dipstik Posted March 21, 2015 Share Posted March 21, 2015 (edited) Lets hope they dont delete this one: updates: 10/8/15: revised to correct for using incorrect uptime for shrouded crusader relic. removed b-mod math. 4/14/15: updated to include the returned boss. the N/A bug for best relic should be fixed. now refering to spike as sigma. 4/12/15: new bmods link. all numbers updated fixed a huge bug where i was missing a "1 - " for squish calcs. for example, kinetic weapon used to be: kw_squish(d,s,a)= (0*d+[(1-d)*(1-s)+(1-d)*s*(1-a)]*(dr_k+dr_w+dr)) , but should have been kw_squish(d,s,a)= (0*d+[(1-d)*(1-s)+(1-d)*s*(1-a)]*(1-(dr_k+dr_w+dr))) now using a swing timer of 0.8333 seconds and a kineticx bulwark regression specific to that timer. new regression is absorb= (-13.901*p^2 + 19.443*p- 0.9247)/100 . spike is now set to the square root of the previous version of spike. spike is now calculated with squish values independent of shield, defense and resistance, which are taken into account in the probability portion of the spike calculation. paid closer attention when determining best relics. due to the way in which stats are optimized, only the relic which was reported as being the best was being calculated for, but changing the rating could make one relic better than another. 3/24/15: added lettered mod numbers. trade 162 mitigation for 288 endurance. i now have 2 tables and 2 spreadsheets to keep the separated. 3/23/15c: added spike calculations. basically a weighted variance. does not take relic procs into account. 3/23/15b: updated for new bulwark math. added boss dps required to make 2nd highest proc relic better than warding relic. Added base squish (includes static defense from relics). 3/23/15: new bulwark math regression: T is timer, p is shield chance effective_absorb=(19.169*p^2-16.317*p+.8901)*T-29.966*p^2+33.146*p-1.4262 note that bulwark math does not include dynamic relic contributions. 3/22/15b: updated math for reactive warding contribution. optimal stats/relics determined by damage taken, rather than squish directly. fixed some bugs: fixed: Absorb&shield from shield offhand counted twice for jugg and pt. updated for updated damage profiles. updated for 1928 shield upper bound. updated for defense from stim (added to defense minimum and pool) 3/22/15a: fixed typo pertaining to reactive warding relic (thanks namesaretough)... received updated damage profiles (thanks Methoxa). These are suggested defense, (d), shield (s) and absorb (a) ratings for full revenite gear. boss_damage*squish - self _heals*heal_buff - heals_received*healer_buff*heal_buff - self_bubble - relic bubble = damage taken. Squish is the fraction of damage taken, so the lower the better. I have assumed 6,000 dps for these calcs. Just copy the coded sections and paste into notepad or excel (and delimit for spaces). it will look alot better. or just download the spreadsheet. I optimized the stats for the relics which provide best mitigation. Relic shorthand used: D: Fortunate Redoubt (defense) A: Shield Amplification (absorb) S: Shrouded Crusader (shield and absorb) W: Reactive Warding (bubble) reactive warding relic best if boss dps < relic_bubble/((squish_with_relic - squish_without_relic)*cooldown). I have =ward in the tables that shows what the boss dps has to be to make the 2nd best relic for that fight better than the reactive warding relic. Please find my errors. Bugs Need to sort out which abilities get multiplicative dr buff. Need to sort out b mods and endruance stuff. i need pre mitigation damage for each boss, and the fight length. also want a spike matters term between 0 and 10 for each boss. and a post mitigation spike of interest, and an amount of health your healers would like you to have after such a hit. (x factor, spike, buffer values to optimize with) need to find absorb minimum for some spike calcs. Using excel solver... i try to use different seeds if i see funny data UNLETTERED sin Malfar Squad Undrlrk Cmndrs Retrnd Sparky Bulo Torque Mstr Cortani d 1567 185 997 605 185 185 823 1142 185 185 s 1200 1928 1574 1676 1901 1901 1558 1496 1928 1928 a 1236 1890 1432 1722 1917 1917 1622 1366 1890 1890 base 0.43128 0.2641 0.2234 0.2825 0.3977 0.3471 0.2764 0.3376 0.3464 0.4328 D 0.42892 0.2603 0.2188 0.2780 0.3945 0.3431 0.2720 0.3343 0.3434 0.4305 A 0.42919 0.2590 0.2186 0.2785 0.3945 0.3436 