HUGE flaw in the latest average kill rate stats
So as confirmed, DC's were excluded to calculate the average killrate. We all know that a survivor DC'ing is very common, but lets be generous and lets say only 1 survivor in 10 games will DC. This can mean 2 things depending on how bhvr excludes DC's from the data.
A) What if bhvr exluded only the survivor that DC's from the data? games where a survivor DC's almost always ends up in the other 3 survivors dead --> thats 3 free kills, which makes the released data alot higher than in reality (considering 1 DC each 10 games with 3 other survivors dying or giving up would give us around a 10% higher killrate). This would mean that the cenobite killrate is not 60% but in reality it is close to 55%, same for nurse this would mean her killrate is not 43 but 38%.
BUT
B) what if bhvr excluded the whole round from the data when a survivor DC's? as these games end mostly in 3-4k, thats alot of kills that are not added to the data. considering 1DC each 10 games this would make that the cenobite killrate not 60% but in reality it is 70%, nurse killrate not 43% but 53%.
If they used this method (which is probably the case) the data that is shown to us gives a good indication of the game balance, however all these survivors not added to the data have experienced a dead in the game and are thrown away like they didnt even happen. (this would mean that the current data shown to us gives more info about game balance instead of the actual kill rates)
Now what if we are less generous and more realistic where 1/5 games have a person DC? this means the data shown to us is !!20%!! skewed to a side (depending on if bhvr used methode A or B)
How are we supposed to read the data if we dont know how the DC's are excluded???
I think this data shows us the biggest problem of dbd in its current state. The actual balance of the game seems to be getting close to the 50% all across the board (+rep for that bhvr), however, the experienced number of deads that survivors have is ALOT higher than 50%.
The balance has gotten alot better to a point that DC'ing has become the main issue in how the game balance is experienced/perceived ... DC games are too common. In the current state of the game the influence of DC's seems to be getting underestimated by alot.
Comments
-
Hopefully they just didn't count any games with a DC in them, but even so, it hardly matters. This gives a general idea of the kill rate relative to other killers and that is it. It is objectively useless data from the get go. So it doesn't matter too much.
0 -
It's logic to use complete trials to judge the balance of a game. Stastistics taking in account the trials where a survivor DC would have a bias.
So when you do stats, you exclude everything that is not how it should be and that's a good thing.
And 50% kill rates mean nothing about balance as a 50% kill rate with a deviation of 50 would be totally unbalanced. The deviation is important
1 -
These stats are also flawed on the endgame end. Plenty of survivors feed kills in the EGC trying to save their friend. If every survivor just bounced as soon as possible, the kill rates would be a lot lower
0 -
I'm sorry were the devs telling that these stats were incredibly, intrinsically flawed and so you shouldn't draw conclusions from them the first four times not enough?
2 -
Needs to be at least 5 times and with scrolling flashing text all over the screen.
0 -
What about games where survivors don't DC, but they just kill themselves on first hook leaving their team to die? This is super common and it pretty much dooms the entire team if done early enough. I would bet that these are counted.
0
