by 10 (0 to 100 for pro. standings, remember?). To continue with the hit ability example, an avatar with 0 HA doesn't miss every single shot and with 10 HA it won't hit every time. The actual percentage ranges from about 80% to 92%. It is reasonable to assume that evading will be similar: even a level 0 Evader will successfully evade some attacks, and a level 100 Evader will still take some hits. Just how much difference a high professional standing in Evader makes is a bit hard to measure, but i did some tests to get an idea.


The Effect of Evader Level

Here is a plot of my results. There are two actual data points on the plot, one for an avatar with Evader level 0.8 and one with Evader level 27.75, along with a linear extrapolation to 0.0 and 100.0 level Evader. The experiment was conducted by wearing the same armor (pixie, which protects about as much as the minimum damage of the mob so as to decay the same every hit) and being attacked by the same mob (Tantillion mature) for a period of 10-20 minutes, followed by a careful determination of the exact armor decay per minute using the sweat/fruit method. This was repeated at least five times for both avatars.

200702-PE-Evaderextrapolation.png

Figure 1a: armor decay[ped]/minute taken from tantillion mature on avatars with 0.8 and 27.75 Evader professional standing with a linear extrapolation to level 100; note that progression must not be linear because the extrapolated effect is negative

The results came as a surprise to me: with a Tantillion mature, the first 27-3/4 levels of Evader out of the 100 available give a 39% increase in evaded attacks. It is immediately apparent from the extrapolation that there must be a nonlinear increase in ability so as not to cross the origin -- in other words, the linear fit is not a good one. The good news seems to be that the first levels of Evader give the most significant improvement. Note that different mobs and mob maturities might be affected differently by one's Evader standing (that was not tested nor confirmed by my experiments).

Using an approximate match by eye between my data and a scaled version of Steffel's data (see later post), i created a composite normalized plot and fit a nonlinear curve to it -- a logarithmic curve. This rather nice fit demonstrates the point made above and shows that, much like skill gains themselves, there is a continuing diminishing return as the level gets higher.

200702-PE-Evaderlogarithmicextrapolation.png
Figure 1b: approximately combined data from Steffel, normalized to relative hits/minute from a hypothetical low-level mob on avatars between 0.8 and 27.75 Evader professional standing with a logarithmic extrapolation to level 100

The relationship of a mob's hit ability to an avatar's level of Evader continues to appear logarithmic. Here is a plot of all the test data to date (2008/3/25):
Evadervs.mobs.gif
Figure 1c: combined plot of known mob hit ability versus Evader level data

Misconceptions

One common misconception about defensive skills is that they decrease the amount of damage a mob does when it does hit, an understandable confusion when considering the many traits and abilities that can accomplish this in other RPGs. It just doesn't make sense in the context of the EU system, but to make sure i measured the damage taken from a mob (again Evader level = 27.75):

200702-PE-Evadermobdamagedistributionlvl27evader.png

Figure 2: Normalized damage from a tantillion mature in chronological order; mean = 0.75

As can be seen, the damage ranges between the mob's maximum and 50% of its maximum, just like an avatar with a maxed ability on a gun. If higher evading ability decreases the frequency of high damage hits, the average (mean) of this data should be lower than the midpoint between the maximum and the minimum damage. No such effect is seen: Evader level does not affect the amount of damage taken in a hit, only how often a mob successfully hits.


The Uber Unlocks

Quickness and Intuition are considered to be Evader unlocks, but there is a lack of information about what level is required to obtain them. Dreicc recently reported that Quickness comes at level 55. It is reasoned that Intuition unlocks at level 80 because that is the highest level any skill is known to unlock (Kill Strike is an example), and there is reason to believe that is the case. Considering the changes to the skill system over the past year, it is highly unlikely anyone new will be unlocking these now, anyway.

If you find yourself still confused or with questions after reading this, please post so that i can refine it. Thank you.
 
Last edited:
Thank you for your research! A very good one!
 
good post +rep.

let me add a graph from my little evader experiment.

it showes the decay 3 avatars at lvl5,7 and 19 evader had on the same group of mobs for the same time of tanking (decay in %, while lvl5 is 100%).

 
Nice job.

I can confirm combat sense having influence on evader pro standing.
On reach lvl 29 evader the sole skill gain at that moment in time was combat sense. I was kind of surprised at that moment.

Cheers
Richard
(ingame Siam)
 
Thanks Steffel; I remember seeing your thread before. I scaled your data to be the best fit (by eye) with my data to a curve, and then fit that curve (logarithmic) mathematically.