0.2725 0.3341 0.3422 0.4298 S 0.43000 0.2610 0.2206 0.2800 0.3957 0.3449 0.2739 0.3356 0.3438 0.4309 relic WD WA WA WD WD WD WD WA WA WA =ward 18459 10117 8402 9802 12324 10898 10032 11406 12899 16332 sigma 0.29622 0.2110 0.2062 0.2446 0.2720 0.2634 0.2461 0.2753 0.2528 0.2684 jugg Malfar Squad Undrlrk Cmndrs Retrnd Sparky Bulo Torque Mstr Cortani d 2005 505 1644 1564 1146 1139 1663 1707 417 560 s 1200 1868 1200 1200 1527 1422 1200 1200 1915 1839 a 798 1630 1159 1239 1330 1442 1140 1096 1671 1605 base 0.43968 0.2621 0.2262 0.2852 0.4009 0.3498 0.2799 0.3423 0.3457 0.4350 D 0.43758 0.2584 0.2223 0.2819 0.3985 0.3467 0.2765 0.3394 0.3427 0.4328 A 0.43778 0.2581 0.2227 0.2822 0.3983 0.3470 0.2769 0.3397 0.3425 0.4326 S 0.43862 0.2597 0.2240 0.2832 0.3993 0.3480 0.2780 0.3407 0.3437 0.4336 relic WD WA WD WD WA WD WD WD WA WA =ward 20209 10367 11147 12774 15863 14169 12556 14710 12760 17106 sigma 0.290040055 0.1888 0.2054 0.2426 0.2619 0.2568 0.2443 0.2711 0.2339 0.2522 pt Malfar Squad Undrlrk Cmndrs Retrnd Sparky Bulo Torque Mstr Cortani d 1352 185 915 776 458 557 948 819 185 185 s 1495 1928 1836 1910 1928 1928 1818 1778 1928 1928 a 1157 1890 1252 1317 1617 1518 1237 1406 1890 1890 base 0.45885 0.2318 0.2107 0.2755 0.4010 0.3438 0.2723 0.3441 0.3299 0.4362 D 0.45668 0.2285 0.2066 0.2719 0.3985 0.3408 0.2687 0.3408 0.3273 0.4342 A 0.45687 0.2281 0.2064 0.2717 0.3983 0.3406 0.2684 0.3411 0.3268 0.4340 S 0.45756 0.2291 0.2081 0.2732 0.3992 0.3418 0.2700 0.3421 0.3277 0.4346 relic WD WA WA WA WA WA WA WD WA WA =ward 19544 11698 9534 10694 15407 12937 10794 13055 14912 18886 sigma 0.305009542 0.1904 0.1776 0.2378 0.2847 0.2705 0.2386 0.2787 0.2587 0.2849 For an average of the 9 bosses posted, using b-mods will increase damage taken by: sin: 0.0032*dps jugg: 0.0025*dps pt: 0.0026*dps Assumptions and math no cooldowns, just perfect execution. using new set bonuses assuming 75% bland melee/ranged kinetic/energy damage. Pool of 4003 from 1200 shield and 90*12+81*10 defense/absorb, and 52 defense from each relic and 81 defense from stim. Jugg ear of 171 defense taken into account for minimum defense. for b-mods, subtract 162 from defense/absorb pool, and add 288 endurance. heavy armor: 4*758+1212*3 light armor: 4*401+3*642 definitions: defense d, absorb a, shield s, (http://www.swtor.com/community/showthread.php?t=779094) fraction m/r+k/e kw (w for weapon), fraction f/tr+k/e kf, fraction m/r i/e ew, fraction f/t +i/e ef, fraction named fn, fraction bland fb, damage reduction of k/e dr_k, damage reduction of i/e dr_e, damage reduction of m/r dr_w, damage reduction of f/t dr_f, overall damage reduction dr rdd is dynamic relic buff to defense rating, rda is dynamic relic buff to absorb rating, rds is dynamic relic buff to shield rating. rsd is static relic buff to defense parentheses without a * means a function of. like f(x,a,b)=a*x+b kw_squish(d,s,a)= (0*d+[(1-d)*(1-s)+(1-d)*s*(1-a)]*(1-(dr_k+dr_w+dr))) ew_squish(d,s,a)=(0*d+[(1-d)*(1-s)+(1-d)*s*(1-a)]*(1-(dr_e+dr_w+dr))) kf_squish(d,s,a)=(0*r+[(1-r)*(1-s)+(1-r)*s*(1-a)]*(1-(dr_f+dr_k+dr))) ef_squish(d,s,a)=(0*r+(1-r)*(1-(dr_e+dr_f+dr))) squish(d,s,a)= kw*(fn*kw_squish(d)+fb*kw_squish(d+0.1)) + ew*ew_squish + kf*kf_squish + ef*ef_squish def_relic_squish=6/22*squish(d+rdd+rsd,s,a) + ((22-6)/22)*squish(d+rsd,s,a) two relics: 6/22*6/22*both_on+6/22*16/22*one_on+6/22*16/22*other_on+16/22*16/22*both_off similar for other relics. sigma= sqrt(sum(probability[ij]*(squish[ij] - squish)^2) where i runs over kw_named, kw_bland, kf and ef (i thought there was ew damage, but i guess not) and j runs over defended, resisted, shielded and full hit. note that squish[ij] differs from the squish previously calculated as it is independent of defense, shield and resistance. see the spreadsheet for more detail. SHADOW m/r dr: 0.05 slow time active i/e dr: 0.10 willpower buff passive 0.10 technique mastery passive 0.02 old 4 set overall dr: 0.02 elusiveness passive tree 0.02 mental fortitude passive tree 0.04 harnessed shadows active utility 0.0114 new 4 set passive 0.02 old 4 set passive endurance 0.01 companion passive 0.02 mental fortitude passive tree 0.03 shadows training passive 0.05 aim buff passive heals: 0.03 shadow shelter passive utility 0.01 companion passive armor scalar 0.2 elusivness passive 1.3 stance passive resistance: 0.02 shadow sight passive tree defense 0.05 psych passive tree 0.04 double saber tactics passive tree 0.02 shadow sight passive tree 0.05 being a consular passive shield 0.15 kinetic ward active 0.05 shield passive 0.15 stance passive 0.05 old 2 set absorb: A(bulwark) active 0.04 shadowsight passive 0.20 shield passive Bulwark math with set bonus. swing timer T, effective absorb contribution A. > m(T):=floor((23)/(T)): > with(SumTools); > p(d,s,r,kw,kf,fb,fn):=1/(kw+kf)*(kf*s*(1-r)+kw*s*(fb*(1-(d+0.1))+fn*(1-d))): > E(n,p):=Summation(piecewise(i<9,(n!)