peddroelm

test subject was elder lion on POTD 4 DR assumed 4 burn DR 27% might bonus +0 ms |Health Damage 0.000000 // apply ring spell +3601 ms |Health Damage 13.803299 // sabre hit +28 ms |Health Damage 3.990997 // deep wound DOT +1 ms |Health Damage 5.351349 // extra flame damage triggers on DOT tick +2992 ms |Health Damage 3.990997 +0 ms |Health Damage 5.351349 // +2462 ms |Health Damage 3.292877 // deep wound ? +526 ms |Health Damage 5.351349 +1458 ms |Health Damage 2.087082 // deep wound ? why no fire damage ? lion is still debuffed (combusting wounds) 5.351349 How to calculate ? 5 * (1 + 0.27) - 4 * 0.25 = 5.35 // plausible, will raise her might //Might bonus up to 63% 81.5 burn damage for 7.5 seconds (searing flame description) +0 ms |Health Damage 0.000000 // hit searing flames +29 ms |Health Damage 7.153152 // ring combusting wounds damage why !? +0 ms |Health Damage 47.918915 // 30 * ( 1 + 0.63) - 4 * 0.25 (searing flames DOT tick ) +3233 ms |Health Damage 7.153152 // combusting wound +1 ms |Health Damage 47.918915 +1513 ms |Health Damage 7.153152 // combusting wound +0 ms |Health Damage 47.918915 +1495 ms |Health Damage 7.153152 // combusting wound +1 ms |Health Damage 47.918915 +1504 ms |Health Damage 7.153152 // combusting wound +1 ms |Health Damage 47.918915 +1424 ms |Health Damage 6.816086 // why less ring damage (maybe because combusting wounds just expired) +1 ms |Health Damage 45.896530 5 * (1 + 0.63) - 4 * 0.25 = 7.15 OK // Added scion of flame - Monk was granted Scion of Flame +87697 ms |Health Damage 0.000000 // apply ring of combusting wounds (Monk) +4221 ms |Health Damage 0.000000 // miss searing flames +16073 ms |Health Damage 0.000000 // hit searing flames (Monk) +32 ms |Health Damage 7.153152 // no change ( we're the 20% extra burn damage ? ) +1 ms |Health Damage 47.918915 // no change (20% extra burn damage plz ) +1515 ms |Health Damage 7.153152 +0 ms |Health Damage 47.918915 +1516 ms |Health Damage 7.153152 +0 ms |Health Damage 47.918915 +1503 ms |Health Damage 7.153152 +1 ms |Health Damage 47.918915 +1503 ms |Health Damage 7.153152 +0 ms |Health Damage 47.918915 +1369 ms |Health Damage 6.521118 +0 ms |Health Damage 44.126709 30 * ( 1 + 0.63) - 4 * 0.25 = 47.9 .. 5 * (1 + 0.63) - 4 * 0.25 = 7.15 OK TLDR ABSOLUTELY NO CHANGE IN DOT DAMAGE/TICK (searing flames & combusting wounds) OBSERVED AFTER adding tln_scion_of_flame_talent.. The ring of combusting wounds was on the Scion of flame char ( priest named Monk) ..

If you didn't get your numbers by performing tests (ie you made them up), then how are you comfortable using them as examples of why Voltron's numbers are wrong? Saying that something is "false" is not the same as saying that it is imprecise - and trying to support either claim with fabricated numbers would be far worse than posting imprecise test results. I did not made them up lol.. Those are "test" numbers.. But I don't consider that a sufficient amount of data to draw my conclusions on the ability .. Just a quick glimpse before having to leave for work .. "Testing" the ability will require significantly more "test data" . https://www.youtube.com/watch?v=Yq59wttu2Po

Also, you only improved one of the three tests that Voltron did, which means that your test alone cannot replace his tests. I did not "test" Deep wounds.. Those were mere examples of what the proper test should/would look like.. When I do get to testing that particular ability I'll post the results and you'll able to compare conclusions .. I'm not in the habit of insulting random people on the internet .. I "attacked" the flawed methodology and the obviously flawed results , not the man .

as pointed before in another thread by dam .. game files say it modifies stats called Element 1 "Bonus Hit to Crit Percent Enemy Bellow 10 Percent" value 20 .. Element 2 "Bonus Hit to Crit Percent Enemy Bellow 10 Percent" value 0.5 .. So something(s ?) could probably happen when target is bellow 10% endurance .. Testing is possible - I could set an enemy at 11% HP and then at 9% , keep hitting it with a base damage 2 (low value to make sure it survives even a crit) weapon and autoheal him back to 11% / 9% after each hit and what happens .. But seems to much hustle at this point for what seems to be a crappy ability .. Unless somebody shows up claiming amazing returns from this ability I will not bother wasting the amount of time required to testing this properly in the near future ..

