Assassin Skill Comparison - DPS

peckked
peckked Posts: 224 Arc User
edited June 2014 in Assassin
*****Updated******

I wanted a way to quickly compare Assassin skills. Here's what has come together with some suggestions from Olbaze and Selbronne :D

Link here.

Instructions:

Choose appropriate sheet (Sage/Demon) at the bottom left.

Fill in the green cells. Minimum/Maximum Physical Attack, Dexterity, Attack Level, Spirit, and APS can all be found on the Character Info screen. Mastery and passive skill damage bonus can be found on the Skills screen. You can manually enter base weapon damage in lieu of min/max physical attack. If a value other than 0 is entered in the "PVP - Spirit(Target)," the cell will turn red and the results will include the 75% PVP damage reduction.

Reading the results:

Results are split into 4 columns

Skill Damage: This is the amount of damage that a skill does before pvp/pve. This is traditionally referred to as DPH (Damage per Hit), but since some skills are multi-hit now I went with Skill Damage instead.

Multiplier: This is skill damage represented as a multiplier of base weapon damage.

DPS: Damage per second. Factors the channel and cast times of each skill to determine the amount of damage done per second.

MPS: Multiplier per second. This is damage per second represented as a multiplier of base damage. I use this column to quickly see which skill has greater DPS. This is mostly for convenience as it's easier to compare smaller numbers (example: 15.74, 15.16, 15.07 is easier to quickly see which is greater than when ranking 55898,53855,53530)

Considerations:

- It's limited to comparing skills against other skills in terms of normal skill damage. It doesn't touch on costs like chi, mp, or cooldowns, or benefits such as buffs (ie invigorate), debuffs(ie subsea).

- I didn't include damage from DoTs as these do not scale evenly with Attack Levels, etc. Puncture Wound and Toxic Torrent were intentionally omitted (FYI Both are among the worst in terms of DPS).

- Crit%, SS/GoF are not taken into consideration

- New multi-hit Primal skills (Elimination, Life Hunter, and Cursed Jail) have wildly varying damage once you take into consideration crit% and GoF. I would love to hear ideas as to how I might address this discrepancy.
Post edited by Unknown User on

Comments

  • Olbaze - Sanctuary
    Olbaze - Sanctuary Posts: 4,242 Arc User
    edited June 2014
    We already have a ton of these spreadsheets. I remember at least two in the past, one of them being Nowitsawn's PW Damage Calc V2.0. There's really no need for a third one.

    There's quite a few errors in your spreadsheet:
    1. Your Weapon Attack value does not take into account Character Level.
    2. Your unsparked "Multi" values are wrong, as you're doing (damage/base*multiplier) rather than (damage/base). Your sparked Multi column does not have the same mistake.
    3. Your "Multi" values compare the total damage to Base Weapon Damage- While this is a valid comparison, I don't think it's meaningful, as your base damage in combat is your Base Physical Damage. Rather compare those two.
    4. Because you use the Multi column to calculate the Multi/s and DPS figures, 2. means that those are also off on the unsparked variant.
    5. Due to the way Attack Levels and Defense Levels interact, your spreadsheet will give out wrong numbers if Target's Defense Level > Attack Levels.

    I also don't see any reason why the Multi/S column is before the DPS column, since the latter refers to the former.

    I would suggest reading this paper I wrote on damage calculation.
    I am Olba. Not Ol, not Baze nor Blaze. And even less would I go by Olblaze. Please, take a second to read a person's username.
    If you see b:cute be sure to take a second, calm look at anything I said.
  • peckked
    peckked Posts: 224 Arc User
    edited June 2014
    Thanks for your reply.

    My purpose for this wasn't to calculate damage specifically. Rather, my purpose was to compare skills against one another. I was unaware such a spreadsheet had been made. The ones I've found seem to focus on auto-attacks (typically comparing weapons), and those that did allow you to look at skill damage weren't designed to handle multiple skills at once. The one you linked would require I enter in each skill and record the result separately, which I certainly wouldn't want to do every time I consider an upgrade, etc.