/(i!*(n-i)!)*p^(i)*(1-p)^(n-i)*i,i<18,(n!)/(i!*(n-i)!)*p^(i)*(1-p)^(n-i)*8,18<=i,0),i=1..n): > P(m,p):=piecewise(m>=18,(m!)/(18!*(m-18)!)*p^(18)*(1-p)^(m-18),0): > A(T,p):=piecewise(m(T)<18, 1/(m(T))*Summation(E(n,p),n=1..m(T)), Summation((Summation(E(n,p),n=1..18+i)*P(18+i,p))/(18+i),i=0..m(T)-18)+ ((1- Summation(P(n,p),n=18..m(T)))* Summation(E(n,p),n=1..m(T)))/(m(T))): GUARDIAN f/t dr: 0.05 guardian slash active i/e dr: 0.05 inner peace passive tree 0.10 willpower buff passive overall dr: 0.03 warding strike active 0.06 stance passive 0.0075 new 2 set (.02 * 4/12) active endurance: 0.01 companion passive 0.05 aim buff passive heals: 0.01 companion passive self heals: active 1 * BonusHealing + 0.05 * 14520 setting to 1600 every 12 seconds replace with 1.2*(BonusHealing + 0.05 * 14520) if using old 4 piece armor scalar: 0.15 exp protector passive tree 0.60 stance passive defense: 0.05 dust storm active 0.05 blade barricade (riposte) active 0.03 stance passive 0.03 single saber mastery passive shield: 0.04 shield specialist passive tree 0.05 shield passive 0.15 stance passive absorb: 0.03 warding power active 0.20 shield passive POWERTECH m/r dr: 0.05 combust active f/t dr: 0.05 firestorm active i/e dr: 0.10 willpower buff passive overall dr: 0.02 power armor passive tree 0.02 ion screen passive (stance) 0.05 stance passive 0.0083 new 2 set (0.02*5/12) active enduance: 0.01 companion passive 0.05 aim buff passive heals: 0.01 companion passive self heals: 20% health every 90 seconds from shield cannon 2.5% health every aoe, 3 sec cooldown, passive - conditional armor scalar: 0.15 power armor passive tree 0.60 stance passive resist: 0.02 old 4 set passive defense: 0.04 deflective plating passive tree 0.02 old 4 set passive shield: 0.02 shield vents passive tree 0.03 shield enhancers passive tree 0.05 shield passive 0.15 stance passive absorb: 0.1475 heat screen and heat blast active 0.04 alblative upgrade passive tree 0.20 shield passive heat screen: 3% for 6 seconds after heat blast heat blast: during heat blast: (25+0)*1.5+(25+1)*1.5+(25+2)*1.5*1+(25+3)1.5 taking time averaged for both gives (3*6+25*1.5+26*1.5+27*1.5+28*1.5)*(1/12) = 14.75% absorb. Damage weight info used: HM Malfar Squad Undrlrk Cmndrs Retrnd Sparky Bulo Torque Mstr Cortani m/r+k/e 0.526 0.515 0.832 0.704 0.4341 0.5366 0.744 0.646 0.404 0.319 m/r+i/e 0 0 0 0 0 0 0 0 0 0 f/t+k/e 0 0.485 0.168 0.181 0.23 0.2314 0.141 0.104 0.409 0.287 f/t+i/e 0.474 0 0 0.115 0.3359 0.232 0.115 0.25 0.187 0.394 link to spreadsheet: https://drive.google.com/file/d/0B_oeO3cuqFbwWjhWYklyeGtTYzA/view?usp=sharing Edited October 8, 2015 by dipstik Link to comment Share on other sites More sharing options...
Methoxa Posted March 21, 2015 Share Posted March 21, 2015 Force Scream Dipstick : (0.05 * 11180 + BonusHealing) *bonus_healing cs_zoltan : 1 * BonusHealing + 0.05 * 14520 Link to comment Share on other sites More sharing options...
dipstik Posted March 21, 2015 Author Share Posted March 21, 2015 (edited) thanks. so with 3.0 there are now damage reductions for all damage types and attack types. i am assuming they are summed, but i can imagine that overall is multiplied after the sums of the specific types. for example: i have 5% m/r damage reduction, 6% k/e damage reduction, lets say 20% k/e damage reduction due to armor, and 10% overall damasge reduction. for a m/r + k/e hit would squish look like [0*d+(1-d)*s*(1-a)+(1-d)*(1-s)]*[1-(0.05+(0.06+0.2))]*(1-0.1) or [0*d+(1-d)*s*(1-a)+(1-d)*(1-s)]*[1-(0.10+0.05+0.06+0.2)] or they could all be multiplicative... [0*d+(1-d)*s*(1-a)+(1-d)*(1-s)]*[(1-0.05)*(1-0.06-0.02)*(1-0.1)] i seem to remember that debuffs that provide overall damage reduction were multiplied after everything from the tree. Edited March 22, 2015 by dipstik Link to comment Share on other sites More sharing options...
