Pets don't increase Critical Chance


AnElfCalledMack

 

Posted

So to satisfy my curiosity, I asked a friend to bring a lvl 50 Thug MM.

At first I wanted to try with dummies but dummies would only show the initial Critical text from Hidden. It will not show any other critical at all unless you use Placate.


So I side-kicked to my friend's lvl 50 to make sure his pets don't get attacked. I found a group of lvl 43 Malta. He "goto" his pets to the middle and I got rid of my Hide. I basically just count how many critical I score in 20 hits.

Well, in the end, I score 4 critical in 26 hits which is 15%. That's far from 28% that I am supposed to get. 10% + 3x6 = 28%.

The first test I only did 12 hits and I only scored two critical which is 16%.


I just don't see any "obvious" critical increase.


Maybe some of you can try it too with a MM. That's all. I thought pets increase my chance, at least it felt like it. Maybe I should do 100 hits to even out the chances. Oh well. My friend got bored.


What's left is to normalize all Assassin Strikes and improve Stalker's old sets (Claw, MA and EM)! You don't need to bring back the missing PbAoE attack. You just need to make the existing ones better! For example, make Slice a WIDER and LONGER cone.

 

Posted

Statistics is a funny thing. 20 hits is way too small a sample to accurately test any sort of random number generator. For example, that "four crits in 26 hits" thing, with a crit chance of 28%, has a 6.68% chance of occuring. That's not an insignificant chance. If you really want to test your crit chaces, start logging your combat chat, because if you can count the number of trials you've used without some sort of mechanical assistance, you haven't made enough for a decent statistical analysis. You want on the order of tens of thousands of data points, minimum, before you can make decent conclusions.


 

Posted

Quote:
Originally Posted by AnElfCalledMack View Post
Statistics is a funny thing.
This.

A slot machine will "eat" 15% of the money put into it, but you can hit the jackpot three times in a week. That 15% is taken over the lifetime of the machine, which is something like a billion plays.



Put another way, I was soloing last night, and got three criticals in a row on one enemy and he died. Is that a 100% crit rate while soloing? No. I just got lucky.

You need a huge sample size before you can accurately know the results.


Where to now?
Check out all my guides and fiction pieces on my blog.
The MFing Warshade | The Last Rule of Tanking | The Got Dam Mastermind
Everything Dark Armor | The Softcap
don'T attempt to read tHis mEssaGe, And believe Me, it is not a codE.

 

Posted

How do crits display in the combat log? I haven't really bothered to look and see if it says more than "you did double damage!" or whatever with a critical, or if it also adds what your chance for a critical was, kind of like your chance to hit rolls. Might be kind of helpful if it did say this, as Stalkers are supposed to have certain chances at times, and Scrapper criticals change depending on the target as well.


Guide: Tanking, Wall of Fire Style (Updated for I19!), and the Four Rules of Tanking
Story Arc:
Belated Justice, #88003
Synopsis: Explore the fine line between justice and vengeance as you help a hero of Talos Island bring his friend's murderer to justice.
Grey Pilgrim: Fire/Fire Tanker (50), Victory

 

Posted

Quote:
Originally Posted by Grey Pilgrim View Post
How do crits display in the combat log? I haven't really bothered to look and see if it says more than "you did double damage!" or whatever with a critical, or if it also adds what your chance for a critical was, kind of like your chance to hit rolls. Might be kind of helpful if it did say this, as Stalkers are supposed to have certain chances at times, and Scrapper criticals change depending on the target as well.
It looks like this:


Yes, that's a double critical on AS - I took the screenshot a while ago for someone who didn't believe that it could happen (before it became common knowledge); essentially it's just "You land a critical strike..." in the combat log and you normally wouldn't see both in a row, nor does it tell you what the chance to hit the critical was.


Quote:
Originally Posted by PleaseRecycle View Post
it has gone from unconscionable to downright appalling that we have no way of measuring our characters' wetness.
Quote:
Originally Posted by Brillig View Post
It's hard to beat the entertainment value of Whackjob Wednesdays.

 

Posted

I'll try to find a friend to do it again tonight. Maybe we'll aim for 100 hits this time.