ERATA (further tests suggested by temek) revelead that 1.04 OSA damage does scale with the might modifier.. 20 * ( 1 + mightmod ) .. Which for the might modifier of 33% with add 26.6 damage ... at might bonus -15% 10 * ( 1 - 0.15) + 20 * (1 - 0.15 ) = 25.5 // hit 10 * ( 1 - 0.15 - 0.34 - 0.5 ) + 20 * ( 1 - 0.15 ) = 17.1 // carnage graze at might bonus 33% previous tests .. + 20 * (1 + 0.33 ).. at might bonus 90% 10 * ( 1 + 0.9) + 20 * ( 1 + 0.9) = 57 // hit 10 * ( 1 + 0.9 - 0.34) + 20 * (1 + 0.9) = 53.6 // carnage hit

Are you sure that wasn't modified by Might or possibly other damage modifiers as 26.6/20 = 1.33 and in your other tests you had exactly the same +0.33 dmg modifier from might so seems like a too good match to be a coincidence. I will test with different might (I did change the damage mod )..If it changes with might it will be an incredibly strange indirect "influence" since it didn't change with the damage mod which is modified by might stat .. EDIT Turns out you're right .. OSA adds 20 * ( 1 + mightmod ) at might bonus -15% 10 * ( 1 - 0.15) + 20 * (1 - 0.15 ) = 25.5 // hit 10 * ( 1 - 0.15 - 0.34 - 0.5 ) + 20 * ( 1 - 0.15 ) = 17.1 // carnage graze at might bonus 33% previous tests .. + 20 * (1 +0.33 ).. at might bonus 90% 10 * ( 1 + 0.9) + 20 * ( 1 + 0.9) = 57 // hit 10 * ( 1 + 0.9 - 0.34) + 20 * (1 + 0.9) = 53.6 // carnage hit Will edit the rest of my posts on the matter ( wish I didn't make so many ) .Thank you for making me test this !

Firstly - I'll be able to see/learn how the various DOT, HOT abilities work and interact with each other .. (most ability descriptions are off) .. The fact that 10-20 whatever how many old beta players "know" how various mechanics are supposed to work from various old (possible obsolete) DEV posts doesn't mean I know .. No better source to learning than watching directly under the hood ! (not even the source code) .. My experience with testing game mechanics is that there tend to be (sometimes bug) differences between what devs design and what programmers put in the game. And then there's bugs - (like DOTs stacking in ways they're not supposed to ..(incendiary ammo in ME3 MP) ..) . Abilities doing other things that they were supposed to (or nothing at all).. This type if testing is invaluable for cleanly exposing bugs .. If I knew from before what I was gonna find - I wouldn't need to go look for it in the first place no ? Concrete example - forums raving about OSA adding +20 damage per hit.. I can tell you with 100% certainty it adds exactly + 26.6 20 * (1 + mightmod) damage per hit .. Didn't even need breakpoints for this one .. This method should also provide very accurate attack speed measurements without the need for time consuming frame by frame analysis ..

with NO engagement stops unless engaged/threatened upon first .. I also would like a way to select which additional enemies my enemies threaten/engage on top of the unit they're currently attacking .. Hypothetical example with 3 engagement on tank surrounded by 1 bear, 2 wolves and 3 other animals.. Attack bear , threaten the 2 wolves, the other animals currently attacking my tank are free to move away without disengagement attack risk from my tank ..

