How to actually calculate how much a person stacks

[EdgeOfChaos]

Since I've heard wrong formulas for this over and over, here's how to find how much ELO a person plays for on average (so you can see whether they stack). This is not as trivial as people make it out to be. User must have played at least one game for this to work.

e = player elo
w = wins
l = losses
W = average gained elo
L = average lost elo

Set up a system of equations.
W + L = 30 --> We know this because the ELO lost and gained always adds up to 30 (with the exception of 2000+ ELO players which we won't get into now)
e=(Ww) - (Ll) --> This states that a player's ELO is simply (wins * elo per win) minus (losses * elo per loss). Keep in mind, players have 1000 base ELO so you must add this after.

Solve for one (say W)
W = 30 - L

Substitute into 2nd equation
e=(30-L)w - Ll
e=30w - wL - Ll

Subtract 30w, Multiply through by -1
e-30w=-wL - Ll
30w-e=wL + Ll

Isolate L
30w-e=L(w+l)
L=(30w-e)/(w+l)
Then it's easy to find W

Let's try an example, make sure it makes sense.

Spoiler!

w=50
l=50
e=100 (so this player is at 1100 elo)

L=(30(50)-100) / (50+50))
=14 ELO lost
So then ELO gained per game is 16

Let's see what this means. The player won 50 games for average 16 elo. This grants 800 total ELO. They lost 50 games for 14 elo which removes 700 elo. This leaves them with 100 elo which is the correct amount. So this player would usually play games on the disadvantaged team.

So some examples from the playerlist:

Spoiler!

w=283
l=111
e=526.68

L=(30(283)-526.68)/(283+111)
L=20.2 elo
So this user plays on average 9.8/20.2 games with the advantage

Spoiler!

w=49
l=2
e=603.14

L=(30(49)-603.13)/(49+2)
L=16.9
So this user plays on average 13.1/16.9 ELO games with the advantage

EDIT: Here is a Java program to calculate this for you:

Spoiler!

Code: Select all

import java.text.DecimalFormat;
import java.util.Scanner;

public class StackCalc {

    public static void main(String[] args) {
        Scanner in = new Scanner(System.in);
        System.out.print("Enter wins: ");
        int wins = in.nextInt();
        System.out.print("Enter losses: ");
        int losses = in.nextInt();
        System.out.print("Enter total ELO: ");
        double elo = in.nextDouble() - 1000;

        double lossPerGame = (30*wins - elo)/(wins+losses);
        double winPerGame = 30-lossPerGame;

        System.out.println("ELO lost per game: " + lossPerGame);
        System.out.println("ELO won per game: " + winPerGame);
        System.out.println("Stack factor: " + new DecimalFormat("0.000").format((lossPerGame-15)/30));
    }
}



[Sylvanas]

Attempting to read this makes me so glad that I don't actually care.

[Burnt]

Solve for one (say W)
W = 30 - L

I don't think this premise always stands.

For example, if one wins every stacked game (+5 elo), and loses every unstacked game (-5 elo), which are highly likely. Then the W and L for this particular player are both 5, which don't add up to 30.

[warlord1994]

all of u stackers.
im the most solo player on this bot.

[EdgeOfChaos]

The ELO lost and gained and lost for every game totals 30. It will always be a pair of, like, 20/10 or 15/15. This just calculates what elo pair you play for on average - that is, average over all games losses and wins. Obviously you can't find whether someone plays a 25 elo game and then a 5 elo game, but you can look at averages. If someone gained 5 then lost 5, this would state that he plays for 15/15 (perfectly balanced) on average. Not true, but equivalent ELO-wise.

If you want, calculate one of the examples in reverse. That is, find (wins)(elo per win)-(losses)(elo per loss) and compare it to the original ELO.

Spoiler!

Original example:
w=283
l=111
e=526.68

L=(30(283)-526.68)/(283+111)
L=20.2 elo
So this user plays on average 9.8/20.2 games with the advantage

Backtracking:
This user plays 9.8/20.2 games.
ELO gained = (283*9.8) = 2773.4
ELO lost = (111*20.2) = 2242.2
2773.4 - 2242.2 = 531.2
(Essentially the initial amount, with small rounding error)

[Dhamma]

you went all math geek on this lol, but not bad

easiest way for people to calculate the average elo played per game is to plug into excel:
w,l,e into a2,b2,c2 respectively
and paste the below into another cell
=30-(30*(A2)-C2)/(A2+B2)

[SLSGuennter]

Fine thing, yet you guys should keep in mind, that only cuz somebody has an average of like 12gain 18 lose, that may not always be cuz he stacks ... if he got high Elo himself he will kinda always be in the team with more average Elo ^^.