Fighting lower level can be annoying as the mobs keep running away. Maybe I'll find an even level mob this time. I just don't want the pets to get hit and die and I want the master to stay away from me so only pets are counted.


Too bad dummy doesn't show any critical except from Hidden. I wonder why that is. It doesn't even show in combat log. It only shows I do zero smashing and energy damage and sometimes I "stun" it.


What's left is to normalize all Assassin Strikes and improve Stalker's old sets (Claw, MA and EM)! You don't need to bring back the missing PbAoE attack. You just need to make the existing ones better! For example, make Slice a WIDER and LONGER cone.

 

Posted

Quote:
Originally Posted by Jibikao View Post
I'll try to find a friend to do it again tonight. Maybe we'll aim for 100 hits this time.


Fighting lower level can be annoying as the mobs keep running away. Maybe I'll find an even level mob this time. I just don't want the pets to get hit and die and I want the master to stay away from me so only pets are counted.


Too bad dummy doesn't show any critical except from Hidden. I wonder why that is. It doesn't even show in combat log. It only shows I do zero smashing and energy damage and sometimes I "stun" it.
Your best bet may be to make some pitifully weak AE enemies and set up a 1-mission arc for the sole purpose of testing. Make an enemy type that clashes least with your secondary (such as Electric Blast for /Elec), only give them 1 or 2 weak attacks, and go to town. Easy, fast, no risk, will show numbers.


My guides:Dark Melee/Dark Armor/Soul Mastery, Illusion Control/Kinetics/Primal Forces Mastery, Electric Armor
"Dark Armor is a complete waste as a tanking set."

 

Posted

Quote:
Originally Posted by Nalrok_AthZim View Post
Your best bet may be to make some pitifully weak AE enemies and set up a 1-mission arc for the sole purpose of testing. Make an enemy type that clashes least with your secondary (such as Electric Blast for /Elec), only give them 1 or 2 weak attacks, and go to town. Easy, fast, no risk, will show numbers.
Except you don't want to be running all over a map. To maintain an absolutely consistent crit rate (assuming the pets matter) you want the pets huddled around you and passive while you swing your lightest and fastest attack repeatedly. Maybe get a cimeroran mission and go after a boss or LT that has a Surgeon backing him up.


Villains: Annie Alias, Dr. Amperical, Shade Golem, Knight Marksman
Heroes: The Clockwork Mime, Soccerpunch, The Fissioneer, Samurai Houston, Oversteer

Join The X-Patriots on Virtue!

 

Posted

Quote:
Originally Posted by Zem View Post
Except you don't want to be running all over a map. To maintain an absolutely consistent crit rate (assuming the pets matter) you want the pets huddled around you and passive while you swing your lightest and fastest attack repeatedly. Maybe get a cimeroran mission and go after a boss or LT that has a Surgeon backing him up.
Spawn a bunch of 'em then. Just increase their spawns. If they have crap attacks that'll be soaked by your secondary, what's the issue?


My guides:Dark Melee/Dark Armor/Soul Mastery, Illusion Control/Kinetics/Primal Forces Mastery, Electric Armor
"Dark Armor is a complete waste as a tanking set."

 

Posted

Quote:
Originally Posted by Zem View Post
Except you don't want to be running all over a map. To maintain an absolutely consistent crit rate (assuming the pets matter) you want the pets huddled around you and passive while you swing your lightest and fastest attack repeatedly. Maybe get a cimeroran mission and go after a boss or LT that has a Surgeon backing him up.
Yes, in order to get a more accurate finding, I need the master to "goto" the pets when I move and I need him to stay away from me pretty far. And I don't want the pets to get hit and that's why I side-kick to his lvl 50 and I find a lvl 43 group so the pets just stand there and won't die while I do the hitting.

Actually, I should have asked him to use Gang War... lol I forgot about it.


PS: The first time I found a group of even level Malta. I jumped in and died. LOL


What's left is to normalize all Assassin Strikes and improve Stalker's old sets (Claw, MA and EM)! You don't need to bring back the missing PbAoE attack. You just need to make the existing ones better! For example, make Slice a WIDER and LONGER cone.

