Omggg this is FIXED on PTR!!
Tested with a GA w/ a 50% amber gem and minimal focus… did 590 damage
Tested w/ the same GA w/ same 50% amber gem but 280 focus… did ~760 damage.
It’s scaling <3
1 Like
That not what needs testing.
You need to test on wolves with GA with no gem then with a 20% gem then with a 50% gem.
both sets of number need to be white damage and the overall damage of both numbers should be higher on the 50% than on the 20% and overall damage of both gems should be higher than a GA with no gem.
Assuming you have high focus and low str as these are meant to be used on FOC builds obviously.
I did test on wolves. I didn’t test all the cases, but if the bug were still there, increasing focus would have had no effect. And it did have an effect.
Oh right… So we don’t know if 50 vs 20 is fixed. But we can see that focus scaling itself is now applied.
This here was my main issue with amber gems. Hoping this is fixed!
The issue is not whether FOC gives damage to gems. It does.
The issue is that it gives the exact SAME amount of damage regardless of whether it’s a 20% gem or a 50% gem.
If you want to test on PTR test this:
Go to EbonScale and find a Boar that you can easily set into “Retreating”. This way you will always be using the same mob, meaning error doesn’t creep in from hitting different creatures.
Another option is go to Bunny Island in Restless Shore. All bunnies are lvl 1 with no armor - your results will be consistent.
Stick a 20% gem
Set stats to 5,5,5,100 FOC,5 and hit. Record TOTAL damage (phys+nat)
Set Stats to 5,5,5,300 FOC,5 and hit. Record TOTAL damage (phys+nat)
Record the difference between TOTAL damage numbers. - this is your FOC scaling from a 20% gem
Now stick a 50% gem
Set stats to 5,5,5,100 FOC,5 and hit. Record TOTAL damage (phys+nat)
Set Stats to 5,5,5,300 FOC,5 and hit. Record TOTAL damage (phys+nat)
Record the difference between TOTAL damage numbers. - this is your FOC scaling from a 50% gem
Compare the differences (Deltas).
If the Difference for 50% gem >>> Difference for 20% gem, then the scaling changed.
If the Difference for 50% gem === Difference for 20% gem, then it’s still broken
*Note - there might be a 1-2 damage variation due to rounding.
P.S. - this starts to feel more and more like my day to day job where I have to tell testers how to test my software 
2 Likes
Yeah if someone can provide an update on the amber gem scaling in PTR…that would be greatly appreciated because as a healer ive been playing with broken gems since lvl 30…SLAMS FIST ON TABLE
Hi,I kow youve been testing the amber gem on PTR. Could you lmk what the results are for scaling? In live server currently the higher the gem the lower the damage im doing. Thanks
all i know is that i do a lot more damage with 20% amber gem than with the highest tier one. and that shouldn’t be the thing. hence it’s bugged.
if it’s not bugged, then it’s false advertising. and they should just delete the higher tier gems because they aren’t an upgrade at all. they aren’t any more useful than a pet rock at this point.
Is this on PTR?
Haven’t tested anything other than 50% gem on PTR. AGS gives you stacks of all the 50% gems to test with but not any other types and I’m not that interested in playing so much on PTR to go find all the other types.
thanks
base 1090 crit 1379 with amber 2
base 878 crit 1112 with amber 4
pristine gem does ~24% less damage
Thanks
PSA: Stop using mobs for testing!
Near every fort in every zone, there is a camp. In that camp there are dummies. Please do your testing with that as is much more comparable between tests and also they have fixed damage reduction / armor / whatever.
I believe the dummies scale with your Level and amour type as well
I redo some testing on the PTR server.
I only tested 20% and 50% gems. Mainly INT Gems.
Tested on dummies, I plotted the damage done based on how many attributes you allocate on INT, using 20% and 50% INT conversion gems.
From top to bottom:
- Purple is the normal damage of the weapon when scaling with STR.
- Blue is the Total Damage (Normal damage + Gem Damage) for 50% INT Gem.
- Yellow(-ish?) is the Total Damage (Normal damage + Gem Damage) for 20% INT Gem.
- Green is the Gem Damage only, for 50% INT Gem.
