Well, for example if you play a game as sven and your stats at the end of the game are 10-2-10, that would give you 10KDA, your K/D is 5 and your A/D is 5, so total of 10.
i think its smtg like f(a \times b) + f(a) + f(b) = f(a) \times f(b)
ax^2 + bx + c = 0 => kda.
Because they don't want to take the extra time to program taking into account 0 deaths which doesn't really happen anyway (for overall games. I know it happens in individual games)
How KDA is calculated is not as trivial as you guys make it sound.
Q1 how is it calculated PER GAME?
A: Suppose you have K kills, D deaths, and A assists, all nonnegative integers. Define f : RXR -> R as
f(x,y) = x + y. Let D := 0. Then,
KDA = f(K,A)
Now, let D > 0. Then,
KDA = f(K,A)/D.
Q2 How is KDA calculated, using MULTIPLE GAMES?
Suppose you have a collection of n games on the same hero with K_1, K_2, ..., K_n kills; with D_1, D_2, ..., D_n deaths; and with A_1, A_2, ... A_n assists. A_i, K_i, and D_i are nonnegative integers for all i in {1, 2, ..., n}. Put these into vectors, K := (K_1, K_2, ..., K_n), D := (D_1, D_2, ..., D_n), and A := (A_1, A_2, ..., A_n). Define the function f : R^n X R^n -> R as
f(x,y) := [1, 1, ..., 1](x+y)
Let D_1 + D_2 + ... + D_n := 0. Then,
KDA = f(K,A).
But if D_1 + D_2 + ... + D_n > 0,
KDA = F(K,A)/ ( [1, 1, ..., 1]D )
where vectors are columns and [1, 1, ..., 1] is a 1xn row.
In the latter case of multiple games, we could have envisioned KDA working as follows (which it does not):
Suppose you have n games where you had the following per game KDA computed: S_1, S_2, ..., S_n. Then, let the total hero KDA be the simple arithmetic average of these n entries.
@king wacoo
I actually provided new information no one else has at the end of my post. It is possible you are not smart enough to realize it, though.
@vandal
so, KDA with 1 die = KDA with 0 die?
Isn't a bit unfair?
and suck at math =! stupid
@Why do i play this stupid...
It's the same because it's not mathematically possible to receive an exact number by dividing by 0 hence the use of two different scenarios by Vandal.
And basically all he's saying in the second is the sum of all kills and assists divided by deaths is your KDR (and this is again divided into the two different scenarios where you have a situation without deaths and a situation with.
Yep. But having 0 deaths in a game will still improve your overall KDA with the hero because you will have 1 less death :P
It's like this
a = b
a + a = b + a
2a = b + a
2a - 2b = b + a - 2b
2a - 2b = a - b
2(a-b) = 1(a-b)
2 = 1
@Doctor Yolo M.D PhD, given that a-b=/=0 => a=/=b, which is not true, since a = b. Nice try, fool.
Relentless give up from dota,cos he had some abandoned matches and didn`t count for rating or smth and he is not satisfid with his 3400 mmr he said he won`t play anymore. :(
thanks. i just backed it out looking at one of my matches. so funny seeing people be like "your an idiot, how do you not know this? its k+a/d DERP"
We officially use:
float(int(k) + int(a)) / float(d)
If you're seeing something other than this please report it as a bug.
thats it in a nutshell yes. and no, another math fail. something divided by zero is not infite. you cannot divide by zero. such op maths
(Total Overall Kills + Total Overall Assists)/Total Overall Deaths
Like u played 2 games, one had a score of, 5/2/19
The other was, 3/6/13
Then ur KDA ratio would be (8+32)/8=5 KDA ratio.
And not (12+2.67)/2=7.333.
12 is the first game KDA, 2.67 is the second one.
Though I am not 100% sure about it, but I believe it is calculated that way.
[(Total kills Per Game) + (Total Assists Per Game)] / (Total Deaths Per Game)
Total Kills per game = Sum of total of kills over every game played with a certain hero / games played with a certain hero
Total Assists per game = Sum of total Assist over every game played with a certain hero / games played with a certain hero
Total Deaths per game = Sum of total of deaths over every game played with a certain hero / games played with a certain hero
@[o] Capt. Moochi
Thanks so much!! Pretty sure this is how they calculate it and I was trying to figure it out!!
Interesting method.
Zaloguj się aby dodawać komentarze.
Willing to know