Poll: Do you want a 20:10 ELO w/l cap for all LIHL games?

[ILOCOS_NORTE]

Spoiler!

From: ILOCOS_NORTE
Recipient: uakf.b

Hello uak,
one question for LIHL:

Can you alter the formula for ELO win/lose for LIHL only and put a cap to a maximum of 20:10 ELO w/l ?

If you think its possible you can read below, else archive.

We have a little problem with the ELO win/lose calculation for our inhouse games. Problem is that the best 10-20 players have an average win/lose ratio of 55 % Even if a high ELO player is playing vs low ELO players the winning chances may raise up to 60-65% but thats it.
Assuming a game were an average 1100 (1400 + 800) ELO team playing vs 850 (825 + 875) the ELO w/l will be around (assuming, I dont have the correct numbers) 25:5. So the high ELO team has to win 5 out of 6 games for equal ELOs wich will barely happen because you can get "out rolled" (means one team has much better towers with prophet builder dont know how much you are in to legion td).
That would be a 83% win/lose ratio for the high ELO player which no one in LIHL ever had.
You see the ELO w/l doesnt match the actuall winning chances.

If you think you can do something about pm me and ill bring it on the forum.

thx


From: uakf.b
Recipient: ILOCOS_NORTE

That is possible

If I get 75+% "yes" votes and LIHL admin support I will tell uak to change the formula

>For equal ELO on 20:10 ELO games, the high ELO team needs to win 66,6%<

[1005]

You could do +1 = win & -1 = loss system, like Supersexxy suggested.

The other way would do is follow the football system. Win = +3 scores, tie = +1, loss = 0. This will encourage people to play more games.

*Support* this 20:10 w/l cap.

[ILOCOS_NORTE]

You could do +1 = win & -1 = loss system, like Supersexxy suggested.

The other way would do is follow the football system. Win = +3 scores, tie = +1, loss = 0. This will encourage people to play more games.

That system doesnt work because from two guys with 55% w/l ratio the one with more games wins the season.

[aRt)Y]

20/10 seem randomly chosen to me just like 15/15. If you want to go with a standard ELO, then do it with -+1.

Furthermore, I doubt you are in charge of telling uak which things to implement. Please leave that to us.

[Krumme]

achilles 1000 elo but 420/440 w/l games is down 20 games his mate beep 1200 elo is 30/10 w/l games (theoretical) and they play together vs 2x 40/40 and 60/60 games that both are around 1000 elo
Seem fair to you?
Go with the elo it's more balanced.

[ILOCOS_NORTE]

I doubt you are in charge of telling uak which things to implement. Please leave that to us.

First of all, you are of course 100% right. No one is in charge to tell uakf.b what to implement since he will have the final word. So you are correct if you say that I am not in charge to demand changes. But if I have reasonable arguments for uakf.b with a (currently 100%) support from the LIHL community im sure he will consider the changes and come to a positiv conclusion.

Anyway, thank you for trying to help and providing your opinion.

20/10 seem randomly chosen to me just like 15/15.

Secondly, the 20/10 is not randomly chosen. I tried to explain that in the spoiler (the msg I sent to uakf.b, its in the first post in this thread) but that was maybe to difficult to understand for everyone. I try to explain me to you in simpler words.

1.) The ELO win and lose points of a game should mirror the winning chances for each team. If the game has a 20:10 ELO distribution, the mathematical winning chances for team(20) are 33,33% and for team(10) 66,66%
2.) In LIHL, the best "real" (see season stats) win/lose ratios of the top players are all below 66,66%. Even if one top player is playing with 3 bottom players a 2v2 with an ELO distribution of 25:5 (which would mirror a 83,33% mathematical winning chance for the high elo team), the actual winning chances are below 66,66%. So a cap for ELO win/lose at 20:10 is not randomly chosen. It just mirrors a maximum 66,66% winning chance for the stronger team for all games.
3.) The winning chances are below 66,66% because the outcome of a game depends on the one hand on skill and on the other hand on a good tower roll with the prophet builder, which is totaly random for every game. *Edit and the kings spell

If you want to go with a standard ELO, then do it with -+1.