I have added the diagram to the original post. It helps illustrate what i had concluded (and makes me wonder if there's any point trying to skill evade at this point considering the curve is just about flat at 28 and the gains in evade-related skills pathetic due to the slowdown -- someone at ~5 pro standing is already halfway to the benefit of level 28 :eek: ). :(

It wouldn't surprise me if the curve actually approaches 50% of the initial value rather than the lower number suggested by the plot. There's significant margin for error here and that's a much rounder number.

200702-PE-Evaderlogarithmicextrapolation.png


Siam: yeah, i saw your post in the other thread: you're the only witness so far.
 
I am lazy but still can +rep this fine work :yay:
 
...makes me wonder if there's any point trying to skill evade at this point considering the curve is just about flat at 28 and the gains in evade-related skills pathetic due to the slowdown -- someone at ~5 pro standing is already halfway to the benefit of level 28 :eek: ). :(

well, both our tests were with kinda low level mobs, tantillion and rippersnapper. so there is hope that high evade pro stands still show some effect on tanking medium and high end mobs...also, high evader can have an effect in melee pvp, like lvl1 swordsman hits a lvl1 evader as good as a lvl20swordsman hits a lvl20 evader or so...

i tried illustrating this once again, while the green line is the point where you "maxed" the evading of that mob class, which is quite early on low mobs and hard on über mobs.

maybe some übers gonna test tanking falx versus low skilled folks :)
 
well, both our tests were with kinda low level mobs, tantillion and rippersnapper. so there is hope that high evade pro stands still show some effect on tanking medium and high end mobs...also, high evader can have an effect in melee pvp, like lvl1 swordsman hits a lvl1 evader as good as a lvl20swordsman hits a lvl20 evader or so...

i tried illustrating this once again, while the green line is the point where you "maxed" the evading of that mob class, which is quite early on low mobs and hard on über mobs.

maybe some übers gonna test tanking falx versus low skilled folks :)

Yes, that's the biggest hole left to investigate, i think. I was careful to test on the same maturity of the same mob in case mob agility &tc does affect the inflection point, but didn't actually do comparative tests across maturity. Nice cartoons, btw. :thumbup:

To be honest i really doubt MA made it all that complex. I mean, if it works the way you surmise it would mean the skill system would make sense. :rolleyes:
 
Last edited:
I have added the diagram to the original post. It helps illustrate what i had concluded (and makes me wonder if there's any point trying to skill evade at this point considering the curve is just about flat at 28 and the gains in evade-related skills pathetic due to the slowdown -- someone at ~5 pro standing is already halfway to the benefit of level 28 :eek: ). :(

It wouldn't surprise me if the curve actually approaches 50% of the initial value rather than the lower number suggested by the plot. There's significant margin for error here and that's a much rounder number.

200702-PE-Evaderlogarithmicextrapolation.png


but if you move to a bigger mob I think it will be bigger difference. I think you have more use of a higher evade level against bigger mobs.
 
but if you move to a bigger mob I think it will be bigger difference. I think you have more use of a higher evade level against bigger mobs.

Yes, possibly, but that's not proven to my knowledge and i am just reporting results and not hearsay. It is what Steffel and i were just talking about, though. If someone with 35+ Evader standing is willing to do a test on a big mob we can check this with some nicely spaced data points.
 
yo

Ill give you some stats to calculate.

A lvl 5 Evader and a lvl 22 Evader, evading vs a Cournanterion young.

Attacks: 50.

lvl 5 Evader gets hit 44 times.
lvl 22 Evader gets hit 27 times.

I hope this helps your work :)
 
yo

Ill give you some stats to calculate.

A lvl 5 Evader and a lvl 22 Evader, evading vs a Cournanterion young.

Attacks: 50.

lvl 5 Evader gets hit 44 times.
lvl 22 Evader gets hit 27 times.

I hope this helps your work :)




well, it would be required to do some more tests to see the average result, but, if it is like that, then, DAMN, i can see the price on evade related chips going sky high
 
yo

Ill give you some stats to calculate.

A lvl 5 Evader and a lvl 22 Evader, evading vs a Cournanterion young.

Attacks: 50.

lvl 5 Evader gets hit 44 times.
lvl 22 Evader gets hit 27 times.

I hope this helps your work :)

Thanks SirFluidHat

The real problem is that there's a pretty large variation over a few minutes; i had one outlier in each of my 5+ data collected. So, in order for this to be meaningful we'd need a few repetitions of each. Is this from one test or multiple?

In any case i plugged your numbers from a pretty high level mob into the logarithmic fit equation (y = -0.0998 ln (evader) + 0.8728) for the low level mob:
Code:
Level [b]Evader[/b]   5   22
Fit value %   71   56
Your hit %    88   54

We really need one more data point here but you can see that while the level 22 Evader numbers fit surprisingly well, the level 5 is off. Whether that's due to a shift in the curve as Steffel described, a poor equation fit, or statistical impropriety is impossible for me to say, anyway (IANAS). If i ever find a reason to go to CND again i'll do a few trials myself and then we'll have a level 28 data point, too.