    In regards to your points:

    1) I completely overlooked this and will fix it when I get the chance, unless of course a better option for my purpose already exists in which case I won't bother lol edit: After a quick review your paper, I think my best course of action might be to omit any modifiers that involve the target as they would scale evenly across all skills. This would negate the need for character levels as well as attack/defense levels.

    2) I'm gonna blame sloppy cutting and pasting from the original excel file I had made. May have even been from a draft. At this point I'm not sure. However, the result is correct (barring point 1), though the method appears, and indeed is, convoluted.

    The formula I have in the unsparked multi column is

    Unsparked Damage/Base Physical Attack*Base Multiplier

    which, since I didn't use any brackets is the same as

    Unsparked Damage * (Base Multiplier/Base Physical Attack) <-- this would have look nicer though still convoluted

    The portion in brackets is the multiplicative inverse of Base Weapon Damage so that gives us

    (Base Multiplier/Base Physical Attack) = 1/Base Weapon Damage

    and therefore

    Unsparked Damage/Base Physical Attack*Base Multiplier = Unsparked Damage/Base Weapon Damage

    3) The "Multi" column was more of a working column than a useful one. The "Multi/s" column was my real goal as it allows me to quickly see which has greater damage/s, taking into account base damage, weapon damage, and additional damage, without getting tangled up in overly large numbers. I could have done this many ways, this way seemed effective enough.

    4) See point 2

    5) When making the spreadsheet (originally just for me), this had crossed my mind. However, I decided to take the lazy route as it's a rare case that a target's def level exceeds my atk levels. Perhaps I should have addressed this before posting this here. Actually, now that I think about it, the application of atk/def levels isn't really necessary at all considering my original intended purpose.

    As far as the order, I wanted Multi/s as the last column. In the original this wasn't the case. You don't wanna see the original. It turns out that it has less mistakes, but the data is all over the place hehe. See point 3. As it is I have Damage in the first 2 columns then Damage/s in the latter 2, for both unsparked then 3-sparked results.

    I do really appreciate you taking the time to look over this. If indeed there is a tool out there making my spreadsheet redundant with regards to my purpose please let me know. In the meantime, I've left the spreadsheet as is in case you wanted to verify point 2.
  • Olbaze - Sanctuary
    Olbaze - Sanctuary Posts: 4,242 Arc User
    edited June 2014
    peckked wrote: »
    1) I completely overlooked this and will fix it when I get the chance, unless of course a better option for my purpose already exists in which case I won't bother lol edit: After a quick review your paper, I think my best course of action might be to omit any modifiers that involve the target as they would scale evenly across all skills. This would negate the need for character levels as well as attack/defense levels.

    2) I'm gonna blame sloppy cutting and pasting from the original excel file I had made. May have even been from a draft. At this point I'm not sure. However, the result is correct (barring point 1), though the method appears, and indeed is, convoluted.

    The formula I have in the unsparked multi column is

    Unsparked Damage/Base Physical Attack*Base Multiplier

    which, since I didn't use any brackets is the same as

    Unsparked Damage * (Base Multiplier/Base Physical Attack) <-- this would have look nicer though still convoluted

    The portion in brackets is the multiplicative inverse of Base Weapon Damage so that gives us

    (Base Multiplier/Base Physical Attack) = 1/Base Weapon Damage

    and therefore

    Unsparked Damage/Base Physical Attack*Base Multiplier = Unsparked Damage/Base Weapon Damage

    3) The "Multi" column was more of a working column than a useful one. The "Multi/s" column was my real goal as it allows me to quickly see which has greater damage/s, taking into account base damage, weapon damage, and additional damage, without getting tangled up in overly large numbers. I could have done this many ways, this way seemed effective enough.