Why would I want to disable such a core mechanic around which the entire combat system is built up // perks balanced // etc etc .. Why ?? You can't stand combat engagement you go play some other game .. (not my case ..) " ..Doctor, this patient suffers from a infection, how do we treat it ? .. Amputate its HEAD !! (disable engagement mechanic) "

marked prey adds 40% lash damage of the weapon's damage type that like all other lashes goes against 25% of the appropriate target's resist DR ( piercing damage will go against pierce 25% DR , crush vs crush .. There is a bug with dual type of damage weapons - lash will not automatically pick the best ) .. There are multiple numeric examples of Marked Prey's damage in this very thread .. CTRL + F "marked" on the first page ..

That is actually a useless/trap/suicidal option never to be picked (again).. It will make your guys ignore ALL/EVERY dis-engagement threats ..And since you can't really see who's actually engaged to what once combat starts ( they're fighting behind a three//under a rock // building // useless spell effect//in the shade ) there's ALWAYS SOMETHING ON TOP of the VITAL combat indicators (more "basic stuff" such as char positions and engagement indicators ) , each move order has the potential to get your guys killed .. /s Sure, you can't find your characters on screen, so you will usually have no idea what each of them is doing during the fight (who cares about this sort of stuff anyway ? ) - but look at that glowing/pulsating/ .. whatever gorgeous spell effect (our art department worked 6 months on the animation for this particular spell effect) . Isn't it a work of art ? / How does one turn off all/every spell effects ? I might be defective for wanting to be able to see the units I'm supposed to control during combat ? (Not that they obey my commands much when I'm actually able to see them anyway ..) I can always turn off my monitor if I ever feel like fighting with no visual feedback ..

show - combat grid - valid moving hexes - Homm style would be great

And matters gets more complicated when you get to consider +extra engagement limit bonuses.. How do you make use of the extra engagement "threat" without auto-engagement ? Do we need an extra button / Action in interface to make use of such items ? Say a particular char's engagement limit is raised to 3 . Separate button for attack (implicit engage) order/action and one to "engage/threaten" 2 other specific other targets of the 5 surrounding him/her .. How are those two targets selected with auto engagement on? Can I switch (engagement) targets without switching attack target ? ////////////////////////////// /////////////////////////////// And somewhat unrelated - what about the INVISIBLE OBSTACLES scattered on every map (and running numbers on the pathing system ) ? Where's my show combat grid//valid movement "hexes" command ? Could you show in a different color UN-WALKABLE tiles//hexes in combat ? How does one position/plan to maneuver his/her units when there's "invisible forcefields" all over the maps ?

Not enemy time related - have you ever played with 6 + 1 AC melee units ? Moving them all into proper position is currently a huge chore ..Perhaps a video is required to showcase grievance..

I'm whining about behavior that happens on all maps .. (not map specific)

Engaging after being engaged by AI is OK (healthy move.. ) . Problem is when AI is ALREADY engaged to capacity by other units.. Your dude will still stick like a fly to **** instead of moving to flank as ordered WITH ZERO risk of dis-engagement attack . .why !!?!? ? He did not engage me yet (how could he - he's already engaging 1/2/../5/.. however many his engagement limit is ) - MY UNIT DOES NOT NEED TO STOP.. And still it stops and engages AGAINST orders.. Takes a second order to cause him to move into position (and he will move without dis-engagement attack since the target was already engaged) .. /s Question - if I shoot/hang 2 of my units for insubordination - will the remaining units obey my orders then ?? Why isn't this listed under game mechanics in the manual ? s/ What does one have to do to make his units obey simple move commands ?!? . (I thought gaming industry had that already figured in 1992 when they made Dune II . Would viewing some short videos of ordering a few units to move in Dune 2 might help devs familiarize themselves with the basing gaming concept of "moving units" ) ..

Yes (don't ask me for the mathematical proof ..) ..But I fail to see the relevance to this particular thread .. @OP ".. propagation of uncertainty (or propagation of error) is the effect of variables' uncertainties (or errors) on the uncertainty of a function based on them. When the variables are the values of experimental measurements they have uncertainties due to measurement limitations (e.g., instrument precision/ ROUNDING) which propagate to the combination of variables in the function." .. When you calculate derived statistics from that measured value ( ex DPS) the effect of your measuring errors (perhaps initially small) also multiplies ..

