Board Game Arena

And win percentage means nothing. If you play 1 game of something with 5 players in it, and win it, congratulations you now have a win percentage above 200%

Wreckage wrote: ↑18 January 2021, 19:46 In 6 Nimmt,
If a player plays 2-player games and has wins of 90% they can stay about 600 ELO above their average opponent.
If a player plays 5-player games and has wins of 90% they can stay about 100 ELO above their average opponent.

i think to be honest a game of nimmt for 2 players? It might be boring as hell.. and very luck dependent instead of strategy.. but you are right on this point

Caninski_Jen wrote: ↑24 January 2021, 19:56 i think to be honest a game of nimmt for 2 players? It might be boring as hell.. and very luck dependent instead of strategy.. but you are right on this point

The opposite. 2-player is less luck dependent and far easier to predict.

Cheery Dog wrote: ↑22 January 2021, 12:58 And win percentage means nothing. If you play 1 game of something with 5 players in it, and win it, congratulations you now have a win percentage above 200%

True, the % doesn't go from 1-100 like we usually understand % to mean, but it still means the player with the higher number is winning a higher percentage against opponents than someone with a lower number.

N_Faker wrote: ↑25 January 2021, 12:04

Caninski_Jen wrote: ↑24 January 2021, 19:56 i think to be honest a game of nimmt for 2 players? It might be boring as hell.. and very luck dependent instead of strategy.. but you are right on this point

The opposite. 2-player is less luck dependent and far easier to predict.

True, but the game is reduced to a very simplistic form. There are many skills of the game that are lost on a 2 player table. Including predicting situations arising at a much faster rate, minimizing risks when the outcome is uncertain, and getting your card to land in a specific place on a row (to control future plays).

Something to think about the current ELO system;
If all your opponents have the EXACT same ELO as you, and you win a 5 player game, you will go up about 25 ELO.
If all your opponents have the EXACT same ELO as you, and you win a 10 player game, you will go up about 25 ELO.

Simple question, is this fair?

Whether it goes up 1 or 25 or 100 doesn't matter, except for volatility.

This is my first time to use the forum of Board Game Arena. Thanks Wreckage for inviting me to give some comments to this issue.
After reading his post, i totally agree with him. I eventually realize that, why some players top the list who get so high ELO are so unfamiliar. I never play with them. I know the reason now. Although the ELO is not so important for me. I still support Wreckage for a fair ELO system about the # of player in the game.
Thanks you Wreckage again for this action. Support Fairness.

Tulivuori wrote: ↑27 January 2021, 21:41 Whether it goes up 1 or 25 or 100 doesn't matter, except for volatility.

The player number actually matters, not really for volatility but because of opponent selection.
Just check the game records of 'top players'.
You'll find out (not all but) most of them earning ELO from significantly less skilled players only.

ufm wrote: ↑06 February 2021, 20:02

Tulivuori wrote: ↑27 January 2021, 21:41 Whether it goes up 1 or 25 or 100 doesn't matter, except for volatility.

The player number actually matters, not really for volatility but because of opponent selection.
Just check the game records of 'top players'.
You'll find out (not all but) most of them earning ELO from significantly less skilled players only.

Yes, it is easier to select newbies in 1vs1 than in multiplayer. But adding volatility to multiplayer ELO won't change it much.

Ideally, the ELO for multiplayer games should be calculated as if you played 1v1 games with everyone at the table; everyone ranked below you you "beat" while everyone above you you lost to.

That's the only fair way to compare the apples and oranges of 1v1 vs multiplayer. A player who outright wins 4 4p games has done the equivalent feat as a player who has won 12 1v1 games.

First, We talk about <Chadmild>.
(My following opinions in this section are based on <2 players normal set>)
<Wreckage> said <Chadmild> almost only plays with rookies to increase his ELO. That is the truth, but so what ?
There is nothing unfair for the other players , <Chadmild> deserves the ELO. Here are my reasons:

<Chadmild> is very easy to win against the rookies, it is true, but not 100% win ratio.
How is the win ratio? BGA gives us the answer , for example in this gamehttps://boardgamearena.com/table?table=148176381,
the probability of the win is 98%, very high right ? But here is also a cost , If he wins the game , he can only increase 0.32 ELO.
If he win 98 times can only increase 31.36 ELO. But if he lose the game very unluckily , how many ELO will he lose ? At least 20.The 98 wins can not pay for the 2 losses.