    5) When making the spreadsheet (originally just for me), this had crossed my mind. However, I decided to take the lazy route as it's a rare case that a target's def level exceeds my atk levels. Perhaps I should have addressed this before posting this here. Actually, now that I think about it, the application of atk/def levels isn't really necessary at all considering my original intended purpose.

    1) You can drop the Attack Levels and Defense Levels if all you care about is comparing damage between varying skills. However, Character Level functions as an integer addition to your Base Weapon Damage, meaning that leaving it out results in inaccurate results, especially when comparing two skills that have different Weapon Damage% additions.

    2) But what are you trying to use that number for? Your autoattack uses Base Physical Attack, so surely comparing to that would be more meaningful than comparing to Base Weapon Damage, which is only used as a seed value.

    3) While it is true that it can be used to compare the skills to each other, I would still argue that comparing to Base Physical Attack, aka a normal attack, would hold more meaning. After all, your dps with autoattacks is (Attack Rate * Base Physical Damage), thus making DPS comparisons much easier.

    It's additional functionality at minimal cost.

    5) See 1)
    I am Olba. Not Ol, not Baze nor Blaze. And even less would I go by Olblaze. Please, take a second to read a person's username.
    If you see b:cute be sure to take a second, calm look at anything I said.
  • peckked
    peckked Posts: 224 Arc User
    edited June 2014
    I think I'll just get rid of the attack/def levels for now.

    1) Looking at this now, since weapon damage is incorporated into weapon damage, changing the base weapon damage accordingly will return accurate results. Personally, instead of spending the time to find and add up base damage from weapon, shards, gear, engravings, titles, meridian, nuema, and war avatars, I figure out my base multiplier and determine my base weapon damage using the difference in average phys attack buffed vs unbuffed. Adding a separate variable for character level would actually complicate this process. If I were to take that route, why stop at character level? I could put in fields for all of the above contributors, but it wouldn't be any more effective than just adjusting the base weapon damage to compensate.

    As per your paper, 1 character level = 1 base weapon damage, right? Then I haven't made a mistake here either. It's up to the user to determine their base weapon damage, which, as your own paper states, includes character level.

    2) Since all I'm looking to do is compare different skills, I could use several different per-second values.

    I considered using your approach as it would give comparable APS values. In fact, my original spreadsheet had such a column (and was actually one my original considerations, since I'm considering possibly retiring my APS set). However, comparing to APS becomes particularly misleading when we take into consideration the use of Chill of the Deep. Skills retain the same ratios with or without chill. APS does not. Also consider the inconsistency should a person swap out even just one piece of gear such as their cape. These adjustments are more likely to be determined by circumstance and less-so by trying to say optimize damage output over a 5s stun or a 12s spark or an 8s amp. Consequently, I fail to see the value in such a comparison, let alone use APS as a basis for comparison. Nonetheless, I've added auto attacks along with a disclaimer "for additional functionality."

    APS values are fairly well established, and it's too easy to fall into the habit of rounding to common APS values. The values returned are also relatively small. Differentiating say 1k additional damage on one skill vs another appears too minute to matter, yet it's those minute differences which add up and is what makes this exercise worth doing in my eyes.

    I could have simply left it as damage values but as they reach 6 figures it becomes unnecessarily tedious to read.

    I found by comparing against base weapon damage solved all of these problems and serves my primary purpose. The returned values for Multi compares different values on an even metric, is not easily confused with another, and isn't difficult to read while retaining a reasonable degree of fidelity.

    Using MPS(multi per second), you might say, is just as valid as DPS. Both are merely academic as practical application has variables that are unrealistic to account for in a simple A skill vs B skill scenario. In the end, both are just numbers that identify potential and can be used for comparison.