The problem with +/- 1 points per win is that we need a round robin system for that. Means that every vouched player fights every other vouched player 1, 2, 3, ..., n times. Else the points are gained with different difficultis and cant be worth the same. I dont want to go into detail here, but if you dont understand that feel free to pm me or ask specificly in another post.
Also 20:10 is no "standard" ELO. The standard ELO for a ballanced game will remain 15:15 if both teams have the same average ELO. 20:10 is just the maximum for unballanced games.
______________________________________________________

@art)y I hope I could help you with your concerns. If you have other suggestions, feedback other critics feel free to post them in this thread.

[aRt)Y]

The standard ELO for a ballanced game will remain 15:15 if both teams have the same average ELO. 20:10 is just the maximum for unballanced games.

If 15:15 remains for the "balanced" games (and it's basically the same as 1:1), what about the issues you mentioned about the 1:1 elo distribution? Or is there any major difference between these two except that 15 is higher than 1?

[ILOCOS_NORTE]

@greatbeyond @armedanddeadly @hutzu

I would please you to post whether you support this or not. I will care for the rest.

thx

[ILOCOS_NORTE]

If 15:15 remains for the "balanced" games (and it's basically the same as 1:1), what about the issues you mentioned about the 1:1 elo distribution? Or is there any major difference between these two except that 15 is higher than 1?

It does not matter if we play for 1 ELO or for 15 ELO, thats just a multiplication by "15". But all Ent games are calculated on the basis of 15 and a start ELO of 1000. Its better for the "eye" to see 1400 ELO and 800 ELO then 1021, 992, 1013, 1041, 979. We are used to this kind of ELOs. Mathematicly and "ballancewhise" there is no difference if we play for 1 ELO or multiply that by 15.

But we still had to ballance the 1:1 games . If one team has 1020 average vs 990 average it would be something like 1,3:0,7

If we play all games for 1:1 ELO no matter what players are on, a "good" player could easely farm 3 "bad" players by playing 2v2 only. On the long term he can reach an infinit ELO score. You see we need unballanced games to be not 1:1.

Example:
4 high ELO players play a 2s. The winners receive 1 point.
1 high ELO and 3 low ELO playing a 2s. The winners receive 1 point as well, althought it was harder (for the weak team) or easier (for the strong team) to gain that point.
In result its unfair if both points have the same value.

[aRt)Y]

Since the 15 is close to what we currently have, would you think it could be adjusted within the season without disturbing the current stats? Or when did you plan should it be implemented?

[ILOCOS_NORTE]

Since the 15 is close to what we currently have, would you think it could be adjusted within the season without disturbing the current stats? Or when did you plan should it be implemented?

It can easely be implemented within the next days/week since the ELO of top and bottem is currently very low and no 20+ : 10- games happened so far.

[aRt)Y]

Without taking the poll into consideration from this topic (https://entgaming.net/forum/viewtopic.php?f=83&t=54688), do you think your suggestion would solve the fears voiced in that discussion?

[eldryan]

Without taking the poll into consideration from this topic (https://entgaming.net/forum/viewtopic.php?f=83&t=54688), do you think your suggestion would solve the fears voiced in that discussion?

they're voiced off the same topic, and this idea was actually originally suggested by v1rus and recommended by diablo_ in one of these threads off this topic.

20/10 is indeed arbitrary. his comment to uakf.b not withstanding as no 1500 ELO players are with 950 ELO players against 1000-1100 players in 2v2s - it's not viable for ELO gain as he said, so isn't within the statistics. the 66% figure based off high ELO players dodging 2v2s and only playing with low ELO players anyways to keep gains 15/15 doesn't reflect a true ELO (eg. one with scaling limiting the imbalance)

the suggestion is one that will benefit the 2v2 ELO as a viable method of grinding post 1300.

the 20/10 ELO limit could be exploited heavily in 4v4 - if you put a team of 1400+ experienced players on one side, and only players who've never reached 1100 on the other I would expect a win-rate of at least 80%. this won't happen without challenge mode, but wow swap-abuse will be powerful.

[supersexyy]

An easier to understand version of scaling, but an improvement all the same.

[pewpew lasergun]

a 2s at 25/5 elo for example is perfectly balanced. its just means at least one player is very good and others from opposite are bad.
however it is not a game worth playing for the 5 elo team ( risk vs reward ), should they care about elo.

the ilocos suggestion is valid and will make those players with high elo play 2s, since 10 elo would be an acceptable risk/reward.
!support