Methoxa Posted March 22, 2015 Share Posted March 22, 2015 i seem to remember that debuffs that provide overall damage reduction were multiplied after everything from the tree. They are, however active abilites like crushing blow armour buff is additive. m/r+k/e m/r+i/e f/t+k/e f/t+i/e Malaphar 0.526 0 0 0.474 Swd Squd 0.515 0 0.485 0 Sparky 0.722 0 0 0.278 Bulo 0.744 0 0.141 0.115 Torque 0.646 0 0.104 0.25 Mstr/Blstr 0.542 0 0.299 0.159 Those are close to my numbers. However you seem to have wrong values for Sparky. His shoulder Throw and his throw back are ft/ke. In my old values i assumed it would be f/T+i/e. New Values : #1 Sparky : M/R+K/E : 54.24% F/T+K/E : 23.14% F/T+I/E : 23,2% DEF 0.2362 0.2679 0.2226 0.2305 0.2383 0.2573 ABS 0.2370 0.2658 0.2229 0.2306 0.2383 0.2567 SHD 0.2364 0.2655 0.2227 0.2303 0.2381 0.2564 Interesting to see that the Juggernaut is the tank receiving the most dmg especially in fights with a high F/T+K/E amount while his def CD's are not effective against that kind of dmg except for invincible and in some rare cases saber reflect. In case you need dmg values of the bosses for HM : #1 Sparky : M/R+K/E : 54.24% F/T+K/E : 23.14% F/T+I/E : 23,2% #2 Bulo : M/R+K/E: 74,4% F/T+K/E: 14,1% F/T+I/E: 11,5% #3 Torque : M/R+K/E: 65,6% F/T+K/E: 10,4% F/T+I/E: 25% #4 Master/Blaster : M/R+K/E : 40,4% F/T+K/E : 40,9% F/T+I/E : 18,7% #5 Coratanni < Cora + Ruugar> M/R+K/E : 31,9% F/T+K/E : 28,7% F/T+I/E : 39,4% #1 Malaphar : M/R+K/E: 52,6% F/T+K/E: 0% F/T+I/E: 47.4% #2 Sword Squadron : M/R+K/E: 51,5% F/T+K/E: 48,5% F/T+I/E: 0% #3 Underlurker : M/R+K/E: 83,2% F/T+K/E: 16,8% F/T+I/E: 0% #4 Commanders : M/R+K/E: 70,4% F/T+K/E: 18,1% F/T+I/E: 11,5% #5 The Returned : M/R+K/E: F/T+K/E: F/T+I/E : Link to comment Share on other sites More sharing options...
namesaretough Posted March 22, 2015 Share Posted March 22, 2015 (edited) Two comments that have little to do with stat weights: Unless I'm blind there's no way to reach some of those shield values. The upper cap for shield in 198s should be 1928. and The reactive warding relic is Best in Slot (BiS) if Boss_dps > 3000 in all cases, in most cases the number is between 1800 and 2500. I think you mean "less than". I haven't looked through the details, those are just two things that jumped out at me. Edited March 22, 2015 by namesaretough Link to comment Share on other sites More sharing options...
dipstik Posted March 22, 2015 Author Share Posted March 22, 2015 (edited) thanks, i will update the numbers for new boss damage profiles, and set it up so shield cant be more than 1928. i also have to add defense minimum from the fortitude stim. (edit:updated) so would all of the buffs from overall dr be outside the sum and multiplied then? shadows: overall dr: 0.02 elusiveness passive tree 0.02 mental fortitude passive tree 0.04 harnessed shadows active utility 0.0114 new 4 set passive 0.02 old 4 set passive guardians: overall dr: 0.03 warding strike active 0.06 stance passive 0.0075 new 2 set (.02 * 4/12) active Powertech: overall dr: 0.02 power armor passive tree 0.02 ion screen passive (stance) 0.05 stance passive 0.0083 new 2 set (0.02*5/12) active Edited March 25, 2015 by dipstik Link to comment Share on other sites More sharing options...
Magsel Posted April 3, 2015 Share Posted April 3, 2015 Cheers for doing the math for us again dipstik I downloaded the spreadsheet for unlettered mods and noticed some cells (on the shadow/jugg/pt tabs) can be modified with my own dsa ratings. Do these cells need to be modified with my own numbers? I also noticed the Endruance and Armor cell, however the Armor cell contains a formula and I'm afraid I'll ruin the spreadsheet if I put in my own armor rating (if the field is for the armor rating ofcourse). When I change the endurance (4448), armor (rating?) (11111), defense (1492), shield (1042) and absorb (743) values for my jugg I get an error in cell F20 on the jugg tab. The error is #N/B. I'm using Excel 2013. When I change the values to the jugg numbers (2005/1200/798) you use for malfar on the bosses tab I also get the same error. When I change the same numbers for my shadow and sin I do not get that error. Mags Link to comment Share on other sites More sharing options...
dipstik Posted April 3, 2015 Author Share Posted April 3, 2015 when you are changing stuff, dont use E13:F22 block. thats just a summary block. the cell you are talking about does get weird sometimes, i think its when 2 relic combos are really close or something. you can just look at the highlighted relic combo to get that info, A17:A23 will light up, the cell next to it will tell you the best relic combo. if you are entering armor rating, you will want to figure out the pre-buff number to put in there Link to comment Share on other sites More sharing options...
