GF Statistical Data (Avoiding Diminishing Returns)

ferret09

Avoiding Diminishing Returns As A Guardian Fighter

Let me start off by saying that I do not know every single thing about guardian fighters. I have not tried every single GF build, nor do I consider myself professional. I did this on my own time to present my thoughts and opinions about this class to encourage others to maximize their gaming experience.

^_^

With the upcoming patch, I took the liberty of testing all of the attributes on my guardian fighter. All the data was collected using the Neverwinter-Preview server. Gear was selected to boost each stat as high as possible, resulting in a range of values between 0 and 4000. Data was recorded in intervals of 75-100 (30-50 data points). The recorded percentages are directly from the stat bonus. Other modifiers that would eventually boost the final percentages were not taken into consideration. Again, the percentages shown are ONLY from the stat itself, nothing else!

The purpose for this data is to inform anyone with a guardian fighter how each stat performs. The following graphs illustrate which stat bonuses are better to stack, and when you hit diminishing returns.

NOTE: The percentages shown in the following graphs were taken from a guardian fighter ONLY. Diminishing returns for other classes MAY BE DIFFERENT.

*Defense was given its own graph so it wouldn't interfere with the other stats. Graph ends at 6000 defense since the diminishing returns are way too high to ever go higher than that amount.

NOTE ABOUT GRAPHS: All stat bonuses are roughly equivalent below 1000. Only after more stat bonus are introduced do you see any real differences between them. As you increase any stat bonus you start getting diminished returns. Graphs end at 4000 since there is NO rational reason to socket any stat higher than that on a GF.

WHEN DO DIMINISHING RETURNS KICK IN?

A rough outline of when diminishing returns kick in can be determined from the graphs above. I measured diminishing returns to "kick in" when the line on the graph deviates from it's most linear region, normally found between 0-500 stat bonus. These are the rough estimates I took using the method mentioned. Determining exactly when diminishing returns start to take effect is quite hard, and is mostly just a guess on my part. Changes to the following values might be made.

Critical: ~900
Deflect: ~1000
Recovery: ~1300 (Most linear)
Armor Penetration: ~1300 (Least Linear)
Defense: early (no real linear region)

EDIT: Added graphs comparing stat bonus and % bonus given per point.

MY THOUGHTS ON THIS DATA

---

ArmorPen > Recovery > Deflect > Crit

---

Critical Strike:
The observations made in the preview server lead me to believe that crit isn't something that should be stacked over 1650. Anything past a 20% crit rate just goes to waste. On average, armor penetration will be 2-3 times better than crit. Just don't ever go over 1650...
Armor Penetration:
Stack as much as you can. This IS the best offensive stat to stack. Returns will always be good, and it adds straight percent damage increase.
Recovery:
This stat has high return, but I wouldn't focus on it. Armor penetration is much better for offense. Just get recovery from your weapon, shield, and armor. In my opinion it's not even worth having on jewelry.
Defense:
The results I obtained also lead me to believe that anything over 3500 defense is wasted. 3500 Defense will give you 45% - 47.5% damage reduction, which is more than enough. The downside to defense is it's vulnerability to armor penetration. Not worth stacking.
Deflect:
Any defensive slots should be socketed with deflect if you already have 3500 defense. The return for this stat is much greater than defense once your defense is above 3000. It's quite easy for GF to get at least 20% deflect rate, which is basically 10% damage reduction on top of your 45%. This added 10% damage reduction cannot be reduced by an opponents armor pen, an added bonus. If you're not running teneberous, stack as much as you can.



Just thought I'd share my results with the community...

I hope some of you can use this data to your advantage to help create your own GF builds. I would also like to hear feedback on my research, and encourage others to do something similar for other classes. If anyone is interested in viewing the spreadsheet data I used to make these charts check out the link below.

https://docs.google.com/file/d/0B98LzfG4v8byMC1aWk9BUHh1ODg/edit?usp=sharing

I didn't know where else to upload it, so it's on Google docs.

ironwill92

Very good thread, thanks for that one but isn't Armor Pen stacking not that good because a boss mob usually has about 22 % Resistance so you require 2220 Armor Pen to penetrate that? And stacking more is wasted? I think ARP is indeed very good but only until 2,2k.