    So.... ummm... just for clarification... were there actually any real errors in my sheet other than when a target's def level is greater than the attacker's atk levels? I want to make sure to fix it if necessary before I attempt applying this format to another class ... with the obvious necessary adjustments, of course.
  • Olbaze - Sanctuary
    Olbaze - Sanctuary Posts: 4,242 Arc User
    edited June 2014
    peckked wrote: »
    1) Looking at this now, since weapon damage is incorporated into weapon damage, changing the base weapon damage accordingly will return accurate results. Personally, instead of spending the time to find and add up base damage from weapon, shards, gear, engravings, titles, meridian, nuema, and war avatars, I figure out my base multiplier and determine my base weapon damage using the difference in average phys attack buffed vs unbuffed. Adding a separate variable for character level would actually complicate this process. If I were to take that route, why stop at character level? I could put in fields for all of the above contributors, but it wouldn't be any more effective than just adjusting the base weapon damage to compensate.

    As per your paper, 1 character level = 1 base weapon damage, right? Then I haven't made a mistake here either. It's up to the user to determine their base weapon damage, which, as your own paper states, includes character level.

    Well, I was simply suggesting it because unlike shards or +patk on gear such as rings, the fact that Character Level influences your Weapon Damage isn't obvious. It's something you would only know if you know the formulae.

    And of course, you can figure out your Base Weapon Damage as Base Physical Damage/Multiplier, since your Base Physical Damage is the damage displayed in your Character's stats.
    2) Since all I'm looking to do is compare different skills, I could use several different per-second values.

    I considered using your approach as it would give comparable APS values. In fact, my original spreadsheet had such a column (and was actually one my original considerations, since I'm considering possibly retiring my APS set). However, comparing to APS becomes particularly misleading when we take into consideration the use of Chill of the Deep. Skills retain the same ratios with or without chill. APS does not. Also consider the inconsistency should a person swap out even just one piece of gear such as their cape. These adjustments are more likely to be determined by circumstance and less-so by trying to say optimize damage output over a 5s stun or a 12s spark or an 8s amp. Consequently, I fail to see the value in such a comparison, let alone use APS as a basis for comparison. Nonetheless, I've added auto attacks along with a disclaimer "for additional functionality."

    APS values are fairly well established, and it's too easy to fall into the habit of rounding to common APS values. The values returned are also relatively small. Differentiating say 1k additional damage on one skill vs another appears too minute to matter, yet it's those minute differences which add up and is what makes this exercise worth doing in my eyes.

    I could have simply left it as damage values but as they reach 6 figures it becomes unnecessarily tedious to read.

    I found by comparing against base weapon damage solved all of these problems and serves my primary purpose. The returned values for Multi compares different values on an even metric, is not easily confused with another, and isn't difficult to read while retaining a reasonable degree of fidelity.

    Using MPS(multi per second), you might say, is just as valid as DPS. Both are merely academic as practical application has variables that are unrealistic to account for in a simple A skill vs B skill scenario. In the end, both are just numbers that identify potential and can be used for comparison.

    Oh, it was just an opinion. To me, knowing how the damage of a skill compares to my autoattack would be more interesting than a easier-to-read comparison between all of the skills. Just a matter of preference, really.
    So.... ummm... just for clarification... were there actually any real errors in my sheet other than when a target's def level is greater than the attacker's atk levels? I want to make sure to fix it if necessary before I attempt applying this format to another class ... with the obvious necessary adjustments, of course.

    Well, since you explained your reasoning for choosing to use Base Weapon Damage rather than Base Physical Damage in the multipliers, I don't count that as a mistake. If you trust that your user will figure their Base Weapon Damage accurately, there is no need to make an addition for Character Level. And Attack and Defense Levels would only scale the numbers.

    So in summary: The numbers are looking accurate to me. I do think you might want to add some explanation, particularly the details (formulae and reasoning) behind the multi colums, to the original post.