Problem becomes quickly obvious with full melee party . Your units ordered to move behind enemy targets (for flanking)- MOVE STRAIGHT AT THEM and then promptly get stuck into a most often non-flanking position and engage the target (FAIL !).. Is there something FORCING your moving units to auto-engage whatever stumbles into their path while on the move? HOW DO WE DISABLE IT ? DON'T stop and engage random crap along the way unless engaged first or specifically ordered to do so !! Otherwise ignore distractions (try to go around or signal no path found) and move into the flanking position I've ordered you to move to ... THIS IS PROBABLY THE SETTING THAT MAKES THE BRAIN DEAD AI UNITS GLUE TO FIRST UNIT IN MELEE RANGE, but for God-sake don't apply it to my units also ... You can micro around it by always moving your characters in straight lines making sure no hostile enemy unit crosses the planned movement path (because otherwise get stuck in melee in wrong position) .. And only order to attack when "behind target" .. But with 7 melee units doing this every combat gets incredibly frustrating and needlessly time consuming ..

I'm not being offensive - just precise .. Good intentions aren't enough to get accurate test results .. Proper testing methodology is mandatory .. And yes "mathematical rules" say "4 is equal with 3.81" is false.. Logical opposite of bolded "precisely 4,0000000 damage" ..

1. False .. (see previous posts) 2. Every 3 seconds (see above posts) And I haven't tested if it doesn't scale in some way yet .. (char level perhaps ?)

Doubt there's any .. You're not gonna dump DEX on your damage dealer right ? Even with -50% from armor and -20% from modal , pretty sure not even 2h weapons would swing slower than 3 seconds (will do some attack speed tests at some point) (unless you miss/get disabled/interrupted a lot) ..

small price to pay IMHO for all the free attacks .. But description should change to clarify if this is intended behavior ..

not broken - just hard (bordering on frustrating) to use .. I've been using nothing but full melee parties during my PoE play time ..+Ranger Pet .. So it can work..There is also massive room for improvement .. Characters not going behind the target (AS ORDERED) already engaged by something else moving straight at target and engaging in non-flanking position - WHY U NO OBEY MY ORDER GAME !@#!@# !!! 3 workarounds .. Move your units in straight lines that don't cross any hostile target's position (or they'll stick in non flanking position) .. Last move attack target when your unit is already "behind the target" .. It will still get stuck to target - but this time in the right flanking position ... OR - if target was already at engagement limit , once your unit has stuck to target in bad non-flanking spot, is engaging BUT is not engaged back , order again to move "behind target" ..It will move there without triggering dis-engagement and get into desired flanking position .. OR - disable target (prone, paralyze, etc) like loading screen tips suggest to allow your units to move into flanking positions without eating dis-engagement attacks .. Having to do this for with 7 units fight after fight - gets old fast ..

Results are normal but - explaining the cause is a extremely complicated ...(deep into how windows operating system works ..) Chief among things is that at any given time pillars of eternity isn't the only process running .. "In the case of a computer with a single CPU, only one task is said to be running at any point in time, meaning that the CPU is actively executing instructions for that task. Multitasking solves the problem by scheduling which task may be the one running at any given time, and when another waiting task gets a turn. The act of reassigning a CPU from one task to another one is called a context switch. When context switches occur frequently enough, the illusion of parallelism is achieved." My script running each time the target's health changes (takes damage) - takes some processor cycles to do its thing ( .., make windows API function to get current time, display time difference , damage received, healing target back to full to be ready for next beak ..) ..My game doesn't run at 30 fps .. I have no idea what mechanisms the game uses to synchronize his timings with "real time" (I don't make 3d games) .. Given all that, I don't think a few milliseconds are "huge variance" .. ////////// Tl_DR - attach 3rd party debbuger to pillar or eternity process.. Set hardware write breakpoint on the memory address holding the float variable representing test-subject's health/hit points.. Execute a script each time the breakpoint is hit (test subject health changes - gets damaged or healed) (automatically) .. Type to log window damage amount and time difference from last breakpoint activation.. This way you can "catch" effects of multiple DOTs/HOTs running in parallel, damage from each individual pellet of multishot weapons(blunderbuss), etc .. Each and every DOT tick (timing and precise damage amount)! Check the separate breakpoint testing thread for examples .. http://forums.obsidian.net/topic/78410-proper-debugger-breakpoints-dot-hot-tests/

b. in both cases actually .. All slashes go vs 25% of the appropriate target's DR.. Even FoD ..