Eh, it still needs a bit more testing before I want to post it publicly. It worked perfectly for me, but then all the class/race bonuses were jacked up until I had someone test them. (Fixed now). Message me in game and ill send you a test copy if you want.
The monk weight softcap from 1-14 is 14. Because Hstr is primarily an offensive stat not a defensive stat. Hdex as a minor offensive bonus, but is primarily defensive; but you don't see anyone asking why Hdex doesn't appear on dps centric augs.
I'm not convinced on this argument. It still needs proper parsing. Hstr adds +1 dam per 10 and also boosts attack. Hdex boosts crit rate (and +1 dam per 10 to ranged attacks, which barely counts as dps for most of us). Given how weak +1 dam per 10 really is, it boils down to attack vs crit basically. I estimated at something like 100-150 dex needed per 1% crit, but it's pretty hard to pin down precisely. I should try juggling augs & gear more to get a decent (say 100) swing and parse both again. It's been years since I did it before.
As far as I am concerned, given how strong crits are relative to their base hits, any increase in crit rate provides more damage than any amount of HStr attainable on gear, over the course of the fight. It would help, however, if Dzarn would provide us with the relevant combat formulas for this as well
My personal preference is for dex over str too, but it needs parsing out. Obviously it'll also have an impact on the mobs. We know attack can have diminishing returns on trash or lower level mobs, possibly even on raid bosses. While crit rate mods are a flat modifier applied everywhere. Similarly boosts that change your crit multiplier (hello shaman epic) benefit crit mods far more. This makes it a bit tricky to parse given weak AC on arena dummies (we can increase, but it's a bit of guesswork to know how much by).
You're assuming a lot here. It's known that the AC on shields isn't affected by the softcap on AC. What is up in the air, well as far as I've been able to find, is whether or not pre-cap AC and shield AC have a much greater effect than the overcap AC. It'd be risky for a tank to try and go full hStr but it could be interesting to see what effect it may have on mitigation and exactly how much 10 hSTR is worth compared to post soft cap AC But then I don't play a tank Back when the 51/50 server was online I got curious as it was common knowledge that Iksar AC bonus wasn't real. I created a half elf warrior and an iksar warrior and test copied both for some parsing. I couldn't get a decent parse because once you moved the test dummies power up where you could tell the difference the half elf would die to a spike within 10 minutes and the iksar would keep going for however long I wanted him to be online. Possibly for raid tanks with bosses that have strikethrough hSTR may be worth messing with even if it resulted in a loss of some post soft cap AC but that needs to be tested
One thing interesting from the original post is that shields are NOT immune to the soft cap. The raw number on the shield (plus hstr/10) increases the softcap. But the raw number on the shield (plus hstr/10) times 4/3 is added to the AC before softcap. In effect, 1/4 of the shield's AC (and hstr/10) is softcapped. This means that a shield will increase the effective mitigation AC of a warrior more than other classes. It also means that, relative to other slots, the mitigation AC increase from a shield is higher the lower your softcap is. AC from a shield is worth (.75 + .25*softcap%)/softcap% times as much as AC on other slots, not 1/softcap% as was previously believed.
I think you got that wrong Yakk. In his example post with his SK: Mitigation AC: 3413 [Value returned by 'ac compute' function on the server] SK - Level 100: Cap 488: Multiplier: 0.33. (488 * 82 / 100) = 400 488 + 400 = 888 888 + 381 (shield ac) = 1269 softcap 7767 > 1269 so: 7767-1269 = 6498 6498 * 0.33 = 2144 1269 + 2144 = 3413 Dzarn adds his shield ac, getting 1269 (armor before softcap). He then 1. subtracts it 2. Multiplies by softcap (0.33). 3. Adds 1269 back on. So the shield ac was never subjected to the softcap.
Also have one built myself based on a few assumtions. Lv 100, over 100 Agility, lv 18 Armor of wisdom, if monk under soft cap weight, max Combat stability, max Physical Enhancement (if available). trying to post pics below. I also went and allowedyour migitation AC to be calculated based off displaced AC and fewer inputs than the full calaculation. Because for me its a heck of a lot faster to look at displayed AC than total up all the AC i have. below are pics and links to them in case embedding them doesnt work. one is a blank and the other is a test with some fun dummy values that are close to the real thing to give an idea. either the yellow or the orange cells can be filled in (or both) and they output in the area below each. http://tinypic.com/view.php?pic=15yfywp&s=8 [/IMG] http://i58.tinypic.com/11bri3m.png
it was added though A sheild is included in "every peice of gear". As such the sheild AC does get added twice. Once for the AC sum and once for the AC softcap calculation. So were as for AC calaculation all AC is worth about 4/3 (every three AC nets you four AC sum) your sheild acualy nets you 7/3s (every three nets you 7 AC) from an AC migitation standpoint, as displayed AC is returned at step 18.
I still don't see that. Step 22 - subtract all ac under the hardcap (which has the shield ac added into it). a. Dzarn example of 1269 ac. So all your ac from a shield is not present in step 23 (multiply by softcap). There is no ac involved when the softcap is calculation. Its subtracted before you multiply. See Dzarns 2nd example where he uses numbers.
