d2jsp
d2jsp Forums > d2jsp > Ladder Slasher > Critical Strike Mechanics And Data Collection
1234Next
Add Reply New Topic New Poll
Bigheaded
Group: Member
Posts: 10,216
Joined: Sep 17 2005
Gold: 27,578.68
Jun 2 2022 09:13am
Hey,

I've decided it's time to add critical strike to my damage calc, the issue is we only vaguely know how it works. It can be found here:
So it's time for more data collection, arguably all previous data is void due to being on the wrong patch, but depending on the amount of information i get from this topic, i may still need to use it, as a pretty significant amount of data exists for old patches and the mechanics are unlikely to have changed


If anyone wishes to help with data collection, please can we use a google sheet similar to this:
https://docs.google.com/spreadsheets/d/1DDVq5aE4510_jHwJCZ_IcNyz7GfxU-53McfWQiLsm5M/edit?usp=sharing

For those of you with a google account, just go to file -> make a copy and delete my data but keep the header line and rename the tab by double clicking on it.
For those of you without a google account, shoot me a pm and i'll send you a clean copy


I suggest having your own copy for 2 reasons:
1. You know that no-one else is playing with your data
2. I know whos data is whos when it's shared to me



My current guesswork of how crit works:
I am unconvinced your level affects crit, however it might do
Class is unlikely to affect crit
Weapon do affect it, but more to do with Weapon Ratio's than specific weapons actually just critting more often. (I would "expect" with 75 strength and 75 dexterity that all weapons would have the same crit), this is why axes appear to crit more as the ratio is not favourable.
More stats = more crit.
Adding to the "wrong stat" of a weapon increases crit more than using the stat which increases damage more regularly.
the item stat +x Critical Strike gives "roughly" a 1% chance to crit for each + critical strike. More data needed but my own testing and jedimagis data (see below) suggests this

A lot of that can be half verified by this topic: https://forums.d2jsp.org/topic.php?t=80551791&f=272
There's a lot more data here if you're interested in giving it a read and coming to your own conclusions: https://forums.d2jsp.org/topic.php?t=80541143&f=74

However i'm looking to get something which i can insert into the damage calculator with reasonable accuracy and currently I haven't got the data i need to do this.



My recommended method for those collecting is to use the in game stats to keep record of the number of hits:

You then literally count each time you get a red number for your damage (this is the indicator your hit was a crit), insert your numbers into the sheet whenever you fancy, i recommend every 100 hits, although if you're struggling to remember how many crits you did then try every 50 or so.

For those who are using healing, remember to use the dmg number which is the left side number in the brackets on the number of actions line.

There are 2 mobs which cause issues:


Ideally use powerstrike and kill before either of them get their ability off, however this often isn't how this works out.

If they do get their ability off, there's 2 ways of dealing with it:
1. Go back to town and just record your current data, i do not specifically need groups of 100 attacks, but 100 is a good number to not be writing it down too often but regularly enough that you can keep track. Lets suggest you had 70 attacks then fought a goocidic who you hit for 1x 0 damage attack. this should just be recorded as 70 attacks on the spreadsheet. Restart your count from after those mobs die.
2. Count all 0 damage attack hits you deal (i am reasonably sure you cannot crit the azapnid illusions or from melee'ing goocidic whilst he has his ability active) and remove that from your total when you decide to write it down, this is a lot more complicated

I dont believe there are any other mobs who cause issues, as you can still crit the few mobs who can survive on 1 hp no matter what damage you deal to them.

At some point i will also need data on caster critical hits, if you want to collect this data and provide it in advance, i can take this data at this point, shoot me a pm if you're interested in doing this, as obviously we are tracking intelligence and crit instead of strength/dex/crit.


