http://dbd.game/killswitch
"Nightlight isn't a reliable website to get use rates"
Well... look at this:
BHVR:
Nightlight:
Result: 8/10 hits
BHVR:
Nightlight:
Result: 9/10 hits (NOED v. Deadlock is the only difference)
BHVR-Killer-Pickrates:
Nightlight:
Result: 9/10
Free-to-play Wraith and Huntress on top together with the good chapters Wesker and Xeno.
Of course its not accurate for 100%, but the direction is pretty similar I would say.
At the end I want to say: THANK YOU BHVR for giving stats. Stats are beautiful, we all love them.
Comments
-
I mean, if Nightlight's user base is large and diverse enough, it should indeed reflect BHVR's own data.
5 -
Nightlight is an interesting place for information, and isn't bad as a reference. It's bad when people start claiming it's gospel truth and somehow 'more accurate than BHVR'S stats'. (Which is what can happen in some threads).
People who tend to rely on nightlight too heavily also tend to suffer confirmation bias. It's great if the data is close, but you're ignoring the misses as well.
Not even having self care in the top 10, when it's #6 for BHVR, is a pretty gigantic whiff. When it's close to right, great. But when it's wrong, it can be really, really wrong.
16 -
Nightlight is a good indicator, but in terms of numbers still way off as we can see:
For example:
WoO: 28% vs 33%
Mft: 21% vs 25%
Adre: 25,5% vs 20%
3 -
Thats because BHVR uses the worldwide data, which includes asian region. This is where Self Care is meta.
7 -
Exactly. Nightlight is a West-centric English-language website so the stats there will be biased toward NA and EU servers. As far as I know, the only significant difference between Asia meta and ours is Self Care though so yeah, Nightlight statistics are fairly reliable overall.
8 -
Which it does and its funny how people claim its unreliable
No joke someone tried to pull a gotcha with killrates on certain maps with a extremely minor killrate ratio leaning towards killers saying nightlight is clearly innacurate because according to it these maps are killer sided forgetting the fact that new survivors tend to play on maps that they think or heard are safe and "survivor sided" lol
4 -
it is normal that number are different because nightlight in statistical analysis is called sample. BVHR statistic in statistical analysis is called a census. census is true result. samples are not correct because there is not enough information. for the most part, nightlight will provide correct information but it may not be exact numbers.
5 -
Nightlight has proven to be even better, since their data comes from dedicated players it filters out perks that are mostly only used by novice players, such as self-healing and NOED.
0 -
ppl love gathering behind statements of others
My absolute fav will be when DBD devs change their stance on Tunneling, publicly confirming it as, “not a viably healthy strat,” and Tunnelers everywhere, no longer having their BHVR given entitlement to posture themselves upon, actually execute their right to remain silent… 6 feet above their highest esteemed tunneling narrative retort.
2 -
A lot of changes would have to come before anything is done to tunneling. The game isn't built to go for multiple chases on most of the killer roster at the current state.
0 -
Sure, it is a sample, thats why i think it is only a good indicator, and not the reality.
We can only assume that it has something to do with that. Or do we know the european/american data aswell?
0 -
Thats not how statistic analisis are done tho, you will never have the complete result, most analysis are done with samples. You cant phisically account for all individuals in a population to get data on, data science uses sample estimates to get close to the actual estimate, this is quite literally what data analysis is about, if nightlight shows to be statistically significant of a sample to predict the metrics of the overall population then statistically you can infer behaviours with the sample. You absolutelly do NOT need to get the comokete population in order to analyse something, if you did we would not have antibiotic treatments, quantum physics and so on, and as weird as it is this being a videogame the data is still analizable in scientific manner.
0 -
No. We just know that Self Care is a popular pick un asian regions
0 -
It seems like nightlight users are more polarized towards being players who have and use a lot of the popular choices, so it definitely can show popularly for the top things but as you go further down the data will be less 1 for 1 when applied to the larger population of dbd players who don't use nightlight
4