Japina Posted April 8, 2015 Share Posted April 8, 2015 So my Juggernaut has base stat 3803 separated for defence 1640, shield 1175 and absorb 988. I read in other KBN thread that ideal would be 1191, 1287 and 1322 for base stat 3800. Clearly I have too much defence now and I should go for more shield and absorb. Am I a bad tank if I don't have ideal stats? What I did, I bought some absorb and shield mods and enhancements for balancing my stats. This lowered my base stats so I had thoughts it cant be right. I replaced my all defence augments with absorb augments but im not event close to ideal. Does anyone has to say for this? Link to comment Share on other sites More sharing options...
dipstik Posted April 9, 2015 Author Share Posted April 9, 2015 i see kbn replied in another thread, but i will reiterate,. getting more stats is better than getting optimal stats Link to comment Share on other sites More sharing options...
Jethsidi Posted April 11, 2015 Share Posted April 11, 2015 Just re-read your number man, good stuff. Going to use some of it in my next project. Writing something that factors in tank cooldowns, going to take a while. Link to comment Share on other sites More sharing options...
dipstik Posted April 12, 2015 Author Share Posted April 12, 2015 Just re-read your number man, good stuff. Going to use some of it in my next project. Writing something that factors in tank cooldowns, going to take a while. good luck! ive gotten 10% of the way on that trip a few times. Link to comment Share on other sites More sharing options...
Jethsidi Posted April 26, 2015 Share Posted April 26, 2015 Just a fast question: Is your math based on the amount of damage going out per damage type of hit? Or the amount of hits going out per damage type? Link to comment Share on other sites More sharing options...
dipstik Posted April 26, 2015 Author Share Posted April 26, 2015 (edited) its based on amount of damage going out, where hits and damage types are coupled, and frequency is modeled by the weights of those couplings. Edited April 26, 2015 by dipstik Link to comment Share on other sites More sharing options...
Jethsidi Posted April 26, 2015 Share Posted April 26, 2015 Ah. Thanks. Mine is mainly based in here's the number of hits in the fight, here's how many of each Type of hit there is, and here's the stats that can be used to counter that type. The one I'm working on for cooldowns is weighted on the side of how much they hit for, but they are about to nerf the ops so everything is going to change. Link to comment Share on other sites More sharing options...
dipstik Posted October 9, 2015 Author Share Posted October 9, 2015 (edited) first pass at 4.0 numbers, with a bunch of missing info. the squish numbers are going to be off due to not having the armor rating figured out. very little difference between high melee/ranged and high force/tech stat distributions. this assumes unlettered purple 44 gear with sturdiness/immunity numbers. sin Malfar Squad Undrlrk Cmndrs Retrnd Sparky Bulo Torque Mstr Cortani d 2915 2915 2915 2915 2915 2915 2915 2915 2915 2915 s 1374 1374 1364 1366 1371 1369 1364 1363 1374 1374 a 1328 1328 1338 1336 1331 1333 1338 1339 1328 1328 base 0.4421 0.2803 0.2159 0.2775 0.4007 0.3478 0.2693 0.3310 0.3608 0.4421 D 0.4412 0.2789 0.2136 0.2756 0.3996 0.3463 0.2673 0.3292 0.3597 0.4412 A 0.4396 0.2762 0.2128 0.2746 0.3983 0.3451 0.2666 0.3288 0.3574 0.4396 S 0.4398 0.2766 0.2130 0.2749 0.3985 0.3453 0.2668 0.3289 0.3576 0.4398 relic WA WA WA WA WA WA WA WA WA WA =ward 26966 16200 21325 23251 27265 24512 24306 29287 19656 26966 sigma 0.2744 0.2360 0.2223 0.2605 0.2835 0.2788 0.2597 0.2858 0.2651 0.2744 jugg Malfar Squad Undrlrk Cmndrs Retrnd Sparky Bulo Torque Mstr Cortani d 2915 2915 2915 2915 2915 2915 2915 2915 2915 2915 s 1408 1408 1329 1329 1408 1408 1329 1329 1408 1408 a 1294 1294 1373 1373 1294 1294 1373 1373 1294 1294 base 0.4372 0.2665 0.2179 0.2785 0.3988 0.3460 0.2724 0.3357 0.3502 0.4372 D 0.4363 0.2649 0.2154 0.2765 0.3976 0.3444 0.2702 0.3338 0.3491 0.4363 A 0.4355 0.2636 0.2156 0.2765 0.3971 0.3440 0.2705 0.3340 0.3479 0.4355 S 0.4355 0.2637 0.2156 0.2765 0.3972 0.3441 0.2705 0.3340 0.3480 0.4355 relic WA WA WD WD WA WA WD WD WA WA =ward 23403 14089 17292 19050 23052 20510 19659 23541 17152 23403 sigma 0.2597 0.2105 0.2146 0.2515 0.2709 0.2667 0.2527 0.2789 0.2465 0.2597 pt Malfar Squad Undrlrk Cmndrs Retrnd Sparky Bulo Torque Mstr Cortani d 2915 2915 2915 2915 2915 2915 2915 2915 2915 2915 s 1609 1609 1609 1609 1609 1609 1609 1609 1609 1609 a 1093 1093 1093 1093 1093 1093 1093 1093 1093 1093 base 0.3462 0.2465 0.2071 0.2737 0.4048 0.3462 0.2688 0.3405 0.3428 0.4447 D 0.3449 0.2453 0.2051 0.2721 0.4037 0.3449 0.2670 0.3390 0.3419 0.4439 A 0.3438 0.2430 0.2042 0.2711 0.4026 0.3438 0.2662 0.3384 0.3400 0.4426 S 0.3439 0.2432 0.2043 0.2712 0.4027 0.3439 0.2663 0.3385 0.3401 0.4427 relic WA WA WA WA WA WA WA WA WA WA =ward 16797 11705 13933 15406 18951 16797 15826 18911 14267 19426 sigma 0.2775 0.2016 0.1929 0.2492 0.2885 0.2775 0.2494 0.2870 0.2601 0.2830 Edited October 9, 2015 by dipstik Link to comment Share on other sites More sharing options...