Maybe you will say that 98% is not correct , the real possibility of win is higher , such as 99.9%.
As a high level player ,I think:

• Please respect the win ratio calculation system designed by BGA, it is scientific.

• 98% = <ELO 716> vs <ELO 0>, so if you think the real win ratio is higher, maybe you are right! But it doesn't means the system is unfair, just the opposite, it means <Chadmild> still hasn't reached the ELO which fits his real ability.

And in my own experience, it is really impossible to totally avoid loss in <6 nimmt!>. There are 2 reasons:

• This game has a percentage of luck. If your cards in hand is always very bad, for example <1,4,7,13,22,88,92,97,101,104>, It is very hard to win unless your oppenet is totally a real rookie.

• A low ELO player is maybe not a real rookie , even maybe he is a new account of a very strong player. So sometimes the tiny ELO does not deserves the risk.

In my opinion, to play with the players with 200-300 ELO and 100% Reputation is a better idea. the <risk-reward ratio> is higher. There are also 2 reasons:

• when the ELO of a new account of a strong player is at 200-300, his Reputation is impossible 100%.When he reaches 100% Reputation, his ELO must already be higher than 300. So I can absolutely avoid to meet them.

• Can get much more ELO(at least triple) per win , but only a little higher of risk, at least the risk deserves the ELO I can earn. And for me , there is almost the same difficulty between win a normal player and a rookie.

Second, We talk about The Comparing Between <2 players set> And <5 players set>.
I do not think it is unfair for the players in <5 players set>.Because the ELO calculation system in a <5 players set game> is actually almost same with a <2 players set game>.If you are a participant in a <5 players set game>, at the sight of system , you are playing 4 independent games with 4 players at the same time.
For example in this gamehttps://boardgamearena.com/table?table=138855369
<bluen602> has 355.49 ELO before the game, then be the 4th place and lose 21.6 ELO, how is the 21.6 composed with? The system told us the details:

• <bluen602> has 80% probability to win <niunia0722(116.88 ELO)> but <bluen602> 's place is lower than <niunia0722>, so system judged <bluen602> lost the independent game between these two players and his ELO -15.96

• <bluen602> has 57% probability to win <Vinicios Chen(306.85 ELO)> but <bluen602> 's place is lower than <Vinicios Chen>, so system judged <bluen602> lost the independent game between these two players and his ELO -11.39

• <bluen602> has 55% probability to win <samazul44(323.54 ELO)> but <bluen602> 's place is lower than <samazul44>, so system judged <bluen602> lost the independent game between these two players and his ELO -10.92

• <bluen602> has 81% probability to win <TiDan(98.1 ELO)> and <bluen602> 's place is higher than <TiDan>, so system judged <bluen602> won the independent game between these two players and his ELO +3.7

-15.96/1.6 - 11.39/1.6 - 10.92/1.6 + 3.7/1.6 = 21.6
Maybe you will ask: what is the "1.6"? It is just a coefficient, and the coefficient is fair to every player whoever winners or losers.
As the same way of calculation: <samazul44> = 4.67/1.6 + 9.52/1.6 + 10.92/1.6 + 4.29/1.6 = 18.37
See? very fair, so do not care of the coefficient please.
Then next step, look at this game:https://boardgamearena.com/table?table=133796443
The system show us about the two players:

• <bluen602> has 351.69 ELO before the game (very close to 355.49)

• <BigVerm> has 100 ELO before the game (very close to 98.1)

• <bluen602> has 81% probability to win <BigVerm> and <bluen602> won, then his ELO +3.8(very close to 3.7)

So the finally conclusion is:
The only difference between <2 and 5 players set> is: Each independent game has a lower ELO variation.
What does it mean? It means if you consider one <5 players set>game as 4 independent 1vs1 games,:

• you will win more ELO when you win a game.

• you will lose less ELO when you lose a game.

And the proportions of increase and decrease are the same (1.6), so , very fair.