[fgfe]

It's always funny getting accused of beeing a stacker, only because you have a high elo.. mostly games get even more inbalanced because so many people jump on the side where the high elo player is. After that usually players from the lower elo side leave, new ones join and complain about the one high elo player to stack.

[Dhamma]

It's always funny getting accused of beeing a stacker, only because you have a high elo.. mostly games get even more inbalanced because so many people jump on the side where the high elo player is. After that usually players from the lower elo side leave, new ones join and complain about the one high elo player to stack.

Got news for you bro, If you are playing alone, and you stay on that team and don't move over to the lower elo side, you got no cojones and are a stacker.

Say you're 1900 elo, playing alone and join a game. A bunch of people jump to your side, mostly around or over 1.1k. The opposite side has a few 800s and suddenly the game is worth 4 elo.

This is when you join the side with the low elo, carry them, and gain 10-12.

This is not always possible when one is high elo, so many times one is forced to play for around 8 elo when they get to 1900+. But having a 600 and a 700 on your team even when you're 22k can make a 15 elo game.

[Sylvanas]

It's always funny getting accused of beeing a stacker, only because you have a high elo.. mostly games get even more inbalanced because so many people jump on the side where the high elo player is. After that usually players from the lower elo side leave, new ones join and complain about the one high elo player to stack.

I've been accused of stacking both for having high elo and for having low elo. I can't read math runes, so take anything I say with a grain of salt, but I think the truth is that it's impossible to track down the real "value" of all games played just by the # of wins/losses and current elo. In fact, the average elo of both teams in one specific game often can't even tell you which side is stacked. You can have a game where you risk losing 20 elo to win 10 and still have the worst team because of smurfs and new/reset accounts, which has been happening quite a lot lately.

[fgfe]

It's always funny getting accused of beeing a stacker, only because you have a high elo.. mostly games get even more inbalanced because so many people jump on the side where the high elo player is. After that usually players from the lower elo side leave, new ones join and complain about the one high elo player to stack.

Got news for you bro, If you are playing alone, and you stay on that team and don't move over to the lower elo side, you got no cojones and are a stacker.

So joining a game and being alt tabbed until it starts without switching teams once makes u a stacker? Makes sense mate.

[Sylvanas]

This is not always possible when one is high elo, so many times one is forced to play for around 8 elo when they get to 1900+. But having a 600 and a 700 on your team even when you're 22k can make a 15 elo game.

To "counter" a 2.2k elo (which I assume is what you meant), you need 1200 negative elo on your team, so that would be more like a 300 elo and a 500 elo. Not only will they most likely be too much dead weight to carry, but you'll actually have to make the effort to SEARCH for those elusive, bumbling idiots to afflict yourself with (and then lose).

To be fair, god awful players aren't that rare, but few of them stick to the same account for hundreds/thousands of games until they have so little elo left. Of course, this doesn't get taken account of either; the 10% win rate player with 300 elo counts as an average 1000 elo player once he starts making a new account every day.

Honestly, the elo system is just bad and was always bad. I don't know why people try to make sense of a system that never worked.

[Dhamma]

So joining a game and being alt tabbed until it starts without switching teams once makes u a stacker? Makes sense mate.

That's just not something i do on a high elo acct i'm building. One can end up playing for 2 elo like that, simply a waste of time.

To "counter" a 2.2k elo (which I assume is what you meant)

Huh?

Im talking about finding games that give you more than 5 elo per game, once you hit 2k, i guess you can't relate.

[Sylvanas]

To "counter" a 2.2k elo (which I assume is what you meant)

Huh?

Im talking about finding games that give you more than 5 elo per game, once you hit 2k, i guess you can't relate.

No, I can't relate, but I don't have to?

This is what you said:

But having a 600 and a 700 on your team even when you're 22k can make a 15 elo game.

What I assumed was that you were saying a 600 and a 700 along with a 2.2k elo in a team evens out the team's average elo for a +15 on win, which wasn't true unless you add in other elements, especially considering very high elo reduces elo won/loss either way.

Calm your 2.2k elo cojones.

[divinerape]

ELO (rank): 1622.93 (100)
Games: 605 W/L: 388/217 (64.13%)

w=388
l=217
e=622.93

L=(30(388)-622.93)/(388+217)
11640-622.93/605
11017.07/605
L=11.79
So this user plays on average 11.79/18.21

Is my math correct boss?