Thoughts/comments/helpers welcome

This post was edited by Bigheaded on Jun 2 2022 09:40am
izParagonzi
Group: Member
Posts: 35,719
Joined: Apr 29 2006
Gold: 823.45
Jun 2 2022 10:16am
Azapnid Mirage monsters CAN be CRIT HIT for 0
Soften the Sneaky shell CAN be CRIT HIT for 0

and yes this is within the last 2 weeks of 2022 Ladder Slasher.
______________


Okay... 100 actions with base stat points (don't add any melee stat points to strength/damage) will only have a Critical Hit chance based on THAT characters starting stats... so you have to identify the "Ratio".

When you are wanting to find out about the "Critical Hit Chance" WHILE using +X Critical Strike items during 100 actions, YOU have to ensure you are comparing it with the SAME base stats with your previous 100 actions etc.

Try NOT using +X% Enhanced Effect weapons so that you can RECORD your 100 actions quicker, EVEN though it may take more time to kill of a monster or wave.

The BEST testing monster IS Sneaky... the so called monster defense is null and void because you are testing the amount of "Critical Hit Chance"... NOT the damage.

Previous to the Updates and testing I and others HAVE done... ARE compatible to this testing stage you are requesting, however, you haven't bothered to disprove previous testing because you think the updates made a change... you are 50% correct on that fact.

I am so glad I actually kept ONLY +X Critical Strike equipment which will be tested with BASE starting stats (best is Guardian 30/10 ratio, but this changes based on weapon type) to be more effective for testing.. hey Guardian can use EVERY Heavy weapon which means EVERY weapon with a base starting stat of 30/10 (75/25 Ratio) right? easily changed with Stat gear right etc).

Up to you bud... but for myself personally... Guardian and Heavy Weapons and low base stats will give you more comprehensive data.
Bigheaded
Group: Member
Posts: 10,216
Joined: Sep 17 2005
Gold: 27,578.68
Jun 2 2022 10:30am
Quote (izParagonzi @ Jun 2 2022 05:16pm)
Azapnid Mirage monsters CAN be CRIT HIT for 0
Soften the Sneaky shell CAN be CRIT HIT for 0

and yes this is within the last 2 weeks of 2022 Ladder Slasher.


Will investigate, makes things easier if this happens. Reasonably sure i tested over 50 hits on each, i know you can 0 crit sneaky but he's using different mechanics.
______________


Quote
Okay... 100 actions with base stat points (don't add any melee stat points to strength/damage) will only have a Critical Hit chance based on THAT characters starting stats... so you have to identify the "Ratio".

When you are wanting to find out about the "Critical Hit Chance" WHILE using +X Critical Strike items during 100 actions, YOU have to ensure you are comparing it with the SAME base stats with your previous 100 actions etc.

Try NOT using +X% Enhanced Effect weapons so that you can RECORD your 100 actions quicker, EVEN though it may take more time to kill of a monster or wave.

The BEST testing monster IS Sneaky... the so called monster defense is null and void because you are testing the amount of "Critical Hit Chance"... NOT the damage.


I am hoping to get over 100,000 attacks worth of data at various values. I cannot get all the data i need from level 1 characters, secondly as i stated, i do not believe level makes a difference, however i have evidence which shows it does as well as evidence shows that it doesn't. So in short i do not have enough data.
Also the entire point here is not to be 100% collecting data, you should still be climbing and having fun, this data collection is more of a on the side "might as well do it as if that Bigheaded guy can use the data to figure out how it works, then that data would be useful to me", rather than i'll spend 100% of my time collecting data and not getting anywhere.

I'm not 100% sure i wish to test on sneaky as he is not a standard mob and i'm unsure that he doesnt have some built in modifiers which increase/decrease the amount of crits he receives.

Quote
Previous to the Updates and testing I and others HAVE done... ARE compatible to this testing stage you are requesting, however, you haven't bothered to disprove previous testing because you think the updates made a change... you are 50% correct on that fact.

I am so glad I actually kept ONLY +X Critical Strike equipment which will be tested with BASE starting stats (best is Guardian 30/10 ratio, but this changes based on weapon type) to be more effective for testing.. hey Guardian can use EVERY Heavy weapon which means EVERY weapon with a base starting stat of 30/10 (75/25 Ratio) right? easily changed with Stat gear right etc).