freehugs9

Ferret09, I'm very interested in the data you gathered from the preview server. Is it possible for you to post or upload the raw numbers?
I'd love to check it against the formula's I've derived from the live servers:
http://nw-forum.perfectworld.com/showthread.php?301761-Calculating-stats-and-their-effects

- Freehugs

uncag3d

ferret09 wrote: »

The results I obtained also lead me to believe that anything over 3500 defense is wasted. 3500 Defense will give you 45% - 47.5% damage reduction, which is more than enough. .

i have 4628 def gives 36.9% DR your numbers are off

dkcandy

ferret09 - Big Thanks. Love #'s

bananachef

What with crit diminishing quickly, doesn't this make base crit chance from dex very valuable?

ferret09

ironwill92 wrote: »

but isn't Armor Pen stacking not that good because a boss mob usually has about 22 % Resistance so you require 2220 Armor Pen to penetrate that? And stacking more is wasted?

2.2k Is incredibly easy to obtain with a passive companion, for pvp 1500 is a decent amount to aim for. Armor pen, even after 2000+ will still be 3 times better than crit.

uncag3d wrote: »

i have 4628 def gives 36.9% DR your numbers are off

My spreadsheet has 4628 Defense giving ~36.8%. The 45%-47.5% is the GF final damage resistance with feat passives, etc. Also I was using T1 Armor and shield, so you could potentially get more damage resistance.

c45hcr0p

Don't think GF get crit % from any stats @ all.

dkcandy

c45hcr0p - Correct GF does not get % from any attributes like other classes.

badxideas

Thank you for starting this thread and getting this kind of dialogue going. I have a lot to learn before I can contribute but this kind of info definitely gets me thinking along the right lines as well as helping me to work out my priorities more effectively.

So yeah, thanks!

ferret09

freehugs9 wrote: »

Ferret09, I'm very interested in the data you gathered from the preview server. Is it possible for you to post or upload the raw numbers?
I'd love to check it against the formula's I've derived from the live servers:
http://nw-forum.perfectworld.com/showthread.php?301761-Calculating-stats-and-their-effects

- Freehugs

I Uploaded the spreadsheet data I used onto Google documents, enjoy.

Also after viewing your thread I decided to plot graphs of the %Boost you gain per point for offensive stats... I put the new graphs in my original post.

sanctumlol

This is correct, but Power>Recovery because it gets doubled for us. The DRs are the same for every class (where did you read that they're not), it's about how every class uses the stat and turns it into dps. Also, DR isn't the spot where they stop being nearly linear, it's the point of maximum concavity.

briseda

Oh great post Suki!

freehugs9

ferret09 wrote: »

I Uploaded the spreadsheet data I used onto Google documents, enjoy.

Also after viewing your thread I decided to plot graphs of the %Boost you gain per point for offensive stats... I put the new graphs in my original post.

Thanks for this. I've checked your numbers against my formulas and it looks like they haven't changed from live, which makes my life easier. Plus you've got so much data there I think I can refine the Armor Penetration and Deflect formulas, which is also great.
Don't suppose you feel like checking Life Steal, Regenerate and Movement as well? XD

ferret09

If the preview servers go back up and I have nothing better to do I guess I could look into those stats. I excluded lifesteal and regen from my research solely because no GF should ever stack them.

I am however interested in movement, and how that will turn out.

warfurybladez

ferret09 wrote: »

If the preview servers go back up and I have nothing better to do I guess I could look into those stats. I excluded lifesteal and regen from my research solely because no GF should ever stack them.

I am however interested in movement, and how that will turn out.

Bro, you're awesome.

Any idea of the flat damage difference between stacking the ArPen &/or Crit vs Tenebrous procs? It would be interesting to see what the actual % of Tenebrous procs are since it's not outright stated in the description, and whether as "true" damage they have a diminishing return or higher proc rate based on stats.

we7reaver

Very nice thread, gratz mate.

dkcandy

sticky this please.

ferret09

warfurybladez wrote: »

Any idea of the flat damage difference between stacking the ArPen &/or Crit vs Tenebrous procs? It would be interesting to see what the actual % of Tenebrous procs are since it's not outright stated in the description, and whether as "true" damage they have a diminishing return or higher proc rate based on stats.

