and
"How to perform incompetent data analysis on tanks"
I'll make this as quick as possible, so probably bullet points will do best.
Version 1.0 as of 19.06.19
(not sure if I missed some changes of the new patch though)
What is this?
Where did I get it from?
But why?
Structure of the file
IMPORTANT! Read before you complain: Weaknesses of the calculation:
I will not put this into spoilers for obvious reasons.
Sheet 1:
- Due to my incapability of getting some data from the files (like the reference points for accuracy/penetration far/mid/near etc, this calculation will always compare performance at max range. If you compare a SU76 vs a P4, then you scew your data to the disadvantage of the SU76, as it has better accuracy and pen than used in the calculation.
- It does not take into account anti-infantry capability. So if a unit performs bad in the analysis, then ask yourself if it makes up for it in some other category, like AI, but also speed etc
- Rate of fire is a bit buggy, since there are more factors to the units than raw reload time and I don't know where to get them from. But it can be an okay-ish estimate. If you know where to find them in the Tools-App, please write me a PM.
Sheet 2:
Oh boy, where do I begin?
- no weighting of stats that might objectively be more important
- no weighting of extreme outliers (i.e. a tank with 0 damage will not be punished as hard by the model as it should be)
- negligence of marginal utility (e.g. the damage increase from 159->160is objectively better than the increase form 160->161, but the model does not account for that)
- negligence of anti infantry performance
- model quality depends on the number of stats used
- negligence of positioning/flanking and utility abilities
- negligence of alpha damage, time-to-kill etc
- negligence of collision-based hits