 

Posted

Quote:
Originally Posted by Jibikao View Post
Yes, in order to get a more accurate finding, I need the master to "goto" the pets when I move and I need him to stay away from me pretty far. And I don't want the pets to get hit and that's why I side-kick to his lvl 50 and I find a lvl 43 group so the pets just stand there and won't die while I do the hitting.

Actually, I should have asked him to use Gang War... lol I forgot about it.
Ah, fair enough. In that case, go find something notoriously tough. Spawn an AV in a throwaway AE mission and get two MMs; one Pain, one FF. Have the FF bubble you up, and the Pain give you every buff he can and watch your HP for when stuff breaks through. Aim the AV away from the pets but stay near 'em.


My guides:Dark Melee/Dark Armor/Soul Mastery, Illusion Control/Kinetics/Primal Forces Mastery, Electric Armor
"Dark Armor is a complete waste as a tanking set."

 

Posted

Quote:
Originally Posted by Dechs Kaison View Post
Maybe create an AV with only one single target attack? It can't kill you while you rotate attacks with pets nearby?
Great minds think alike. Funny enough that's how I tested my Kin/Regen Stalker's crit rate with two Corruptor buddies near me. It was fun!


My guides:Dark Melee/Dark Armor/Soul Mastery, Illusion Control/Kinetics/Primal Forces Mastery, Electric Armor
"Dark Armor is a complete waste as a tanking set."

 

Posted

Since you're doing this for testing and not for xp you can give your AE enemies zero attacks. A small map like Johnny's Hell or the Tiny sized sewer or abandoned office will limit the ability for the enemies to run. In fact, it might be best to use a custom group of Council, Crey, or 5th Column snipers since they are immobilized by default unless they are hit by outside buffs and the attacks are ST as opposed to DE sniper summons and Nemesis sniper AOE.


 

Posted

Quote:
Originally Posted by Nalrok_AthZim View Post
Ah, fair enough. In that case, go find something notoriously tough. Spawn an AV in a throwaway AE mission and get two MMs; one Pain, one FF. Have the FF bubble you up, and the Pain give you every buff he can and watch your HP for when stuff breaks through. Aim the AV away from the pets but stay near 'em.

This is so involved!!! lol I haven't created any AE at all. I don't think I'll go with the AE route.

You guys can try the test too you know? :P


What's left is to normalize all Assassin Strikes and improve Stalker's old sets (Claw, MA and EM)! You don't need to bring back the missing PbAoE attack. You just need to make the existing ones better! For example, make Slice a WIDER and LONGER cone.

 

Posted

Quote:
Originally Posted by Jibikao View Post
This is so involved!!! lol I haven't created any AE at all. I don't think I'll go with the AE route.

You guys can try the test too you know? :P
Well, we could, but we probably can't be bothered. Personally, I'd rather play the game than get all caught up in whether the Stalker crit boost is working as intended. You're the one who was interested, and the Stalker scaling crit rate needs very tightly controlled circumstances to keep all the data comparable. In short, if you're out to test a probability down to anything like 3%, it's going to take a lot of work. If you don't feel like putting in the work, you don't have to, but we're not going to be that interested in "I only crit 4 times in 26 hits! Something's broken!", because testing it *does* take that much work.

As foar as knowing how to do it, though, there have been some rather impressive results from things like this. Arcanaville and others have found some really significant bugs, like defense stacking, through very similar methods. For example, to test the randomness of the RNG, there were a number of people logging their combat chat as they played and sending them to Arcanaville. She ran an analysis on millions of to-hit rolls and found that there was a small bias in the fourth decimal place, which barely ever matters. If you fine-tune your accuracy really carefully, you can sometimes squeeze an extra 0.01% out of the system, or maybe be robbed of 0.01% you'd "paid for", but nether of these effects is significant compared to the minimal effort required to cap your to-hit. But to get decent data, they needed millions of rolls. To spot a 3% difference, you're going to need at least tens of thousands of hits.

Frankly, I think an easier test would be to get a combat pet of your own, line up against an AV with one attack that you resist strongly, and set your T1 attack to autocast, then go have lunch. For a comparison, dismiss the pet, then repeat. That'll let you generate thousands of crit chances without taking up anyone else's time, though you're only looking for a 3% boost, so you might need mroe data.