I can easily use the data I collected to produce a graph comparing crit vs. armor pen. Teneberous is based off hp and should be a linear 1%, 2%, 3%... etc of your HP. From what I tested on the preview server, tenebereous builds are completely overpowered for single target scenarios... For PvE AOE situations having 6 procs of teneberous giving 5-7k extra damage every 7 seconds isn't that much of an increase for the cost of AD. Tenebereous is best for PvP, not as good in PvE, but still exceptionally better than crit/armor pen.

EDIT: Does this answer your question? lol ... In my first post I mentioned that you shouldn't go over 1650. If you're talking about PvP I wouldn't stack it at all, just get crit from jewelry... for PvE get a passive companion, that will get you to 1650 quite easily.

Graphs only have 10 data points, so these are rough outlines of how they would perform. Percent damage from crit was calculted like so... 10% crit chance from your critical strike stat bonus, therefor you have 10 * 0.75 = 7.5% damage increase. Armor penetration was calculated as straight damage increase, as you will be penetrating armor with it.

dkcandy

Greater Tene is hands down the largest DPS increase of any enchant.

zakaray

if i transfer my character to the test server, does my respec token come with me? I ask because I want to see what kind of damage I can get going tactition before I respec. Also, if so - is it easy to trasnfer character there? Thanks.

ferret09

No you don't get a free respec token on the preview server. Transfering there is quite easy. I have been working on movement speed in the meantime and got some results...

MOVEMENT

Movement stat bonus is completely linear. The data collected was from the preview server with no buffs/modifiers.

Movement

---

% Increase
0

---

0
65

---

0.3
120

---

0.6
185

---

1.1
210

---

1.3
288

---

1.9
348

---

2.4
413

---

3
438

---

3.3
465

---

3.5
530

---

4.1
595

---

4.8
620

---

5
649

---

5.3
714

---

5.9
760

---

6.3
825

---

6.9
871

---

7.3
936

---

7.9
961

---

8.1
986

---

8.3
991

---

8.4
1016----8.6
1076----9.1
1091----9.2
1116----9.4
1161----9.8
1186----10
1246----10.5
1316----11


Graph of the above data:

zakaray

im asking can I transfer my respec with me to the test server?

briseda

If you have the diamonds, you can just transfer your char and use the diamonds, it wont affect your char on the regular server.

synozeer

zakaray wrote: »

im asking can I transfer my respec with me to the test server?

Just transfer your character to the test server and see - the hardest part is the time it takes to download the files.

dartakx

briseda wrote: »

If you have the diamonds, you can just transfer your char and use the diamonds, it wont affect your char on the regular server.

Exchange market is disabled. You can't buy zen with AD.

jfdufour

Thx for a lot for doing all of this and sharing it with the comunity!

demonsunder

Been looking at the data produced by this testing and its great good work, but I'd like to add saying a stat isn't good because it doesn't stack well isn't a good way of representing the data you need to factor is alot more variable's, for example Rec eff's is on high c'd abilities more than low cd's, so Rec is more valuable to High cd abilities than low cd's. Or having High AP is great but you need to factor in the base weapon damage as well there no point in having 22% ap if your weapon only does 200 damage.

There are only a few Non-modifiable stat's, there are the only stat's you directly affect without a in-game variable.

Deflection, lifesteal, Movement speed, power and regeneration


The data is fantastic and very very useful, but its only part of the picture.

EDIT: And yes I understand this thread is to avoid diminishing returns, but others may not understand that a little more of "Stat" with diminishing return's can better than adding another stat because of "variable A"

freehugs9

ferret09 wrote: »

Movement stat bonus is completely linear. The data collected was from the preview server with no buffs/modifiers.

Thanks Ferret09, used those numbers to check that the movement formula has not changed on the test server (it hasn't) and to improve the formula slightly. Here's the result:
http://nw-forum.perfectworld.com/showthread.php?301761-Calculating-stats-and-their-effects
Movement Tooltip = 31*Movement^1.4/(42300+Movement^1.4)

It's not actually linear, just close to it.

zakaray

Ferat any idead on percent increase to damage when usi g villianes menace? Would appreciate feedback

ferret09

Any armor penetration you add, the more damage you will do. If you ignore 15% of a targets damage resistance, you in essence deal an addition 15% damage on every skill. The only stats that will effect how much damage Villain's Menace will do is power and base weapon damage. I produced this graph earlier showing how much power effects every GF skill. You need (slightly on average) 25 points of power to increase a skills damage by 1.