Up to you bud... but for myself personally... Guardian and Heavy Weapons and low base stats will give you more comprehensive data.


As i said, i may use the old data, we both agree it "should" be the same.

I cannot get enough data simply from level 1 guardians with critical strike gear. It may be the best way of testing what the item mod critical strike does, for you are correct he has terrible base strength/dex so the effect of those stats should be minimal.
However as i stated there's evidence out there that +1 crit is close to +1% chance to crit.

The main point though is i need the baseline of how strength/dex impacts crit. Until i have that formula there's no real point in getting any other data.
izParagonzi
Group: Member
Posts: 35,719
Joined: Apr 29 2006
Gold: 823.45
Jun 2 2022 11:39am
Quote (Bigheaded @ Jun 3 2022 04:30am)
deleted:

As i said, i may use the old data, we both agree it "should" be the same.

I cannot get enough data simply from level 1 guardians with critical strike gear. It may be the best way of testing what the item mod critical strike does, for you are correct he has terrible base strength/dex so the effect of those stats should be minimal.
However as i stated there's evidence out there that +1 crit is close to +1% chance to crit.

The main point though is i need the baseline of how strength/dex impacts crit. Until i have that formula there's no real point in getting any other data.


If you look at my old testing, you will notice that I was NOT testing Critical Hit Chance (based solely on Ratios per weapon type) AT level 1... I did my testing at HIGHER levels.

Not sure if this is new to you or others... the Monster Difficulty while normally leveling up DOES NOT affect a Critical Hit chance... a ChC does NOT alter based on the monster level or your character level, this is the foremost thing you need to get out of your test... YES at level 1 to 15... you can still 1 hit kill using a Tier III item and your characters Base stat(s), HOWEVER, the ChC you are testing doesn't change outside of weapon "ratios" (this is where stat values CAN make a difference and should be recorded.)

Let me put it this way ... if you are a Guardian, and you have access to every Heavy weapon when encountering Sneaky @ level 30-55+ ... you will obtain all the data you require for ChC and Critical Strike items (if you have those on hand as well).

But, like you, I would want to test this out for myself instead of taking someone elses word... so... get hold of all Heavy weapons (without EE) and armor / main charm with + CS and hunt for Sneaky solo'ing above level 30... damn, last night I got Sneaky @ level 65 solo'ing ... presently I haven't even found Sneaky @ level 35+ ... If you want... I can easily record information like I did in previous tests. Let me know... Note: I like testing when I am drunk because I am numb to the boredom :P
Bigheaded
Group: Member
Posts: 10,216
Joined: Sep 17 2005
Gold: 27,578.68
Jun 2 2022 11:56am
Quote (izParagonzi @ Jun 2 2022 06:39pm)
If you look at my old testing, you will notice that I was NOT testing Critical Hit Chance (based solely on Ratios per weapon type) AT level 1... I did my testing at HIGHER levels.

Not sure if this is new to you or others... the Monster Difficulty while normally leveling up DOES NOT affect a Critical Hit chance... a ChC does NOT alter based on the monster level or your character level, this is the foremost thing you need to get out of your test... YES at level 1 to 15... you can still 1 hit kill using a Tier III item and your characters Base stat(s), HOWEVER, the ChC you are testing doesn't change outside of weapon "ratios" (this is where stat values CAN make a difference and should be recorded.)

Let me put it this way ... if you are a Guardian, and you have access to every Heavy weapon when encountering Sneaky @ level 30-55+ ... you will obtain all the data you require for ChC and Critical Strike items (if you have those on hand as well).

