League rules should be changed

[artmas]

In my opinion, I think elo system for lihl League must be removed.
Just, I think the odds for the outcome should be introduced.
Due to the elo system, many users should avoid 3s or 2s games.
What do you think about this?

[supersexyy]

Remove elo for win/loss.
Will make captains mode more fun rather than choosing low elo players.
Alternatively scaling can be changed (and should be changed).

[pewpew lasergun]

yet a better idea is to introduce a point system:

win = +1
loss = -1

lets try a 2s old system:
https://entgaming.net/findstats.php?id=5709333

ilocos_norte 1378.64 (132/104)
tastay 780.28 (29/45)

tictactoe 975.12 (249/256)
achillesgr 860.16 (352/367)

average Team A 1079.46
average Team B 917.64

even with a low elo player ilocos lost 21.87

eventual new system
ilocos_norte +28
tastay -16

tictactoe -8
achillesgr -15

ilocos would lose -1 to be +27

off topic: i think ilocos is the best player of the league, others players should get inspired from him. ^^

[SLSGuennter]

like i said on diablo's post before, a new elo-scaling would be legit imo

[aRt)Y]

Well, draft some proposals then. We will gladly consider implementing them.

[Diablo_]

Can we get the current function which calculates the elo gain/loss?

[nabo.]

@uakf.b

[Diablo_]

Plus would it even be possible to change it only for lihl

[eldryan]

even if you can't, you can just add 3 ELO branches instead of 4v4/3v3/2v2 all being same branch. the strategies vary pretty drastically anyways, kind of like dota 3v3 compared to dota 5v5.

[supersexyy]

3 Elo branches is a horrid idea but most people can agree scaling is required, or at least an Elo gain/loss cap.

[nabo.]

How does win/loss =+1 -1 differ from elo 15/15?

[supersexyy]

Its the same thing so why multiply everything by 15 when there's no need to

[eldryan]

3 Elo branches is a horrid idea but most people can agree scaling is required, or at least an Elo gain/loss cap.

3 ELO branches creates actual balance. win/loss record is stupid as is elo gain/loss cap. will you just have whoever wins at a 51% clip and plays 1000 games the winner of the season? seems kind of basic. to a lesser effect win/loss capping will create this effect - effectively scaling ELO so that after a certain point it means nothing, as everyone above 1300 gets the same gain/loss anyways and it will just be win/loss for everyone over or below a certain ELO.

scaling will make it very easy to gain elo in 2v2 in theory as 1100 ELO and 1000 ELO would effectively gain the same amount which wouldn't be true in 4v4. at higher ELOs scaling will just do the same thing as win/loss cap. trying to "scale" 2v2 to match the 4v4 bracket will always be flawed, as they're not directly equatable.

[supersexyy]

I never said anything about scaling 2v2 and 3v3 differently. You can change the multiplier to scale differently for the elo difference.

[HealByColor]

just make everygame = 15 elo but do auto elo balance still