I will possibly at some point ask for volunteers to come get hit by hogglo while i fap to test this but right now i'd rather play than spend another evening doing tests. ;) The real challenge is finding someone with a significantly higher level Evader professional standing who is willing to participate.
 
Very good, informative thread Doer.

The shape of the evade effectiveness vs evader pro level graph is quite interesting, and quite comforting for those of us who will never unlock stuff like quickness & intuition (unless medical science makes giant leaps forwards in longevity treatments over the next few decades) that progressing up to level 20,25,30 is quite worthwhile.

It all sounds about right to me going by my own experiences. I certainly get hit noticeably less than I used to by mobs like feffs and corns (the only two mobs I hunted this time last year I can think of that I still hunt occasionally). Having had recent experiences of getting myself and my fap mauled to death by fast attacking mobs like little-cornoanterion and middle-aurli which are not doing me huge amounts of damage but are hitting me with scary regularity, it'd be interesting to find out if as suggested by SirFluidHat the graph looks similar for big mobs (ie. they just hit more in general, evade still works the same) or whether a higher level of evade is required for them as suggested by Steffel.
 
Nice article. If you dont mind ill add it to wiki too.
 
I will possibly at some point ask for volunteers to come get hit by hogglo while i fap to test this but right now i'd rather play than spend another evening doing tests. ;) The real challenge is finding someone with a significantly higher level Evader professional standing who is willing to participate.

try getting starman out there...i think he qualifies...or skalman...vixen....stryker...*(some other hunter)
 
try getting starman out there...i think he qualifies...or skalman...vixen....stryker...*(some other hunter)

You put the emphasis in the wrong place of that quote, though. I would want volunteers, and typically uber players are busy off doing their own uber things, understandably enough, and i don't want to ask. When someone qualified has sufficient interest in the topic to volunteer his time we might procede.

Thanks Jimmy for your comments, and Witte you are welcome to.
 
You put the emphasis in the wrong place of that quote, though. I would want volunteers, and typically uber players are busy off doing their own uber things, understandably enough, and i don't want to ask. When someone qualified has sufficient interest in the topic to volunteer his time we might procede.

Thanks Jimmy for your comments, and Witte you are welcome to.

no dare no gain ;)


however, i think the research you are doing is very important, and you help alot of people out, i think.
 
I fapped some people on globster the other day and as far as i could tell globsters never missed them (it kept me damn busy fapping, that's for sure. :laugh: ). It would be very useful to know if someone has a high enough Evader level to successfully evade globster attacks. Please PM me or post your results here if you fall in that category, or if you have over level 30 Evader and are willing to do a few tests.

Also, a bump for visibility.
 
There's always a neanderthal around

I took advantage of testing caperon damage today for another thread to gather some data for a high level mob evader efficacy comparison. I collected it from a caperon young, one of very few i could find in the herd, which i led back closer to the tp to facilitate the tests. Between the second and third avatar test another player showed up and i asked him to please not attack the caperon because i was performing evade tests. He hesitated and then began killing the mob, dying in the process i might add. He came back to finish off the caperon, and then PKed me (him in vigi and i wearing my shorts and a shirt) and ran off.

My friend Skippie, who was kind enough to agree to be a subject for the really boring test, helped me find another young and we did some tests, leading it back toward the revive a bit to speed up the process. About 2/3 of the way through the test our friend the neanderthal showed up again and did another RT kill of our young. Skippie had had enough, and frankly, i had, too.

This neanderthal character is called Cosmos, and i ask for an explanation and an apology from Cosmos' society, the GPS Hunters. The CEO is gps4prizes and Cosmos has apparently risen to Department Manager. AdjustedShadow is an employee. I would appreciate a PM from one of them regarding this behavior, and will remove this when that has been resolved.

As you can see, the data i collected is rather incomplete and will remain so until i get some resolution and the loathing for everything EU out of my system.

PS Thank you jaba and Skippie for your help.

200702-PE-EvaderlogarithmicextrapolationwCaperon.png

Caperon young test data scaled and interposed on the logarithmic fit to the low mob data; there is either a shifted inflection point as suggested by Steffel, or (more likely) the data for Evader=21.8 (n=2) is insufficent due to external intervention
 
Last edited:
well, both our tests were with kinda low level mobs, tantillion and rippersnapper. so there is hope that high evade pro stands still show some effect on tanking medium and high end mobs...also, high evader can have an effect in melee pvp, like lvl1 swordsman hits a lvl1 evader as good as a lvl20swordsman hits a lvl20 evader or so...

i tried illustrating this once again, while the green line is the point where you "maxed" the evading of that mob class, which is quite early on low mobs and hard on über mobs.

maybe some übers gonna test tanking falx versus low skilled folks :)

Very interesting, could you also post what are the attributes of a mob that give it's uber-ness, is it HP alone or some of the other mob attributes?