- Red is the Gem Damage only, for 20% INT Gem.
Base damage remained constant as weapon scales off STR not INT.
The dashed line is at 150 INT, when the perk +15% More Elemental Damage procs, so we see an increase in damage.
This was tested using the exact same weapon. No perks altered the results and the damage recorded was consistent after every hit. The last INT perk (more damage against full HP target) was ignored by doing damage first.
Couldn’t test 30% and 40% gems due to workstation problems on PTR (all of them are lv2)
In my opinion, there are one single bug: being how a 20% gem does more total damage than a 50% one and how the 50% gem outdamages the 20% one after the 150 INT landmark. This behavior is the same for FOC Gems but it doesn't get "corrected" at any attribute bracket.
Other comments in this thread bring other points as bugs but I rather read them as misconceptions from what the gems actually do. As the tooltip wording is poor one can only assume things.
Some says that the better the gem, the more the scaling but they expect to be reflected on total damage. Others says that the weapon itself should scale with INT, like a fire staff would do.
As far as I can desglose from the tooltip itself, the intended behavior is that a percent of weapon damage is taken out and turned into elemental damage, which scales based on int. That is happening as showed by the green and red lines in the graph.
As a percentage is taken out from the damage, the total (the 100%) will always be the same no mater the % of the gem.
So the only advantage would be do more damage to a mob whos weak against that element type. I tested this briefly against the same Ancients mob using a Hatchet, an Ice Gauntlet and a Warhammer with 20% an 50% Light conversion and the result was a 12% more total damage with the 50% gem at most. Some giving 8% more or even 5% less.
Given how little this % was, I might say that there’s another bug but it isn’t clear from the tooltip description. 12% more damage for a T5 gem is useless in every scenario (even when the gem is scaling off STR). So I see no reason at all for using any other than a 20% gem, no matter what you are doing.
I personally believe that the problem lies in the actual scaling of each gem and that it must mirror the weapon scaling mechanics in a way. Currently, the more base damage a weapon has, the bigger the return of each attribute on weapon damage. Gems does not do that, instead, the elemental damage alone between different gems tier gives a "flat" amount of damage (because is taken as percentage from the base weapon damage). One would expect that more INT in a better gem translates into a damage curve being more positive sloped than a lower tier gem.
Some here have stated precisely that as the bug, but it ain’t clear from the tooltip if that is the intended effect of the gems or if just giving us at most 12% more damage is what the devs had in mind.
Thank you for coming to my ted talk.
1 Like
YellowMembrillo
I love the analysis, but I disagree with a few of the conclusions.
-
The “Total” dmg lines should be within +/- 2 dmg. This is due to rounding. Most likely the before/after 150 int isn’t actual difference in scaling as it is just rounding issues.
-
Your tests only had pure INT builds. If you add some STR to the tests (ex. 50-100 str) you will see that the total damage decreases as the gem quality goes up.
-
Previous tests with STR suggest that:
- At 80/20 split the Phys portion of the weapon damage gets scaling from 80% of STR, while the Nature portion gets scaling from ~28% of FOC.
- At 50/50 split the Phys portion of the weapon gets scaling from 50% of STR, while the Nature portion Remains at ~28% of FOC.
Based on this STR scaling observation - The calculation for Physical damage does in fact look at what % of base is attributed to that Stat, and then this % is applied to the scaling calculation.
This however does not happen for FOC, as it’s stuck at some arbitrary %, which is not tied to how much of the weapon base has been converted.
The worst case scenario is if you have 145 STR and 155 FOC (similar to the paladin builds). This is where you will see a very large drop in total damage as you go from 20% gem to 50% gem.
- Last time I used those dummies, I got some very strange armor scaling when using low GS weapons. I don’t trust the dummies, which is why I stick to repeating tests on either the same mob (letting it go into “retreating” to regen hp) or on lvl 1 bunnies that have no armor.
1 Like
bump. can i get a refund on the pristine ambers i bought for my weapons due to misleading info? thanks devs.
The wolves outside WW are just perfect for amber gem testing. If you get them to attack you once and hit them once in the face they will give white numbers for both your base attack number and your FOC nature attack number.