Board Game Arena

Wreckage wrote: ↑19 December 2025, 23:19 DBJ is the Goat, I’m sure you think you are the best player. You may be. Your stats are impressive. If you play Baptlens 60 games, and won 35 and lost 25, you would gain 100 ELO above your current score. If you did the same thing against Chadmild, you would be up nearly 200. If the top 10 all play each other 60 games each, and anyone wins 60% of those games, they would easily be 500 ELO higher. The leader among you could probably maintain ELO in the 1200s or higher, unlike those of us playing 5p, whose scores are in the 500s.

So, we know how high players can get in 5p for sure, we are seeing it continually. We absolutely don’t have any idea how high 2p scores can go. I suggested if the top 10 play each other, the leader could be 500 higher, but what if there were 20 or 30 of you, and you play each other? A modest win % would be enough to go even higher.

I don't think the players' ratings would have changed much if they were playing each other, and certainly not up to 1,000 elo. Getting a rating when playing with average players will not differ much from the rating when playing only with strong players, and in theory it should not differ at all. Some difference may appear when playing with very weak players, and in this case, when playing with strong players, the rating will only drop.

Andrei3009 wrote: ↑20 December 2025, 00:17
I don't think the players' ratings would have changed much if they were playing each other, and certainly not up to 1,000 elo. Getting a rating when playing with average players will not differ much from the rating when playing only with strong players, and in theory it should not differ at all. Some difference may appear when playing with very weak players, and in this case, when playing with strong players, the rating will only drop.

I said they would gain 100 elo by beating another 700 player 10 more games than they lost. Which is absolutely correct. In a 2p game, if you beat a player with the same elo as you, you will gain about 10 elo for that 1 win. And since Baptlens has higher elo, DBJ would get more like 13 elo for that first win.

Wreckage wrote: ↑20 December 2025, 02:10

Andrei3009 wrote: ↑20 December 2025, 00:17
I don't think the players' ratings would have changed much if they were playing each other, and certainly not up to 1,000 elo. Getting a rating when playing with average players will not differ much from the rating when playing only with strong players, and in theory it should not differ at all. Some difference may appear when playing with very weak players, and in this case, when playing with strong players, the rating will only drop.

I said they would gain 100 elo by beating another 700 player 10 more games than they lost. Which is absolutely correct. In a 2p game, if you beat a player with the same elo as you, you will gain about 10 elo for that 1 win. And since Baptlens has higher elo, DBJ would get more like 13 elo for that first win.

I'm not arguing with the elo calculation formula, but the key word is "if"". What makes you think he can beat this player with such a win rate, and if he could, why didn't he already get that 100 elo by playing with average players?

You are absolutely right. I did not make it clear enough in my last post that I did not want to argue 2p was easier to climb higher in than 5p. However, it is the why that is tripping me up so much. Your claim is that it is because of the system, but I don't think that is necessarily true.

Random thought: I actually joined Baptlens game once, but was declined. The fact we don't play does not really benefit 5p players much though, as you explained with our peaks being higher, and whoever drops would play to get back to the same spot. The fact we don't play each other is a valid point, but there is nothing I can do about it alone.

I do not consider myself the best overall. As you mentioned, you play the best of the best. It is very difficult for either side to argue they are better in my opinion. The 100 elo I "don't have" is simply because I had to stop playing. I never was in a slump or hardstuck at any elo point. My win rate was more than high enough for me to consistently gain until I was at least 850 imo. For the record, I do think I can beat Baptlens at a rate of at least 35:25

Why it is not the system that is the issue: As you mentioned, your win rate against these extremely good players is high, and is not 90% for reasons that are not necessarily your fault. However, because of that specifically, you can't really say the system needs to improve to make lower win rates result in higher rewards. The problem is this is not a "league" where the best is showcased out of the games displayed, which is what 5p experts would be experiencing if they were the top. Elo is an evaluation of skill, or rather a way to calculate one's chance to win against another elo, which means while win rates like 55-65% against strong players in 5p is certainly impressive, it still means you are expected to win 55-65% of the time, and your elo will reflect that.

I want to apologize, I took the situation too abstractly and didn't study the players enough, DJB is da GOAT and really Goat with an incredible percentage of wins and he can certainly rise even higher.

I think you focus a lot on how many ELO you win, but the problem is the probability of wining in a Row, and that why we can't, as 5p players, be permanently at the top.

Wreckage is one (and for me probably) the best 5p player, and he can't stay in the 5p 3th leaderboard. Sometimes we play together, and just that fact can make us fall together in ELO.
And today i'm 7th on the leaderboard with some works... and the 6 on top are 1v1 players, and i know i will never reach top 3, even if i play against small players cause i can't win 80% of my games.

look achievements of 5p players, i'm not sure there is a single one with a 10 wins streak. Look now 1v1 players... WE can't beat that

Congrats Alesk for reaching #1! It shows you as 6th, but you are actually at the top. well done.

Always the same complains here with calculations ans statistics in part wrong

Here are my reply to your ideas/comments :

- "the fact they are "master" is because they're playing against low ranks (<200 ELO)" ===> Personnaly, i played several big tournaments with good players so i didn't totally build my Elo thanks to weak players.
This sentence is wrong. You can only play 1 vs 1 against good players and get a 700 Elo if you are a strong...You earn Elo points depending on the % chances of win. The expected value is the same against a bad player or a strong player if the % chances of win calculated by BGA is fair. If I lose against a 150 Elo player, i'll need to play and beat this players 21 times in a row to recover my points...
I would even say that's disadvantageous to face this kind of players. A BGA player could be 150 after few games only, he can know very well and be good at this game in real (6nimmt is a well-know game all around the world), so my real % chances of win is overestimated. When you face a 300 Elo player, this guy has already played lots of game so his Elo reflects +/- his real level at this game.

- "if they stay at top is because they reached it a long time ago, they disapear in 3 months, play a single game, winning as always, and dont touch the game anymore for the next 3 months" ===> this is a all-time ranking so that's logical to display the players who get the most points ever
BGA changed the rule 2 years ago i think with the necessity to play a game every 3 months to appaear in the standings.
What do you want ? A weekly standing ? A monthly standing ? A year standing ?

- Create 2 standings (1 vs 1 and 3+ players) : lots of games here have different strategies/rules depending on the number of players. Do you want BGA to create multiple standings for each extension, game etc ? I want a standing for 3p games in Catan, 4p games in Quoridor etc

Don't sweat your elo and just play the game to enjoy it. If you want a more competitive mode, create a tournament or try arena.

Elo is a flawed system for games with a luck component, although this game is relatively high skill. Elo is also a flawed system for a multiplayer game, because it can lead to weird incentives like playing for second.