But, like you, I would want to test this out for myself instead of taking someone elses word... so... get hold of all Heavy weapons (without EE) and armor / main charm with + CS and hunt for Sneaky solo'ing above level 30... damn, last night I got Sneaky @ level 65 solo'ing ... presently I haven't even found Sneaky @ level 35+ ... If you want... I can easily record information like I did in previous tests. Let me know... Note: I like testing when I am drunk because I am numb to the boredom :P


Mis-understood what you meant with guardian, i just presumed level 1 as fighting level 30+ mobs with base strength/dex guardian stats sounds like torture to me, so presumed you'd have statted strength or dex by that point which would have ruined the ideal minimal stats which we specifically wanted a guardian for.

As i mentioned i am unconvinced sneaky has other side mods as he is a boss mob.
I do not believe you have enough data to confirm whether or not your level/mob level has an impact or not.
Lastly i'd rather you didn't attempt to do counting when you're drunk. My counting isn't the best at the best of times but i can assure you whilst drunk i'm lucky to count to 10 once, let alone on many different occasions. Or in the case of sneaky 50+ or however much data you decide to collect. So i'd potentially not use data from drunk rex.

I ideally need over 10,000 attacks per weapon to get reasonable data to form a hypothesis and to then apply that hypothesis to other weapons. Should probably do the test with daggers seeing they are the most commonly used weapon and even if the hypothesis is wrong for other weapons, if it works for daggers i'd have something somewhat useful.
Currently doing polearms seeing it's pretty fun to use them and this is data which is of some interest to me seeing i have a set on HC.
slushslush
Group: Member
Posts: 46,989
Joined: Apr 8 2007
Gold: 0.42
Jun 2 2022 11:59am
This will be fun :)
LabattBlue
Group: Member
Posts: 35,284
Joined: Nov 2 2007
Gold: 3,629.64
Jun 2 2022 12:05pm
wouldnt you want to use crit gear for testing this? i just looked at your spreadsheet and it showed 1 or 0 cs from gear
Bigheaded
Group: Member
Posts: 10,216
Joined: Sep 17 2005
Gold: 27,578.68
Jun 2 2022 12:14pm
Quote (LabattBlue @ Jun 2 2022 07:05pm)
wouldnt you want to use crit gear for testing this? i just looked at your spreadsheet and it showed 1 or 0 cs from gear


I need the baseline for crit before anything else, how would i know how much crit gear has without any other data?
Lets suggest i crit 16% of the time with +10 crit gear with 10,000 attacks, this means absolutely nothing as I dont know how often you would have crit when you have no crit gear at all.
Knowing with 10,000 attacks that i have a crit of 8-12% with polearms depending on how many stats i was using would give a baseline specifically for polearms and given my knowledge of how damage calc works i could potentially suggest a formula for how strength/dex affects crit based on weapon ratio. Then i could apply the 10,000 attacks with 10 crit gear and suggest how much 10 crit affects your chance to crit.
So as i said ideally i need data with 0 crit gear first, but having the data with crit is also helpful as i'll need it later on anyway as that's one of the most important stats items can have

That's partly why the link to jedimagi's testing is interesting is it does 3 different scenarios with wildly different crit chances for similar stats on the same weapon.
I'm wouldn't trust 500 attacks of each type too much but it certainly proves a point that we don't know enough and strength/dex significantly impacts crit chance.
LabattBlue
Group: Member
Posts: 35,284
Joined: Nov 2 2007
Gold: 3,629.64
Jun 2 2022 12:15pm
isnt the baseline just the normal calculator or does that not take into effect standard crit hits
Bigheaded
Group: Member
Posts: 10,216
Joined: Sep 17 2005
Gold: 27,578.68
Jun 2 2022 12:17pm
Quote (LabattBlue @ Jun 2 2022 07:15pm)
isnt the baseline just the normal calculator or does that not take into effect standard crit hits


it does not take into effect standard crit hits.


only implemented quickdraw and power shot so far as it's presumed each 1% of each literally gives you 1% to perform it and this is pretty easily calculated.
Go Back To Ladder Slasher Topic List
1234Next
Add Reply New Topic New Poll