This would make it easy when choosing a mob, since you know what armors you have, how good are they, what decay will the mob usually do, and now, just how many times it will attack you for your Evade.
 
Some months ago I also did a hitrate test with different low lvl mobs.
What I saw was there seems to be a different hitrate for EVERY mob and maturity, so that might be investigated also more.
If we can get data for multiple evader vs mobs hitrate we could enter those results into wiki in the number of threath, because threath is not only the comparing of max dmg and hitpoints, but the hitrate is very important.
 
@mrproper: it was just speculation i did, not facts or a theory based on recorded observations. the definition of a mob class is in each's personal eye...but i would tie it to hitpoints and damage dealt.
 
In case people want to do testing hitrate of each mob/maturity vs evader lvl,
I made something to store it in where all can see the results.
If there is a lot of data I can try some statistics on it to get more accurate data of the hits per minute per mob for each evader level.

http://www.gamezjunkie.com/pe/evader/index.php (no registration needed)
 
God damnit...

You must spread some Reputation around before giving it to Doer again.
 
@mrproper: it was just speculation i did, not facts or a theory based on recorded observations. the definition of a mob class is in each's personal eye...but i would tie it to hitpoints and damage dealt.

I could find no correlation between hitrate and any attribute. Mobs like neconu and formidon hardly hit, and have pretty high attributes. Mobs like phasm and caperon have pretty much the same atributes but hit allot more often. (on ~23 evader).

It would be cool if this kinda info could be seen when scanning a mob. And while at it, they can make bigger mobs harder to scan, making scanning skill usefull too.
 
Some months ago I also did a hitrate test with different low lvl mobs.
What I saw was there seems to be a different hitrate for EVERY mob and maturity, so that might be investigated also more.
If we can get data for multiple evader vs mobs hitrate we could enter those results into wiki in the number of threath, because threath is not only the comparing of max dmg and hitpoints, but the hitrate is very important.

After that data is gathered, a comparison with the scannable attributes of them mobs could provide us with some insights on the world of "mob skills" ... sounds like fun! I will try to add some results into this very interesting thread!
 
One other fact that would help to know is the exact attack rate (attacks/min) of mobs. I have always thought that all mobs except aurlis and kreltins attack at the same rate -- is this true? My tests on the caperons suggest that their attack rate is exactly 1/3.25 s. I would appreciate verification of this observation, and of its applicability to all mobs except the aliens.

Using the assumption that tantillions attack at the same rate (1/3.25 s), and that pixie decays at the rate of 0.495 pecs/hit (for impact only, protection=9) from Scatha's data, i translated my decay/s measurements on the tantillions into hits/s and then into hits/attacks. While it's much easier to collect large data sets using armor decay over time for small mobs, it's probably more accurate and easier to count hits and misses on big mobs. Both my conversion and Scatha's somewhat underestimate the decay on pixie per hit due to the ongoing decay over time, so there is a larger error margin. However, this conversion creates an approximate bridge between the two sets of data:

200702-PE-EvaderlogarithmicextrapolationwCaperon_257958.png


As i described in my last post, the caperon data should be considered incomplete and unreliable due to Cosmos' meddling, but it is sufficient to say that there's a large difference in hit ratio for some mobs compared to others.
 
I could find no correlation between hitrate and any attribute. Mobs like neconu and formidon hardly hit, and have pretty high attributes. Mobs like phasm and caperon have pretty much the same atributes but hit allot more often. (on ~23 evader).

It would be cool if this kinda info could be seen when scanning a mob. And while at it, they can make bigger mobs harder to scan, making scanning skill usefull too.

So is the attack success rate (actual hits/attempts) related to an attribute, or is it completely dynamic and changes as many times as the game engine pleases so?

My tests on the caperons suggest that their attack rate is exactly 1/3.25 s. I would appreciate verification of this observation, and of its applicability to all mobs except the aliens.

All my tests so far show that all mobs hit at the same frequency, about 3 seconds, I would need some video footage to verify the exact timing :)
 
What do u mean by hit frequencey?
Unless mobs also miss their target regulary without causing u to gain evade, the mobs I'm hitting (mostly melee) are certainly not hitting on a regular time interval such as ~3s.
 
Back
Top