If you know how to read them correctly. And you can’t do that objectively without sifting through BR’s and comparing opponent teams. If SPM/Skill was subjective to your overall opponent’s ability then yeah. As is roflstomping a lvl 0 team gives the same amount of SPM as it would playing against equals. 3.0 KPM is easier to obtain against lvl0 vs equals and so forth.
edit: Accuracy in’t a loose stat. That value says tons.
you’re going to be able to get a fully accurate picture, all you’ll ever get is an estimation. there’s too many variables even in a single match to truly make up a picture. over time, the picture gets clearer as consistencies develop. even defining what an “equal” is is hard. you cant really quantify player behavior beyond the aggressiveness that SPMs show.
yeah, you have to know how to read them. but does that really improve by having map specific breakdowns?
it wont change the individual’s understanding of the numbers, all it does is give you different variables to consider, which someone who knows what to look for is already considering.
Totally agree. The variables used and how they are used can at best give you an estimation. But it’s really down to the variables and algorithms.
For example … purely hypothetical…Let’s for argument sake define “equal” as number that is possible to calculate or within a very small margin of error accurate.
if the points you gain for killing player X was based on the opponent’s calculated skill. An equal or closely there to would net the higher xp toward SPM.
So let’s say Player A has a Skill of 800. If Player A kills anyone below a skill rating of 750 he gets less points vs what he would get for killing players above 750. Even more xp for those above 850. Now do the same for KDR and KPM. Lower than 750 the kill doesn’t count toward the SKILL calculation.
You would still be awarded the high kdr/kpm in overall stats, but the SKILL stat would more accurately represent your overall (loosely read) Skill.
Another way to look at it is like Relaa playing against the same caliber of players in every round played. What would his SKILL/KDR/SPM/KPM be compared to him playing against Pub players. Without a doubt it would be drastically different. And that’s the whole argument in a nutshell.
Map based stats provides a more thorough view of an individual’s performance on that specific map without the need to load individual map reports and do the calculations yourself. As is in the example it doesn’t use any additional data than what would be sent in a BR. It just compiles them and displays more.
Map to Map most players do not perform the same. Seeing that difference at a glance could help some players.
that’s essentially just the ELO system they used in BC2 and BF3.
which is broken because it measures head to head confrontatins in a setting where that is not always the situation. if you have 800 skill, and are fighting with 4 people of 400 skill and one with 100 skill, and you kill all four with 400 skill, and die to the last with 100 skill, you would be severely punished despite taking on a 5v1 situation and killing 4 of the 5.
total skill combined divided by your skill (1700/800 = +2.125) added toward Skill stat kdr? There’s a way to calculate a much more accurate representation of skill. It’s a matter of finding the right combination of variables. The current algo isn’t even close to that answer.
That my friend is the 64 million dollar question. I’m sure given all the variables, scenarios etc mathematicians could figure out something respectable.
Arguably given that we keep a stat designated as Skill I’d start by giving each game mode it’s own independent Skill stat solely calculated based on the modes unique variables. TDM Skill is completely different from CQ Skill respectfully. In TDM the arena is smaller, everybody’s focus is strictly set on killing, no vehicles to contend with, spawning is predictable so on and so forth.
Maybe push that further to have both a vehicle and infantry skill for modes that support vehicles.
On the TDM side a hybrid ELO approach could be used.
i think all of those run into the same issue though…i think the game is simply too large to get a truly accurate picture.
if we distilled th game down to 5v5, yes, you can absolutely get more out of it. but 64 players has too many moving parts and variable server size is another issue. how do you punish or reward players for server size?
I wouldn’t have a skill stat if it were up to me. How to balance servers at that point? No friggin clue for CQ or any other objective based mode. TDM/SQDM would be pretty easy and straightforward.
I don’t mind the skill stat, what bothers me is the rapid fluctuations, sometimes on situations you have little control over I.e losing a match. Note I said little control not no control.
The could simply introduce an increase or decrease of max 5 points depending on round performance. With the initial skill to be determined after around 20 hours of gameplay
Nothing wrong with being cautious if you know you cant aim that well. Nothing wrong with flanking rather than going headlong into a choke point.
Nothing wrong with trying to learn each aspect of the game (classes and vehicles). You dont get good with something immediately.
I have a 2.4 k/d (not massively high) and a .85 kpm.
I dont camp.
I’m not concerned with my k/d (it was nearly at 3 when i was going for 100 jet , now i’m trying to master weapons).
I’m not overly cautious (850 spm over 1100 hours)
Just realised if i quote you i can see what you said. Yeah i guess…but I was talking about stars and it censored it with four stars. Guess that’ll have to do.