If you want some rough numbers to play with, in terms of mitigation ac (assuming you are already over the soft cap): shields add roughly (1 + x^2) times its value in mitigation ac, where x is your soft cap return. ac in other equipment adds roughly [(4/3)*x] to your mitigation ac
umm no, you subtract out all your AC over the soft cap, multiply that by your class's AC mod and then ADD BACK IN your soft cap to put it into one line. Truncate[Step3 + Step20+ ({Step18 - Step3 - Step20}* ClassMultiplier)] were step 3 is your Sheild AC, step 18 is your total AC (not displayed AC) and step 20 is your softcap AC.
The benefits of players being able to puzzle something out only exists if it's possible to puzzle them out. War, auth Magelo sorted by AC the Top 50 is a 159ac range from items/augs/food/drink (5562-5721) atm and the hStr range is 45 (282-327) for those same 50. 45 hStr is equal to ~10ac or 6.3% of the difference between the #1 and #50 Warrior by AC. Even if every Warrior posted a long mitigation parse the value of hStr would be essentially impossible to separate from RNG fluctuations in "only" 100 hour mitigation parses. Let alone other minor variables like Heroic Fortitude ranks. If Test had a vendor with non-Lore +50 Heroic augs then players would have the opportunity to test out blocks of 1000 hStr which is roughly the value of 222ac and have a reasonable chance to solve for the value of hStr while wearing a Shield. Instead it was simply impossible for players to determine Lifeweaver's Symbol (38ac/7hStr) was better than a 39ac with no hAgi/hStr aug and worse than a 40ac aug for mitigation purposes for a Tank with a Shield equipped.
5470 AC from all items in inventory window -- this includes the shield 350 base shield ac + 31 shield ac from heroic strength 0 from food, drink, tributes, trophies (5470 * 4) / 3 = 7293 note that we multiplied all displayed AC (including shield AC) by 4/3 If our shield AC went up by 30 points, the number above goes up by **40** points. We then add a bunch of other stuff to get displayed mitigation AC contribution. We get 7767 total. Again, if we upgraded our shield item AC by 30, this value would go up by 40 to 7807. Then we get to the softcap calculations: SK - Level 100: Cap 488: Multiplier: 0.33. (488 * 82 / 100) = 400 488 + 400 = 888 888 + 381 (shield ac) = 1269 softcap Note that 381 is your shield AC, not 4/3 times your shield AC. So if your shield AC went up by 30, the softcap goes up by 30. In short, the pre-softcap AC goes up by 40, and the softcap goes up by 30, when you increase your shield AC by 30. We then apply the softcap. (7767-1269)*.33 + 1269 = 3413 Suppose our shield gained 30 AC. Then the numbers change as follows (7807-1299)*.33 + 1299 = 6508*.33 + 1299 = 2147 + 1299 = 3446 Your effective mitigation AC goes up by 33 when your shield AC goes up by 30. Not the 40 that your displayed AC goes up by, nor the 30 that your shield AC changes by, but rather (shield AC change + shield AC change * 1/3 * softcap%). In short, the real mitigation AC you gain from a shield's AC changes based off of your classes softcap%, because only 3/4 of its contribution to your mitigation AC boosts your softcap!
byte, short, int, and long are only a concern for fitting the number in the register and has no impact aside from speed of calculations and storage. public static void main(String[] args) { double v = (9d/2d); // the d suffix is forcing the values to be treated as double. int value = (int)v; System.out.println("Float value is: " + v + " but cast to int is: " + value); } output: "Float value is 4.5 but cast to int is: 4 Since he stated that they are using integer math (unless otherwise stated) you can see the effects here public static void main(String[] args) { int v1 = 9; int v2 = 2; double v = (v1/v2); int value = (int)v; System.out.println("Float value is: " + v + " but cast to int is: " + value); } output = "Float value is: 4.0 but cast to int is: 4" Meaning, the int(v) primitive type cast is not required if you are dividing integers. There is a precision loss, but the computations happen much quicker with integer math when you take into account server loading. When I code things like PID and Servo feedback simulations, I don't use integer math, and I don't cast the double value until all calculations are done. I need the precision or I'll risk hitting a limit switch or a hard stop (system fault, power down and cycle stop = bad) and I can do it because any given system I'm running only has maybe 50 feedback loops per PC tops. I can also delay trend data in a queue and take an average within sample window for most applications. Not sure if EQ does this too.... probably does.
Repeating something you said earlier does not make it any more correct. Worn AC (including shield AC) is multiplied by 4/3 before being used to calculate displayed AC. Shield AC increases the softcap **without** being multiplied by 4/3. So if you have a shield with 99 AC, and without it equipped you had 8000 displayed mitigation portion of AC, afterwards your displayed mitigation portion of AC is 8132. If before your softcap was 700 and your return after the cap was 25%, your real mitigation AC is (8000-700)*.25 + 700 = 2525 After equipping the shield your softcap becomes 799. So (8132-799)*.25+799 = 2632.25 (rounded down). This is +107 actual mitigation AC, or 99 + 99/3 * .25 = 107.25 increased actual mitigation AC. Not 99 + 99 * .25^2. There is no portion of the calculation that I can see that involves multiplying your returns past the softcap with your returns past the softcap.