    If it were me, the next step would be making a Demon version of it as well. Easiest way to do that would be to copypaste the current one into a separate sheet inside the same document and change all of the relevant multiplier to their Demon variants. Of course, it wouldn't allow comparing skills between cultivations due to the subtle differences between the cultivations (Dagger Mastery, CotD, Wolf Emblem). But that comes with certain issues, such as how to interpret crits and Wolf Emblem, so I wouldn't worry about that yet. But I would like to see you go there eventually, it would help settle the age-old argument about which cultivation deals more DPH.
    I am Olba. Not Ol, not Baze nor Blaze. And even less would I go by Olblaze. Please, take a second to read a person's username.
    If you see b:cute be sure to take a second, calm look at anything I said.
  • peckked
    peckked Posts: 224 Arc User
    edited June 2014
    Thanks for the suggestions. When I find some down time at work I'll update the first post with a better explanation and tweak my spreadsheet. I'm thinking I'll include Attack Levels again with appropriate conditions/formulae should the targets Defense Levels exceed a players Attack Levels.

    Going back to autoattacks, I'm still at a loss as to how to make a valid comparison without creating a redundant sheet. I agree it would be useful and I would really like to find a succinct method of comparison. It's a question of what might be the best way to go about it. If one doesn't take into consideration changing gear or employing chill, the comparison is easy. As is, my spreadsheet can already do this but the results are very predictable. In APS gear, APS will out-DPS skill spamming. In full r93 or nv3 skill spamming will out-DPS APS. I spent some time a couple months ago playing around with partial sets. While I don't have any numbers from it, the results were basically, you could find a middle ground where APS roughly had the same DPS as skill spamming, but DPS was significantly lower than going with either full APS or full DPH for their respective strengths. Going from APS to partial was sacrificing a fair chunk of DD for a little bit more defense. Going from DPH to partial just made skill spamming pointless, in which case it's better to just go full APS. Swapping 6-7 pieces of gear just gets tedious.

    BTW do you happen to know how the anti APS buff works? Is it even consistent?
  • WannaBM - Archosaur
    WannaBM - Archosaur Posts: 1,984 Arc User
    edited June 2014
    peckked wrote: »
    BTW do you happen to know how the anti APS buff works? Is it even consistent?

    It is a buff that reduces mellee auto attack dmg. I think by a factor 4 but i am not sure. Also not sure if this factor may be different from 1 boss to the other.

    It cannot be purged and bosses that are already there when you meet them, you cant purge it. On the bat boss in FSP you can interupt it casting the buff if you stand and wait where he spawns.
    Everything i write is from PvE perspective unless mentioned otherwise.
  • peckked
    peckked Posts: 224 Arc User
    edited June 2014
    Hehe thanks WannaBM... but I'm afraid you didn't answer my question. You've done a really good job at answering "What is it?" but my question was "How does it work?"
  • Selbronne - Heavens Tear
    Selbronne - Heavens Tear Posts: 79 Arc User
    edited June 2014
    peckked wrote: »
    Hehe thanks WannaBM... but I'm afraid you didn't answer my question. You've done a really good job at answering "What is it?" but my question was "How does it work?"

    The next attack that hits you will trigger an effect
    that reduces all regular attack damage by 75% for a brief period.

    http://www.pwdatabase.com/pwi/items/34647

    That's all the info i found about it sadly. It dosen't have explanation on pwi-wiki debuff guide b:angry

    btw. gj with that Comparison, i like it. Sucks that atack/def lvls and spirit stat dosen't work there. All i can think of about helping u apply apply att/def & spirit lvl to it is giving u this link cause they updated spirit on wiki:

    http://pwi-wiki.perfectworld.com/index.php/Damage#Spirit

    maybe u will figure sth out with this.

    Edit: tyvm for changing it :D
  • peckked
    peckked Posts: 224 Arc User
    edited June 2014
    Updated the spreadsheet and the first post.

    Spreadsheet now includes:

    Sage and Demon
    Attack Levels and Defense Levels (uses appropriate formula if def level > atk level)
    Spirit (both PVE and PVP application)
    PVP 75% damage reduction (if Spirit(Target) >0)
    Automatic calculation of Base Weapon Damage with Min/Max Physical Attack Values

    First post now has instructions and a basic explanation of different output values.