dipstik Posted October 19, 2015 Author Share Posted October 19, 2015 (edited) here are my assumptions going into 4.0 no cooldowns, just perfect execution. using new set bonuses assuming 75% bland melee/ranged kinetic/energy damage. Pool of 4003 from 1200 shield and 90*12+81*10 defense/absorb, and 52 defense from each relic and 81 defense from stim. Jugg ear of 171 defense taken into account for minimum defense. for b-mods, subtract 162 from defense/absorb pool, and add 288 endurance. heavy armor: 4*758+1212*3 light armor: 4*401+3*642 definitions: defense d, absorb a, shield s, (http://www.swtor.com/community/showthread.php?t=779094) fraction m/r+k/e kw (w for weapon), fraction f/tr+k/e kf, fraction m/r i/e ew, fraction f/t +i/e ef, fraction named fn, fraction bland fb, damage reduction of k/e dr_k, damage reduction of i/e dr_e, damage reduction of m/r dr_w, damage reduction of f/t dr_f, overall damage reduction dr rdd is dynamic relic buff to defense rating, rda is dynamic relic buff to absorb rating, rds is dynamic relic buff to shield rating. rsd is static relic buff to defense parentheses without a * means a function of. like f(x,a,b)=a*x+b kw_squish(d,s,a)= (0*d+[(1-d)*(1-s)+(1-d)*s*(1-a)]*(1-(dr_k+dr_w+dr))) ew_squish(d,s,a)=(0*d+[(1-d)*(1-s)+(1-d)*s*(1-a)]*(1-(dr_e+dr_w+dr))) kf_squish(d,s,a)=(0*r+[(1-r)*(1-s)+(1-r)*s*(1-a)]*(1-(dr_f+dr_k+dr))) ef_squish(d,s,a)=(0*r+(1-r)*(1-(dr_e+dr_f+dr))) squish(d,s,a)= kw*(fn*kw_squish(d)+fb*kw_squish(d+0.1)) + ew*ew_squish + kf*kf_squish + ef*ef_squish def_relic_squish=6/22*squish(d+rdd+rsd,s,a) + ((22-6)/22)*squish(d+rsd,s,a) two relics: 6/22*6/22*both_on+6/22*16/22*one_on+6/22*16/22*other_on+16/22*16/22*both_off similar for other relics. sigma= sqrt(sum(probability[ij]*(squish[ij] - squish)^2) where i runs over kw_named, kw_bland, kf and ef (i thought there was ew damage, but i guess not) and j runs over defended, resisted, shielded and full hit. note that squish[ij] differs from the squish previously calculated as it is independent of defense, shield and resistance. see the spreadsheet for more detail. SHADOW m/r dr: 0.05 slow time active i/e dr: 0.10 willpower buff passive 0.10 technique mastery passive 0.02 old 4 set overall dr: 0.02 elusiveness passive tree 0.02 mental fortitude passive tree 0.04 harnessed shadows active utility 0.0114 new 4 set passive 0.02 old 4 set passive endurance 0.01 companion passive 0.02 mental fortitude passive tree 0.03 shadows training passive 0.05 aim buff passive heals: 0.03 shadow shelter passive utility 0.01 companion passive armor scalar 0.2 elusivness passive 1.3 stance passive resistance: 0.02 shadow sight passive tree defense 0.05 psych passive tree 0.04 double saber tactics passive tree 0.02 shadow sight passive tree 0.05 being a consular passive shield 0.18 kinetic ward active 3% more in 4.0 0.05 shield passive 0.15 stance passive 0.05 old 2 set absorb: A(bulwark) active 0.04 shadowsight passive 0.20 shield passive Bulwark math with set bonus. swing timer T, effective absorb contribution A. > m(T):=floor((23)/(T)): > with(SumTools); > p(d,s,r,kw,kf,fb,fn):=1/(kw+kf)*(kf*s*(1-r)+kw*s*(fb*(1-(d+0.1))+fn*(1-d))): > E(n,p):=Summation(piecewise(i<11,(n!)