Hello, held off on posting this for quite awhile (ever since p2) because I just assumed I have bad luck when it comes to crit chance. However, with this being the end of classic so to speak, I wanted to post this to generally ask if anyone else has experienced this or what specifically is going on.
The issue is this: I have an absurdly high crit chance on paper, but when raid logs are done, my actual crit rate is sometimes off by up to 25% less than what’s on the paper. I have a nice screenshot of my huntard having 48.7% crit, and then warcraft logs showing my actual crit rate at 31.4%/33.3%/37.5% Autoshot/Multi/Aimed. Even with the 5% crit suppression on bosses, that’s almost 5% unaccounted for. It’s a bit confusing to say the least, and that’s one of the more generous ones. I’ve had 45%+ crit before, and only parsed at 20%.
Is there a bug? Is this just bad rng? Is the addon displaying crit improperly (CharacterStatsClassic)? Anyone have any experience with this? It’s like, the more crit I stack, the less I actually crit. I realize this all seems anecdotal, but I do have screen caps showing logs vs actual crit chance.
Sorry for the wall, but would definitely appreciate some perspective on this.
That’s against mobs your level. Bosses are lvl 63 which gives crit suppression due to their defense being 315. There is also a 1% to miss and misses can not crit. Then RNG…
Yea, thats in the second paragraph lol even accounting for 6% (1% miss + 5% crit suppression) there is still a variance in some cases of 25%+. It’s rough to see 45%+ crit on paper, and then show an average rate of 12% on a boss. That’s what I’m wondering about.
Definitely know about the crit suppression tho - thats why I get songflower (I’m also surv spec for 3% crit, and have a feral druid with an additional 3%…)
I mean, I raided in retail through BC and WotLK and never experienced a crit variance like this… There’s really just RNG at play when you parse 12% crit on a boss with 45% crit on paper? thats such a huge difference…
EDIT: Like I really am not trying to start a topic about RNG, it just seems like such a vast difference that would be mathematically improbable to crit that low.
well lets see… your log for 12/20 you did not have a songflower and you crit 39% on multishot and 30% on others. averages right out.
Yea, I’m just generally referencing it here. Most of the screen grabs and such have been from previous phases - as we really don’t have much raid time in p6.
If you want to look at a weird one, 11/15 P5 in AQ on C’Thun. 26%/34.7%/33.3% - Base 44% crit. 44-6=38-26=12. Where is that extra 12%? that’s in the realm of RNG?
There are actual parses if you’re looking, that will show a normal crit rate. However, I’ve experienced this abnormal crit rate quite frequently (too many times to go back through and pick all the examples lol).
Definitely appreciate the response tho. I just don’t wanna believe its rng because of how stark the difference is xD
You did not have a song flower for that fight and you averaged like 37% on ranged attacks.
You missed 4 auto attacks…
You missed 6 aimed shots.
Yea, like I was saying, it’s not always something that happens, just that it happens frequently enough to be a concern. Since we don’t have much raid time in p6 the sample size is much smaller, so comparing averages isn’t really fair yet
Yeah it’s RNG but it should average out.
Lol i literally dont even know how that’s possible to miss shots at hit cap xD Never even realized. Kinda funny really - and in a weird way illustrates my point. Weird.
I guess I’ll take the RNG answer, but it still seems odd af
That hit cap is for yellow attacks and there is always a 1% miss for all attacks. On mobs like eye stalks things can be batched and the shot goes off but the mob is already dead, that is counted as a miss.
1 Like
Ahhh that makes sense with the batching. Probably what happened since those things go down super fast. Good info, tyvm!
I know this seems like a dumb thing to talk about in P6, but classic has had a ton of bugs for hunters, so i wasn’t really sure where the “not a bug” and “is a bug” line was
only casters got that 1% miss chance you cant avoid
I honestly never played vanilla, so learning these little intricacies has been an adventure.
I’ve had fights where I crit 40% of Backstabs at over 80% crit for Backstab.
1 Like
Instead of looking ona boss by boss basis look at your overall damage for an entire night.
Looking at your log on dec the 13th, I can see quite abit of natural variance on your abilities over the entire raid and my incredibly rough eyeball math says you were sitting at about 30% over the entire raid.
So I’d write off that log as a “bad” night and just wait for a better night.
However much you want to discount RNG it is something that you can try to minimise but at the end of the it has you by the short and curlies every time you take a swing a mob/boss.
EG. I’ve had some absolute shocking boss fights where I’ve done 180dps thanks to windfury not proccing,an absurdly high amount of glancings or attacks being dodged/resisted left right and centre. On the flipside I’ve done over 800 DPS when the stars align.
It’s just something you have to deal with.
2 Likes
Another reason could exist. Let’s not forget that parsing is part of a meta-game - a competition of value. DPS is valued by damage done, and that is divided by either total time or active time to result two DPS scores. Since dividing by active time allows cheating, we only view DPS as having been divided by total time. That gives only one possible means to cheat the numbers. Game hacking, as is done daily for farming, gathering, and rarely in PvP has proven to be possible in WoW, so it might also be possible to alter combat results given by WoWCombatLog.txt. This could be done in one of three ways.
- Manual edit of the text file.
- Intervening the server log write, so that a name attributed to damage is changed to another participant.
- Intervening the server log write, so that fictitious damage is added.
If it is #2 or #3, then the question then turns toward who is able to intervene server function to alter the log.