/(i!*(n-i)!)*p^(i)*(1-p)^(n-i)*i,i<20,(n!)/(i!*(n-i)!)*p^(i)*(1-p)^(n-i)*10,20<=i,0),i=1..n): > P(m,p):=piecewise(m>=20,(m!)/(20!*(m-20)!)*p^(20)*(1-p)^(m-20),0): > A(T,p):=piecewise(m(T)<20,1/(m(T))*Summation(E(n,p),n=1..m(T)),Summation((Summation(E(n,p),n=1..20+i)* P(20+i,p))/(20+i),i=0..m(T)-20)+((1- Summation(P(n,p),n=20..m(T)))*Summation(E(n,p),n=1..m(T)))/(m(T))): this assumes 23 second ward duration, 20 charges, 10 stacks of bulwark. i get the following table, which gives me the regression absorb% = 4.87 + 10.6*prob - 4.43*swing_time prob time time p 1.035 0.85 0.35 4.003 4.793 0.355 4.058 4.856 0.36 4.112 4.915 0.365 4.167 4.975 0.37 4.221 5.034 0.375 4.275 5.09 0.38 4.33 5.149 0.385 4.383 5.205 0.39 4.436 5.261 0.395 4.489 5.313 0.4 4.542 5.369 0.405 4.595 5.422 0.41 4.645 5.474 0.415 4.7 5.526 0.42 4.75 5.575 0.425 4.8 5.628 0.43 4.85 5.682 0.435 4.9 5.728 0.44 4.95 5.778 0.445 5 5.824 0.45 5.05 5.872 GUARDIAN f/t dr: 0.05 guardian slash active i/e dr: 0.05 inner peace passive tree 0.10 willpower buff passive overall dr: 0.03 warding strike active 0.06 stance passive 0.0075 new 2 set (.02 * 4/12) active endurance: 0.01 companion passive 0.05 aim buff passive heals: 0.01 companion passive self heals: active 1 * BonusHealing + 0.05 * 14520 setting to 1600 every 12 seconds replace with 1.2*(BonusHealing + 0.05 * 14520) if using old 4 piece armor scalar: 0.15 exp protector passive tree 0.60 stance passive defense: 0.05 dust storm active 0.05 blade barricade (riposte) active 0.03 stance passive 0.03 single saber mastery passive 2% per tick of ravage, lasts for 6 seconds in 4.0. ataking 2 ticks at timer=0 and last tick at 2.7 sec we get (4*(2.7-0)+6*(8.7-2.7)+0*(18-8.7))/18=2.6% defense shield: 0.04 shield specialist passive tree 0.05 shield passive 0.15 stance passive absorb: 0.03 warding power active 0.20 shield passive POWERTECH m/r dr: 0.05 combust active f/t dr: 0.05 firestorm active i/e dr: 0.10 willpower buff passive overall dr: 0.02 power armor passive tree 0.02 ion screen passive (stance) 0.05 stance passive 0.0083 new 2 set (0.02*5/12) active enduance: 0.01 companion passive 0.05 aim buff passive heals: 0.01 companion passive self heals: 20% health every 90 seconds from shield cannon 2.5% health every aoe, 3 sec cooldown, passive - conditional armor scalar: 0.15 power armor passive tree 0.60 stance passive resist: 0.02 old 4 set passive defense: 0.04 deflective plating passive tree 0.02 old 4 set passive shield: 0.02 shield vents passive tree 0.03 shield enhancers passive tree 0.05 shield passive 0.15 stance passive absorb: 0.1475 heat screen and heat blast active 0.04 alblative upgrade passive tree 0.20 shield passive heat screen: 3% +5% for 6 seconds after heat blast. 5% for 4.0 heat blast: during heat blast: (25+0)*1.5+(25+1)*1.5+(25+2)*1.5*1+(25+3)1.5 taking time averaged for both gives ((3+5)*6+25*1.5+26*1.5+27*1.5+28*1.5)*(1/12) = 17.25% absorb. Edited October 19, 2015 by dipstik Link to comment Share on other sites More sharing options...
dipstik Posted October 19, 2015 Author Share Posted October 19, 2015 (edited) and here are numbers assuming 4.0 changes at 6 piece set bonuses (and immunity type numbers)... please check my work. the pt nubers not changing makes me think i did something wrong, but they change if i set the stat pool to a higher value. sin Malfar Squad Undrlrk Cmndrs Retrnd Sparky Bulo Torque Mstr Cortani d 2915 2915 2915 2915 2915 2915 2915 2915 2915 2915 s 1363 1364 1338 1341 1353 1349 1338 1336 1366 1363 a 1339 1338 1364 1361 1349 1353 1364 1366 1336 1339 base 0.4369 0.2716 0.2095 0.2717 0.3957 0.3422 0.2638 0.3264 0.3536 0.4369 D 0.4360 0.2703 0.2073 0.2698 0.3945 0.3408 0.2618 0.3247 0.3525 0.4360 A 0.4343 0.2674 0.2063 0.2687 0.3932 0.3394 0.2610 0.3241 0.3500 0.4343 S 0.4345 0.2677 0.2066 0.2690 0.3934 0.3396 0.2612 0.3243 0.3503 0.4345 relic WA WA WA WA WA WA WA WA WA WA =ward 26045 15645 20645 22506 26367 23713 23532 28357 18978 26045 sigma 0.2758 0.2320 0.2180 0.2585 0.2841 0.2784 0.2577 0.2854 0.2646 0.2758 jugg Malfar Squad Undrlrk Cmndrs Retrnd Sparky Bulo Torque Mstr Cortani d 2915 2915 2915 2915 2915 2915 2915 2915 2915 2915 s 1408 1408 1329 1329 1408 1408 1329 1329 1408 1408 a 1294 1294 1373 1373 1294 1294 1373 1373 1294 1294 base 0.4340 0.2613 0.2095 0.2715 0.3945 0.3407 0.2650 0.3292 0.3462 0.4340 D 0.4331 0.2598 0.2071 0.2694 0.3932 0.3391 0.2628 0.3273 0.3450 0.4331 A 0.4323 0.2585 0.2074 0.2696 0.3928 0.3387 0.2631 0.3276 0.3439 0.4323 S 0.4324 0.2586 0.2074 0.2695 0.3929 0.3388 0.2631 0.3276 0.3440 0.4324 relic WA WA WD WD WA WA WD WD WA WA =ward 23878 14367 17979 19754 23684 21130 20453 24531 17475 23878 sigma 0.2627 0.2124 0.2151 0.2537 0.2741 0.2697 0.2547 0.2819 0.2491 0.2627 pt Malfar Squad Undrlrk Cmndrs Retrnd Sparky Bulo Torque Mstr Cortani d 2915 2915 2915 2915 2915 2915 2915 2915 2915 2915 s 1640 1640 1640 1640 1640 1640 1640 1640 1640 1640 a 1062 1062 1062 1062 1062 1062 1062 1062 1062 1062 base 0.4422 0.2424 0.2036 0.2706 0.4022 0.3433 0.2657 0.3379 0.3394 0.4422 D 0.4415 0.2412 0.2017 0.2689 0.4012 0.3420 0.2640 0.3364 0.3385 0.4415 A 0.4401 0.2388 0.2006 0.2679 0.4000 0.3408 0.2631 0.3357 0.3365 0.4401 S 0.4402 0.2390 0.2008 0.2680 0.4001 0.3409 0.2632 0.3359 0.3367 0.4402 relic WA WA WA WA WA WA WA WA WA WA =ward 19027 11464 13647 15089 18561 16452 15501 18522 13973 19027 sigma 0.2853 0.2029 0.1932 0.2502 0.2905 0.2792 0.2503 0.2883 0.2622 0.2853 link to spreadsheet: https://drive.google.com/file/d/0B_oeO3cuqFbweUd2YUJZLWoyZzg/view?usp=sharing Edited October 19, 2015 by dipstik Link to comment Share on other sites More sharing options...
KeyboardNinja Posted October 19, 2015 Share Posted October 19, 2015 (edited) and here are numbers assuming 4.0 changes at 6 piece set bonuses (and immunity type numbers)... please check my work. the pt nubers not changing makes me think i did something wrong, but they change if i set the stat pool to a higher value. The defense minima on the new gear virtually eliminates gear customization, since there are relatively few effects which have a significant impact on the relative value of shield and absorb (due to the way that they trade off against each other). So it doesn't surprise me that the numbers are fairly constant. It is worth noting that my (very preliminary!) work produces slightly different results for shadows (lowMR followed by highMR): {{defense -> 2931, shield -> 1207, absorb -> 1495}, {defense -> 2931, shield -> 1202, absorb -> 1500}} These results don't take into account the new Kinetic Ward or the extra Kinetic Bulwark stacks, which obviously has an impact on things. EDIT Actually, an even bigger omission from my calculations is the fact that I'm using unlettered mods for both stat budget and defense min. There's literally no question that B mods will be BiS by a mile, so I need to make that adjustment. Edited October 19, 2015 by KeyboardNinja Link to comment Share on other sites More sharing options...
dipstik Posted October 19, 2015 Author Share Posted October 19, 2015 The defense minima on the new gear virtually eliminates gear customization, since there are relatively few effects which have a significant impact on the relative value of shield and absorb (due to the way that they trade off against each other). So it doesn't surprise me that the numbers are fairly constant. It is worth noting that my (very preliminary!) work produces slightly different results for shadows (lowMR followed by highMR): {{defense -> 2931, shield -> 1207, absorb -> 1495}, {defense -> 2931, shield -> 1202, absorb -> 1500}} These results don't take into account the new Kinetic Ward or the extra Kinetic Bulwark stacks, which obviously has an impact on things. EDIT Actually, an even bigger omission from my calculations is the fact that I'm using unlettered mods for both stat budget and defense min. There's literally no question that B mods will be BiS by a mile, so I need to make that adjustment. i will end up making two tables again, for lettered and unlettered. my guess is that trading 400 defense for 4500 health is probably worth it after you hit 2500 defense stat. Link to comment Share on other sites More sharing options...
Recommended Posts